key: cord-0006404-w5xaq43c
authors: nan
title: Deutsche Gesellschaft für Pharmakologie und Toxikologie Abstracts of the 34th Spring Meeting 16–18 March 1993, Mainz
date: 1993
journal: Naunyn Schmiedebergs Arch Pharmacol
DOI: 10.1007/bf00871888
sha: c0ebd086b04e72135b905a174c10892de417d29e
doc_id: 6404
cord_uid: w5xaq43c

nan

The measurement of certain model reactions is a simple, very sensitive and a relatively specific method for the determination of the activity of different P-450 forms. Among 7alkoxycoumarins, 7-ethoxy-and 7-propoxycoumarin have mostly been used as model substrates. In our experiments 7-ethoxy-, 7-propoxy-and 7-butoxycoumarin O-dealkylation have similar characteristics: the same induction by phenobarbital (PB) and B-naphthoflavone (B-NF) in 10-and 18-day-old rats, whereas in older ones the induction by ~-NF prevails over that by PB. The rates of 7-methexy-, 7-pentoxy-and 7benzyloxycoumarin O-dealkylation, however, are preferentially induced by PB in all age groups, but a certain induction by ~-NF is also detectable, at least in older rats. Depending on the position, additional groups at the benzene ring of 7-benzyloxycoumarin modify 7-O-dealkylation rate. 7-isoalkoxy-and 7-n-alkoxycoumarins also differ in their metabolization. The introduction of an additional 4-methyl group causes the most important alterations: it changes the basic metabolic rate and the inducibility of O-dealkylation considerably. The induction by B-NF is not possible, whereas PB markedIy increases dealkylation rates. Some 7-alkoxy-4-methylcoumarins have advantages over classical substrates for the specific detection of the induction type. Recent studies have demonstrated that ethinylated synthetic steroids inhibit in vitro human liver cytochrome P-450 3A enzymes [Guengerich FP (1990) Chem. Res.Toxicol.3:363; BScker R et al., (1991) Advances Contracept. 7:Suppl 3,140; BOcker R et al. (1992) Klin.Pharmakol.aktuell 3:56]. Among these steroids gestodene is the most potent inhibitor in vitro with an apparent IC5o concentration of less than 4 ~M (in some cases). Systematic investigations on the additional influence of various substituents at 3 distinct positions of the synthetic steroid (derived from norethisterone) demonstrated the exceptional inhibitory potency of gestodene. These studies were done as described earlier to investigate a mechanism based inhibition and they indicated an inhibitory effect of the 18-methyl moiety and the 415 double bond of the 17c(-acetylenated compounds when both structural elements are present in the molecule. When position 17c( is substituted by a cyano-methyl group like in dienogest an inhibition of the P-450 3A activity cannot be seen even at very high concentrations in the preincubation mixtures. The effect of TCDD and of benzo[a]anthacene (8A) on the transcription of the CYPIAI gene was studied in extrahepatic cells derived from ram seminal vesicles (SEMV) which contain high levels of PHS (Freyberger et al, 1987, Mol. Toxicology ~:503-512) . Induction of CYPIAI was measured by ethoxyresorufin-O-deethylase (EROD) activity. SEMV cells lack constltutively expressed monooxygenase activity, and EROD activity was barely detectable (~ 0.05 pmol.per minx I0 B cells) in uninduced cells. EROD activity Is dose-dependently induced by TCDD and BA between 0.01-10 nM and 1-10 ~M, respectively, albeit the maximal activities remain low (0.5-0.? pmol per min x 10 B cells). The concentration dependent effect of the two inducers suggests that the induction process is mediated by the arylhydrocarbon (Ah) receptor. After treatment of SEMV cells with 3H-TCDD for 2 h, the Ah receptor was detectable in the nucleus by sucrose density gradient centrifugation. Incubation of SEMV cells with TCDD up to 10 nM showed no sign of toxicity, e.9. inhibition of cell proliferation. In contrast, proliferation of SEMV cells was moderately stimulated by TCDD at concentrations found to induce ER00 activity. We conclude from our study that despite the ability of SEMV cells for EROD induction its drug-metabolizing potency is dominated by PHS-mediated oxidation reactions. On the other hand, because of the presence of the Ah-receptor SEMV cells seem to be a useful tool for studying the effect of TCDD and related compounds on the regulation of PHS. A cDNA encoding the human CYP 1A1 was recombined with a SV40 early promotor containing plasmid DNA for gene transfer and expression in V79 cells. CYP 1A1 is of particular toxicological relevance as it metabolically activates procancerogenic polycyclic aromatic hydrocarbons. Several cellular clones were obtained and identitied for containing the CYP 1A1 cDNA in their chromosomal DNA. Expression of this cDNA was detected by Northern blotting. Activity in these V79 derived cell lines was checked by ethoxyresorufin-O-deethylase (EROD) and aryl hydrocarbon hydroxylase (AHH). Activities ranged from 5 to 90 pmol/min/mg total protein for EROD and from 3 to 30 pmol/min/mg total protein in the AHH assay depending on the cell clone tested. The ratio between EROD / AHH was found to be a constant of about 2.5. HepG2 and other human cell lines revealed a ratio ol EROD / AHH of 1 indicating another cytochrome P450 expression and not just human cytochrome P450 1A1. The human CYP 1A1 expressing V79 ceils were validated by cytotoxicity testing and micronuclei formation by benzo [a]pyrene, trans-7,8-dihydroxy-7,8-dihydrobenzo[a] pyrene, and 1,3dinitropyrene. Effects were observed starting at a concentration as low as 0.01 I~M. A protecting effect was observed in cytotoxicity and micronuclei formation by piperine contained in pepper species such as black pepper. Further studies are being performed to demonstrate differences of species-specific CYPIA1 activities between rat and man. Carbonyl group containing compounds are widely distributed in organisms and appear as steroid hormones, mediators, transmitters and intermediates. In addition, many xenobiotic carbonyl compounds, including pharmacologically relevant substances, are ingested by humans and have to be metabolized and eliminated. The reductive biotransformation of these compounds to the respective alcohol metabolites is carried out by oxidoreductases which are localized in cytosolic and/or microsomal fractions of the cell.

In previous investigations we established the involvement of several steroid dehydrogenases in xenobiotic carbonyl metabolism, which belong to the short chain alcohol dehydrogenase (SCAD) family. This family comprises up to date about 20 known pro-and eucaryotic members, which can be deafly distinguished by the absence of Zn++-ligands and smaller subunit size from the other known alcohol dehydrogenase families. Originally they seem to fulfill catabolic functions in the respective organism but during evolution their function seems to have changed to a regulatory role in hormone and signal physiology.

In this study we report on investigations with purified 3c~-hydroxysteroid dehydrogenase (HSD 28) from Comamonas testosteroni. This steroid inducible enzyme performs reversible reduction of the oxo group at C3 of the steroid nucleus in a great variety of steroids with different steric conformations. In addition, kinetic analysis shows that HSD 28 is also capable of promoting the reversible oxidoreduction of xenobiotic carbonyl compounds as well as being able for alicyclic hydrocarbon degradation.

In conclusion, the occurrence of this procaryont in the environment as well as in the gut of vertebrates, together with its ability for enzymic degradation of steroids, alicyclic hydrocarbons and xenobiotic carbonyl compounds suggest a significant role of HSD 28 in detoxication processes.

Dept. Pharmacol., Philipps-University, Lahnberge, D-355 Marburg. OF Cytosolic epoxide hydrolase (cEH) is an enzyme involved in the detoxication of a variety of potentially carcinogenic epoxides including transstilbene oxide. It displays a low level of expression in the livers of untreated rats. Efficient induction of cEH may be achieved by hypolipidemic agents such as clofibrate and tiadenol. Attempts to clone cDNAs coding for cEH using specific antibodies as probes for immunoscreening have failed in our hands. In order to obtain useful DNA probes for screening we have isolated cEH protein from livers of tiadenol-treated Sprague Dawley rats and analyzed part of the amino acid sequence of peptides obtained by endoproteinase Glu-C digestion. The resulting sequence information served as the basis for the construction of sense-oligonucleotides that were employed as specific primers for the amplification of cEH cDNA fragments from reverse transcribed liver mRNA from tiadenol-treated rats. In each case, dT-homooligomers were taken as antisense-primers. One of the experiments resulted in the specific amplification of a 680 bp fragment that encodes for the carboxyterminus of cEH as evidenced by sequence analysis. This fragment was subsequently used to screen a cDNA library constructed from the reverse transcribed mRNA (see above), and a specifically hybridizing 1992 bp cDNA clone containing an open reading frame of 1662 bp was isolated. The calculated molecular weight of the encoded protein is 62.3 kDa, which is in good agreement with the reported molecular weight of cEH of about 59 kDa. No homology to other published cDNA sequences was apparent on homology search of the EMBL data bank. An especially interesting feature of the deduced protein sequence is an imperfect peroxisomal targeting sequence at its carboxyterminus, that perfectly explains the simultaneous localization of cEH in both, the cell cytosol and the peroxisomes of rodent liver. 

Except oxidation also reduction plays a vital role in xenobiotic biotransformation. Whereas nitro and azo compounds are reduced via the microsomal monooxygenase system, a separate group of enzymes, the aldo-keto reductases, is known to mediate the carbonyl reduction of aromatic and aliphatic aldehydes and ketones, including quinones. In addition, dihydrodiol dehydrogenase, NAD(P)H:quinone oxidoreductase and several hydroxysteroid dehydrogenases (3a-,31~-, 17B-) were described to be involved in reductive xenobiotic carbonyl metabolism. In this study we purified and characterized a carbonyl reducing enzyme (34 kDa) from mouse liver microsomes. The subcellular localization in the endoplasmic reticulum, the molecular mass of 34 kDa and the ability of this enzyme to catalyze the reversible l ll3-oxidoreduction of glucocorticoids suggest this enzyme to be l l/~hydroxysteroid dehydrogenase, the physiological function of which is glucocorticoid activation and inactivation. Enzyme kinetic studies revealed that the xenobiotic carbonyl substances pnitroacetophenone, p-nitrobenzaldehyde and metyrapone are also potent substrates of this enzyme. Until now, several steroid dehydrogenases, which catalyze the hydroxyoxidation or oxo-reduction at C-3 and/or C-17 of the steroid nucleus are described to be also involved in reductive xenobiotic carbonyl metabolism. This is the first report of 11J~-hydroxysteroid dehydrogenase being capable of xenobiotic carbonyl reduction.

Taken together, a membrane-bound carbonyl reducing enzyme has been purified from mouse liver endoplasmic reticulum and is demonstrated to catalyze the reversible 11B-oxidoreduction of glucocorticoids. On the other hand, this enzyme mediates the reductive metabolism of xenobiotic carbonyl compounds, and probably plays a role in detoxication processes. Thus, it shows properties similar to the previously described 3c~-, 3~-and l%-hydroxysteroid dehydrogenases and contributes to an expanding list of pluripotent enzymes involved in reductive xenobiotic carbonyl metabolism as well as being specific towards their physiological steroid substrates. Dept. Pharmacol. ,Philipps-University,Lahnberge,D-355 Marburg.

1. Brockmtller, R. Kerb, R. Kaiser, B. Staffeldt, and I. Roots Many genetically polymorphic foreign compound metabolizing enzymes can now be studied on the molecular genetic basis. Furthermore, mutations in oncogenes or in the tumor-suppressor gene, P53, may represent fingerprints of initial mutational events, reflecting exposure to certain carcinogens, but also reflecting individual differences in activating or detoxifying enzymes. These new possibilities were included in a case-control-study on the impact of polymorphic enzymes of foreign c/ompound metabolism on the individual susceptibility towards urinary bladder cancer. By enzymological and genetical techniques, defidency of glutathione S-transferase class Mu (GSTM1, with allelic variants la and V) was detected in 61% of 193 bladder cancer patients contrasting to a proportion of 51% in 240 reference patients (p=0.025). Thus, deficiency of GSTM1 may predispose to bladder cancer. Phenotypically and genetically characterized deficiency of debdsoquine hydroxylase (CYP2D6) was not associated with the overall bladder cancer risk. The Table  shows 

As an alternative method to study the induction of Cytochrome P450 enzymes in vivo, we investigated the in vitro induction of Mixed Function Oxidases (MFO) in isolated hepatocytes. The cells were culivated under different conditions, such as Williams Medium E (WE), Williams Medium E with 0,5 mmol/L Metyrapon (WE/M) and modified Earls Balanced Salt Solution (EBSS).MFO activities were determined with the substrates ethoxyresorufin (detecting P450-IAI), pentoxyresorufin (detecting P450-IIBI), and benzyloxyreeorufin (detecting phenobarbitone-and 3-methylcholanthrene inducible isoenzymes).The rate of ethoxyresorufin O-deethylation (EROD) found in cultured hepatocytes after 24 h was the same under all culture conditions. Pentoxyresorufin O-depentylation (PROD) and benzyloxyresorufin O-debenzylation (BROD) were weak in WE cultured hepatocytes after 24 h; in the presence of WE/M the activities were nearly unchanged as compared to the control (untreated hepatocytes). Furthermore, isolated hepatocytes did not loose the ability to respond upon inducing agents, such as phenobarbitone (PB) and 3-methylcholanthrene (3-MC). -Hepatocytes cultured in WE with 1 and 2 mmol/L PB show an increase of EROD (1,9-and 1,4-fold) , of BROD (8,2and 6,5-fold), and of PROD (l,4-fold and no increase). -Hepatocytes cultured in EBSS show a similar but not identical picture, whereas in hepatocytes cultured in the presence of 1 and 2 ~mol/L 3-MC the EROD and BROD are increased; however, the PROD remained unchanged after this treatment.

Department of Pharmacology and Toxicology, University of Ulm, Albert-Einstein-Allee II/N-26, D-7900 Ulm/Donau, Federal Republic of Germany Activities of epoxide hydrolases (EH) are usually determined by a radiometric assay. As an alternative an assay for the determination of EH activities in microsomal preparations using gas chromatography/mass spectrometry (GCMS) was developed. As substrate unlabelled styrene oxide (SO) was used, which is hydrolyzed by EH forming phenylethylene glycol (PG). Microsomal incubations with styrene oxide were performed at 37°C in phosphate buffer for i0 min prior to addition of phenylpropylene oxide (PP0) and phenylpropane glycol (PPG) as internal standards. PG. and PPG were extracted in ethyl acetate, followed by the extraction of SO and PPO in n~hexane. The organic phases were combined and concentrated and the amount of SO and PG determined by GCMS in the electron impact mode with single ion monitoring, thus allowing for the quantitative determination of SO and PG in a single run.

The reproducibility of the GCMS method itself resulted in a coefficient of variation of 0.7 ~ for the determination of SO and 1.5 Z for the determination of FG. The reproducibility of the complete method, including the extraction steps, resulted in a coefficient of variation of 2.9 ~ for the determination of SO and 7.0 Z for the determination of PG. The detection limits for SO and PG were about 5 nmol and I0 nmol/vial, respectively.

The activities were determined in microsomes isolated from male Fischer-344 rats, pretreated with phenobarbital (PB), acetone or saline. The activities were found to be 10.20, 3.66 and 5.71 nmol SO hydrolyzed/(min*mg protein), respectively. The corresponding values for the formation of PG were found to be 9.01, 4.28 and 4.59 nmol SG/(min*mg protein), respectively, thus demonstrating a good correlation with the rate of SO hydrolization. 3-Methylcholanthrene-inducible phenol UDP-glucuronosyltransferase UGTIA1 appears to be present in a variety of tissues . To investigate constitutive and TCDD-inducible expression of UGT1A1 a selective cDNA probe for its exonl was utilized. UGT activities toward 4-methylumbelfiferone and 6-hydroxychrysene were used as functional probes. Male and female Wistar rats (200g) were treated with TCDD (10#g/kg, dissolved in corn oil,s.c.) and organs were isolated after 7 days. Constitutive expression of UGT1A1 was low in fiver, but high in ovaries, testes, epididymides and kidneys, based on polyA RNA samples. Low UGT1A1 mRNA levels were also detected in lung and brain. UGT1A1 mRNA levels were clearly increased in liver (10-fold) and kidney (2-fold) after TCDD-induction. UGT1A1 activity toward 6-hydroxychrysene better reflects the induction of UGT1A1 than that toward 4-methylumbelliferone. The results substantiate the wide distribution of UGT1A1 in rat tissues. Previous studies suggested an interaction of neighbouring microsomal UGTs in diglucuronide formation of 3,6-dihydroxybenzo(a)pyrene and 3,6-dihydroxychrysene. To substantiate this suggestion mono-and diglucuronide formation of these compounds by 3-methylcholanthrene-indncible phenol UGTs were studied in situ utilizing radiation-inactivation analysis. Microsomes of 3-methylcholanthrene-treated rats were irradiated with high-energy electrons from a 10 MeV linear accelerator at doses from 0-25 megarad. UGT activity toward the diphenols of polycylic aromatics were determined by HPLC methods after detergent activation of UGT activity. Radiation-inactivation analysis revealed relative molecular masses of 48 ± 11 and 145 ± 8 kDa for mono-and diglucuronide formation of 3,6-dihydroxybenzo(a)pyrene, respectively. With 3,6-dihydroxychrysene its 6-monoglueuroulde (M6) and 3-monoglucuronide (M3) could be separated. Target sizes of 103 5:15 and 91 5:9 kDa were found for monoglucuronide formation of M6 and M3, respectively. However, a target size of 191 ± 34 kDa was found for dighicurohide formation. Target sizes for UGTs catalyzing glucurohidation of standard substrates such as 1-naphthol and 6-hydroxychrysene were 76 ~ 4 and 107 ± 8 kDa, respectively. Although radiation-inactivation analysis cannot determine Irue molecular weights, the results suggest 2-to 3-fold larger target sizes for UGTs catalyzing diglucurohide formation, compared with monoglueurohide formation. They confirm the hypothesis that diglucuronide formation is facilitated in microsomal membranes by interacting UGTs in protein clusters. In the present study we have extended our characterization of the human alpha class GST (glutathione S-transferase) gene family. We have utilized a rat Ya cDNA clone to identify the human GST A2 subunit gene. Nucleotide sequence analysis of the clone indicated that the gene spans a coding region of about 13 kb and consists of seven exons interrupted by six introns. The transcription initiation site has been determined by primer extension analysis. A "TATA"-sequence was found 26 nucleotides upstream from the transcription start site. It could be shown by reverse transcription assays, that the GST A2 subunit gene is expressed in human liver and HepG2 cells. A comparison of the structure of the A2 subunit gene with that of the A1 subunit gene revealed significant sequence identity between the two genes. Southern blot analysis of human genomic DNA was performed using exon 4 of the GST A2 subunit as a radiolabeled probe. Digestions with different restriction enzymes that did not cut in exon 4 showed several hybridizing fragments, indicating the presence of multiple genes for the alpha class. The present study was aimed at identifying the enzyme(s) which mediate(s) the activation of nitro aromatic and nitro heterocyclic compounds in mammalian cells.

TO this end we have used a panel of cell lines (chinese hamster lung V79-NH, mouse hepatoma BWIJ, rat hepatoma H4IIEC3 and 5L) which were assembled according to their specific expression of xenobiotic metabolizing enyzmes such as nitroreductases cytochromes P45~, or N-acetyltransferases. In these cell lines we determined cytotoxicity of the nitro compounds by measuring the uptake of neutral red.

The test cell lines showed a specific pattern in their response to some of the nitro compounds, in particular nitro polycyclic aromatic hydroearbons, indicating the nature of the enzymes involved inthe activation process. Thus compounds typified by 1,3-dinitropyrene appeared to be dependent on the presence of cytochrome P450 (IAI) for their activation, whereas others typified by 1,6-dinitropyrene required the activity of a specific nitroreductase and acetyltransferase. With a few exceptions, the nitro benzenes and nitro heterocyclics tested did not cause strong differential effects within the test cell lines. Apparently, these chemicals are activated by an enzyme, or enzymes, common to all four test cell lines. This excludes the enzymes involved in the activation of the nitro polycyclic aromatic compounds described above. Furthermore, it excludes xanthine oxidase and DT-diaphorase, which are either lacking in some of the test cell lines or differ in their activities by two orders of magnitude~ respectively. A possible candidate for the enzyme in question is the microsomal cytochrome c reductase which has been found in all established cell lines tested so far. In order to reduce the n~ of ~ni~qls, commonly used in biotransformation studies of xenobiotics, we compmed microsomes and hepatocytes from pig livers with microsomes and hepatocytes from livers of Wistsr-rats for their cyt. P-450 contents and their cyt. P-450 enzyme activities. Pig livers were obtained from castrated pigs (4-6 months old) at the local slaughterhouse, immediately after sacriflcation. Pa~s of the liver lobes (60-80g) were cut off, washed ~d perfnsed with ice-cold University of Wisconsin (UW)-solution. After ~ansport to the laboratory the pig he~tocytes were isolated by using a three step perfnsinn technique, based on the method described for the isolation of rat hepatocytes (P.O.Segleu, 1972, Exp.Cell.Res. 74, 450-454) . Microsomes from both species were prepared according to Guongerich (~.P. Guongerich,1982, in Principles and Methods of Toxicology, edt. A. Wallace Hayes, Rave~ Press, New York, 609-634) . ~ae specific cyt. P-450 contents of liver microsomes were lower in pigs than in rats: 0.53 + 0.15 nmol P-450/mg micvesomal protein for pig liver microsomes (mean + SD, N=7) and 0.64 + 0.17 nmol P-4501mg microsomal protein for rat liver microsomes (mean + SD, N=I0). This diife~m~e was more pronotmced when the cyt. P-450-and protein-content was measered in freshly isolated helmtocytes of both species: 0.18 ± 0.035 nmol P-450/106 viable pig hepatocytes and 1.10 + 0.12 mg protein/lO 6 viable pig hepetocytes (mean ± SD, N=8 live~s). For comparison the content of rat hepatocytes: 0.32 + 0.029 nmol P-450/106 viable cells and 2.01 ± 0.35 mg i~o~in/lO 6 viable cells (mean :h SD, N=8 rat livers). The enzymatic activities of 3 different P-450 enzymes (P-450 3A, P-450 2El and P450 2D) were measured for liver microsomes and hep~tocytes of both species by us~g specific substrates: 4-phonyl-2,6-dimethyl-3,5-bis~(methoxycarbonyl)-1,4-dihydro-pyridins (substrate concentration: 25-200~M), chlorzoxazone (125-10001~M) and dextromethorphane (12,5-100~M), respectively. Kinetic analysis of microsomal incubations showed 30% and 35% higher enzymatic activities (V n~x) for P-450 3A and P-450 2El dependent activities in microsomes of rat, compared to pig. The K M values of these mononxygensse activities were similar for both species and ranged between 100~M and 300~M. The kinetic parameters Vmax and KM of dexfromethorphane Odemethylafion by pig liver microsomes were about lO-fold increased in comparison to rat liver microsomes (Vmax: 32,5 nmol dextrc~pliane/nmol P-450*miu. (pig) |mean, N~] and 3,3 nmol dextrorplume/nmol P-450*min. (rat) [mean, N~] ; KM :190 ~ (pig) [me~m, N=5] and 18 ~M (rat) [mean, N~] . The results of this study indicate that microsomes and hepatocytes from pig livers may be suited as alternative in-vitro models for fmlher investigations in the biotransfonnation of xenobiotics.

Lehrstuhl flu" To~dkniogie und Pharmakologie dex Universltiit Erlangan, Universlt~t~stralle 22, D-g.~20 Erteng~n The industrial solvents n-hexane (HEX) and n-heptane (HEP) are metabolized to -/-diketones. Such compounds react with primary amino groups leading to pyrroles. HEX causes polyneuropathy which is attributed to the reaction of its metabolite 2,5-hexanedione (HDO) with amino groups of neurofilamental protein.

We have shown, that administration of HEX and HDO to rats resulted in the urinary excretion o! substances which alter addition of p-dimethylaminobenzaldehyde (DMAB) gave absorption spectra characteristic for pyrmles (pyrrol-like substances excretion, PLS). PLS can be considered to be a measure of the neumtoxicity of HEX [Kessler et al (1990) Amh Toxico164:242] . In this work we studied PLS in rats after inhalation of HEP with the goal to link PLS with the pharmacokinetics of HEX and HEP and to use PLS for comparing the neurotoxicity of both compounds. Pharmacokinetcs of HEX and HEP in male Wistar rats were already studied [Kessler et al (1989) NATO ASI Set A 159:123; Csan~dy et a~ (1992) 32 Jahrestag Arbeitsmed Gentner, Stuttgart:613]: For both compounds two different ways of metabolism were distinguished, one characterized by relative high affinity to the substrate and low capacity and a second one characterized by relative low affinity and high capacity. For studying PLS after inhalation of HEP, male Wistar rats were exposed over 8 h to constant HEP concentrations of 0, 2750, 4400, and 8650 ppm. Urine was collected for 72 h following start of exposure. PLS were determined photometrically after incubation with DMAB at the absorption maximum according to Kessler et al (1990) . As a quantitative measure for PLS the sum of the products of optical density and urine excreted [AE526 x ml urine] was used. PLS depended on HEP exposure concentration and correlated with the way of metabolism of HEP characterized by low affinity and high capacity. In contrast, P/S after exposure to HEX correlated with the way of HEX metabolism characterized by high affinity and low capacity [Kessler et a11990] , Compared to identical exposure conditions of HEX and HEP, PLS after exposure to HEP was distinctly smaller: Exposure to 2750 ppm HEP resulted in PLS of 0.095 [AE526 x ml urine], a value which was about hall of that obtained after exposure to HEX at its current MAKvalue of 50 ppm. Oxidative attacks at C2 ol HEX and C3 of HEP appear to be the rate limiting steps leading to 7-diketones and pyrroles. The metabolic way characterized by high affinity to the substrate and low capacity seems to represent the C2 hydroxylation, the metabolic way characterized by low affinity to the substrate and high capacity the C3 hydroxylation. From the PLS studies, we conclude that the neurotoxic potency of HEP in rats is by far lower than that of HEX. Currently, the transferability bt these results to the human situation is being investigated. GSF-Institut ffir Toxikologie, Ingolst~dter Landstr.1, D-8042 Neuherberg

M. Korn 1, W. GfrSrer~, U. Knecht 2 A high workplace concentration of ethylbenzene is a known health hazard to workers. In many countries its air concentration is limited to 100 ppm, which should be controlled by ambient air monitoring and personal air sampling. Determination of the inhaled solvent vapor equivalent in blood or in alveolar air or determination of the major metabolites mande/ic acid or phenylglyoxylic acid in urine samples are recommended for biological monitoring in man.

In urine samples of exposed persons mandelic acid occurs as enantiomeric Rand S-mandelic acid. The average amounts of R-mandelic acid and S-mandelic acid resulted in a S/R ratio of about 1 : 18. R-mandelic acid and pt3enylglyoxylic acid are preferably formed and excreted in humans. S-mandelic acid seems to be a minor metabollte in man. S-mandelic acid and R-l-phenylethanol are excreted in similar amounts. The height of S-l-phenylethanol excretion was only one third of the R-enantiomer.

In this study we analysed urine samples from occupationally exposed workers and experimentally exposed test persons with respect to phenylethylene glycol. Urine samples were taken at the end of ethylbenzene exposure. We determined the concentration of these metaholites by gas chromatographic procedures using a glass capillary column coated with the chiral stationary phase ChirasiI-L-Val. Details in sample preparation -f.i. the enzymatical hydrolysis -have been published previously (Kern et al., Arch Toxico158:110, 1985) . Urinary excretion of R-phenylethylene glycol and R-l-phenylethanol were below 10 mg/L The height of S-phenylethylene glycol excretion is smaller than that of the R-enantiomer, the S/R ratio is about 1 : 7. Formation and excretion of phenylethylene glycol enantiomers confirm the postulated scheme of biotransformation of ethylbenzene in man (Kern et al., Int Arch Occup Environ Health 64: 75, 1992) . Physiological pharmacokinetic models are used to calculate tissue burdens of xenobiotics and their reactive metabolites. Such models require knowledge of the partition coefficient tissue/blood (Pt/b) for the tissue or organ of interest. Our goal was to determine PCo-s for styrene-7,8-oxide (SO) in diverse tissues and organs of mouse, rat, and man. SO is a reactive, direct alkylating intermediate metabolite of styrene, which is an important industrial chemical. The volatility of SO is very low. Therefore, we developed an in vitro method allowing the direct determination of PCo-s for non-volatile substances. Laboratory animals were pretreated i.p. with diethyl maleate (DEM) (600 mg/kg) 15 rain before sacrifice to reduce GSH levels. Prior to incubation, tissues with the exception of fat from mice and rats were inactivated by heating to 80 °C for 10 min. Blood and tissue samples of man were additionally preincubated with l,l,l-trichloropropene oxide (10 lamoFL) in order to inhibit epoxide hydrolases. Blood samples were preincubated for 30 rain at 37°C with DEM (10 mmol/L). Then, to one aliquot (I ml) only SO was added, to the second one (1 ml) beth SO and a suitable dssue sample (fat, oesophagus: 0.1 g; other tissues: 0.5 g) were given. Concentration-time courses of SO were monitored in both aliquots following extraction with n-hexaoe using gas chromatography/mass speclrometry. Pt/b-s were calculated by means of a two compartment pharmacokinetic model from the resulting concentration-time courses. The resulting Pt/b-s of SO are summarized in the The Pt/b-s obtained from different tissues were similar among the species. From the similarity of the values it can be concluded that SO distributes nearly homogeneously within the body and accumulates only slighdy within fat. Incorporating the Pt/b-s into a physiological pharmacokinetic model allows to predict the pharmacokinetics of styrene oxide in mouse, rat and man for various exposure conditions. Experiments in vivo are indispensable for toxicokinetic reseamh. Such studies can result in stress to the animals and have increasing administrative requirements. To minimize these problems attempts were made to gain kinetic parameters using microsomes and cytosol of the metabolizing organs. However, it became evident that metabolic parameters, especially the Michaelis-Menten constant, determined in subcellular fractions are less suitable for the correct prediction of in vivo conditions than intact cells. Therefore, the goal of this work was to study the toxicokinetics of styrene (S) in hepatocytes of rats and mice and to compare the results with data obtained previously in vivo.

Metabolism of S is catalyzed by hepatic cytochrome P450 dependent monooxygenases. Therefore, hepatocytes of male Sprague-Dawley rats and B6C3Fl-mice were freshly prepared and suspended in Krebs-Henseleit buffer (5ml) containing 1% (w/v) albumin (pH 7.4). Incubations were done at 37 °C in gaslight glass vials of 25 ml after administration of defined amounts of S. Initial gas concentralions were between 43 and 2950 ppm. The elimination of S was followed withdrawing samples from the gas phase, which were analyzed by gas chromatography. The concentration-time courses of S in the atmosphere of the vials were kinetically analyzed lor the maximum metabolic elimination rate (Vmax) and the apparent Michaelis-Menten constant (Km) by means of a two compartment pharmacokinetic model [Filser (1992) Arch Toxico166:1]. Vmax and Km in hepetocytes of rats and mice, respectively, were (~ _+ S.D.) 3.1 + 0.3 [nmol/min/5 x 106 viable cells] and 25.5 + 7.8 [ttmol/L] and 11.4 + 0.7 [nrnol/min 5 x 106 viable cells) and 11.7 _+ 0.3 [ttmol/L]. Using the Vmax values the maximum melabolic elimination rates of S in vivo were calculated considering the amounts of cells in the incubations, the cell densities in livers of mice (103x106/g) and rats (143xl 06/g) and the liver weights of 1.4 g in a 25 g mouse and of 10 g in a 250 g rat. The scaled in vivoVmax values for one rat and one mouse were 53. Both hematotoxic and leukemogenic effects of benzene may result from the generation of benzene metabolites in liver, their transport to the bone marrow, entry into marrow target cells, and further metabolic activation. Since mice are more sensitive to benzene toxicity than rats, we measured the release of benzene metabolites from isolated bepatocytes of each species to determine whether differences in hepatic metabolism might ¢outribute to the difference in sensitivity. Total metabolism of 0.5 mM benzene in hepatocytes from Swiss-Webster and NMRI mice was 5 times greater than in those of Wistar rats. In mouse hepatocytes, a significantly higher rate of hydroquinone (NQ) formation and the formation of two polar metabolites not detectable in rat cells was observed. They were labeled by incubating with [~sS]-sulfute and were identified by mass spectrometric methods as HQ sulfate and 1,2,4-trihydroxy-benzene (THB) sulfate. Incubations of I00,000 x g supernatant of mouse liver with either THB or HQ in the presence of [~sS]-3" phosphoadenosine 5' phosphosulfate yielded labeled products which had HPLC retention times equivalent to the ['~SO4]-Iabeled metabolites detected in hepatocytes. Sulfates may play a critical role in benzene-indnced myelotoxieity as semistable transport forms of reactive phenols, which may be released in bone marrow by arylsulfatase, and oxidatively activated to yield toxic and leukemogenic metabolites.

Prostaglandin-H synthase (PHS) can activate the food pyrolysis product IQ and s t r u c t u r a l l y related heterocyc]ic aromatic amines to products which are mutagenic in Salmonella typhfmurium (Wi]d& Degen, 1987, fiarcinogenesis ~:541-545). The s t r u c t u r e -a c t i v i t y relationships and the response pattern of d i f f e r e n t S. typhimurfum strains are similar to those observed with monooxygenase activation yet a r e l a t i v e l y stab]e mutagenic a c t i v i t y is generated by PHS. To gain a better understanding of the mechanisms involved in the mutagenic activation of IQ we have now studied i t s PHS-dependent metabolism and characterized the products. When radiolabeled IQ Is incubated with microsoma] PHS supplemented with arachldonic acid, I~ is metabolized to protein-bound products. Noreover, incubation extracts analyzed by HPLC or TLC reveal that PHS oxidlzes IQ to a pink product (not observed in monooxygenase incubations). This major product has been i d e n t i f i e d by means of UV/VIS spectroscopy and mass spectrometry as.azo-I~, and i t also formed upon chemical oxidation of IQ with sodiumperiodate. Isolated azo-l~ is mutagenic in S. typhimurium strains TAg8 and more strongly In ¥GI024. In addition, trace amounts of another apparently very potent mutagen have been detected in PHS-incubations. Due to i%s chromatographic properties and high mutagenicity this product most l i k e l y represents nitroso-IQ which has not been described so far; also formation of n i t r o -I ¢ cannot be ruled out. Together these da%a point to PHS-actlvation as an interesting a l t e r n a t i v e mechanism which may be relevant f o r tumor induction by food-borne arylamines in extrahepatic tissues. . a distinct tumor induction in the urinary bladder which is related to formation of to-carboxylated metabolites. These polar intermediates are subject to low rate a-oxidation [Janzowski C., Jacob D., Henn I., Zankl H., Pool-Zobel B.L., Eisenbrand G., Toxicology (1989) 59: 195] or might act as substrates for mitochondfial I~-oxidafion. The DBN metabolite butyl-3carboxypropylnitrosamine (BCPN) is biotransformed into butyl-2-oxopropylnitrosamine (BOPN) when incubated with rat liver mitochondrial fraction (7000xg pellet) and ATP. Without addition of ATP, mitochondfial rates are drastically reduced; other cofactors (NAD, CoASH) do not exert significant effects. BOPN formation is inhibited in the presence of octanoic acid, supporting involvement of fatty acid metabolizing enzymes. Subfractionation of the 7000xg pellet and monitoring of mitochondrial and peroxisomal marker enzymes revealed BOPN to be almost exclusively formed in mitochondria. Rat kidney mitochondrial fractions also transformed BCPN to BOPN. Bladder carcinogenic methylalknitrosamines are supposed to undergo oxidative degradation via ~-oxidation and subsequent I~ -oxidation cycles. Two intermediates, methyl-3-carboxypropylnitrosamine and methyl-3-carboxypentylnitrosamine (MCPeN), are metabolized to methyl-2-oxopropylnitrosamine by rat liver mitochondrial fractions, MCPeN beeing more readily I~-oxidized.

In conclusion, II-oxidation by mitochondfial enzymes represents a relevant activation mechanism for highly water soluble carboxylated nitrosamines, generating Iipophilic metabolites readily activated by P-450 mediated monooxygenation to genotoxic electrophiles.

Department of Chemistry, University of Kaiserslautern, Erwin Schr6dingerstr., W-6750 Kaiserslantern, Germany

The cytostatic activity of 5-fluoro-2'-deoxyuridine (FUDR) is often limited by primary or acquired chemoresistance of tumor cells. This effect may be caused by a limited conversion to the active nucleotide metabolites because the major fraction of FUDR is cleaved to 5-fluoruracil (FU) and finally to the inactive compound o~-fluoro-g-alanine (FBAL), We compared the intracellular pattern of FUDR and metabolites in freshly isolated rat hepatocytes, Novikoff hepatoma and the human colorectal carcinoma lines, Cole 205 and WIDR. The cells were incubated with 300 I-tM 3tt-FUDR for 0.5 -180 rain. In hepatocytes the intracellular radioactivity almost exclusively consisted of FBAL indicating that the catabolic cleavage of FUDR by far exceeded the rate of uptake. In Novikoff hepatoma cells, however, significant amounts of FU (67 -76 %) and FUDR (9 -19 %) were observed intracellularly. Also in WIDR cells a substantial fraction consisted of FU (16 -38 %) and unchanged FUDR (27 -60 %). In Cole 205 tumor cells the intracellular composition of uachanged FUDR and its catabolic metabolites resembled those in WIDR cells. In rat hepatocytes, the degradation of FUDR could effectively be blocked by 5-diazouracil while d-inosin and benzylacyclouridine (BAU) had only minor effects. In Novikoff hepatoma cells BAU caused a 15 -20 fold increase of intracellular FUDR levels. Finally, in WIDR cells 5-diazouracil was the most effective modulator of FUDR metabolism. In summary, the balance between FUDR inflow and in~racellular metabolic elimination is different between hepatocytes and tumor ceil Iines and the modulation of FUDR catabolism might be a useful approach to increase the efficiency of cytostafic chemotherapy. Pharmacokinetic and toxicokinetic modelling is usually done without a detailed statistical treatment. Specifically, procedures of fitting models to experimental data like the frequently used "curve striping" often rely on the subjective experience of the user. An insight into the problems involved was sought by application of statistical procedures on the two-compartment model for inhaled gases and vapours by Filser (1992) : "The closed chamber technique-uptake, endogenous production, excretion, steady-state kinetics and rates of metabolism of gases and vapors". Arch.Toxicol. 66, 1-10. For volatile chemicals the "closed chamber technique", as reviewed by Filser (1992) , allows investigation of kinetics of volatile substances in vivo. The experimental design is based on a closed inhalation system where during the exposure period the atmospheric concentrations of the substance are measured. Based on this closed chamber technique, concentration-time data obtained by inhalation experiments in rats with propylene were analysed by Golka, K., H. Peter, B. Denk and J.G. Filser, (1989) : "Pharmacokinetics of propylene and its reactive metabolite propylene oxide in Sprague-Dawley rats". Arch. Toxicol. Suppl. 13, [240] [241] [242] The two-compartment model in the context of compartmental analysis can describe uptake, endogenous production, excretion of the (unchanged) substance and its metabolic elimination. The mathematical treatment of dynamic systems modelled by compartments typically yields systems of differential equations. Based on their solutions, the concentration-time data obtained by the inhalation experiments in rats with propylene under firstorder kinetics were analysed again using the methodology of nonlinear regression. The individually estimated values for the parameters by this analysis and the values of the literature differ considerably. To which extend this may be explained by insufficiency of the models, wrong supplementary assumptions, or by biological variability, remains to be elucidated.

Institut fiir Abeitsphysiologie an der Universit~t Dortmund, Ardeystr.67, W-4600 Dortmund, FRG

N. AI-Obaidi, U. Kastner, S. Klug, and D. Neubert We have developed a culture method which allows simulation of most processes involved in the closure of secondary palate (mouse and rat) in vitro using a chemically-defined, serum-free medium and suspension culture conditions. Craniofacial regions were dissected on gestational day 13 for mice, on day 15 for rats and were then cultured for 3-4 days. In order to assess its applicability with respect to testing compounds for their teratogenic potential, it is neccessary to obtain results from investigations with substances known to induce cleft palates (CP) in the rat and mouse. Therefore, some typical CP-inducing substances were tested and data obtained with all-trans retinoic acid (RA) and a triazole (2-(l-Methylsulfonylcyclopropyl)-1-(4-chlorphenyl)-3-(1,2,4-triazol-l-yl)-propan-2-ol (MP) in mouse palates are presented here: In order to elucidate the mechanism underlying selective pancreatic B-call toxicity of alloxan, we have used the inhibition of spermine-induced mitochondfial calcium transport in isolated permeabilized pancreatic Bcells as an indicator system for cytotoxicit~. Pre-incubation of isolated intact pancreatic islets from ob/ob mice wtth alloxan inhibited subsequent rnitochondrial calcium transport from permeabilized cells with a half maximal inMbitory concentration of 3.4 raM. Glucose and 3-O-methylglucose protected the B-cells against the inkibitory effect of alloxan. RINm5F insulin secreting tumour cells, which in contrast to normal pancreatic B-cells do not express the low affinity GLUT-2 glucose transporter in the plasma membrane, are resistant towards alloxan, The concentration of alloxan required for inhibition of spermine-induced mitochondsial calcium transport from these tissue culture insulinoma cells was 115 raM. When the effect of alloxan on calcium transport was studied in RINm5F cells transfected with the human GLUT-2 glucose transporter, these insulinoma cells regained much of the alloxan sensitivity as shown by a half maximal inhibitory concentration of 24 mM for alloxan. When we studied the effect of N-alkyl-substituted alloxan derivatives, we observed a decrease of the resistance of the RINm5F cells towards inhibition of mitochondrial calcium transport in parallel to the increase of the chain length of the substituent. Incidentally the increase of the chain length of the substituent is parallelled by a decrease of the hydrophilicity of the derivatives. As alloxan is a very hydrophilic compound the results support the contention that alloxan due to a molecular shape similar to that of glucose is transported into the pancreatic B-cell through the low affinity GLUT-2 glucose transporter, where it exlfibits its toxic action.

Institute of Clinical Biochemistry, Haunover Medical School, Konstanty-Gutschow-Str. 8, W-3000 Hannover 61, Germany 32 TOXIC EFFECTS AND PHARMACOKINETICS OF GANCICLOVIR IN THE TESTES OF MALE RATS Annette Klug, Ute Rahm, Gerd Bochert An adverse effect on male reproductive organs is one of the important problems regarding toxicity of xenobioties. The question of a sufficient dose to induce toxic effects in the testes and of the reversibility of such changes have to be taken in account when assessing the risk of an exposure. We performed a study with Ganciclovir (Gan), which is an acyclic nueleoside analogue with antiv/ral activity. The study was aimed at answering the following questions: -Is it possible to induce adverse effects in the testes by short exposure ? -Is there a reversibility of such an adverse effect? Adult male rats were treated with three single doses of 100 mg Gan/kg body wt s.c. on one day. The males were divided into six groups (n=ll). Weights of testes and number of spermatids per testis were investigated at 2, 4, 6, 8, 16 and 24 weeks after treatment, respectively. Controls were investigated in the same way. The table shows that the absolute and relative (% body wt) weights of testes and the number of spermatids per testis were significantly reduced from the fourth week on after treatment, when compared to the control animals. This effect was mostly pronounced eight weeks after treatment, and was still observed after 24 weeks. Pharmacokinetics were studied with a single dose of 100 mg Gan/kg s.c. The concentration of Gan was measured in plasma and testes. The highest concentrations in plasma were found between 90 and 120 min after treatment. At this point the amount organ in the testes was about 13 % of the plasma concentration. It can be concluded that even a short exposure of a substance can induce adverse effects in the testes which are only partially reversible. Institut flit Toxikologie und Embryopharmakologie, FU Berlin Garystr. 5, W-1000 Berlin 33, Germany 1:19

Adverse effects on male fertility may manifest themselves as a decrease of the daily sperm production (dSP), an increase in the portion of abnormal sperms and a low pregnancy index. Male rats, treated with Ganciclovir (see abstract Klug et al.) , were investigated in regard to these three variables. All animals were mated before their investigation for two weeks with untreated females (1 male/1 female). The dSP was calculated as the sum of the number of spermatids in both testes divided by the factor 6,1. Alterations of sperm heads and signs of cellular degeneration (kinky sperm tails, seperated heads and tails) were considered abnormal. The pregnancy index was calculated as the ratio of pregnant females to "sperm positive" females.

The 24.9 ± 6,2 13.8 ± 3.1 73 6

9.6 ± 7,5 26.4 ± 11.3 64 8

1.0 ± 0.6 95 9 16

15.4 -. % 13.1 94.2 -+ 5.0 27 24

29.4 ± 11.1 20.3 + 5.6 67 Controls(n=66) 50.5 ± 13.9 I 6.6 ± 3.2 88

The effect on all three variables was obvious after 4 weeks, and the most pronounced effect appeared 8 weeks after treatment. The data recorded after 16 and 24 weeks point to a possible reversibility. It can be concluded that a short exposure of Ganciclovir induces remarkable effects on male fertility which manifest themselves first after 4 weeks. It has to be emphasized that even after three cycles of spermatogenesis the dSP and the pregnancy index are still lower, and that the ratio of abnormal sperms is higher in comparison to the controls. Therefore, we recommend an investigation of male fertility for an adequate duration after treatment. The purpose was to characterize the activation pattern of the NMDA/non-NMDA-receptor complex in the neuronal hybrid cell NG 108-15. In difference to neuronal primary cell cultures a hybrid cell offers more reproducible conditions, but it seems to be, that the extracellular conditions are the critical moment in the receptor function and in the receptor mediated calcium influx into the cell. In the first step we estimated the optimal test conditions: l.We find the pH value of 8.5 as suitable for the test, because the hydrogen concentration influences the receptor binding and/or the kinetics on the calcium channel. 2.The extracellular calcium concentration must be increased. 3.For activation of the glutamate-calcium-cascade well defined glutamate concentrations are necessary.

As it is known glutamate activates the main binding sites of the NMDA-receptor (only in connection with a glycine activation) and the non-NMDA-receptor (the kainate-and the quisqualate-binding site). We observed, that the glutamate induced reactions is blocked by 5-methyl-10,11-dihydro-SH-dibenzocylohepten-5,10-imine maleat (MR 801) in a unspecific mode on the level of calcium channels. The competetive antagonist 5-aminophosphonovaleriate (5-APV) is able to reduce partially the aspartate/ glycine reaction. Magnesium like other divalent ions is able to block partially the calcium influx. Barium mediates a block of the sodium influx and therefore the potential-regulated calcium channels are not opened in the same extent like under conditions without barium. Nimodipine is able to block the potential dependent calcium influx only to a little extent. The NMDA/non-NMDA-test system would be a suitable in vitro model for brain injuries as such caused by drugs and poisons to understand the mechanisms of actions of drugs. Besides it is possible to use this test system for screening of neuroprotective substances, e.g. nootropics or calcium overload blockers.

Institute of Pharmacology and Toxicology, Medical Academy 'Carl Gustav Carus'Dresden Karl-Marx-StraBe 3, 0~8080 Dresden, Federal Republik of Germany

Peroxisome proliferation is a phenomenon frequently observed in the livers of rats and mice in response to the administration of a specific class of non-genotoxic carcinogens, the so-called peroxisome proliferators. This process has recently been shown to be receptor-mediated. We have examined the effect of oral administration of the strong peroxisome proliferator fenofibrate on the hepatic expression level of the peroxisome proliferator activated receptor (PPAR) in rats. Immunoblots of rat liver cytosols and nuclear extracts using antibodies raised against recombinant PPAR/l~-galactosidase fusion proteins revealed a pronounced increase in the mount of PPAR protein in response to fenofibrate treatment. This induction could be confirmed on the level of RNA by Northern blotting. A time course investigation showed a delayed accumulation of mRNA in, response to the treatment, starting on day 2 after a latency period of at least one day. In contrast, mRNA of Acyl CoA oxidase, the key enzyme in peroxisomal l~-oxidation of fatty acids, was found to be increased already after one day of treatment. The induction of the PPAR as a response to peroxisome proliferators represents one important dimension of the pleiotropic effects of peroxisome prolfferators. in an unspecific mode of action but also modifies some functions of definite cell surface structures, e.g. the NMDA/non-NMDA-receptor complex or the GABA-receptor complex. Modifications of the ion channels regulated by these receptors are also decribed. The purpose was to characterize the dependence of ethanol toxicity on time, pH value and calciumconcentration in NG 108-15 with or without participation of NMDA/non-NMDA-receptor activation. Under the conditions of participation of the NMDAor the kainate-receptor binding site the tests were carried out with a subtoxic concentration of ethanol. We compared the NMDA and the kainate dependent reactions under ethanol incubation.

Our results indicate, that ethanol does not modify the channel conditions in NG 108-15, but does influence the receptor plasticity. The aniline derivative procainamide (PA) is known to cause drug-induced systemic lupus erythematosus (SLE) in up to 30% of the patients treated with this drug. Indirect evidence suggests that metabotization of PA is involved in the pathogenesis of this condition in that N-hydroxylation of the arylamine group of PA favours the development of drug-induced SLE whereas N-acetylation prevents it. If this is correct, one would expect hydroxyl~mine-PA (HAPA) to be immunogenic, whereas N-acetyI-PA should be non-immunogenic. In the present investigation, _using the popliteel lymph node assay (PLNA) in mice, this wasfound to be the case. Injections of both PA and N-acetyI-PA (1.6 mg/mouse) into C578L/6 mice completely failed to induce a reaction in the direct PLNA. By contrast, the same amount of HAPA induced a vigorous PLN reaction, and there was an excellent dose-effect relationship in the range of 0.2-1.6 rag/mouse. Using the adoptive transfer PLNA, splenic T cells of mice that had received three injections of HAPA were shown to be sensitized to this metabolite, but not to PA or N-acetyI-PA. In this system, an anamnestic T cell response could-also be elicited when homogenized peritoneal cells of mice that had been treated with PA for 4 months were used as the challenging antigen, indicating that the peritoneal cells of PA-treated animals contained the reactive intermediate metabolite HAPA. This finding supports the concept that reactive intermediate metabolites, such as HAPA, are generated by phagocytic cells outside of the liver and that such metabolites may be immunogenic for T cells. Antiestrogens are known to inhibit proliferation of estrogen-dependent tumor cells. In the present study, we investigated the induction of cell death in cultured mammary carcinoma cells (MCF-7) by Tamoxifen (TAM) . In concentrations between i0 ~8 to 10 -6 M TAM induced a dose-dependent increase of cells with nuclear condensation and fragmentation suggestive of apoptosis; the maximal number of dead cells were found 6-8 days after TAM. The nuclear changes preceded the release of lactate-dehydrogenase (LDH) by about 24 hours. Electron-microscopical investigations showed that organelles and cell membranes are well preserved during early stages of cell death. These studies also suggested the involvement of autophagic processes in cell death. Furthermore, TAM induced cell dea£h could be inhibited by 10 -3 M estrogen. In conclusion, the morphological and functional features of cell death after i0 -U to 10 -6 M TAM suggest the occurrence of apoptosis~ After i0 -~ M TAM, however, all MCF-7 cells died within 24 hours as indicated by release of (LDH) and trypanblue-exclusion test. Light-and electronmicroscopical investigations revealed ly~is of mitochondria and cell membrane. TAM (i0 -~ M) induce@ cell death could not be inhibited by 10 -9 to i0 -~ M estrogen. These findings are indicative of necrosis. The underlying pathway(s) leading to cell death after high TAM concentrations remain to be elucidated. Apoptosis occurs in the liver after treatment with drugs, during regression of the hyperplastic liver and during tumor development. We study the mechanism of apoptosis. In an established modell using the regression of hyperplastic liver after treatment with the antiandrogen cyproteroneacetate (CPA) we are able to induce apoptosis by the epithelial growth inhibitor transforming growth factor-B1 (TGF-I~ 1). In a preliminary experiment we could show that this growth inhibitor induces apoptosis only in hepatocytes primed by withdrawal of cyproteroneacetate. We were now interested in the time course and dos e response of the induction of apoptosis by TGFfil. Two hours after injection of TGF-f)I the first morphological signs of apoptosis appear. The process reaches its peak at about 4,5 hours. Thereafter, the formed apoptotic cells are degraded within their neighbouring cells. The induction of apoptosis by TGF-BI is dose dependent. In a preliminary study we have proven that hepatic nuclei treated with Ca 2+ differ from apoptotic nuclei in morphology and DNA content, as shown by H33258 staining, electron microscopy and FACS analysis. Studies applying the method of in situ nick-translation further confirmed that fragmentation of the DNA is not responsible for the chromatin condensation taldng place during apeptosis. Similar results were found in serveral cell lines in which apoptosis could be induced by serum withdrawal or treatment with TGF-131. From these studies it can now be concluded that DNA fragmentation into oligosomes occurs not in all cell types and if so, represents a late stage of apoptosis. HEXACHLOROPHENE CAUSES DAMAGES IN DIFFERENT CELL  TYPES -THE CYTOTOXCITY AND NEUROTOXICITY IS NOT  DEPENDENT ON DEFINITE RECEPTOR OR CHANNEL  STRUCTURES. C.Hohlfeld,Chr.Matthes,K.Andreas

The hexachlorophene-induced cytotoxic brain oedema with particular evident damages in the cerebellum is a suitable in vivo model for the investigation of brain injuries. The vacuolation of tissue is accompanied by an increased water content. This change is reversible. The regeneration may be influenced by cerebropotective substances like nootropics or calcium overload blockers. The purpose of the study was to characterize cytotoxic processes of hexachlorophene on various cell types in vitro: erythrocytes, primary cerebellar cell cultures and the neuronal hybrid cell NG 108-15. Cerebroprotective substances don't have any or only a little protective effect in vitro. It seems, that hexachlorophene like a membrane active agent causes a general cell damage and it's toxicity does not specifically change the activity of special receptors or channels. Many available inbred mouse strains are known to differ with respect to their liver tumor susceptibility. Tumor susceptibility refers to the development of spontaneous liver tumors (without exposure to carcinogens) as well as to the development of chemically induced tumors after treatment with genotoxic or non-genotoxic hepatocarcinogens (i.e. hepatic tumor promoters). We studied the liver growth response after short-term administration of non-genotoxic compounds (phenobarbital and ~-hexachlorocyclohexane) in susceptible (C3H, CF1, CBA, B6C3F1) and resistant (Balb/c, CD1, C57BL, NMRI) mouse strains and compared the results with the liver growth response in rats.

In normal mouse liver, short-term phenobarbital treatment results in increased hepatic weight and centrolobular hepatocyte hypertrophy due to the significant increase in liver protein content. Despite an increase in DNA synthesis and DNA content could be observed, no hyperplasia is morphologically evident. We also compared the liver growth response of rats and mice after a single dose of o~-HCH. Whereas the peak in DNA synthesis is followed by an increased mitotic activity in rats (hyperplasia), no increase in mitoses after DNA synthesis could be observed in mice, suggesting polyploidization of mouse hepatocytes. Comparing susceptible and resistant mouse strains, a strong liver growth response does not correlate with tumor susceptibility. It may be possible that some of these differences are important for the mechanism of tumor promotion and susceptibility of different mouse strains to hepatocarcinogenesis. In previous studies using isolated perfused rat livers, we have shown that hypoxia and reoxygenation as well as partial ischemia result in a marked hepatotoxicity, and that reactive oxygen species are involved in both models of hepatic injury. Albumin was shown to possess strong antioxidant properties due to the direct scavenging of free radicals through the thiol group, the binding of copper or iron ions capable of catalyzing reactions which yield oxygen radicals, or both. in this study we examined the capability of albumin to prevent hypoxic and ischemic damage towards isolated perfused rat livers. Perfusion experiments were carried out at 37°C using a hemoglobin-free medium (Krebs-Henseleit buffer), gassed with carbogen to yield an oxygen partial pressure of 600 mm Hg. Partial ischemia was induced by lowering the perfusate flow rate from 60 to 10 ml/min and kept low for 90 min. Hypoxia was generated by replacing carbogen with a mixture of 95% N 2 and 5% CO 2 for 30 rain followed by reoxygenation for 60 rain. Both, partial ischemia and hypoxia/reoxygenation, resulted in marked hepatic injury as evidenced by an increased release of hepatic enzymes (GPT, LDH), by a strong drop of bile flow and by a decrease in hepatic glutathione levels. With partial ischemia, also ATP-depletion and calcium accumulation were observed in the livers at the end of the experiments, Bovine serum albumin was added to the perfusate at concentrations of 0.03, 0.1, 1, 2 or 5%, and provided protection against liver injury at concentrations of 0.1% and above, The same level of protection was also afforded by sulfhydrylblocked and fatty acid-free albumin preparations. Thus, the protective effect of albumin in our models of oxidative liver injury is not due to free radical scavenging or specific transition metal binding through the thiol moiety, nor due to the presence of oxidizable fatty acids in the albumin fraction. More likely, albumin provides protection by an unspecific binding of redox-active metal ions and, thus, by interacting with the catalysis of reactions leading to the formation of hydroxyl or hydroxyllike radicals. Institut Male mice are sensitive to nephrotoxic effects of several low molecular weight toxins e.g. chloroform and acetaminophen. These compounds are metabolized to reactive intermediates by hepatic cytochrome P-450 2E1. p-Nitrophenol has been developed as an specific substrate for P-450 2E1. To establish a role for 2E1 in the male specific renal toxicity of chloroform and others, we investigated the metabolism of p-nitrophenol in liver and kidney miorosomes from mice and rats subjected to different pretreatments. The results obtained in mouse kidney microsomes show that p-nitrophenol oxidase activity (identical to CYP2E1 ; 3.4 nmol p-nitrocatechol (p-NC)/mg/20 rain) is present in male animals exclusively; its activity is regulated by testosterone. In naive females, no activity was found in kidney microsomes, p-nitrophenol oxidase activity could be induced by pretreatment with testosterone.

In kidney microsomes obtained from male rats and in 3 samples of human kidney microsomes (male donors), no p-nitrophenol oxidase activity was detected, p-Nitrophenol oxidase activity was present in liver microsomes from both male and female mice and ethanol pretreated rats in similar activities (9.1 nmol p-NC/mg/ 20 min). Liver microsomes from naive male rats had a p-nitrophenol oxidase activity of 3.1 nmol p-NC/mg/20 rain; the activities in mouse liver microsomes were not influenced by pretreatment with testosterone. These data suggest that CYP2E1 (or an enzyme with similar substrate specificity and immunological characteristics) is expressed only in male mouse kidney. Since CYP2E1 has been suggested to bioactivate chloroform and acetaminophen, the sexand species-specific expression may explain the unique sensitivity of the male mouse kidney to these compounds.

Institut f0r Toxikologie, Universit&t W0rzburg, Versbacher Str. 9, D-8700 WQrzburg, FRG R12 45

0smolarity of kidney cells is regulated by organic osmolytes like corbitol, inosltot, betain and glycerophosphorylcholine. These organic osmolytes balance the osmolarity of the cells in dependence on the varying osmolarlty of the extracellular fluid as part of the urinary concentration mechanism. In contrast to the kidney, the process of osmoreguiation in the bladder is not clarified in detail.

To examine which mechanisms are involved in this process investigations were performed by using freshly isolated porcine bladder epithelial cells, representing the in vivo status of the organ, compared with primary cultures of these cells under defined culture conditions. The aim of the study was to find out whether this culture model can be used as an in vitro model for the bladder. Based on the situation in the kidney it was investigated whether sofoltol acts as an organic osmolyte in the bladder. Sorbitol content in freshly isolated bladder epithelial cells revealed that intracaituiar sorbltol content depends on udne osmoiarity. When osmolarity of the culture medium was increased in culture, only a slight increase in intracellular sorbltol was measured. The increase of the total amount of sorbltol was too low to regard regulation of this osmolyte as the main component of osmoregulation in these cells. In addition activities of aldose reductase, an enzyme involved in sorbitot synthesis, also increases slightly. When intracellular urea concentration was measured in dependence on the extracellular urea concentration it could be shown that urea permeates the cell membrane of cultured cells and that its intracellular concentration was raised when the extracaitular concentration was increased. From these experiments it was deduced that urea, as an osmoly~io effective component of the urine, may act in the process of osmoregulation in bladder cells. Cultured porcine bladder epithelial cells are suited for development of a model for the investigation of the influence of toxic agents on osmoregulation and, vice versa, for studies on the influence of intracellular osmolarlty on the effects of toxic and genotoxic agents like aromatic amines in the bladder.

Institut for Arbeitsphysiologia an der Universit&t Dortmund, Ardeystr. 67, 4600 Dortmund 1, F.R.G.

ADULT AND JUVENILE RATS C. F5rster., I. Baumann-Wi)schke, and R. Stahlmann

Ouinolones exhibit cartilage toxicity in/uvenile animals and are thus contraindicated in children and adolescents. Very little is known with regard to differences in the toxic potential of the different quinolonas available, We initiated in vitro binding studies to compare the affinity of three fluoroquinolones to cartilage tissue. Samples of cartilage tissue (sternal cartilage or femoral heads) from adult and juvenile Wistar rats (age: 4.5 m and 34 d) were incubated in solutions of 5 or 50 mg quinolone/l PBS buffer (pH = 7.2) at 37 °C, The ratio of the tissue and the buffer was t:10 (w/v). Drug concentrations in the medium were measured spectrofluorometrically in 0.1 M glyclne-HC[ buffer (pH = 3.01 at the beginning of the incubation period and after 24 h. Binding rates were calculated as the percental decrease of the quinolone concentration during the incubation period. These rates varied with the age of the animals investigated and depended considerably on the quinolone tested. Binding rates were as follows (drug concentration: 5 mg/I PBS buffer; data are mean values + SD, n = 3 to 8 tissue samples): All quinolones bound to a higher extent to cartilage samples collected from juvenile rats than from adults and showed higher affinity to femoral heads than to sternal cartilage. Binding rates were lower with higher concentrations (e.g. ciprofloxacin 50 mg/I PBS buffer: 38.0+ 1,0%; sternal cartilage, juvenile rats). In our opinion, these results supply valuable information on differences between the drugs with respect to their affinity to cartilage. This approach might also help to clarify some open questions concerning the mechanism of cartilage toxicity. Institut f£)r Toxikologie und Embryopharmakologie der Freien Universit~t Berlin, Garystr. 5, 1000 Berlin 33 46

C. Guhe, I. Grabke and W. F~itmann

Dividing long-term monolayer cultures of porcine urinary bladder epithelial cells were obtained by isolating cells from bladders of freshly slaughtered pigs. The cultures were investigated morphologically and characterized according to their growth characteristics and enzymatic functions. Ultrastructurai investigations over a period of up to 12 weeks demonstrated that the cells regain their in vivo polarisation with apical situated membrane vesicles and tight junctions between neighbouring cells when they have built up a confluent monolayer. Membrane integrity and high cell viability was indicated by measuring low levels of lactate dehydrogenase activity released into the medium during the culture time. The chromosome set of the cells was stable during the first five weeks of culture. Under standard culture conditions activities of the enzymes alkaline phosphatase, a marker for luminal membranes, and acid phosphatase, a marker for lysosomes, were stable over a period of four weeks. Gamma-glutamyl transpeptidase activity, indicating fetal regression, did not increase. Activity of NADH-dehydrogenase, a marker for the endoplasmatic reticulum, decreased in culture compared to freshly isolated cells indicating a decrease in ER as a consequence of adaptation to the culture condition. By measuring a concentration dependent increase of sister chromatid exchanges (SCE) induced by an alkyiating agent (N-methyI-N'-nitro-N-nitrosoguanidine, MNNG) the suitability of this model for genotoxicity tests was estimated. These results indicated that the model can be used as a target for bladder cancer inducing agents to identify the effects of such agents on the level of the target organ in further genotoxicity studies.

Institut for Arbeitsphysiologie an der Universit&t Dortmund, Ardeystr. 67, 4600 Dortmund 1, F.R.G.

Usually prenatal toxicity of xenobiotics is evaluated after exposure during the organogenesis phase. Disturbance of the development of the embryo is routinely examined on day 21 of gestation. A possible postnatal manifestation of skeletal alterations as a result of the influence of xenobiotics administered only in the last third of pregnancy is not investigated in normal cases. We studied whether the skeleton may be influenced even after organogenesis has been completed. C¥c~ophosphamide was chosen as model substance. The prenatal toxicity of this alkylating agent has been described repeatedly in animal studies and a few case reports also suggest a possible teratogenicity in man. We explored whether cyclophosphamide is able to induce alterations of the skeleton during late pregnancy in rats and whether these alterations persist during postnatal development. Pregnant Wistar rats were treated with 15 mg cyclophosphamide per kg body wt orally from day 17 to day 19 of gestation. The pups were sacrificed on day 1, 7, 14, 21 and 49 postnatallv. After fixation in formalin the cartilage was stained with alcian blue, subsequently the calcified bone with alzian red. The treatment with cyclophosphamide during the last third of pregnancy led to pronounced defects of the skeleton in the pups. At all time points investigated there were shortened and bent bones of different parts of the skeleton. Remarkable were fractures of the hollow bones, the pelvis, the rips and the spine scapulae. Obviously, cyclophosphamide caused a disarrangement of the ossification of the skeleton also at this late stage of development. It can be concluded that even after exposition in the last third of pregnancy certain substances are capable of inducing adverse effects that manifest themselves and persist postnatally. Institut for Toxikologie und Embryopharmakologie der Freien Universit~t Berlin, Garystr. 5, 1000 Berlin 33 49

Stephan Klug, Ursula Kastner, Nadim Obaidi, Helga St~rje

Cleft palate is one of the spontaneous congenital malformations found with relatively high frequency in mammalian newborn, furthermore, this malformation can be induced in mice by various compounds. In order to obtain additional information on how certain teratogens produce this birth defect, we have established an in vitro system of culturing palates of mouse and rat embryos in a suspension culture technique which allows simulation of the main processes of palate closure under standardized conditions.

Studies with varying culture conditions revealed: 1.

The gestational age of embryonic palates set into culture is of great importance for the fusion rates (day 14-.5=46%; day 15=81%).

Minimal differences in the dissecting technique cause fusion rates between 0 and 81%.

The speed of rotation of the culture bottles has a remarkable influence on palate fusion rates (25 rpm =31%; 12 rpm = 81%).

The oxygen concentration with which the palates are gassed, substantially determines the rate of fusion (30% 02=0%; 60% 02=63%; 95% O2=81%).

Some further characteristic behaviors of the cultured explants concerning: the ratio of wet and dry weight and the ratio of dry weight and protein content, are described. The release of arachidonic acid (AA) from membrane phospholipids is part of a signal system with which cells respond to environmental changes of various kinds. Already Stark et al. [In: Goldberg, A.M (1983) , Alternative Methods in Toxicology 2, Product Safety Evaluation, Liebert, NY:179-203] demonstrated for the irritants nbutanol and isopropyl myristate the enhanced release of radioactive material from [3H] -AA-prelabelled hepatoma ceils. Using the promyelocytic cell line U937 we studied the kinetics of AA release in [3HI -hA-labelled U937 cells under the influence of lysolecithin, sodium dodecyl sulphate (SDS), chlorarnine, formamide, 2,4dichlorophenoxyacetic acid (2,4-D), and dimethyl sulfoxide (DMSO). Membrane-toxic agents like lysolecithin and SDS cause pronounced enhancement of [3H] -AA release in U937 cells. The effect occurred immediately after exposure and proved to be dose-dependent, nonenzymatic and irreversible. In comparison with untreated controls, DMSO up to 100 mg/rnI did not cause any membrane damages. Like this compound, formamide up to 10 mg/ml, chloramine up to 500 ~tg/ml, and 2,4-D up to 100 ~tg/ml were well tolerated by the cells. AA release is suggested as a highly sensitive method for detecting membrane-toxic and irritating agents and a promising alternative for the eye irritation test. It is known that the intracellular caldum concentation is a critical factor in mitosis. DES induces mitotic disturbances resulting in aneuploidy; the mechanisms of these events are still largely unknown. We now have tried to identify the calcium source for the DES~mediated calcium rise. For the determination of the intraceflular calcium content suspended ceils were incubated with Fura-2 (1 h) and subsequently analysed for intracellular fluorescence with a spectrofluorimeter (340/385 nm ratio). Our recent results show that depletion of intracellular calcium stores by thapsigargin or ionomycin (1 p~M) in the presence of calcium-free medium almost completely diminish the DES-induced calcium rise. In addition, the DES mediated calcium signal is dependent on the extracellular calcium concentration. Furthermore, addition of DES shortly after generation of the ionomycin induced calcium rise slows down the decay of the signal. Experiments with isolated rat liver mitochondria showed that DES does not cause a calcium release from these stores. In conclusion, DES releases calcium from /23 and GTP sensitive stores and inhibits extrusion of calcium from the cell. Further studies are aimed at the analysis of a possible interrelationship between the described calcium changes and the induction of aneuploidy. The activity of Calcium-transport ATPase of human erythrocyte membranes is sensitive against the action of a large number of xenobiotics including industrial chemicals and biocides. Therefore, we tested a series of these compounds with the aim to develop a simple, fast and unexpensive test system for screening the acute toxicity of xenobiotics. The compounds were tested at concentrations between 1 ~mol/L and 3 mmol/L. -In addition to the investigation of a series of chlorinated phenols, the higher chlorinated compounds showing a high degree of inhibition, we tested several derivatives of aniline and nitrophenol. Although the effects were less pronounced than in the case of the phenol series, some of the derivatives of 2,4-dinitrophenol showed moderate inhibitory effects (expressed as the concentration [in ~mol/L] yielding half maximum inhibition): 2,4-dinitrophenol: 3 300, 2,4,6-trinitrophenol (picric acid): 3 000, DNOC (dinitroortho-cresol): 2 500, dinoseb: 700, dinoterb: 900. In contrast, the parent compounds phenol, aniline, and their derivatives containing a methyl-group or a nitro-group or an amino-group or a chlorine atom show less than 20 % inhibition up to a conc. of 1-3 mmol/L. -In most cases, a fairly good correlation (spanning some orders of magnitude) between the inhibitory potency of the compound and its octanol-H20 partition coefficient (Pow-value) can be seen.

Abteilung Pharmakologie und Toxikologie der Univer-sit~t Ulm, 7900 Ulm, Federal Republic of Germany OF TOPICALLY ADMINISTERED TETRA-CHLORODIBENZO-P-D1OXIN IN C57BL-M1CE. Kietzmann, M., C. Dasenbrock ~ and D. Lubach ~ Tetrachlorodibenzo-p-dioxin (TCDD) was administered topically on the inner ear skin surface and the tail skin of female C57BL/6J-mice. Administered to 6 mice per group, the dosages of TCDDD which was dissolved in acetone were 0, 0.01, 0.1, 1, 5 and 10 nmol/0.25 cm ~ skin. The animals were sacrified three days later by cervical dislocation. Tail and ear skin were dissected. The epidermis was separated from the remaining tissue by dispase incubation. The incorporation rate of thymidine triphosphate into DNA and of leucine and histidine into protein were measured. Additionally, keratin proteins were separated by polyacrylamide gel electrophoresis. TCDD induced a dose-dependent increase of the thymidine triphosphate incorporation rate in tail and ear epidermis. An enhanced leucine incorporation was found also in both regions, while the histidine incorporation increased only in the ear epidermis. In contrast to ear epidermis, a dose-dependent increase of the intensity of keratin protein fractions in the molecular range of 45 to 70 kDa was found in the tail epidermis. Similar changes were found also after repeated oral administration of TCDD (Kietzmann, M. et al., Naunyn Schmiedeberg's Arch. Pharmacol. 344 (Suppl.) , R113, 1991) . From these results it is suggested that the determination of epidermal metabolic parameters is a sensitive tool to demonstrate effects of chlorinated hydrocarbons. We have shown that arsenic inhibits glucose uptake in Madin-Darby canine kidney (MDCK) cells (1). In addition to the known impairment of cellular energy metabolism, this inhibition may contribute to the acute toxicity, especially of trivalent organic arsenicals, like oxophenylarsine (PhAsO). We have now investigated whether the inhibited glucose uptake may be reversed by 2,3-dimercaptopropanol (BAL) and other sulfhydryl reagents. The culture medium (DMEM/F12) of confluent MDCK cells (3-4 d) was repIaced by Hanks's buffered salt solution containing 10 ~moI/I glucose. 30 rain later 2 Ixmol/1 PhAsO, and again 30 rain later 2, 20 or 200 ~tmol/] of various thiol compounds were added. At timepoims 5, 10, 20, 30, 40, 60, 90, 150, 180 min after arsenic addition an uptake assay with [6-t4C]-glucose was performed as described (1). In the preseuce of PhAsO glucose uptake was reduced to about 50 % of the controls within 30 rain. This inhibition could be reversed in a time-and concentration dependent manner when BAL or related dithiol compounds, e.g. 2,3-dimercaptopropanesulfonate (DMPS) and meso-2,3-dimercaptosuccinic acid (DMSA) were added after 30 rain. The water-soluble dithiols DMPS and DMSA required higher concentrations / 1ouger incubation times for complete reversal of the As-induced inhibition. On the other hand, monotbiols, e.g. 2-mercaptoethanol, could not restore impaired glucose uptake. It is concluded that 1) inhibited glucose uptake in MDCK ceils is reversed by antidotes which also effectively antagonize the metabolic effects of arsenic and that 2) MDCK cells may provide a valuable experimental system for the investigation of potential arsenic antidotes. In rats, dietary arsenic leads to a concentration-and timedependent accumulation of copper exclusively in the kindney, preferentially in the cortex. Compared to controls with renal copper levels of about 5 ug Cu/g wet wt, feeding of a diet with 60 ~g As/g of diet (as NaAsOs) for 2 wk increased renal copper levels to about 45 Dg Cu/g wet wt. D~e to the high retention of arsenic in the erythrocytes, the rat is considered to have a special metabolism of arsenic which could have been responsible for the observed effect. Therefore, we looked for a corresponding effect of dietary arsenic in mice and guinea pigs. Each species was divided into 4 groups of animals according to the diets fed which contained increasing concentrations of NaAsOs (0, i0, 30 and 60 pg As/g of diet). Animals were killed by exsanguination after feeding periods of I, 2 and 3 wk. The liver, kidneys and small intestine were removed and analysed for arsenic and other trace metals (Cu, Fe and Mn) by atomic emisssion spectrometry. An increase of renal copper levels by increased dietary arsenic concentration, to an extent similar to that in rats, was only observed in guinea pigs which, however, did not accumulate arsenic in the blood. Renal copper accumulation was time and concentration dependent as in rats. Feeding a diet with 60 ,g As/g for 3 wk increased renal copper level from about 5.5 to 40 ~g Cu/g wet wt.

Renal copper levels in mice as well as other trace metal levels in both species were only slightly altered by dietary arsenic. The study shows that the renal copper-arsenic interaction is not restricted to the rat and is not due to a high blood retention of arsenic. Since in rats and guinea pigs, but not in mice, arsenic accumulated in the kidney to a much higher extent than in other organs, it is suggested that copper and arsenic accumulate synchronously. In a recent paper we reported the reaction of various reducing xenobiotics with the oxy-tbrms of myoglobin and hemoglobin thereby producing low-level chemiluminescence (I).

A stro.ng, oxidant was assumed to be responsible for the light emission. To prove this we have extended our investigations in order to identify the corresponding reaction mechanism. Using a series of effectors we were able to exclude free hydroxyl radicals and superoxide. The use of a singlet ox?fgen trap in combination with HPLC allowed us also to eliminate singlet oxygen as an alternative light emitting species. Our results strongl~ support that the light emission stems from an electron transter to the berne-bound hydrogen peroxide via a globin peptide. Therby a protein radical and the ferryl tbrm of heine were formed. This was concluded by EPR spectroscopy in combination with visible spectroscopy. A tentative reaction mechanism is discussed.

( The fatlure of oNdoxime in reactivating significantly acetylcholinesterase (ACHE) blocked by soman (S) was the reason for investigating further oximes, in this study the effects of HL5 7 and HI 6 were tested in mica poisoned by sublethal doses (50% LD~) of soman. Nine groups of 10 white mate NMRt-mice, each, were built: group 1: (control) saline 0.9% s.cJ i.p. group 2: S 0.1 m~kg s.c. + s~ine i,p., group 3: S ÷ atropine (A) 10 mg/kg i.p., group 4]5: S + A + HI 6 or HL5 7 10 Fmo~g i.p. group 6/7: S + A + HI 6 or HLb 7 30 i~mo~/kg and group 8/9: S + A + HI 6 or HL5 7 150 i~mol/kg, After the injection(s) the motor performance was tasted on a rotating mash wire drum (diameter 20 cm, 14 rpm) by determining the running time (rt), After the running period the activity of AChE in bleed was measured photometricatly (Ellman GL, Courtney KD, Andres V, Featherstone RM (1961): A new and rapid colorimetric determination of acetylcholinesterase activity, Biochem Phermacol 7: 88-95). S reduced significantly rt and AChE activity. A improved significantly rt but not AChE activity. In combination with A, dose depending improvement of rt and AChE activity compared to the S and S + A groups was observed. After the highest investigated dose of the oximas (150 ~mol/kg) HL5 7 is significantly superiar to HI 6 in rt and AChE activity improvement. The consequences of oxidative stress were investigated in normal and glutathione-<leplated rat hepatocytes. Glutathione(GSH)-depletfon was achieved by way of GSH-reductase inhibition (75 pM BCNU) or GSHconjugation with phorone (250 pM). Oxidative stress was subsequently induced by incubating the cells in an atmosphere of humidified air with 500 pM nitrefurantoin (NFT). Under these conditions NFT enters a redox-cycle which generates toxic reactive oxygen species. When cytoprotection was hampered due to the lack of GSH, NFT was highly detrimental as demonstrated by trypan blue exclusion, lactate dehydmgenase leakage, decrease of the intracallular K+-concentration, plasma membrane blebbing, lipid perexipation and loss of reduced protein thiol groups. However, the sensitivity of BCNU-pretreated cells, which still disposed of reduced GSH in the early phase of NFT-exposure, was much higher compared to pharone=pretreeted cells raising the question whether GSH might be partly involved in the cytotoxic process. The importance of GSH and protein dedved thief groups was strengthened by the fact that thiol reducing agents (i,e, dithiothreitol) as well as GSH precursors (i.e. N-acetyl-cysteice) diminished or postponed cellular injury, A protective effect could be achieved even after prolonged incubation when most of the cells already exhibited extensive plasma membrane biebbing. This indicates the promising use of certain thiol-reactive substances in the defense of oxidative stress. Since thiol homeostasis is believed to interfere with the cellular Ca ++compartmentation cytosolic free Ca++-concentrations were measured and compared with the degree of cytotoxicity. Ca++-alterations are supposed to tdgger cell damage, thus antagonists of Ca ++ -activated processes were investigated with regard to their cytoprotective potential. FU Berlin, Koserstr. 20 There is only restricted information about transfer of mercury from the mother to the fetus. We have compared mercury concentrations in maternal blood vs. umbilical blood and placenta. 186 women (16-42 years) from South Germany (Ulm area) and their newborns were included in the study. Possible sources of exposure to mercury (fish consumption, occupation, etc.) and health status were assessed by questionnaire. Amalgam status was quantified by dental inspection. Blood samples were taken in K-EDTA tubes. The samples were digested with nitric acid in the Perkin-Etmer Autoclave-3 under pressure and at increased temperature. The determination of total mercury concentration was carried out by Cold Vapor AAS (Perkin-Elmer Atomic Absorption Spectrophotometer A.AS 430, MHS-20). The median values for mercury concentrations were for maternal blood 0.63 (0.1-7 3) tag/t, for umbilical blood 0.78 (0. I-5.0) gg/t and for placenta 4.09 (0.71-25.0) tag/kg ~vet weight. Number of amalgams ranged between 0-18 (mean 7.3) and surface area between 0-770 mm 2 (mean 278.3). The women with amalgams (n=166) had, on average, higher level of mercury in blood 0.66 (0.1-7.3) lag/1 than the women (n=20) without amalgams 0.39 (0.1-1.6) gg/L Though mercury concentrations are only moderate correlated with the number and surface of amalgams. The results demonstrate that mercury concentrations in fetal blood exceed those in maternal blood and that dental amalgams of the mother contribute to the mercury exposure of the fetus. Inhalative zinc chloride poisoning is a rare, but life-threatening calamity seen after short-time exposure to zinc chloride smoke. An effective therapy is not established and lethality is high. Acute respiratory distress syndrome (ARDS) is the most prominent clinical feature of this intoxication and death from respiratory insufficiency usally ensues several days later. In order to establish an in-vitro model to investigate acute toxic effects of zinc salts we have studied several cell lines derived from human lung. The cancer cell line A549, and fibroblast-like 8Lu, 11Lu and 16Lu cells were grown in DMEM/F12 medium supplemented with fetal calf serum (10% vlv) at 37°C in moist atmosphere containing 5% CO2. After reaching confluency the cells were exposed to various concentrations of zinc chloride. Cell viability and morphology were assessed and angiotensin-converting enzyme (ACE) activity was measured as a biochemical parameter. When Zn concentrations exceeded 100 gmol/1 the growth of the cultures was progressively depressed. The formation of blebs and the detachment from substratum were indicators of impaired cellular viability. In Lu cells ACE activity increased and showed a maximum after 16 h exposure to zinc. No ACE activity could be detected in A549 cells. When corticosteroids (< 10 ~tmol/l of hydrocortisone, beclomethasone, budesonide) were administered prior to the addition of zinc no prophylactic effect could be observed. ACE activity increase and cell death were similar in treated cultures and in controls (without steroids). On the other hand, known antidotes against heavy metal intoxication like 2,3-dimercaptopropanesuffonate, Dpenicillamine, and 2,3-dimercaptopropanol (BAL) effectively prevented zincrelated toxic effects including the ACE activity rise when given in 10fold molar excess over zinc. Our data suggest that fibroblast-like Lu cells provide a suitable in vitro model for the study of zinc-related toxic effects and the evaluation of prophylactic measures. Hepatic Cu-levels are elevated in the newborn and dramatically change during development. Cu in the particulate fraction of neonatal liver has been reported to be partly localized in lysosomes. This Cu has been assumed to be bound to aggregated metallothionein (MT).

The aim of the present study was to estimate the accumulation and distribution of Cu, Zn and MT in isolated hepatic lysosomes of neonatal male Wistar rats of various ages. Lysosomal fractions were obtained by differential centrifugation followed by isopycnic centrifugation in Percoll ® gradients. Lysosomal fractions were identified by activity of acid phosphatase and separated into light and heavy fractions according to density. Cu and Zn were quantified by AAS or ICP-AES. Zn-and Cu-containing MT was analyzed by the Thiomolybdate(1)-assay. The amount of oxidized MT was determined after treating the samples with 2-mercaptoethanol. Whereas Zn was detected in both, the light and heavy lysosomal fractions, Cu was almost exclusively found in heavy tysosomes. Cu and MT levels in heavy lysosomes were high during the first 13 days of life and showed a characteristic decline thereafter. MT, however, was almost totally oxidized suggesting its inability to bind metals. Therefore, MT seems not to represent the major Cu-binding protein in lysosomes of neonatal rat liver.

(1)Anal. Biochem. 189, 35-39 (1990) * The transepithelial electrical resistance (TER) has been used to assess the functional integrity of epithelia. Madin-Darby canine kidney (MDCK) cells are well-known for their cell polarity when grown in culture. For example, virus budding has been shown usually to occur from the apical membrane face for many viruses, but from the basolateral domain for Vesicular Stomatitis Virus (Rodriguez-Boulain et al., J. Cell Biol. 96 (1983) 866-874) . We have studied the effect of zinc, cadmium and mercury added to the apical or basolateral side of MDCK cell cultures on the time course of TER. Wild-type MDCK cells were grown on permeable supports in DMEM/F12 medium at 37°C in moist atmosphere containing 5% CO2. Cells developed TER values above 2000 ~cm 2 within 2 -3 days after reaching confluency. The addition of zinc, cadmium and mercury acetates on both sides of the monolayer led to a concentration-dependent TER decrease within 24 h. The decrease in TER was followed by detachment of the cells from the support and other signs of impaired cellular viability. Equipotent concentrations for almost complete abolition of TER within 24 h were 10 pmol/1 HgAc2, 50 gmol/l CdAc2 and 500 gmol/1 ZnAc2. When the metal salts were added on one side of the monolayer zinc (500 txmolft) decreased the TER only when given to the basolateral face. Cadmium was also more effective when added to the basolareral compartment. However, mercury.salts completely abolished the TER when added to either side. The series of increasing toxicity in the order Zn < Cd < Hg was paralleled by a decreasing membrane domain selectivity. Our data suggest that the inherent polarity of cultured MDCK cells might also play an important role in the cytotoxicity of heavy metals. It is tentatively assumed that the observed polarity corresponds to an asymmetric distribution of cellular targets responsible for the binding of the heavy metal ions. (b.w.: 200 g) . The animals (6 per group) were sacrificed under diethyl ether narcosis 0.5; 1.0; 2.0; 3.0; 6.0; 12,0 hrs after Au applic.. Zinc (Zn), copper (Cu) and Au were measured by atomic absorption spectrometry, and the metallothionein content was determined by gel chromatography.

The main purpose of this study was to find out, whether the induction of metallothionein synthesis is caused by Au itself or by intracellular mobilization of Zn or Cu by Au.

I. As first event an accumulation of Au was observed in the kidney already 0.5 hrs after Au applic.. 2.) This event was followed by a delayed increase of Cu and metallothionein concentrations 6 hrs after Au applic.. 3.) The Zn content in the kidney did not change significantly during the observation period of 12 hrs.

These results indicate that the first event is an increase of Au concentration in kidney cytosol. The second step seems to be an induction of metallothionein production by Au, which is followed by an increase of Cu binding to this metallothionein.

Department of Toxicology and * Department of Pediatrics, University of Kiel, D-W-2300 Kiel 1

In the last years, metallothionein (MT) became of increasing interest not only for the investigation of heavy metal toxicology but also for studies on tumor resistance to anticancer drugs, especially C/splatin and its analogues. However many investigations are limited, since MT quantification is rather difficult. Often the 2°JHg-binding assay and the Cd/hemoglobine affinity assay are used which are relatively unspecific and do not allow differentiation between various isoproteins of MT. Therefore two chromatographic methods are presented and compared, allowing an exact chemical quantification of different MT-isoproteins. Commercially available MT from rabbit liver was used as reference substance and sample preparation was done by solid phase extraction with RP-18 cartridges. First a conventional HPLC-analysis was performed using an acetonitril/water-gradient. Separation of MT-isoproteins from rabbit liver requires a 80 rain. run with a detection limit of 2/~g/ml MT. Secondly, MT analysis was done by capillary zone electrophoresis using a phosphatebuffer-system of pH 2,5 containing methylcellulose to reduce electroosmotie flow. Complete separation of MT-isoproteins was achieved after 20 min with a detection limit of 0,2 /zg/ml MT. For the future, capillary eleetrophoresis seems to be an appropriate method to achieve a fast and high resolving quantification of MT-isoproteins. Learning disorders and changes in behaviour have frequently been observed in children and in experimental animals as the result of chronic exposure to lead. The molecular causes of these neuretoxic effects are still largely unexplained. Damage to central receptor systems is assumed. As the NMDA (N-Methyl-D-Aspartate)receptor, a subtype of the glutamate receptor, is probably involved in cognitive functions in the CNS, our investigations were intended to clarify to what extent this type of receptor in the brains of mice was influenced by lead in vitro and in vivo. The investigations were carried out on adult and juvenile NMRI raice. Female mice were given 100 or 1000 ppm lead as nitrate in their drinking water tbr 30 and 90 days. The water was available ad libitum. Juvenile animals were exposed pre-and posmatally by providing pregnant mice from the 5th post conception day with 100 or 1000 ppm lead in their drinking water. Characterisation of the NMDA-receptors was carried out ex rive using binding studies on homogenates from the frontal cortex of the brain with the allosteric NMDA antagonist 3H-MK-801. In vitro, complete inhibition of the radioligand binding was found with an ICs0 value of 19.7 -+ 2.6 ~M (SEM). In comparison, the lead content in frontal cortex of juvenile mice treated with 1000 ppm lead was determined to be 10.0 +-. 1.8/.~mol/kg W.W. (SEM). In the presence of amino acids and of divalent cations, such as calcium or magnesium, low lead concentrations led to a significant increase in receptor affinity. Analysis of the saturation experiments, which were carried out on brain homogenates of lead-treated animals, showed a slight increase in receptor density of 13 or 15% with an unchanged KD value only in the adult animals treated with 100 ppm lead. If radioligand binding was stimulated with glycine and glutamate, this effect could no longer be observed. No changes in receptor affinity or receptor density were found in the neonatal lead exposed juvenile animals. Modulation of radioligand binding by lead, which was demonstrated in vitro, is not associated in the same manner with changes in receptors in the in vivo lead-exposed animals. In summary it can be said that on the basis of the in rive findings obtained in our experiments on mice, the neurotoxic effects of lead are not attributable to impairment of cortical NMDA receptors. 

C a d m i u m adversely affects the energy metabolism and thereby impairs the calcium homeostasis. To clarify the effectfi of Cd 2 + on myocardial cells we measured the intracellular free Ca z + concenIration over a longer period by means of fura 2. The influence of Cd z + was monitored during the systolic and diastolic phases of isolated cells of an adult guinea-pig heart. The tiara loaded cells were placed in a temperature controlled (35 ° C) chamber (0.5 ml volume) on a Zeiss IM 35 fluorescence microscope. The chamber was perfused with 0 2 saturated modified Krebs Hense-]eit Buffer (KHB; 1.2 m M Ca2+; 1 m l / m i n ) with or without CdC12. The fluorescence of one cell was selected with an adjustable rectangular diaphragm. The A m k o muttiscan-2-fluorimeter was used for excitation and data aquisition. _The ratio of 340/380 nm excitation represents the free intracellular Ca z + . After addition of 5 ~M Cd 2+ at 0.5 Hz stimulation frequency, we found a decrease of the ratio by 49% during 40 s. Thereafter only a slight further decrease of the calciumtransient but a m a r k e d increase of the diastolic Ca 2+ concentration has been observed. After 5 rain the cell was superfused with KHB without Cd 2 + . Then the transient increased but the Ca 2+ concentration during the diastole remained ~onstant~ These effects were als9 seen to a smaller extent at 1 ~M Cd 2+ . These results show that Cd 2+ has two effects on the free intracellular [Ca 2 + ]i" The first is a calciumantagonistic effect at the outside of the membrane. It corresponds to the results of Lansman et al. (J. Gen. Physiol., 88, 321-347, 1986) . They reported about blocking of the Ca current through single calcium channels in guinea-pig ventricular cells by Cd 2+ . The second may be an intracellular effect which leads to an increase of [Ca 2 + ]i in the diastolic phase. Further investigations are necessary to elucidate the latter effect. Institut fiir Toxikologie, GSF-Forschungszentrum ftir U m w e l t u. as injurious substances as they vary in their mechanism and cellular location of radical induction. In order to correlate the radical formation with the exerted cytotoxic effects the extent of membrane damage, the thiol and energy status, lipid peroxidation, membrane blebbing and disturbance of intracellutar calcium homeostasis were determined. Alterations in intracellular free calcium are supposed to be an important trigger of the cascade of pathobiochemical reactions in radical induced cell death. Nevertheless a correlation between the quantity of lipid peroxidation and call damage in general with changes in intracellular free calcium, as indicated by the fluodmetdc determination with Quin-2-AM could not be found in all cases. BrCI3C and CumOOH induced a rise in intrecellular calcium whereas the other substances used did not. On the other hand deterioration of call viability was supreme in AA treated calls. This marked discrepancy leads to the conclusion that cytosolic free calcium oscillations are not the pivotal triggering impulse in radical induced call death. Oximes are used in organophosphate (OP) poisoning to restore neuromuscular block (NMB). Because of lacking efficacy in some types of poisoning and inherent toxicity the therapeutic effect of obidoxime chloride (OBI), pralidoxime iodide (2-pAM) and the newer oximes HI 6 (l-(((4-(aminocarbonyl)pyridinio)methoxy) methyl)-2-(hydroxyim/no)methyl)pyridinium dichloride) and HL5 7 (l-(((4-(aminocarbonyl)pyridinio) methcxy) methyl)-2,4-bis((hydroxyim/no) methyl) pyridinium dimethanesulfonate) on paraoxon induced NMB of the mouse phrenic nerve-diaphragma was investigated in vitro. In a continuously perfused bath, muscles were stimulated indirectly in 30 sec intervals with 20 Hz (T20), 50 Hz (T50) and i00 Hz (T100) for 2 sec. The generated force within the stimulation period was measured (results as percent of control): firstly for control (100%); secondly for OP effect after a 25 minute incubation period with paraoxon (0.01 mmol/l), followed by wash out of the ~oison after NMB was achieved; thirdly for oxime effect estimation after a further 15 minute incubation period with different oximes (0.5 mmol/l, each; n = 6/group). Type XIII AChE (Sigma, St. Louis) was inhibited by paraoxon or monocrotophos for 30 minutes at pH 7,4 and 37,0°C and then reactivated by HL6 7, HI 6, obidoxime or pralidoxime (i0 ~mol/l, each) for 60 minutes.

The AChE activity was measured photometrically.

After incubation with paraoxon AChE activity decreased to 2.1% ~ 0.2% (mean ± SEM) of control.

The subsequent incubation with HL6 7 caused an increase to 70.2% ± 1.2%, with obidoxime to 72.4% ± 1.7%, with pralidoxime to 29.9% ± 2.1% and with HI 6 to 20.5% ± 1.0%. With monocrotophos the activity decreased to 9.5% ± 0.3% and increased with HL6 7 to 69.0% ± 1.1%, with obidoxime to 66.1% ± 2.0%, with pralidoxime to 58.4% ± 1.1% and with HI 6 to 37.8% ± 3.6% of control. The data show a reactivation of AChE activity with all oximes after inhibition by both organophosphates, where HL6 7 was at least as effective as obidoxime. Pharmacol. 339: suppl. RI8, 1989 ) the combined application of dexamethasone and activated charcoal markedly reduces lethality in acute poisoning with T-2 toxin (3~-OH-4B,15-diacetoxy-8~-(3-methyl-butyryloxy)-12,13-epoxy-trichothec-9-en) a trichothecene mycotoxin. We investigated wether the treatment alters the amount of metabolites excreted or the metabolite pattern. Tritium labelled T-2 toxin (0.66 mg/kg) was injected into the tail vein of 12 male Sprague Dawley rats. After 30 min., three rats each received activated charcoal (i g/kg, po; group II) or dexamethasone (8 mg/kg, iv; group III) or the combination of both (group IV) or only the vehicle (group I). T-2 toxin and its metabolites in urine and feces were determined by HPLC and LC-MS. During the experimental period of 50 hours, less than 1% of the administered dose appeared as unchanged T-2 toxin. In the urine, 2-4% of the administered dose was present as deacetylated T-2 toxin (HT-2), and about 10% as 3'OH HT-2 in all experimental groups. In the feces, 3'OH HT-2 was the main metabolite, (15% of the administered dose in group I and 20% in group II-IV) followed by 3'OH T-2 (1.2-2.8%). In minute amounts only (<1%) HT-2 was detected in group I-III, nothing in group IV. Furthermore, in group I less than 0.5% of the dose was found as HT-2 glucuronide, whereas in group IV about 4.5% was found as HT-2 glucuronide. It is concluded that the treatment with dexamethasone and charcoal did not alter markedly the elimination pattern of T-2 toxin and its metabolites. Chem. 251 289 (1970) . After improving the glassware and the experimental set-up we were able to concentrate ethyl acetate or acetone solutions of different compounds under partial reflux from the ppb-to the 10-ppm-range for quantitative GC: W. Diinges "Pr~i-chromatographische Mikromethoden" 1979.

Recently we have applied this principle to the analysis of lower boiling and more dilute solutions. By optimizing heating and cooling parameters, stable temperature conditions in the reflux zones of the concenlxating vessels were achieved. This standardized procedure allows the optimal performance of up to 12 parallel experiments without individual adjustment of the vessels. A considerable gain in practicabilty and reproducibility results.

The stream of the constantly evaporating solvent p[events backdiffusion of air into the vessels. Solvents and solutes which are prone to autoxidation are thus protected from oxygen.

We can now reduce e.g. 5 ml of ethyl ether containing "PCB"s or N-pesticides in the 100-pg-amounts to 5 td with minimal losses. The same holds for all investigated polar or nonpolar ~xaces in ether, whereas in pentane intermediately polar compounds as esters or chlorinated pesticides as endrine are retained poorly.

The method is the basis for quantitative GC of pesticides (pptrange) from 20 ml of water and of volatiles (ppm-range) from 10 /~1 of blood. Silicosis is a progressive fibrotic lung disease caused by inhalation of silica (SiO2) particles. There is suggestive evidence that, given a comparable dust exposure, genetic factors decide on whether silicosis develops or not. For investigation of this different genetic predisposition to fibrotic reaction a mouse model was developed with Balb/c as high-responder and DBA/2 as Iow-responder strains. Silica particles are taken up by macrophages and presumably enhance the production by these cells of fibrogenic cytokines, such as IL-11?,, TNF~,, and TGF-fL if true, genetic differences in susceptibility to silicosis should be reflected in a different cytokine pattern long before formation of pathological alterations. To test this hypothesis Balb/c and DBA/2 mice were injected i.p. with 0.Stag SiO 2 or, for control, the inert dust TiO 2, Using PCR, mRNA expression by peritoneal cells of the IL-lf$, TNFc¢, IL-6, and TGFf~ 1 was determined after 5,8,11,14 and 17 days, respectively. One group of both strains remained untreated. Following injection of SiO 2 and TiO 2, respectively, non-responder DBA/2 mice failed to show mRNA expression of TNFo,, TGFI~I, and IL-6. IL-1I~ was expressed constitutively, but dust treatment failed to enhance expression of this cytokine. In the highresponder Balb/c, IL-I£; mRNA expression showed a strong increase after exposure to both TiO 2 and SiO 2 ( days 11, 14, 17). TNFe¢ expression was enhanced in SiO2-treated mice (days 11, 14, 17) compared with the TiO2-treated and the untreated groups. As in DBA/2 mice, TGFI~ 1 expression was undetectable. Similar to IL-11~, expression of IL-6, an inflammatory but not a fibrogenic cytokine, was elevated in both TiO 2-and SiO 2-treated Balb/c mice. These genetic differences in SiO 2-induced cytokine expression are consistent with the results of previous studies showing an essential role of TNFct in the pathogenesis of silicosis. Furthermore, this animal model may be used as a test system for antifibrotic and anti-inflammatory drugs. Several members of polycyclic aromatic hydrocarbons (PAH) containing the structural element of a "fjord-region", e.g. the pentacyclic benzo[c]chrysene (BcCh) and benzo[g]chrysene (BgCh) and the hexacyclic dibenzo[a,l]pyrene (DBalP), exhibit mutagenic and carcinogenic properties. In particular the environmental pollutant DBa/P is the strongest carcinogenic PAH ever tested.

As in the case of PAH containing a "bay-region" it seems reasonable to assume that the genotoxicity of BcCh, BgCh and DBa/P is due to theh" enzymatic conversion to "fjordregion" dihydrodiol epoxides.

To examine this hypothesis, it was necessary to synthesize the putative active metabolites of all three PAH. A novel synthetic approach was devised to construct the aromatic system of DBa/P and regioselectively the desired derivatives.

Investigations on the raicrosomal metabolism of "pre-fjord" dihydrodiols of the three PAH, applying reversed-phase HPLC, indicated the format/on of tetrahydrotetraols which derived non-enzymatically by hydrolysis of the "fjord-region" anti-dihydrodiol epoxides. No evidence was found for the metabolic formation of We syn-diastereomers.

In line with the results for BcCh and 8gCh (H.R. Glatt et al., Cancer Res. (199t) ~1, 1659-1667) the "fjord-region" dihydrodiol epoxides of DBa/P were highly mutagenic as determined by reversion of the histidine-dependent mutant strain TA 100 of Salmonella typhimurium; the anti-~astereomer surpassed the bacterial mutagenicity of the corresponding "bay-region" isomer of benzoIa]pyrene by a factor of ten thus representing the dibydrodiol epoxide with the strongest mutagenic properties of all PAH investigated so far. In addition, investigations on the DNA-binding in vitro of the "fjord-region" dihydrodiol epoxides of DBa/P by the P-32 postlabeling method exhibited multiple nucleobase adducts thus showing a remarkable and polyregional affinity of these electrophilic molecules to DNA.

These observations lead to the conclusion, that the strongly carcinogenic DBa/P is possibly activated by microsomal anti-epoxidation of the trans-11,12-dihydrodiol in the "fjord-region" to highly genotoxic metabolites. Dibenz[,~,h]anthracene (DBA) exhibits two bay-regions at distal sites of the molecule. Many polycyclic aromatic hydrocarbons bearing a bayregion are metabolized by P-450 enzymes into possible DNA binding diol epoxides, Also from DBA bay region diol epoxide-DNA adducts were found after metabolic activation and the absolute stereochemistry of these adducts was assigned. Furthermore highly polar DNA adducts were obtained. These polar DNA adducts were also formed from /razz.ca'arzg-3,4; 10, ll-tetrahydroxy-3,4; 10,11 -tetrahydro-DBA (DBA-3,4;10,11bisdiol) as well as from a,'at~,s-3,4-dihydroxy-3,4-dihydro-DBA (DBA-3,4-diol) after incubation with calf thymus DNA in presence of liver microsomes form rats treated with Aroclor 1254. Due to the symmetry of the molecule only three enantiomers of DBA-3,4;10,11-bisdiol are conceivable. Highly polar DNA adducts were only observed for metabolites of those DBA-3,4;10,11-bisdiols, 3A',4A';10A',llA'-tetrahydroxy-3,4;10,11-tetrahydro-DBA and 3A',4A';10S,11,.¢-tetrahydroxy-3,4;10,11-tetrahydro-DBA, having at least one R,A'-configurated dihydrodiol moiety. No DNA adducts which could be assigned to metabolites of 3~,4d'; I 0~,11S-tetrahydroxy-3,4; 10,11-tetrahydro-DBA were found. These findings were congruent with the observation that bay-region diol epoxide-DNA adducts were predominently formed from a metabolite of the h',A~-configurated enantiomer of DBA-3,4-diol after incubation of DBA with DNA in presence of the metabolizing system. Because of this similar stereoselectivity it is probable that the same isoenzyme which is responsible for the further biotransformation of DBA-3,4-diol to the diol epoxide is also responsible for activation of 4; 10, We have developed an assay in which specific repair endonucleases (formamidopyrimidine-DNA glycosylase, endonudease III, exonudease ]II) are used to quantify various oxidative modifications (8-hydroxyguanine, 5,6-dihydropyrimidines, sites of base loss) in mt-DNA. The assay makes use of the fact that an incision by repair endonucleases at modified DNA sites gives rise to a relaxation of the supercoiled mt-DNA molecule that can be easily detected and quantified. The mt-DNA was isolated from mitochondria of rat and porcine liver and kidney by a mild lysis protocol; supercoiled DNA from bacteriophage PM2 was added before lysis and served as internal control for damage during the workup procedure. The results indicate that the mt-DNA of the animals contained less than one 8-hydroxyguanine residue per 105 base pairs. No base modifications sensitive to endonuclease III (5,6-dihydropyrimidines) could be detected. It has to be concluded that the steady-state level of oxidative modifications in mt-DNA of young healthy animals is low. Chloroprene (2-chloro-1,3-butadiene) is used for the production of chemical resistant plastics and glues. The substance has not been classified by the IARC (International Agency for Research on Cancer) in regard to its carcinogenic potential. Since the chemically closely related substances 1,3-butadiene and vinyl chloride are proven carcinogens, and previous data on chloroprene are inconsistent, experiments were performed to investigate the mutagenicity of chloroprene. Pure chloroprene is unstable. It was purchased from Matthey Co., FRG, as a 50% solution in xylene and freshly distilled prior to each experiment. Due to the low boiling point of chloroprene, the Ames Test with S. typhitnurium TA100 was performed as a preincubation test in gas tight screw cap vials. Propylene oxide served as positive control. Cytotoxicity and mutagenicity were investigated in parallel experiments. Freshly distilled chloroprene was not mutagenic, neither as solution in ethanol, nor in DMSO, nor with the addition of $9 mix. However a mutagenic effect occured with increasing age of the distilled substance. The mutagenicity of aged chloroprene was the same whether it was tested under gas tight conditions or not, it doubled when $9 mix was added. Glutathione, in physiological concentrations, reduced the mutagenic and cytotoxic effects, both with and without metabolic activation. Analysis by gas chromatography revealed six substances, which were not found in freshly distilled chloroprene.The mutagenicity of aged chloroprene correlates linearily with the appearance of four of these substances. Under pure oxygen, the increase of these byproducts was faster than in samples kept under air. Under argon, two new substances are formed, which were also found to be mutagenic in the Ames Test.The formation of these substances was slow at -20°C and rapid at room temperature Analysis by GC/MS revealed the presence of peroxides and dimers in chluroprene kept under aerobic conditions, and the presence of dimers in samples kept under argon.

The results show that reaction products ofchloroprene, but not the substance itself, are responsible for the mutagenicity in the Ames Test. (1992) Chem Ree Toxicol 5: 470]. The formation of 8,9-dihydro-8-(NT-guanyl)-9-hydroxy -AFB~ as the initial lesion of DNA from the reaction of DNA with the ¢xo-epoxide has been described previously [Baertschi et al. (1988) JAm Chem Soc 110:7929] . The question arises whether both or only the ¢xo-epoxide are responsible for the DNA adduct formation and by that also for the species differences in sensitivity for AFB 1 induced cancer.

The 8,9-epoxide stereoisomers were tested in the Ames assay with S#lrnonell~ O~plz#nuHunt TA 100 and in the SOS-urnuC-assay with S. typh. 1535/pSK 1002. The exoepoxide was isolated from the epoxide mixture by crystallization. Unfortunately the sndoepoxide could not be separated completely from the epoxide mixture (ratio axe-shale 4:1). While the ¢xo-epoxide has a very short half life in 50% acetone/water at room temperature (0.7 rain) the ¢ndo-epoxide is more stable under these conditions (t 2:30 rain). Prior to the assays with the endo-epoxide, the mixture was allowed to hydrot/y~e for 15 min in 50% acetone/water. During this period the amount of exo-epoxlde was reduced 103 -105 fold while the amount of sndo-epox[de was decreased by only 25%. 250 ~ 20 0.5 13 * Numbers represent 3 experiments with duplicales ea.ch; no metabolic activation system was added: concentrations of sn~/o-epoxide are prior to hydrolysis cy~olox city appeared at highest dose of each compound The results in the table show that in both test systems only the ¢xo-epoxide shows rnutaagenic response. The endo-epoxide produced neither an increase of the revertant rate of E. /yph. TA 100 in the Ames assay, nor expression of P.-galactosidase in the SOS/~rn~Cassay. ~hese results suggest that the species differences in the sensitivity of AFB~ tumorlgen~city are due to the amount of the exo-epoxide produced and to the capacity ot ¢xo-epoxide detoxificstion by g[utathione conjugation of the species but not to the effects of the stz#/o-epoxide of AFB 1.

Center in Molecular Toxicology, Vanderbilt University, Nashville, TN 37232, USA

Electrostatic interactions between histones and poly(ADPribosyl)conjugates are involved in nucleosomal unfolding of chromatin required for DNA repair, DNA replication and initiation of gene transcription. Thus poly(ADP-ribosyl)ation, which is induced by DNA strand breaks, may be an important link between DNA damage, alteration of gene expression and malignant transformation. The present study investigates the induction of poly(ADP-ribosyl)ation by three renal carcinogens with different mechanisms of action. Dichlorovinylcysteine, the metabo~ite responsible for the nephrotoxic and nephrocarcinogenic effects of trichloroethene, is a potent inducer of Cag+-dependent DNA double strand breaks; application of dichlorovinylcysteine to male Wistar rats in vivo results in a dose-and time-dependent increase in the amount of poly(ADPribosyl)conjugates in the rat kidney cortex. A similar increase was observed with KBrO3, whose nephrotoxicity and nephrocarcinogenicity is probably resulting from increased formation of reactive oxygen species. Administration of a single dose of dimethylnitrosamine to rats, maintained on protein-free diet, has been shown to induce kidney tumors in absence of gross nephrotoxicity. However, the identical exposure regimen did not influence the poly(ADP-ribosyl)transferase activity in the renal cortex nuclei from treated rats. The results indicate that poly(ADP-ribosyl)ation may be an important aspect of the organ specific effect for nephrocardnogens that mainly act by inducing mediated DNA damage.

Institute for Toxikologie, Universit&t W~rzburg, Versbacherstr. 9, D-8700 W=3rzburg, FRG. The possible role of araehidonic acid metabolism in the tumor-promoting activities of TCDD and TCB was studied i~n vitrq. (i) TCDD and TCB enhanced the DMS0-induced differentiation of the erythroleukemia cell line F4-6, as measured as an increase of the cellular hemoglobin content. This effect was significantly reduced by indomethacin, an inhibitor of cyclooxygenase, or dexamethasone, an inhibitor of phospholipase A 2. (2) TCDD enhanced the malignant transformation of M2 mouse fibroblasts induced by Nmethyl-N'-nltro-N-nltrosoguanidine or 3-methylcholanthrene. This effect was abolished by the addition of either indomethacln or hydrocortisone. Furthermore, M2 flbroblasts were labelled with 3H-araehidonic acid and, thereafter, the calcium ionophore (A23187)-stimulated release of radioactivity into the culture medium was measured. This release was enhanced after TCDD-or TCBtreatment of these cells. Thus, the results suggest that the metabolism of arachldonic acid is increased by TCDD or TCB and that metabolites of the aracidonic acid may be involved in the tumor-promoting activities of TCDD or TCB i~n yitro, !-~., the enhancement of differentiation and of cell transformation. and chemically-induced severe liver injury. It has been suggested that these cells can give rise to cholangiocellular as well as hepatocellular carcinomas. In order to test this hypothesis established oval cell lines were transformed in vitro either by leaving the cells for long time periods in confluence or by exposing the cells to N-nitro-N'-nitro-N-nitrosoguanidine and injected s.c. in newborn rats. The obtained tumors were poorly to well differentiated cholangiocellular carcinomas, being solid, cysticpapillary and glandular in structure; no areas of hepatocytic differentiation were visible. Whereas the tumors were albumin-, alpha-fetoprotein-and ferritin--negative, they stained positive for cytokeratins 7, 8, 18 and 19. In an attempt to analyze the plasticity of the cultured oval cells these were exposed to dimethyl sulfoxide (DMSO) or sodium butyrate (SB), two differentiation promoting agents, for up to 6 days and the expression of a number of bile duct epithelial and parenchymal cell markers was studied. After addition of DMSO or SB 35-40% of the cells reexpressed albumin, glucose-6-phosphatase activity was enhanced and tyrosine aminotransferase activity remained undetectable; furthermore, SB strongly increased gamma-glutamyltranspeptidase and alkaline phosphatase activities. In conclusion, the results presented herein show that: i) oval cells are the precursor cells of cholangiocellular carcinomas; 2) certain traits of mature liver parenchymal cells can be rexpressed in oval cells after exposure to DMSO or SB; however, taking into account the morphology of the cells before and after treatment with DMSO or SB one cannot conclude that these substances are supporting a general program of hepatocyte differentiation in oval cells. Toxicol. 56, 1-6 (1984) ). With polyfunctional compounds, several adducts can be formed, and their analysis provides information about major and minor metabolic activation pathways, which may be used to assess the metabolic status of the individual. We administered single doses of 1,3-diaminobenzene (1,3DAB), l-amino-3-nitrobenzene (IA3NB) and 1,3dinitrobenzene (I,3DNB) to female Wistar rats, isolated hemoglobin after 24 h, determined the hydrolysable adducts and calculated a hemoglobin binding index (HBI = Binding (ntmol/mol Hb)/Dose (mmol/kg)). At least two cleavage products were obtained by hydrolysis of the sulfinamide type adduct. The high HBI of 1,3DNB corresponds well with the known capacity of met-Hb formation. It is concluded that (i) nitrogroups in 1,3DNB are to some extent reduced to the amine , but l-nitroso-3-nitrobenzene is predominantly bioavailable, Ethylene oxide (EO), a proven carcinogen in animals, is a major plastic monomer and a sterilizing agent for heat sensitive materials. Due to its chemical reactivity, it is able to react directly with nucleophilic sites in proteins and nucleic acids. During the past years, some research has been done on the determination of these adducts in animals and humans. Both have been discussed as possible parameters for a biological monitoring of exposed workers and linear dose-adduct relationships have been investigated. In a previous study, we determined DNA adducts both in lymphocytes of unexposed rats (base levels) and in liver of rats exposed to unlabelled EO. Furthermore, Potter et al. [1] determined the relationship between the alkyladon of DNA of various tissues (and hemoglobin) using 14C-labelled EO. On the basis of these results, however, a comparison ofadduct levels between exposed and unexposed animals is not possible. Male Sprague Dawley rats (5 animals per dose) were exposed in a closed exposure chamber (10.3 1) to 0, 30, 60, and 125 gmol 14C-labelled EO. In addition, 5 animals received 460 gmol unlabelled EO in a similar exposure system. The EO concentration was monitored by GC and the exposure was terminated at a final concentration lower than 1 ppm. The animals were sacrificed andliver and spleen were extracted from each animalJ Lymphocytes were isolated from all animal blood sample. DNA was prepared from these samples, hydrolysed in 0.1 N HC1, and 7-(2-hydroxyethyl)-guanine levels were determined by HPLC in the case of the labelled compound and by GC/MS in the case of unlabelled EO. Prior to the determination, the GC/MSsamples were silylated by N-methyl-N-trimethylsilyltrifluoroacetamide. The investigation of labelled 7-(2-hydroxyethyl)-guanine shows a linear dose-dependency that is comparable to the results of Potter et al. [1] . The amount of this guanine adduct is approximately 20 % of the whole total activity bound to liver DNA. The results allow a comparison with the results from lymphocyte and liver DNA previously investigated.

[1] Potter et al.: Arch. Toxicol. Suppl.: 13, 254-257 (1989 Methyl bromide (MBr) is widely used as a fumigant because of its fungicide, nematocide, and insecticide properties. Since the substance has a high acute toxicity and a systemic genotoxic potential in laboratory animals, monitoring of human exposureto MBris necessary. This monitoring cannot be performed by the measurement of the concentration of the substance in ambient air or by its quantitation in blood or urine due to interindividual differences in metabolism. Prelimin ary in vitro experiments showed that an enzyme polymorphism for GSH conjugation in erythrocytes influences the individual uptake and distribution of MBr in human blood. A determination of the biological effective dose by measurement of alkylation adducts with blood macromolecules may therefore be better suited for biological monitoring. On the basis of these in vitro experiments, a method was established for the parallel quantitation of serum albumin and globin adducts as biomonitoring parameters. First, the proteins were isolated from whole human blood. After acid hydrolysis (6 N HCI, 1100 C, 24 h), amino acids were de rivatized with FMOC, yielding fluoresce nt adducts. In the next step, HPLC amino acid analysis was performed using a RP 8 column, a binary gradient system, and a fluorescence detector. Levels of S-methylcysteine were determined in both proteins. Compared to unexposed controls, significantly higher levels of cysteine adducts were measured in both albumin and globin of MBr workers. Additionally, the importance of the enzyme polymorphism was outlined. As a result, the method presented here is suitable forthe biomonitodng of MBr-workers and shows a good comparison between the common target proteins in human blood. , one of the carcinogenic heterocyclic arylamines found in fried meat, fish and meat extract, is metabolically activated to DNA-binding arylnitrenium ions by N-oxidation and a subsequent acetylation. Using cell cultures we have recently shown that human N-acetyltransferases (NAT1 and NAT2) contribute to the activation and DNAbinding of IQ and probably have an associated O-acetyltransferase (OAT) activity (Probst et al., Carcinogenesis 13, 1713 ft., 1992 . Here we report experiments performed to test a correlation of N-acetyltransferase (NAT) activity in different organs of rats and the occurrence of IQ-DNA-adducts. Two rat strains (F344, WKY) were used. NAT activities with the substrates sulfamethazine (SMZ), p-aminobenzoic acid (PABA) and IQ were determined in cytosols from liver, lung, pancreas, and intestine. After a single oral dose of IQ (30 mg/kg) IQ-DNAadducts were assayed by means of 32p-postlabelling. The PABA-NAT activities in liver and other organs were 3-5 times higher in F344 than in WKY rats. Thus the F344 rats are rapid and the WKY rats slow acetylators. However, the SMZ-NAT activities differed only slightly and the IQ-NAT activities were identical in the two strains. Of all organs, the intestine had the highest PABA-NAT activity (5-10 times more than liver); liver generally had the highest SMZ-NAT and IQ-NAT activities, in F344 rats, liver had also most IQ-DNA-adducts (304 adducts/108 bases); the adduct frequencies correlated with the IQ-NAT activities in all organs (r2=0,9) . Adduct studies with WKY rats are in progress. The above data show that the level of genotoxicity of IQ depends on the acetylator status. This is probably relevant also for human slow versus rapid acetylators.

( A number of heterocyclic amines (HA) formed in protein-rich food upon cooking under household conditions have been shown to be strongly mutagenic and carcinogenic in rodents; following metabolic activation to hydroxyl amines HA form covalent DNA adducts. 32P-Postlabelling analysis of HAmodified DNA has resulted in characteristical patterns of up to five adduct spots with considerable inter-laboratory variations.

Incubation of HA with rat hepatocytes or photoreaction of DNA with azido derivatives of HA yielded similar patterns upon conventional a2p-postlabelling analysis. By introducing an additional Nuclease P1 hydrolysis step following the labelling reaction, we obtained a simplified pattern of 5'-monophosphate adducts: One major adduct (>95%) was observed for IQ, 4,SDMIQx, 7,8DMIQx and PhiP-modified DNA, respectively, while two major adducts (>95%) were observed for both MIQ and MIQx.

Thus, we reduced the number of spots to eight major 32p. labelled 5'-monophosphate adducts formed by six very similar HA. While these were not ressolved on ion-exchange TLC, HPLC on phenyl-modified reversed-phase columns yielded eight distinct peaks of radioactivity.

Since man is exposed to mixtures of HA, this new method will be a useful tool in analysing DNA from human biopsy samples and towards the estimation of the carcinogenic risk imposed by these compounds. In addition to ct-C-hydroxylation and subsequent bioactivation, carcinogenic N-nitroso compounds are also denitrosated by a Cyt. P-450 dependent reaction. NO is generated by this pathway and it is converted to NO2" and NO3" by the intermediary formation of NO2. It should be mentioned that NO is also generated via the L-arginine pathway as a 2nd messenger in different mammalian cells. Previously we could show that gaseous NO: was able to induce DNA single strand breaks (SSBs) in Chinese hamster lung (V 79) fibroblasts and that physiological antioxidants like 3'.-toeopherol, Bcarotene, ascorbic-and uric acid were able to inhibit these NO~-induced genetoxic effects. Possibly the denitrosation pathway may contribute to the careinogenicity of N-nitroso compounds. Quite probably NO2 has some relevance for the observed lipid peroxidation induced by some Nnitrosamines.

To investigate further the significance of NO and its oxidation product NO2 with respect to possible genetoxic effects, we studied SIN-I which is the active metabolite of molsidomine, a drug against angina pectoris. As NO is generated from S1N-I the denitrosation reaction of N-nitrosamines could be simulated in metabolic inactive V 79 cells. Therefore V 79 cells were treated with millimolar concentrations of SIN-I and micromolar concentrations of various antioxidative vitamins for different hours. The rate of SSBs was measured by the alkaline elution assay, the amount of DNA by a fluorimetric assay, We found considerable amounts of SSBs by treating V 79 cells with e.g. 3 rnM SIN-I for 12 hours. These SSBs were effectively inhibited by addition of e.g. 100 pM ascorbie acid. The antioxidative vitamin ~,-tocopharol did not inhibit SIN-l-induced SSBs. The fact that SIN-I induced SSBs in V 79 ceils, which were inhibited by aseorbic acid supports the hypothesis that NO~ and/or other radicals are involved in the SIN-1 dependent generation of genetoxie effects. Methylene chloride (CH2Cl 2) has been classified by the European Communities as a compound suspect of carcinogenic activity (R40). A glutathionedependent metabolic pathway leads to the formation of formaldehyde (HCHO). In earlier experiments with methyl chloride (CH3C1), we could ascribe the resulting DNA-damage due to the action of HCHO, an intermediate formed in CH3CI metabolism which may react with nucleic acids and protein.

In the study presented here, the induction of DNAdamage was investigated in the DNA of lung and liver of male mice (B6C3FI) after exposure of the animals to 4000 ppm CH2CI 2 for 7h. Immediately after exposure, lung and liver cells were isolated and subjected to a modified alkaline filter elution (AFE). To assay for cross-links, aliquots of cell nuclei were x-ray irradiated at 600 rad with or without proteinase K treatment prior to AFE.

The elution curves showed the occurrence of single strand breaks in liver DNA but not in lung DNA of mice. DNA-DNA-(DDCL) and DNA-protein-cross-links (DPCL) were obtained in both lung and liver of mice. The elution profiles showed that the DDCL and DPCL are more pronounced in lung than in liver.

Our data point to an involvement of HCHO in CH2CI 2 induced genotoxicity in the male B6C3FI mice.

Institut f~r Arbeitsphysiologie an der Universit~t Dortmund, Ardeystr. 67, D-4600 Dortmundl Commercial grade toluene diisocyanate (TDI) is widely used in the industry especially for the production of polyurethanes and for various glues, paints and dyes. Inhalation of TDI vapours is associated with immediate type hypersensitivity, direct toxic effects and allergic alveolitis. TDI is known to induce chromosome aberrations, base-pair substitution and frameshift mutation after metabolic activation.

The immediate effect of toluene diisocyanate (TDI) upon genomic DNA of sheep and rabbit white blood cells has been investigated. The results of FPLC indicate that TDI induces DNA double-strand breaks in sheep white blood cells. Following denaturation of TDI treated DNA, the results demonstrated that at least one DNA fragment, representing about 10 -20% of total DNA, is cross-linked by TDI or TDI metabolites. About 39% more of double-stranded structures are reconstituted in DNA extracted from sheep white blood cells three hours after incubation with TDI.

In addition, chromosomal DNA from rabbit blood was investigated with "agarose-plugs" method b~¢ pulsed-field gel electrophoresis (PFGE) two hours after incubation with TDI, and the agarose gels were quantitatively evaluated with a laser densitometer. Further studies with prednisone have been conducted to correlate these results with the phenomenon of apoptosis. The in vitro results show that TDI can induce degradation of mitochondrial DNA and large DNA fragments caused by apoptosis to small DNA fragments.

Research Institute of Occupational Medicine at the Ruhr-Un/versity of Bochum, Gilsingstr. 14, D-4630 Bochum 1, Germany MUTAGENICITY OF SENNOSIDES AND RHEIN U. Mengs*, C.D. Stotzem* and A. Heidemann** Sennosides and rhein did not increase the number of revertants in five Salmonella typhimurium strains when tested up to 5000 gg/plate in the presence and absence of metabolic activation. Sennosides and rhein were tested for their ability to induce mutations at the thymidine-kinase (TK +/') locus of mouse lymphoma L 5178 Y cells both in the presence and absence of $9 mix. Up to the highest concentrations tested, there were no biological significant increases in mutation frequencies with sennosides or rhein. The chromosomal aberration test was carried out with cultured Chinese hamster ovary cells (CHO). From the results, there was no evidence for a clastogenic activity of sennosides and rhein either in the presence or absence of metabolic activation. The micronucleus assay was performed in mice at oral dose levels of 100, 500 and 2,500 mg/kg (sennosides) or 1,500 mg/kg (rhein). The test compounds did not increase the number of micronucleated polychromatic erythrocytes. These results suggest that sennosides and rhein have no mutagenic potential in vitro and in vivo. is the most widely used antituberculosis drug. The hepatic toxicity of INH was noted early and there are reports of increased risk of liver reaction when INH-treatment is combined with rifampicin, another antituberberculosis drug, or with alcohol consumption. Mutagenic activity of INH was found in the salmonella-mutagenicity-assay which has been proposed due to its metabolic conversion into hydrazine, a well-known microbial mutagen. Rifampicin and ethanol are known to be chronic inducers of naicrosomal enzymes. In this study investigations on the interaction of INH, rifampicin and ethanol were performed in order to detect if induction of microsomal enzynaes may increase the mutagenicity of INH. Mice were pretreated orally with 1NH (66 mg/kg b.w.), rifampicin (88 mg/kg b.w. in drinking water ad lib) or ethanol (10 % w.v.) alone, or with a combination of the substances mentioned above. The effect of pretreatment was tested on induction of microsomal enzymes (Burke et.al., Biochem. Pharmacol.: 34, 3337, 1985) and on mutagenicity in the salmonella-mammalian microsome assay (Ames-test) according to Maron and Ames 1983 (Mut. Res.: 113, 173) . Liver-S9 of differentIy pretreated animals was used as metabolizing systena and was tested on its enzyme activity. The data demonstrate that the oxidative metabolisna of INH can be induced by pretreatment with INH and rifampicin and/or ethanol and is increased especially by pretreatment of female animals with ]NH, rifampicin and ethanol. In the Ames-test INH demonstrated a slight mutagenic activity with TA 1535, no nautagenicity was found with TA 100 and TA 98 tester strain of salmonella. However, compared to control mouse liver-S9, the metabolic activatioo mediated by pretreated mouse liver-S9 caused neither enhancement nor reduction of the response. The results postulate that concomitant treatment with rifanapicin or alcohol consumption do not increase the mutagenic effect of INH. In a previous study, it was shown that 2-naethylpropene (isobutene), a gaseous compound widely used for the production of rubber chemicals, plastics, adhesives, sealing compounds, fuel additives and antioxidants, is metabolized to the epoxide 2-methyl-1,2-epoxypropane. Furthermore, the mutugenic activity of 2-naethylpropene and its naetubolite was tested in several systems. In contrast to the parent compound, for which the results were negative, the epoxide showed amutagenic effect in the Ames test.

In the experiments presented here, the nautagenicity ofthesetwo compounds was tested in the ,,in vitro" micrunueleus test using human lymphocytes. Cultures (5nal), were prepared in gastight test flasks coutaining4 ral medium (Ham F10 with Hepes buffer), 0,8 ml foetal calf serum, 0.1 nal phytohaemagglutiuine and 0.4 nal whole blood. 2-Methylpropene and its epoxide were injected with a gaslight syringe in different concentrations. Since 2-naethylpropene is practically insoluble in water, each concentration was tested in double -once by shaking the cultures to maximize exposure to the cells, and once without shaking. After incubation at 37°C for 44 h, cytoeholasine B was added to a final concentration of 61.11/ml, and after 72 h the cultures were harvested. The lymphocytes were isolated by density centfifugation with Ficoll-Paque and subjected to hypotonic treatment 0tPMl:water, 1:3). The sample was again centril~gated and the pellet was applied to slides as spots. The slides were subsequently fixed in 100 % naethanol for 10 nfin and stained with 5 % Gienasa. 1000 binucleate cells were counted at a magnification of 1000. The test results show that 2-methylpropene induced no naicronuclei in contrast to its epoxide 2-methyl-1,2-epoxypropane. This epoxide shows an approximate dose response in the frequency of micronuclei, Shaking of the cultures had ilo influence on the mutagenic effects, neither for 2-naethylpropene nor for its epoxide. The toxicity test parameters concerning the chick embryo comprise, generally, lethality, developmental retardation, teratogenicity, hematological criteria and organ histopathology (see, e.g., Kemper and Luepke, Fd. Chem.Toxic. 24, 647-648, 1986) . Additionally, some biochemical and physico-chemical methods may provide an excellent means in screening for nucleotoxicity. Therefore, some DNA and/or chromatin interactive agents were tested in vitro and/or in vivo by scheduled (SDS) and unscheduled (UDS) DNA synthesis, RNA synthesis, viscometry of alkaline cell lysates and nucleoid sedimentation with freshly isolated liver and brain cells from the 15th day old chick embryo. As compared to manunalian cells, the physico-chemical tests reveal a high spontaneous -possibly apoptotic -rate of DNA strand breakage in the embryonic cells which is biochemically confirmed by the high basic activity of some DNA repair enzymes and/or phenomena, such as poly(ADP-ribose)polymerase and hydroxyurea resistant DNA synthesis. Because of the high spontaneous DNA fragmentation in primary cultures of chick embryo cells some DNA strand breaking agents may be detected with lower sensitivity than in mammalian cells. An additional disadvantage is the innate inability to metabolize various indirectly acting agents. Apart from this, similar dose-effect relationships were obtained with directly acting genotoxic agents in chicken embryo and rat cells. The gp 80 giycoprotein complex is the canine homolog of the rat TRPM-2 (Testosteron-repressed message-2), the htmaan Apolipoprotein J and complement lysis inhibitor (CLI). There is direct evidence from avian and circumstantial evidence from mammalian systems that the gp 80 gene is regulated by oneogene products. In quail neuroretinal cells gp 80 expression is enhanced after transfeetion with Rous Sarcoma Virus. In mammalians gp 80 is induced during apoptosis in the regressing prostate of the rat after castration and in rat kidney after ureter obstruction. In both cases the expression of c-myc is increased. We have shown, that gp 80 mRNA is enhanced in human kidney clear cell tumors and in rat phaeochromocytoma cells PCI2 during NGF induced differentiation, which is known to be mediated by the action of ras oncogene. We have isolated a mouse genomie done carrying 10 kbp of 5"-ftanking sequences, the 5"-untranslated region and most of the coding part of the gp 80 gene. We will use these sequences for mobility shift assays, footprinting experiments and transfection experiments with reporter genes in suitable acceptur cells, such as PC12 and Rat-1 cells expressing the e-myc gene under the control of a regulatable promotor, to characterize the DNA elements involved in gp 80 regulation. Cellular oncogenes (protooncogenes) are involved in growth and differentiation of cells. Protooncogenes of the raf-family are located in the cytoplasm and mediate signal transfer between membrane bound receptors and nuclear effectors. An altered expression has been found in tumors and transformed ceils. We studied the expression of c-raf in early stages of chemically induced rat liver tumorigenesis using juvenile female or adult male Sprague-Dawley (SD) rats: After initiation with diethylnitrosamine, for promotion polychlorinated biphenyls (PCBs) or phenobarbital (PB) was applied to female rats, to males PCBs only. The incidence of preneoplastic loci was evaluated histochemically by demonstrating deficiency in adenosine-5'-triphosphatase (ATPase) and emergence of gammagiutamyltranspeptidase (GGTase). C-raf expression was measured biocherrdcally in /iver tissue containing preneoplastic foci and in isolated hyperplastic nodules up to 36 weeks. In female rats foci numbers amounted to 60-70/cm 2 liver section with both promoters and both histochernical markers. In male SD rats foci numbers reached about 20-40/cm 2 with both markers and with PCBs as promoting agent. Foci area developed more rapidly in female rats. In female rats c-raf expression foil-owed generally the time course of foei development and was highest in large nodules (> 4 mm in diameter) with up to 3-fold in PB-treated and 10-fold-in PCB-treated animals compared to untreated controls. In male rats craf expression was marginal and nodule incidence was neglectibly low.

In conclusion, c-raf oncogene expression correlated with the incidence of foci and nodules, female rats being more sensitive than males.

GSF -llnstitut ftir Toxikologie, 8042 Neuherberg GSF -2Abteilung ffir Zellchemie, 8042 Neuherberg Overexpression of p-glycoprotein encoded by fine mdrl class of genes confers multidrug resistance towards a variety of cytotoxic drugs since p-glycoprotein functions as an energy dependent efflux pump of amphipathic polycyclic compounds. In hepatocytes which exhibit physiologically high expression of p-glycoprotein, the protein is located in cytoplasmic membranes exposed to biliary canaliculi indicating its significance as an excretion system for toxic exogenous or endogenous substances into the bile. In order to investigate a possible influence of tumor promoters such as phenobarbital or of hepatotrophlc growth factors such as epidermal growth factor (EGF)or transforming growth factor-a (TGF-a) on p-glycoprotein expression, primary murine hepatocytes isolated by collagenase perfusion and maintained in serum-free culture medium were incubated with EGF {10 pM-10 nM) or 1.5 mM phenobarbital for up to 3 days. Subsequently, p-glycoprotein expression was determined by Western blot analysis of cytoplasmic membrane fractions utilizing enhanced luminescence detection. Treatment of murine hepatocytes with EGF had no significant effect on p-glycoprotein levels although partial hepatectomy is known to induce mRNA expression in rats and the TGF-a/EGF receptor is thought to play a role in mediation of liver regeneration following partial hepatectomy. However, culture of hepatocytes led to a decrease in detectable p-glycoprotein indicating that pglycoprotein expression might act as a marker of differentiated hepatocyte function. , 1929, 1989) . Hepatocytes isolated from carcinogen-treated rats immediately after termination of PB-promotion showed a dye coupling of only about 50%. Putative preneoplastic hepatocytes in these cultures were visualized by indirect immunofluorescence using anti-z-glutamyltranspeptidase (r-GT) antibodies. Dye coupling in r-GT-positive hepatocytes was significantly lower amounting to only 20%. Due to their low capacity to communicate, preneoplastic liver cells may escape from the growth control of surrounding normal cells. Enzyme-altered putative preneoplastic liver foci were induced in Wistar rats by application of diethylnitrosamine (10 mg/kg/day) for 5 days, followed by treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD; 100 ng/kg/day) or 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (HCDD; 5/zg/kg/day) for up to 17 weeks. Groups of animals were killed 5, 9, 13, or 17 weeks after start of promoter treatment. For evaluation of DNA synthesis, 5-bromo-2'deoxyuridine (BrdU) was administered 24 hours prior to sacrifice. While TCDD concentrations in livers reached a steady state after 9-13 weeks of treatment, HCDD concentrations increased steadily. Quantitation of adenosine-triphosphatase-deficient liver foci revealed that the number and volume fraction were significantly increased in TCDD and HCDD-treated rats, indicating comparable promoting activity. Using a double-labeling technique, glutathione-transferase (GST)-P-positive liver cells were found to show an approximately 10-fold higher labeling index (LI) than normal cells at all periods of investigation. When examining the time course of DNA synthesis, a time dependent decrease of BrdU-labeling was found in loci of all treatment groups. In TCDD-treated rats, the initially high rate of proliferation appeared to persist for a greater length of time. Assignment of liver loci into 4 size classes revealed that after 5 weeks of promoter treatment foci in the stablest size class had a very high LI, whereas foci in larger size classes had moderate to low Lls. At subsequent time points, Lls of foci in the smallest size class decreased progressively. LIs in the larger size classes varied considerably without obvious correlation between LI, size of the lesions or time point of investigation, indicating heterogeneity in the growth behaviour of liver lesions. Overall, HCDD had only marginal effects on LIs, while TCDD seemed to slightly enhance the LI of liver foci. Adenosine shows a multiplicity of actions which in regard to the systemic circulation are negative chronotropic, negative dromotropic, negative inotropic and vasodilator effects. Metabolically stable, potent and highly selective A 1 and A 2 adenosine receptor agonists and antagonists provide useful pharmacological tools in separating different receptor responses in vivo. The quantitative in vivo characterization of the relationship between pharmaeokinetics and pharmacodyuamics using an appropriate, clinically relevant pharmacodynamic variable then allows evaluation of the possible role of purinurgie drugs for ultimate use in man. The selective adenosine agonist ~-Cyclopentyladenosine (CPA) and antagonist Cyclopentyltheophylline (CPT) have proved to be valuable tools in the characterization of the A1 adenosine receptor subtype in vivo in rats. The pharmacokinetic and pharmacodyaamic interaction between CPA and CPT was quantified by standard pharmacokinetic parameters, a reduction in heart rate and a decrease in mean arterial blood pressure. Four groups of eight rats each received 200 ,ug/kg of CPA intravenously in 5 minutes at a steady-state infusion of CPT at a rate of 0, 57, 114 and 228 ,ug/kg/h, respectively. The blood concentration -heart rate relationships of CPA during the three different CPT steady-state concentrations were derived from each animal. A two-exponential equation was used to quantitatively describe the blood concentration -time profiles of CPA. The pharmacokinetics of CPA did not change in the presence of CPT. Similar pharmacokinetie parameter estimates of CPA were obtained as compared to control administration with mean half-lives of 6.4 + 0.23, 5.9 + 0.42, 7.0 ± 0.20 and 7.1 ± 0.26 rain and mean volumes of distribution of 269 + 16, 264 ± 6, 268 -+ 6 and 314 ± 26 ml/kg with raising CPT steady-state concentration. The concentration -heart rate relationship of CPA was quantitatively characterized fitting the sigmoid Era~x model to individual data. At the steady-state concentrations investigated CPT showed no intrinsic efficacy with respect to heart rate. The mean baseline heart rate measurements E 0 and mean maximal effect values Ema x of CPA were comparable among the four groups. CPT did not influence the shape of the concentration effect profiles of CPA as well. However, the three different CPT steadystate concentrations caused the pharmacodynamic parameter estimate ECs0 of heart rate to increase from 2 ± 0.7, 15 -+ 2.0, 38 -+ 8.6 upto 91 -+ 12 ng/ml with increasing antagonist concentration (0, 14 ± 1, 32 -+ 1, 59 ± 5 ng/ml, respectively). Simultaneous pharmacoklnefic -pharmacodynumic modelling allowed estimation of the pharmacodynamic parameters for CPT as well. The apparent parallel shift in the concentration -heart rate relationships of CPA with increasing CPT eoneentratian indicates a competitive agonist-antagonist interaction at the A 1 adenosine receptor subtype in vivo in rats.

Center for Bit-Pharmaceutical Sciences, University of Leiden, P.O.Box 9503, NL 2300 RA, The Netherlands Glimepiride is a new hypoglycaemic agent of the sulfony~urea group. Its main biotransformation products are the metabolites M1 (hydroxyglimepiride) and M2 (carboxyglimepidde). The aim of the study was to investigate the pharmacokinetics of a 1.5 mg dose of M1 given intravenously. Pharmacodynamics were compared to placebo. Twelve healthy males, mean age 26 years (22 -33) and mean weight 75 kg (69 -85) received 1.5 mg M1 or 0.9 % NaC} in a randomized cross-over design as i.v. injection. On the days before dosing, food and fluid intake were standardized. On the tdal days no food was permitted until 10 h after medication. Concentrations of M1 and M2 in serum and urine (HPLC), glucose, insulin and C-peptide in serum were determined.

The following means + SD were obtained for MI: AUDC = 349 + 110 ng.h/ml, tl/2. z = 1.2 + 0.4 h, CL = 78 + 24 and CL R = 44 +_ 16 ml/min. Udnary recovery was 70 + 7 % of dose (as sum of M1 and M2 after molecular weight correction). ANOVA and confidence intervals were calculated for the profile parameters of the pharmacodynamic variables. M1 reduced glucose by 12 % (p _< 0.05) and increased insulin and C-peptide by 4 % (n.s.) and 7 % (p _< 0.05) respectively. The drug was safe and well tolerated. 

Annette Berndt, K, Richter, R. Oertel and Th. Gramatt6

Tiracizin (Bonnecor ~, AWE) GmbH Dresden) is a newly developed antiarrhythmic drug used predominately in the treatment of ventricular and supraventricular tachycardias. Within the comprehensive pharmacokinetic characterization of tiracizin and three of its metabolites the serum concentration -effect relationships were investigated regarding changes of the ECG of 8 healthy volunteers, Investigations were carried out after single oral administration of 50, 100, 200 mg and during steady state (50 rag, b.i.d.), respectively, Heart rate increased in a dose dependent manner. Average peak rate was measured 3 h after drug administration, correspondir~g to tmax of the main metabolite. Qre remained unchanged. However, there were dose dependent prolongations of PQ and QRS. These effects culminated at tmax of the parent compound (1 h after dosing). For tiracizin concentrations above 80 ng/ml the log concentration-effect curves displayed a linear increase without reaching any plateau within the concentration range observed. For this reason enhanced conduction defects might not be excluded after repeated administration of higher doses. In some volunteers AV block I* and pattern of bandle branch block could be observed temporarily after single administration of 200 mg tiracizin, Taking into account the tiracizin-related extensions of PQ and QRS as well as the extensive hepatic metabolism of this drug (first pass effect 60-70 %), tiracizin is suspected for inducing conductivity disorders after doses mentoined above in patients with decreased hepatic blood flow (i.e. heart failure; liver disease, especially if it is complicated by shunting). x 1005 is an acidic quinoline type inhibitor of 5-1ipexigenase with high oral activity in different species and currendy in phase-II of clinical development. Three issues were investigated: 1. The relative bioavailability of a tablet with 250 mg BAY x 1005 vs. an oral solution as reference (randomized cross-over with 12 male volunteers). 2.The in-fluence of an antacid (Maalox®) on the kinetics of BAY x 1005 (2x250 mg as tablet, randomized cross over with 8 volunteers). 3. The possible effect of the simultaneous intake of an american breakfast on the kinetics (cross-over, 4 volunteers, 2x250 mg ). BAY x 1005 was analyzed by HPLC with heart-cut column switching, separation in a narrow-bore RP-18-cohilnn and UV-detection at 305 nm. The method shows a high precision and accuracy of below 5% (calculated as r.s.d, and deviation from conventional true value, resp.). The limit of quantification was 3 ng/ml. Relative bioavailability of the tablet is about 100% with a distinct higher Cmax and more rapid absorption after administration as solution. The slight antacid interaction (33% increase in Cmax) may be caused by a lowering of the gastric pH and thereby increased solubility of BAY x 1005. The strong food effect on Cmax is interpreted in terms of an increase of the in-rive solubility, too. In all studies BAY x 1005 was well tolerated. Oral low dose aspirin has been used for the prevention of vascular occlusion and reocclusion due to its selective, presystemic effect on platelet cyclooxygenase (COX) in the portal circulation. As this treatment takes 3 -5 days to become effective, patients with unstable angina or on thrombolysis have been given high oral or intravenous loading doses to achieve a rapid onset of platelet inhibition, thus suppressing the favourable vascular effects of prostacyclin and risking higher rates of untoward effects. Therefore we have developed a regimen for a single intravenous low dose of aspirin (LDA, 50 rng, 60 min) and compared its effects on platelet aggregation and prostanoid formation with high dose aspirin (HDA, 500 mg i.v.) and placebo in 10 healthy male volunteers. The infusion rate for LDA was adjusted to previously determined venous blood flow rates in order to avoid systemic inhibitory plasma levels of aspirin in the systemic Circulation by means of dilution in the venous circulation; HDA was infused within the same time period. The urinary metabolites of TXA2 (2,3Ktinor-TXB2), prostacyclin (2,3-dinor-6-keto-PGPI~), and urinary PGE2 were determined by GC-MS/MS. Urinary PGE2 is solely derived from the kidney and was taken as an index of systemic COX activity. Platelet aggregation was similarly inhibited by LDA and HDA by more than 85% at 30 rain and remained suppressed for 24 h. Platelet TXA2 synthesis was maximally inhibited at 60 rain after LDA (-93.8%); it was below detection limit at 10 rain after HDA. Urinary 2,3dinor-TXB2 was equally suppressed by HDA (-83.2%) and LDA (-67.4%; n.s.). The urinary excretion of 2,3-dinor-6-keto:PGFla and of PGE2 was also markedly decreased by 79.2% and 63.5%, respectively, after HDA, whereas LDA suppressed 2,3dinor-6-keto-PGFla excretion significantly less (-30.3%; p<0,02) , and had no inhibitory effect on PGE2 excretion. The ratio of 2,3-dinor-6-keto-PGFla / 2,3-dinor-TXB2 was unaffected by HDA or placebo (from 1.0 to 1.2 and 1.1, respectively), but was markedly increased after LDA (to 2.1). These data show that the clinical pharmacological adjustment of LDA infusion rates may beneficially influence its effects on platelet aggregation and TXA2synthesis while sparing systemic COX activity (as detemined by urinary PGE2). However, a slight inhibition of urinary 2,3-dinor-6-keto-PGFlc~ excretion seems to be an unavoidable consequence of platelet COX inhibition by aspirin, and is probably due to the inhibited transfer of platelet-derived endoperoxides to the endothelium, thus depriving it of one of its sources for prostacyclin synthesis. Dynamic light evoked pupillometry has proved to be a reliable method to evaluate pharmacodynamic effects of CNS active drugs. To investigate a possible correlation between the route of administration and the pharmacodynamic effect of the neuroleptic drug chlorprothixen a fourfold crossover tdal was carried out. 3 different oral galenical formulations of chlorprothixen (film coated tablet, suspension, aequous solution) were applied in dosages of 100 mg each to 8 healthy male volunteers and compared to 100 mg of the i.v. formulation. Pupil diameter was determined using noninvasive dynamic light evoked pupillometry (light stimulus with intensity of 22 cd/qm, duration of 0.1 s). Sedes of 10 repeated single measurements (duration 3 s, interval 30 s) were performed in supine position, before and 1, 2, 3, 4, 5, 6, 8, 10, 24 and 48 h after drug administration. Initial, minimal and final diameter as well as reflex amplitude and time to minimal diameter were recorded and used for further evaluation. Chlorprothixen serum concentrations were determined at the above mentioned time points. Results: 2 x 4 fold ANOVA indicates (p<0.01) that all four formulations can be differentiated on the basis of reflex amplitude, initial, minimal and final diameter, the reflex amplitude (p<0.001) appearing to be the most sensitive indicator. Furthermore, relationship between plasma levels and mathematically transformed reflex amplitude was analyzed. Up to a concentration of ~10pg/L changes the of pupil diameter corresponds to the course of plasma levels. Conclusions: Our results indicate that pupillometry is a suitable method to characterize the different formulations of chlorprothixen and may lead to an estimation of plasma levels within a limited range. The non-selective muscarinic ag nists acetylcholine (ACh) and methacholine {MCh) are known to induce vasodilatafion in the forearm vascular bed. At least part of this response apparently seems to be dependent on the release of nitric oxide (NO). Animal experiments indicate the involvement of the M3-receptor in endotheliumdependent relaxation of vascular smooth muscle. To further assess the vasodilatation induced by ACh and MCh in the forearm of normotensive volunteers (n=10), we infused ACh (dose range 1 -1000 ng/kg/min) in the presence of saline, and MCh (dose range 0.1 -1000 ng/kg/min) in the presence of saline, the non-selecfive muscarinic antagonist atropine (dose range 0.6-6 ng/kg/min) or Ms-receptor selective pirenzepine (dose range 0.8 -800 ng/kg/min), and the gyanylate-cyclase and NO-scavenger methylene blue (1000 ng/kg/min), using automated R-wave triggered venous occlusion plethysmography. By calculating the plasmaconcentrafions of the infused compounds, we were able to obtain ECso-Values and perform 'Schild analysis'. ACh and MCh both caused dose-dependent vasodilatation (ECso-Values of 1.34 5:0.62 ,uM and 64.8 + 13.5 riM, respectively). The dose-response curve (DRC) of MCh was shifted to the right by both atropine and pirenzepine (apparent pA2-values of 7.92 5:0.24 and 6.70 5: 0.04, respectively; slopes not different from unity). Methylene blue did not affect the DRC of MCh. The apparent 20-fold higher potency of MCh compared to ACh, can be explained by the rapid degradation of ACh in the plasma. The pA2-value of pirenzepine is comparable to values reported by in vitro experiments. The apparent pA2-value of atropine is relatively low compared to literature. This can be explained by the fact that atropine is known to require a long time to reach a steadystate equilibrium. Unlike our previous finding that methylene blue potentiates ACh-induced vasodilatation in the forearm, methylene blue did not influence the vasodilatation to MCh. The human forearm is a well established model for resistance vessels. The present technique enables us to perform muscsrinic receptor characterization using different competitive antagonists in the vascular bed of the human forearm. The antiarrhythmic drug propafenone (2'[2 hydroxy-3(propylamino)propoxy]-3phenylpropiophenone) is subject to extensive and variable metabolism in man. Variability is mainly introduced by genetically polymorphic, CYP2D6 mediated 5-hydroxylation, which is absent in 7% of a Caucasian population. This subset of patients (poor metabolisers; PM) has a higher incidence of side effects as compared to extensive metabolisers (EM) and hence knowledge of phenotype would be desirable. Conventional phenotyping during therapy is impossible, since propafenone is a strong competitive inhibitor of metabolism of probe drugs such as sparteine. While EM excret most of propafenone as gincuronide of 5hydroxypropafenone, PM eliminate propafenone by formation of glucuronide of parent compound. Thus determining the excretion of propafenone glucuronide (PPFG) in urine should allow an assignment of phenotype.

We therefore developed a method to phenotype patients during chronic therapy using the urinary excretion of PPFG. PPFG was quantified by HPLC using propafenone glucuronide-O-methyloxime as internal standard. Urinary excretion of PPFG was determined in eleven patients (3 female, 8 male; age 61 5: 8.4) during dosage interval and related to the metabolic ratio (MR) of sparteine (ratio of sparteine and the hydroxy metabolites in the 8 hour urine after administration of 100 mg of sparteine), which had been determined before initiation of therapy. Total excretion of PPFG ranged from 8 to 117 mg (3.6 to 51.5 % of dose) and was correlated to the MR of sparteine (r=0.95, P< 0.001).

In conclusion, urinary excretion of propafenone glucuronides reflects sparteine phenotype and may be useful for phenotype assignment during propafenone therapy. Clinical therapy with prostacyclin or its chemically stable analogs is limited by the dose-dependent occurrence of unwanted drug effects, like hypotension, flush, headache and diarrhea. Therefore, in order to maintain the platelet antiaggregatory effect but to avoid these unwanted drug effects we studied the augmentation of platelet inhibitory effects of iloprost by acetylsalicylic acid (ASA). In a placebo controlled, randomized trial ASA (0, 10, 30, 100 or 300 mg/d) was administered to five groups of eight volunteers each for seven consecutive days. On day eight, exactly one hour after the intake of the last dose of ASA, intravenous infusion of iloprost was initiated with increasing doses of 0.5, 1 and 2 ng/kg/min for one hour each. Blood was withdrawn on day 1, day 4 and day 8 before and after the ASA intake and after 60 minutes of each iloprost infusion dose. Serum thromboxane was dosedependently reduced by ASA for 48, 92, 98 and 99% of control following intake of 10, 30, 100 and 300 mg of ASA. Collagen induced platelet aggregation was reduced by 62% following intake of placebo and infusion of iloprost at 2 ng/kg/min. When the volunteers were on ASA, already 10 mg ASA/day were sufficient to augment the inhibitory potency of iloprost, resulting in a 57% inhibition during infusion of 0.5 ng/kg/min of iloprost. Thus, the platelet inhibitory potency of iloprost was significantly enhanced already by very low doses of ASA, allowing for an effective platelet inhibitor therapy with low iloprost infusion rates with a low rate of unwanted drug effects.

Klinisch-Pharmakologisches Labor, II. Medizinische Klinik, Johannes Gutenberg -Universit~t, 6500 Mainz, Germany

N. Drakoulis, C,R. Gross, I, Roots

As cytochrome P-450 plays an important role in the drug oxidation system, the attention of our group has been focused on the relation between genetic polymorphism and interindividual difference of sesitivity to drugs and carcinogens. Both members of the CYPIA gene family (CYPIAI and CYPIA2) are inducible by polycyclic aromatic hydrocarbons and polychlorihated aromatic hydrocarbons. However the degree of induction is known to be genetically variable. Two point mutations have been recently reported to be responsible for high inducibility of CYPIAI [Nakachi K et al (1991) Cancer Res, 51~ 5177]. Mutation ml resides in the 3'-flanking region following exon 7, and provides a new MspI-site. The other (m2) results the substitution of Ile for Val in residue 462 in the heme binding region of CYPIAI protein, 85 unrelated German healthy volunteers were recruited from students and staff. Mutation ml was detected by digestion to completion with MspI of a 340bp PCR-amplified gene fragment including the polymorphic site. Allele specific PCR was applied to detect the CYPIAI point mutation in exon 7 (m2). As to MspI-polymorphism (ml) 83.5% of the individuals were homozygous wild-type and 16.5% were heterozygous while no homozygous mutated person was found. 91.8% were homozygous carriers of the point mutation within exon 7 (m2). Only 8.2% were heterozygotes for this mutation. No homozygous m2 was detected. All individuals being heterozygous m2 were heterozygous ml whereas 7 persons being heterozygous ml were not heterozygous m2 indicating close but not absolute association. Allele frequency of ml in the German population screened correlates with data published for a Norwegian population. In the Japanese population both ml and m2 alleles are reported to be more frequent than among Europeans [Tefre T et al (1991) Pyridostigmine (PYR), a quaternary ammonium salt, is poorly absorbed from the gastrointestinal tract and eliminated mainly by renal excretion. Renal clearance values measured previously in a few subjects had exceeded those of creatinine pointing to a role for tubular secretion. Since chronic administration of PYR for symptomatic treatment of myasthenia gravis often necessitates comedication with acid secretory inhibitors like ranitidine and pirenzepine, a kinetic interaction seemed perceivable due to competition of these drugs with PYR for the renal cation carrier. Renal clearances of PYR (CLp) and creatinine (CLio) were measured in 6 volunteers and "19 myasthenic patient~"without and with ranitidine or pirenzepine coadministration. CL# amounted to 350 ± 90 % of CLcR, and both clearances decreased with increasing age. Based on the assumption that the glomerular filtration rate is CLrQ/I.15, the contribution of tubular secretion to the PYRV~learance was calculated to be 75 %. Ranitidine (150 mg every 8 h) and pirenzepine (50 mg every 12 h) reduced CLp by 8 % (p < 0.05 by one-tailed paired t test) and 19 % (p ~ 0.002), respectively, in patients and volunteers combined and decreased the clearance by secretion to a similar extent. PYR plasma concentrations were, however, not affected by the comedication. An influence of the gastric protectants on total PYR excretion could not be detected either. The metabolism of many antiarrhythmic drugs, especially of the IC subgroup, is markedly influenced by the sparteine/debrisoquine polymorphism as their oxidative biodegradation is catalysed by the polymorphic cytochrome P-450 206 (CYP2D6). Aprindine (AP), a group IC antiarrythmic drug, is extensively metabolised in humans. Beside Ndealkylation, aromatic hydroxylation is the major metabolic pathway and two hydroxy metabolites are excreted in urine. In vitro experiments using human liver microsomes were performed in order to investigate whether AP metabolism is catalysed by CYP2D6. The formation of the major hydroxy metabolites of AP followed Michaelis-Menten enzyme kinetics in two microsome preparations of livers of extensive metabolisers of sparteine. In marked contrast, both metabolites were not detected in incubations with liver microsomes of a poor metaboliser, AP N-dealkylation, however, was observed in all microsome samples suggesting that this minor metabolic conversion of AP is mediated by another eytochrome P-450 isoenzyme. AP is also a potent competitive inhibitor of sparteine and propafenone metabolism which is catalysed by CYP2D6. In addition, LKM-1 autoantibody positive serum which selectively immunoprecipitates CYP2D6 strongly inhibits the formation of AP hydroxy metabolites. Both hydroxy metabolites of AP were absent in the urine of a poor metaboliser after oral intake of 50 mg of AP hydrochloride. Therefore, hydroxylation of AP which represents its major metabolic pathway is catalysed by CYP2D6. As a consequence of this finding, poor metabolisers should exhibit manifest alterations of pharmacokinetic parameters. Severe side effects found in some patients under AP treatment might be a result of the impaired metabolism in poor metabolisers and its narrow therapeutic index. By applying the so-called Bayesian approach pharmaeoldnetic dsta can be generated during routinely performed TDM for different patient populations. In a prospective study we have tested in each of 12 hospitalized patients treated either with digoxin, theophyllLne or gentamicin whether such data based only on one trough steady state plasma level (and on one peak level in the case of gentamicin) are reliable and comparable to pharmacokinetic calculations (model-free and compartmental analyses) based on 8 samples drawn during a dosing interval of 8 h (gentamicin 0.5 h infusion), 12 h (oral slow release theophylline) or 24h (0.2 rug acetyldigoxirt/die). The results (mean ± SD) are summarized in the Antipyrine ® (INN: Phenazone) is used as a probe for the oxidative drug metabolism in humans. Its predictive value is limited due to the fact that the P450 enzymes involved in the formation of norantipyrine, 4-hydroxyantipyrine and 3-hydroxymethylantipyrine have not been identified so far. Methods: A highly sensitive and selective coupled gas chromatographic/ tandem mass spectrometric method, based on stable isotope labelling of the 3 major antipyrine metabolites as internal standards was developed. Human livers which had been characterized by means of immuno blotting for their contents of CYP1A2, CYP2C, CYP2D6, CYP2E1 and CYP3A were used for microsomal incubations. The Michaelis-Menten kinetics for the formation of the individual metabolites were studied in the microsomal fraction of human livers. Results: "]7he maximum rate of metabofite formation (Vmax) of norantipyrine, 4-hydroxyantipyrine and 3-hydroxymethylantipyrine in different livers ranged from 0.05 to 0.583, from 0.634 to 1.249 and from 0.216 to 0.495 pmol.l~g-l.min "1 respectively. The velocities of formation of 4-hydroxyantipyrine and 3-hydroxymethylantipyrine were correlated to the contents of CYP1A2 and CYP3A. Conclusions: Evidence for the involvement of CYP1A2 and CYP3A in the formation of 4-hydroxyantipyrine and 3-hydroxymethylantipyrine ~vas found. The identification of the enzymes involved in antipyrine metabolism will allow a more rational use of antipyrine as a probe drug for human drug metabolism. During recent years an increasing amount of evidence suggests a role of leukotrienes in psoriasis (P) and atopic dermatitis (AD). LTB4 and cysteinyl leukotrienes have been demonstrated in skin punctures from patients with P. In addition, LTB4 has also been demonstrated in skin from patients with AD using suction blister sampling technique. However, these techniques are invasive and an artifactual synthesis can not be excluded. Pharmacokinetic studies in man have shown that LTC4 is rapidly metabolized to LTE4 which in part is directly eliminated into urine. We have measured urinary LTE4 to assess cysteinyl leukotdene synthesis in these patients. LTE4 was isolated from urine by solid phase extraction followed by RP-HPLC. We conclude that bimakalim induced a substantial vasodilation associated with a rise in heart rate and enhanced cardiac performance. Bimakalim furthermore increased plasma renin activity accompanied by a mild antidiuratic effect. The aim of the studies was to investigate the safety, tolerability and pharmacokinetics of rivastatin (BAY w 6228) which is a novel and enantiomerically pure HMG-CoA reductase inhibitor. Animal studies have shown that rivastatin is about 100 fold more potent than lovastatin. Therefore a starting dose of 20 #g per subject was chosen.

In ascending single dose studies 106 healthy male volunteers were studied in a dose range from 20 #g to 400 #g per subject. Rivastatin or placebo was administered as a solution (20-50 #g) or as an immediate release tablet. The design was placebo-controlled crossover or parallel. Multiple parameters were measured to prove safety and tolerability including physical examination, special eye inspection, adverse events, hemodynamics including ECG, heart rate and blood pressure, hematology, clinical chemisay and urinalysis. Clinical chemistry inc/uded standard parameters for liver, pancreas, kidney and muscels as well as LDH and CK isoenzymes. Urinalysis included standard parameters as well as renal marker enzymes.

Results: There was no difference in pre-and after study examination of body and eyes. ECG, heart rate and blood pressure showed no relevant clinical differences between verum and placebo. Adverse events in the active drug groups were all mild and evenly distributed between active and placebo. Transient, minor changes in clinical chemistry tests including urinalysis were seen generally evenly distributed between verum and placebo at all dose levels. Therefore it was concluded that rivastatin given as a single dose up to 400/~g was well tolerated and no safety concerns were raised. Although the liver is the major site of CYP3A mediated metabolism extrahepatic expression of this enzyme family may result in degradation and/or bioactivation of drugs in other organs or tissues. We therefore developed an immuno-histochemical method (Avidin-Biotin-Complex) using both polyclonal and monoclonal anti CYP3A and applied it to 158 routinely fixed and paraffin embedded human tissues, which were filed in the department of Pathology. The different antibodies used showed identical results. In addition to liver CYP3A was detected in various extrahepatic tissues (such as stomach, duodenum, pancreas). High abundance was observed in the gastrointestinal tract with staining intensity decreasing from stomach to rectum. In summary the method is suitable to detect and localize the expression of CYP3A within an organ. Quantitative immunohistochemistry may allow for an assessment of relative abundance in various organs. The method will be applied to other cytochrome P450 enzymes as well. Bay u 3405 [(3R)-3-(4-Fluorophenylsulfonamido)-1,2,3,4-tetrahydro-9-carbazolepropanoic acid] is a thromboxane-A2-receptor antagonist synthesized by Bayer AG, Wuppertal. In a placebocontrolled phase-I study involving 14 young and 14 elderly healthy male volunteers the anti-aggregatory efficacy of repeated doses of 20 mg q.l.d, over 12 consecutive days were investigated, It was found that Bay u 3405 left the bleeding time largely unaffected but proved to be highly potent to reverse the platelet aggregation induced in vitro by the stable thomboxane-agonist U 46619; the action lasted longer under steady state than under single dose conditions and in the elderly than in the young subjects.

The pharmacodynamic effect was quantified as the area under the curve representing the relation between platelet aggregation (in percent) and agonist concentration (the larger the area the lower the anti-aggregatory effect); for each blood saraple taken after medication this relation was established using five individually selected concentrations of U 46619 between 100 and 1000 ng/ml. By allocating these areas to the pertinent plasma concentrations of Bay u 3405 and plotting them jointly the relaUon between pharmacodynan~c effect and antagonist concentration was demonstrated independent of the agonist concentration. Moreover, by means of a curve fit according to the least squares principle a mathematical description of the relation was obtained. The metabolism of dihydropyridine calcium channel blockers and of caffeine in volunteers is inhibited by grapefruit juice. Naringenin, the aglycone of the juice component nadngin, was shown to inhibit the cytochrome P450 isoforms responsible for the metabolism of these drugs, i.e. CYP3A and CYP1A2, respectively. Since verapamil metabolism also depends on these isoforms, we examined the effect of naringenin and nadngin on the different pathways of verapamil metabolism in human liver microsomes. Microsomes of two donors (R1 and R6) were prepared by standard methods. Incubations and measurement of verapamil metabolites were as repoded [Kroemer et al. (1992) To study a pharmacokinetic hUeraction regarding the resorption of lansoprazole enteric coated capsules (L) after co-administration of magaldrate (M), an open cross-over trial in 12 healthy male volunteers was conducted. They received either 30 mg L only, or 30 mg L together with 2400 mg M (Riopan®). The wash-out period was 7 days between the two treatments. L was analysed by an HPLC method with a limit of detection of 5 ng/ml. The extent of absorption, exemplified by AUC~,~, was not significantly different between the two treatments: 1369.84 (745.15) ng x ndax h, (mean (SD)) after L only, The pharmacokinetics of roxatidine acetate (rex) in healthy subjects are well described. In patients with renal insufficiency (CLc/~ < 20 ml/min) a reduction of the dose from 150 rag/24 h to 75 rag/48 h is recommended. However, pharmacodynamic data are lacking to address the clinical problem whether during the second night acid secretion is sufficiently suppressed for treatment of peptic ulcer disease in renal failure. 6 and 8 patients (mean age 51, range 22-63 years; duration of renal failure 97, 20-240 mouths; CLcR 13, 6-18 ml/min) received 75 mg rex orally every 48 h (group I) or every 24 h (group II) at 22.30. Using ambulatory intragastric long-term pH-metry pH values were monitored over 24 and 48 h without and after rex treatment for 12 days. Gastrin and rex plasma levels were measured by RIA and GLC respectively. Animal studies suggest that the peptides growth hormone-releasing hormone (GHRH) and somatostatin (SRIF), besides their opposite regulatory influence on growth hormone (GH) release, exert distinct effects on sleep electroencephalogram. When applied centrally m animals, growth hormone-releasing hormone (GHRH) stimulates slow wave sleep (SWS), whereas somatostatin (SRIF) increases REM sleep. We investigated whether these peptides also affect the sleep EEG in humans when given intravenously. In addition nocturnal cortisol and GH release were examined. Seven male normal controls (20 -30 years) entered three protocols with iv administration of four boluses of (1) placebo, (2) 50 #g GHRH and (3) 50 #g SRIF at 22.00, 23.00, 24.00 and 1.00 hours according to a randomized schedule, Sleep EEG was recorded from 23.00 to 7.00. In addition, we collected a blood sample through a long catheter every 20 minutes from 22.00 to 7,00 and measured plasma cortisol and growth hormone (GH) levels. After GHRH administration a significant increase in SWS was found, but no change after SRIF (percent spent in SWS per night: 14.0 + 5.6 SD after placebo, 20.2 + 6.6 after GHRH and 15.1 + 8.2 after SRIF). Application of SRIF was accompanied by a trend toward increased REM density (2.5 + 0.7) compared to GHRH (2 + 0.8) but not by other changes of sleep EEG variables. In comparison to SRIF and placebo, GHRH predueed a significant increase in plasma GH concentration throughout the night (mean + SD: 10.8 + 2 ng/ml after GHRH; 3.0 + 1.7 ng/ml after SRIF and 3.2 + 2.0 ng/mi after placebo). Nocturnal cortisol secretion was signiftcantly blunted after GHRH but remained unaffected by SRIF (61.4 + 2.9 ng/ml after placebo, 46.6 + 19.7 ng/mi after GHRH and 70.8 + 12.6 after SRIF). The effects of episodic GHRI-[ administration upon SWS, GH and cortisol secretion were opposite to those previously reported for corticotropin-relsasing hormone (CRH) [Holsboer et eL, Neuroendocrinology 1988; 48: 32-38] , which supports the view that neuroregulation of human sleep involves an interaction of central GHRH and CRI-I. Supported by a grant of the "Deutsche Forsehungsgemeinsehaft" (Ste 486/1-1). The adrenergic system and the renin-angiotensin-aldosteron system are important factors for the genesis of hypertension. Since several years the Na/K-ATPase system (NKA) and the so-called 'digitalis-like factor' (DLF) are known as factors which also induce hypertension. We investigated the changes in the NKA activity of erythrocyte ghosts and in the plasma level of DLF in 60 patients suffering from hypertension (stage I and II WHO) as compared with that of 30 healthy subjects before and 4 weeks after treatment with nifedipine (30 mg), the 13-receptor-blocker ddazolol (40 mg) or with the ACE inhibitor captopril (75 mg). in untreated hypertensives the mean NKA activity (0.28 vs. 0.45 pmol Pi/mg protein/h) was depressed and the DLF plasma level enhanced (0.188 vs. 0.105 ng-Equ digoxin/ml) as compared with normotensive controls. After a 4 week treatment with nifedipine the NKA activity was enhanced (0.40 vs. 0.32 IJmol Pi/mg/h) and the DLF plasma level was lowered (0.111 vs. 0.142 ng-Equ/ml). In the cases of ridazolol and captopdl the NKA activity was not restored (0.31 and 0.33 vs. 0.35 and 0.30 IJmOl Pi/mg/h; 0.159 and 0.192 vs. 0.132 and 0.211 ng-Equ/ml, resp.). The three antihypertensives lowered the systolic and diastolic blood pressure significantly after their intake over 4 weeks. The data of nifedipine cannot be interpreted as a reaction of the calcium channel blocker on the NKA activcity. First results indicate that therapeutic relevant concentrations of the calcium channel blockers verapamil and diltiazem had no inhibitory activity on the NKA, but they inhibited the activity of the Ca2+-sensitive ATPase of erythrocyte ghosts to 15 to 30 per cent. Nifedipine was not included in the enzymatic studies because of its light-sensitivity. H2-receptor antagonists have been shown to differ in their cardiovascular effects (1, 2). As e.g. for nizatidine cholinergic side effects have been reported, question arose how these differences in the haemodynamic effects of the antagonists could be explained. Therefore in a randomized placebe-controlled study 10 healthy subjects were treated after randomized ailecation with single intravenous doses of ranitidine 50 mg, famotidine 20 mg and an oral dose of nizatidine 300 mg (not i.v. available) with and without autonomic blockade (propranolol 0.2 mg/kg b. w. plus atropine 0.04 mg/kg b.w. (3) both intravenously). Additionally systolic time intervals were determined. The cardiovascular effects of ranitidine seen with and without autonomic blockade did not differ from those observed with placebo. Famotidine led to a decrease in stroke volume and cardiac output accompanied by a rise in systolic and diastolic blood pressure, both with and without autonomic blockade. The changes caused by famotidine differed significantly from those on placebo. The more recent H2-antagonist nizatidine lowered heart rate (p<O.05 with and without autonomic blockade). In conclusion, results of the present study indicate that haemodynamic effects of nizatidine and famotidine are mediated rather by histaminergic than by cholinergic or ~-adrenergic mechanisms. , range 4? to 77 yrs.} with pain due to different kinds of metastases were treated for 2 days with D (4 times 50 mg per day). A randomized double-blind sequential trial (Bross 1952) was followed in order to compare D with SR (2 times 100 mg plus 2 times placebo per day). Pain was recorded 4 times daily for two days on a 100 mm visual analogue scale (VAS) and on a 5 step verbal rating scale (O=none, 1=slight, 2=moderate, 3=strong, 4=worst pain). Mean pain scores at study entrance in the D group were VAS= 52.8 and VRS=2.5?, and VAS=62.8 and VRS=2.57 in the SR group. During the sequential trial the mean pain scores were much better in the D group (VAS=23.5, VRS=1.11) compared with the SR group (VAS=42.B, VRS= 1.89). The higher analgesic efficacy of D vs. SR was statistically confirmed (2a=0.10, 1~=0.05) after 7 treatment pairs. The treatment in the D group was rated better in the patients' final global evaluation.

We conclude that in spite of the probability of lower compliance plain dictofenac is preferable to SR diclofenac in cancer pain. In recent years, paramedical methods have been promoted by incompetent politicians and have gained public acceptance with support from newspapers and television and their correspondingly incompetent writers and television managers. However, the unrestrained recommendation of methods like homeopathy even by medical doctors has to be taken seriously into consideration. This fact signals that a great number of medical students do not accept the more or less "theoretical" ballast in the field of general pharmacology, For this reason the students direct their attention towards the more practically-oriented paramedical methods. However, in the course of their education, they have missed the background which is necessary to interpret the beneficial paramedical healing-effects that are occasionally observed.

We are well aware that homeopathic products will never meet the high standards worked out by critical pharmacologists for acceptance of data providing drug efficacy. On the other hand, we can never agree with those pharmacologists, who attribute the beneficial "healing effects" of paramedical treatment to placebo alone. It is our aim to discuss the dose-response-relationships of effective drugs compared to placebo, and the somewhat different results of homeopathic treatment. Bay x1005 is a quinoline substituted arylcarboxylic acid, which has demonstrated its ability to inhibit the leukotdene synthesis in animal models as well as in in vitro studies. On the basis of pharmacological properties Bay x1005 provides advantages in the treatment of inflammatory or allergic diseases. Phase I studies for the investigation of tolerability and pharmacokinetics have been performed in healthy male volunteers with an increasing single oral dose regimen from 50 to 500 mg and with multiple doses up to 500 mg bid over 5 days. A study to evaluate the relative bioavailibility has also been performed. 121 volunteers have been included in 12 partially placebo controlled double blind studies. Tolerability: Bay x1005 was subjectively and objectively well tolerated in each dose step after single and multiple dosing. No major adverse events occured, laboratory findings and vital signs were not affected.

Dose ( The antiarrhythmic propafenone (PPF) is marketed as a mcemate. While the class 1C antiarrhythmic activity is mediated by both enantiomers, 13blockade resides in the S-PPF. In vitro lines of evidence indicate metabolism of S-PPF to be inhibited in presence of the R-enantiomer. Such a metabolic enantiomer/enantiomer interaction should lead to increased concentrations of S-PPF and hence more pronounced 13blockade after administration of racemic PPF as compared to separate administration of the enantiomers. We tested this hypothesis in a placebo-controlled, single blinded, randomized study. Medication consisted of racemic PPF, S-PPF and R-PPF (150 mg q 6 hours for 4 days to 7 male healthy volunteers). Maximum exercise heart rate (HRmax) and QRS intervals were evaluated after 75 hours of treatment and compared to pretreatment data (* P < 0.05).

Racemic PPF 158_+10 148-+11 * 100+-9 109-+10" S-PPF 158-+11 153-+10 94+-10 107-+11" R-PPF 157_+11 155_+11 96_+11 108+-14" Placebo 156_+6 157-+11 97+-9 96_+8 HRma x was significantly reduced only after administration of racemic PPF, whereas S-PPF, R-PPF and racemic PPF significantly prolonged QRS. These data are in agreement with the hypothesis of a metabolic interaction of S-and R-propafenone leading to higher concentrations of S-PPF after administration of racemic propafenone which in turn provokes enhanced 13-blockade. Thiazide diuretics are one of the basic principles of antihypertensive therapy. Yet, in spite of their importance especially in the treatment of elderly patients, only limited reformation is available about their pharmacokinetic and pharmacodynamic properties with regard to old age. This is also true for the frequently used thiazide diuretic bemetizide. Therefore the aims of this study were to obtain data about potential alterations of the pharmacokinetics of bemetizide in old age, and to evaluate, whether there is a need for a dose reduction. Ten healthy young volunteers of both gender, as well as ten male and ten female elderly volunteers (__~: 75 years) were given one tablet containing 25 mg of bemetizide and 50 mg of triamterene. The tablet was administered on day 1, and days 3-8, at 9.00 a.m., 45 minutes after a standardized breakfast. Plasma and urine concentrations as well as volume and electrolyte excretion were measured up to 36 hours after dosage. A stereospecific HPLC method was developed that included extraction with ethylacetate and separation on a C-8 reversedd-phase column with methanol and phosphate buffer (5:6, by vol).

In contrast to earlier investigations, in the course of which concentrations had only been measured up to 12 hours after administration [1], a longer terminal half life and higher maximum plasma concentrations were found. This holds true for the elderly as well as for young volunteers, especially on day eight of the study, when steady-state conditions were achieved. The trough concentrations on da), 8 were significantly higher in the elderly than in the young subjects.

The available data suggest that some accumulation of bemetizide occurs upon repetitive dosage.

[ -cyclopropyl-7-(2,8-diazabicyclo[4.3 .0]nonane)-6-fluoro-8chloro-l,4-dihydro-4-oxo-3-quinolinecarboxylicacid) is a novel 4-quinolone with outstanding broad-spectrum antimicrobial activity. It is active against most gram-negative and gram-positive bacteria in low concentrations 0zg/L). The drug might be a substantial contribution to the progress in antiinfective therapy. We investigated the safety, tolerability and pharmacokinetics of BAY y 3118 after oral single (50-300 mg, n=42) and multiple dosing ( In recent years inhibitors of cholinesterase have been tried with some success in senile dementia of Alzheimer's type. Galanthamine, an alkaloid isolated from snow drops is a 100 times better inhibitor of acetylcholinesterase in brain and erythrocytes as of butyrylcholinesterase in plasma.

The effect of such compounds on the human EEG-power spectrum may serve as indicator of the therapeutic efficacy and also to differentiate responders from nonresponders. Previous work in our group has shown that Galanthamine does increase electric power in the alpha band of the EEG as shown with conventional EEG recording. In the present study we are able to record the power-spectrum under 17 electrodes in 6 different frequency bands continuously by the electrodiagnostic system CATEEM and to relate the increase of power and the effects on peak and mean frequency to the Galanthamine-level as well as to the inhibition or actual activity of acetyicholinesterase measured in erythrocytes during and after a 15 rain. infusion of 0,2mg/kg Galanthamine in comparison to electrolyte solution without the alkaloid. It is the purpose of this study to elaborate concentration-effect-relations with regard to the EEG-power spectrum and also the relation between inhibition of acetylcholinesterase and changes in prwer and frequency in the resting EEG. Results are given in absolute figurs and also in relation to those during the 15 minutes before the infusions of Galanthamine or placebo were started, single blind and in a randomised order. Power in the alphal-or alpha2-frequency band increased gradually with a lag time of 7 to 15rain up to 200%. Changes of the peak and mean frequencies in the alpha-band were on the same time schedule. No changes at other frequency bands did occur . The changes described were most striking in the temporal, parietal and occipital regions. The intention of GCP in clinical pharmacological research is to improve quality of experimentation, data sampling, and data organization, biometric analysis and reporting. Normally quality assurance represents a control process in parallel to the study for detecting and correcting faults. This strategy of quality assurance we define a passive one.

An active quality assurance based on a computer control system avoids faults just at the point of investigation, that is: a PC based control system organizes in real time all important processes of measurement, case reporting, data sampling, data organization and biometric analysis. This paper presents and illustrates a system for active quality assurance, based on IBIS, which was developed by Forschungsgruppe IBIS, Bochum. Up to now IBIS is also used in clinical pharmacological studies for psychometric investigations. It represents a user programming tool like 4th generation programming languages do. The investigative process can be modelled with this software tool. Peripheral devices like systems for automatic measurement of ECG, blood pressure, biochemical analysis etc. may be controlled as well as the timing of investigators activities.

Prof. Dr. Friedhelm Meier, Forschungsgruppe IBIS, H~Isbergstrat~e 77a, D-4630 Bochum, Germany

Arylamine N-acetyltransferase (NAT2) exhibits an inherited polymorphism that might influence the response of patients to arylamine and hydrazine drugs and other chemicals. Previous studies documented the clinical importance of individual patient-genotype for appropriate medication of drugs to be acetylated.

In the European population phenotypical slow and fast acetylators distribute about equally. Slow acetylators prevail among bladder, gastric, or other cancer patients presumably due to inefficient detoxication of arylamine carcinogens via N-acetylation. Lately, multiple point mutations were found revealing distinct NAT2 variants. This allowed the development of specific genotyping procedures in humans that can be now offered to the clinicians [Grant DM et al (1992) Xenobiotica 22: 1073]. 302 healthy German volunteers were studied. Allele specific nested PeR was applied to detect mutations causing slow acetylation. To avoid false amplification of the highly homologous "monomorphic" N-acetyltransferase gene or the arylamine N-acetyltransferase pseudogene, patients DNA was first amplified with the NAT2 specific primers P55 and P56 and then reamplified using P56 as common downstream primer and mutation specific upstream primer for each of the six point mutations (C282~T, T341~C, C481~T, G590~A, AS03~G and G857~A) in the coding region of the NAT2 gene responsible for impaired acetylation. Mutation C282~T results loss of FokI-site, C481~T loss of KpnI-site, G590~A loss of TaqI-site, A803~G gain of DdeI-site and G857~A loss of BamHI-siteo T341~C is the most common mutation in the German population (44.0%) followed by G590~A (28.6%). Wild-type alleles represent 26.2% and the G857~A mutation is very rare in Germany. Frequencies of mutations C282~T, C481~T, and A803~G have not been yet calculated. Slow acetylator genotypes may result by single or combination of above mutations. Calcium antagonists are known to increase plasma levels of digoxin dose dependently, thereby leading to an alteration of the cardiovascular effects of this combination therapy (1). Re 40-5967 ((1S, 2S)-2-(2-)(9-(benzimidazolyl)propyl)-6-fluoro-l,2,3,4,-tetrahydro-l-isopropyl-2-naphthylmethoxyacetate dihydrochloride)

is a new verapamil-like calcium antagonist. 42 healthy male subjects (mean age 26 + 0.6 yrs., body weight 77 + 1.4 kg, height 181 + 1.2 cm; X ~ SEM) were investigated in-a crossover study. F~llowing an oral loading dose of digoxin 0.375 mg thrice daily, subjects were simultaneously treated with digoxin 0.375 mg once daily for 7 days. Afterwards subjects were given digoxin 0.375 mg once daily plus Re 40-5967 50 mg once daily (group 1, 14 subjects), Re 40-5967 100 mg once daily (group 2, 14 subjects) and Re 40-5967 150 mg once daily (group 3, 14 subjects). Mean maximum plasma concentrations (Cmax) after one week's treatment with digoxin alone were 2.1 + 0.15 ng/ml (X ~ SEM). Administration of 50 mg Ro 40-5967--1ed to an increase of digoxin Cmax to 2.6 + 0.17 ng/ml, of 100 mg Ro 40-5967 to 3.0 ~ 0.12 ng/ml and ~f 150 mg Re 40-5967 to 3.0 + 0..22 ng/ml. These kinetic alterations were associated with--changes of the PQ time in ECG. PQ time was 157 + 5.3 msec (baseline), 166 + 5.6 msec at Cmax of digoxin atone, 168 ~ 3.9 msec at Cma~ of digoxin plus 50 mg, 179 + 4.2 msec at Cmax of digoxin plus 100 mg and 190 + 7.6 msec--at Cmax of digoxin plus 150 mg Re 40-5967. In conclusion, the verapamil-like calcium antagonist Re 40-5967 led to a dose dependent increase of digoxin plasma concentrations, whereby a "plateau effect" was observed. Nitrendipine is a 1,4-dihydropyridine Caa+-antagonist which is clinically used as a racemate. The S-enantiomer is at least 10 times more potent than R-nitrendipine with respect to its Ca2+-antagonistic properties ~. The first-pass metabolism of nitrendipine is stereoselective with the R-enantiomer being preferentially metabolized. Since liver cirrhosis can affect the stereoselective first-pass metabolism of high clearance drugs we have investigated the stereoselective disposition and action of nitrendipine in six patients with liver cirrhosis of Child B and in six age matched healthy controls which received a commercially available tablet (20 mg nitrendipine). Arterial blood pressure, leg blood flow and biochemical parameters were measured. Nitrendipine enantiomers were quantified using a stereospecific method. Liver cirrhosis had a profound impact on the phaymacoldnetics and stereoselective disposition of nitrendipine, see Table. The stereoselective first pass metabolism was less pronounced in patients with liver cirrhosis with a SIR ratio of 1.7 (referred to AUC) compared to healthy volunteers (S/Rratio of 2.4). The pharmacodynamic effects were closely related to the S-concentration. The more pronounced effects in liver cirrhosis are due to the much higher S-concentrations. In addition the free fraction fu was increased by about 30% compared to the healthy controls. Both codeine and morphine seem to occur endogenously in animals and in man. The metabolic pathway for the formation of "endogenous" codeine and morphine in mammals is considered to be the same as in the poppy plant. The O-demethylation of exogenous codeine to morphine is catalysed by CYP2D6 which exhibits a genetic polymorphism. Subjects who do not express this enzyme (PM's) form only trace amounts of morphine after administration of codeine whereas in extensive metabolisers (EM's) upto 9% are biotransformed to morphine. It is possible that also endogenous morphine formation involves CYP2D6. Therefore, we have studied "endogenous" codeine and morphine formation in relation to the phenotype for the sparteine / debrisoquine polymorphism. In 20 age, weight and sex-matched EM and 20 PM subjects the 24 hour urinary excretions of total morphine and codeine were quantified by solidphase extraction, HPLC separation and RIA. No differences were observed between the phenotypes in the excretion of codeine (EM: 1.0 +_ 0.4 nmol, PM: 0.59 _+ 0.32 nmol) and morphine (EM: 2.4 _+ 1.1 nmol, PM: 2.1 +-. t.8 nmol). The excretion of codeine and morphine covaried (r=0.83) and the ratio of codeine to morphine in urine was unimodally distributed. In addition, 4 EM's collected urine on 3 consecutive days and no differences of codeine and morphine excretion were found between the days. Furthermore, these subjects took a single oral dose of quinidine (50rag) which inhibits metabolism via CYP2D6. However, quinidine did not alter endogenous codeine and morphine excretion for upto three days. These results indicate that the formation of endogenous morphine is not mediated by the polymorphic CYP2D6. Med. 42 (1987) , 1369-1372 I Also, we had described in an earlier report that the performance of normal subjects in the K-L-T was only negligibly diminished by 4 mg phenytoin/kg. Differentiated blood level-oriented consideration revealed that the performance was impaired only in subjects with phenytoin levels of less than 2 ~g/ml serum. No differences from the findings obtained with placebo were seen in case of higher blood levels.

Again, this may be interpreted as proposed by DOker. An bei der Erprobung antihypertensiv wirksamer Substanzen eingesetzte BlutdruckmeSsysteme werden hohe Anforderungen hinsichtlich Richtigkeit und Genauigkeit gestellt. Besonders bei multizentrischen Studien wird angestrebt, den Einflu8 des jeweiligen Untersuchers im Me0vorgang zu minimieren und ManipulationsmSglichkeiten auszuschalten. Weitere Forderungen sind leichte Handhabung, Robustheit und Preisw0rdigkeit. Derzeit ist kein allen diesen Forderungen in optimaler Weise genQgendes Ger&t ver-f0gbar. Internationaler und von ZulassungsbehSrden akzeptierter Standard ist auch heute noch das Prinzip des Quecksilber-Sphygmomanometers; zunehmend werden aber auch (halb)automatische Ger&te eingesetzt. In zwei hinsichtlich Versuchsdesign und Datenstruktur sehr &hnlichen Dosisfindungsstudien verwendeten wir zwei MeSsysteme und verglichen die hiermit erhaltenen Blutdruckdaten miteinander. Bei Studie 1 handelte es sich um eine Multicenterstudie, die in 14 Zentren mit insgesamt 110 Patienten durchgef0hrt wurde und in der in 939 MeSvorg&ngen (3 Wiederholungsmessungen zu Zeitpunkten t~, i=1,2,3, im Intervall [t,,t,+~,] mit "~, = 4 min und t2=tl+-~,/2 ) der diastol. Blutdruck DBPi=DBP(t~) mit einem automat. System (Automatic BOSO-Oscillomat) bestimmt wurde. Studie 2 haben wir monozentrisch unter Verwendung eines Quecksilber-Sphygmomanometers durchgef0hrt und dabei in 5142 MeSvorg&ngen (dito, mit "% = 5 min) Werte DBP~ fOr den DBP erhalten. Zun&chst wurde die (absolute) Schwankung r = MAX{ABS(DBPcDBP ) l i,j= 1,2,3} for jedes dieser Tripel bestimmt und die sich so for beide MeSmethoden ergebenden empirischen H&ufigkeitsverteilungen miteinander verglichen. Es zeigte sich eine gerirlgere (p < 0,01%) Schwankung in Studie 2; zum Vergleich (jeweils entsprechende Kenngr68en for die Verteilung von r aus Studie 1 I Studie 2) -Median: 7 I 5, unteres Over the last decade, cyclosporine has become well established in the therapeutic management of organ transplant patients. More recently, it has been introduced in the treatment of a variety of autoimmune disorders including rheumatoid arthritis. Its use in rheumatoid arthritis patients results in its coadministration with nonsteroidal anti-inflammatory drugs (NSAID's). To exclude the possibility of a pharmacokinetic interaction between cyclosporine and commonly used NSAID's a controlled investigation was performed in healthy volunteers. Acetylsalicylic acid was chosen because of its widespread use as a therapeutic adjuvant in the treatment of rheumatoid arthritis. A randomized, two-period crossover investigation was performed in 24 healthy volunteers in which a single oral dose of 300 mg cyclosporine was administered alone and on day 10 of multiple oral t.i.d, dosing of 960 mg acetylsalicylic acid. Serial blood samples were obtained over 48 hours after each cyclosporine dose and over a steady-state dosing interval for acetylsalicylic acid on day 9 (acetylsalicylic acid alone) and day 10 (coadministration of cyclosporine and acetylsalicylic acid). Cyclosporine whole blood concentrations were determined by a specific RIA method and plasma concentrations of acetylsalicylic acid and metabolites by HPLC. Lack of a pharmacokinetic interaction Was conclusively demonstrated for the rate and extent of cyclosporine and acetylsalicylic acid absorption and for the rate and extent of salicylic acid formation after a single dose of cyclosporine was coadministered during steady-state acetylsalicylic acid dosing. Based on these results, no dosage adjustment is necessary when cyclosporine and acetylsalicylic acid are coadministered. The effect of nifedipine [NIF] on hepatic blood flow [EHBF] was investigated in 11 healey male volunteers. Subjects received a 10mg capsule NIF (Cordicant ~ Kapsel) either at 08.00h or 20.00h. HBF was estimated from indocyanine green [ICG] kinetics 45 rain before (control) and 25 rain after NIF. Blood samples were taken prior and 2, 5~R!0 , and 15 min following a rapid i.v. injection of ICG (Cardio Green ). ICG concentrations were measured by spectrophotometry (805nm). Serum concentrations of NIF and its main metabolite were determined from blood samples collected before, 10, 15, 20, 30, 45 min and 1, 1.5, 2, 3, 4, 5 and 6h after NIF using gaschromatography with ECD. Subjects were always studied after 6h of fasting and resting in supine position 45 min before control-EHBF and up to 3h after NIF. Results (mean + SEM) are summarized in the The ACE-inhibitor lisinopril is well established in the therapy of all grades of essential hypertension. It does not require any biotransformation step to become effective and is not significantly metabolised in humans. Absorbed drug is primarily excreted unchanged via urine. Therefore, applying normal doses, impaired renal function is associated with up to 10-fold increased serum concentrations during chronical treatment with lisinopril. In order to avoid side-effects in patients with severe renal failure, an adaptation of the dosage regimen is necessary. For this purpose a clinical study was performed, in which 8 patients with severe renal failure (creatinine clearance < 30 ml/min) received a reduced dose of lisinopril once daily over a period of 15 days. Based on the method of Dettli (1977) the individual dose was calculated based on the respective creatinine clearance. 16 healthy vo~nteers who received 10 mg per day served as control group. During the first and 15 t day of treatment serum and urine samples were taken over a period of 24 and 48 It, respectively. Lisinopril concentrations were determined in both body fluids by RIA. The most important aims of the study were 1. to compare the patients' lisinopril concentrations in serum and their urinary recovery with the respective values of the volunteers. 2. to evaluate safety concerning chronic lisinopril therapy in patients with severe renal failure. Cma x of the patients after 15 days of treatment with the reduced dose were significantly lower than the respective concentrations of the volunteers (29.7 vs 40.7 ng/ml), while the mean patients' AUC-values were in the same range as the volunteers' AUCs (525 vs 473 ng*h/ml). It is known that the terminal elimination halflife of lisinopril is not identical with the relevant halflife of accumulation (tl/2rel). In order to determine tl/2rel and the corresponding ratio of accumulation, an kerative estimation based on th~ amounts excreted via urine (day 1 and 15) was performed. The mean estimated tl/2rel for the volunteers and the patients were 20 and 389 h, respectively. Accordingly tla~ accumulation factors amounted to 1.32 and 9.00. The corresponding mean AUCvalues in steady-state were estimated to be 474 and 1000 ng*h/ml, respectively. Taking into account the inter-and intraindividual variabilities in absorption of lisinopril it can be resumed, that dosage reduction according to the individual CLcR in patients with severe renal failure leads to steady-state blood levels which are not-essentially increased compared with healthy volunteers receiving a standard dose of 10 rag.

Pharmakologisches Institut ffir Naturwissenschaftler, Klinikum der J.W. Goethe-Universitht, Theodor-Stern-Kai 7, D-6000 Frankfurt am Main In the following study liver blood flow was measured by hepatic sorbitol clearance during rest, food intake and bicycle ergometer load. Six healthy male subjects received an i.v. infusion of 40 % sorbitol (50 mg/min) for 10 hours. They were in a recumbent position up to 8 hours after starting the infusion, after 4 hours lunch was taken within 30 minutes. After 8 hours the volunteers performed a bicycle ergometer load of half maximal performance capacity for one hour followed by an one-hour resting period. Sorbitol serum concentrations were measured at predetermined times, urine was collected in three fractions (0 -4, 4 -8, 8 -10 hours). Renal excretion of sorbitol during each collection period was about 11 % of the ' administered dose in the same time interval. The estimated liver blood flow (ratio of hepatic sorbitol clearance over 1-hematocrit) in the six healthy male subjects ranged between 1009 and 1357 ml/min.

During food intake a 45 % increase, during bicycle ergometer load a 25 % decrease of liver blood flow within one hour was observed. The presented data show that physiologic variables have an important influence on liver blood flow. This has to be considered during investigation of elimination kinetics and bioavailability of compounds with high hepatic clearance. Sorbitol clearance allows to investigate drugs for their effect on splanchnic blood vessels and to elucidate the influence of liver perfusion on biochemical parameters. Rivastatin (BAY w 6228) is a novel, synthetic, enantiomerically pure HMG-CoA reductase inhibitor which is about 100 times more active than lovastatim Unlike lovastatin, rivastatin is an open-chain hydroxy acid and thus resembles fluvastatin and pravastatim Single dose studies on safety, tolerability and pharmacokinetics in healthy male volunteers were performed in the dose range 20 to 400 #g per subject. Plasma concentrations were determined by RIA, specific GC with ECD after double derivatization, and by HPLC with fluorometric detection after photochemical derivatization. The pharmacokinetic parameters based on GC and HPLC were similar whereas those by RIA were slightly different. The mean pharmacokinetic parameters obtained by moment analysis and based on HPLC data after a single oral dose of 200 #g (2.8 #g/kg) were as follows (n= 12, geometric means/geometric deviation): peak concentration 2.28/1.31 ng/ml, time to peak 2.6/1.07 h, plasma half-life 2.2/1.12 h, AUC 10.3/1.23 ng*h/ml, CL/f 0.27/1.30 1/kg*h, MRT 4.6/1.11 h. In the dose range studied the pharmacokinetics of rivastatin was linear. Rivastatin was well tolerated, hence multiple dose studies were initiated. (1). Aim of the present study was the in vivo investigation of T metabolism. 106 healthy volunteers recieved a single dose of T (50 mg hydrochloride). T-, M1-and M2-concentrations were determined in 24 hours urine by GC/FID. 81 of these volunteers were phenotyped with respect to CYP2D6 polymorphism using sparteine (100 mg sulfate). Mean cumulative excretion (% of dose) of T, M1 and M2 in 24 hours urine of the 106 volunteers was 11.9, 14.4 and 4.1 respectively. In the 73 extensive metabolizers (EM) of sparteine O-DM (metabolic ratio T/NIl) and N-DM (metabolic ratio T/M2) showed wide interindividual variation: T/MI: 0.04-18.21; T/M2: 0.37-44.5. 8 volunteers were phenotyped as poor metabolizers (PM) of sparteine. Mean T/M1 ratio in the 8 PM of sparteine was considerably higher than in the 73 EM (4.6 versus 0.93, p < 0.0001). In contrast no statistically significant difference was found between EM and PM with respect to N-DM. In the EM metabolic ratio of sparteine was positively correlated to T/M1 (r = 0.584, p < 0.0001). A microparticle formulation of Buserelin, an analogue of luteinisi6~" hormone releasing hormone, was given to 32 healthy females to determine the dose required to suppress ovarian function for 4 to 6 weeks. S.c. doses of 1.8, 3.6 and 7.2 mg were given to 12, 11 and 9 subjects. Serum buserelin, progesterone and oestradiol as well as urinary buserelin excretion were assayed twice weekly to assess the pharmacokinetics and the duration of suppression of steroid hormones defined as number of days from injection until increase of: a) Progesterone to above 8 nmol/I b) Oestradiol to above 300 pmol/l. receiving 40 mg/d OM for one week. 8 of these EMs received 120 mg/d OM during a second test period. The volunteers received 3 mg/kg ~3C-[N3methyl]-caffeine before OM treatment, at the 7th dose, and after a 7-day washout period. The breath test was performed for 8 h, caffeine plasma kinetics (n=14) and urine sampling (n=26) for 24 h. Urine was collected in 3 intervals from 0-5 h, 5-8 h, and 8-24 h. Caffeine and its metabolites were determined by rp-HPLC. The udnary ratios of paraxanthine/caffeine (=ratio) and (1-methylxanthine+l-methyluric acid + 5-acetyl-6-formylamino-3-methyluracil / 17-dimethylxanthine (=index) were evaluated. 13C-exhalation was elevated by 38.0 + 23.5% (mean ± SD, p=0,005, n=6) in PMs after 40 mg OM. In EMs ~3C-exhalation increased by 12.8 + 17.4% (n=12, n.s.) and 27.2 + 16.5% (n=8, p=0.002) after 40 mg and 120 mg OM/d, respectively. The urine collection period from 5-8 h was found to correlate best with the breath test (R= 0.69, p<0.0001, n=26, index). An increase of 44.6 _+ 24.7% (p=0.04, ratio) and 48.3 + 21.0% (p=0.003, index) in PMs after 40 mg OM (n=6) was found. In EMs receiving 40 mg OM (n=12) the respective increases were 9.6 + 25.0% (n.s., ratio) and 16.8 ± 21.1% (p=O.04, index). In EMs receiving 120 mg OM (n=8) elevations of 26.4 _+ 15.7% (p=O.02, ratio) and 34.0 ± 18.6% (p=O.O05, index) were found. These results fit to those observed in the breath test. The overall caffeine plasma clearance in 14 volunteers (3 PMs, 11 EMs) was increased by 33.7 ± 30.6%. Thus, OM elevates CYP1A2 activity in PMs and after high doses in EMs. This is quantitatively equally reflected by the N3-demethylation product in breath, caffeine clearance in plasma, and metabolic ratios in urine. Prednicarbate (PC) applied topically is about equipotent to betamethasonevalerate in inflammatory skin disease but does not induce skin atrophy in non-occluded skin (Korting et al., J. Am Acad. Dermatol. 27:87,1992) . There is, however, still a lack of knowledge on the dose-response relationship of PC. This holds true also with respect to the activity in relation to hydrocortisone double esters introduced into therapy only recently. Therefore we compared the activity of cream preparations with PC 0.025-0.5%, betamethasone-valerate 0.1% (BMV), hydrocortisone aceponate 0.1% (HCAP), hydrocortisone buteprate 0.1% (HCBP), hydrocortisone 1% (HC) and two base preparations in healthy volunteers. The vasoconstriction test (n=24), UV-induced erythema (2 MED; n=lS) and sodium laurylsulfate-induced dermatitis (5%; n=18) served as test models. Erythema/blanching reactions were graded by visual inspection and chromametry, transepidermal water loss by evapodmetry and skin roughness by profllometry.

Under occlusion vasoconstricton was most pronounced following PC 0.1-0.5%, HCAP, HCBP and BMV whereas HC and PC <0.05% turned out less potent. PC activity was less pronounced, if occlusion was not applied. Suppression of LTV-induced erythema was closely related to skin blanching under occlusion. Sodium laurylsulfate-induced dermatitis responded well to PC 0.25% and its vehicle. A difference between PC cream and the respective base preparation, however, was not observed. The results demonstrate a close dose-response relationship for PC when tested in standard models for topical glucocorticoids. PC (> 0.1%) as well as the hydrocortisone double-esters are potent topical steroids. In several experimental animal studies it has been shown that transformation of substituted benzimidazoles (SB) into active compounds in the acidic lumen of parietal cells is necessary to inhibit the gastric proton pump. In this study, we compared the IC50 of three SB of clinical relevanqe -omeprazole (O), pantoprazole (P), and lansoprazole (L) on H~--production in human gastric mucosal cells with the rate of transformation of SB to the active agents in 0.1 M HC1. Human gastric mucosal cells were isolated by successive digestion with pronase and collagenase of 4 to 7 ~ndic biopsy specimens. H +production was measured by detecting C-aminopyrine accumulation in 2 x 105 cells per tube. H+-production was induced by 1 mM ribcyclic AMP with or without different concentrations of SB. The rate of transformation (k) of 0.1 mM SB in 0.1 M HC1 was assessed by time dependent change in the absorbance of the solutions measured at 354 (O) and 332 (P and L) nm at 25 °C. Lansoprazole is the most potent inhibitor of the gastric proton pump among the three SB tested. There seems to be a relation (r = -0.959) between inhibitory activity of SB in the human in vitro system and their rate of transformation into the active compounds at pH 1 (acid activation). In addition, two parameters of proton pump inhibition can be investigated with clinically relevant human material thus avoiding animal experiments. Supported by the Robert-Bosch-Stiftung. *Robert-Bosch-Krankenhaus, Abt. Klinische Chemic, D-7000 Stuttgart 50, Germany. In order to characterize the ~-adrenoceptor (AR) subtype(s) mediating cardiac effects of ~9-AR agonist terbutaline (TER) we compared the effects of TER~in vitro on isolated human right atria (RA) and ]eft ventricles ~-~V) on adenylyI cyclase (AC) and force of contraction (FC) and in vivo in 7 healthy volunteers on heart rate (HR), systoIic~1o--o-~-pressure (P ,,~+) and contractility (assessed as shortening of pre-ejection s~ period [PEP] and QS~-time) with those of isoprenaIine (ISO). In vitro TER concentration-dependently increased RA-and LV-AC ~n = 6.04 and 5.88, respectively) and -FC (pD~ = 6.33 and 5.76~ respectively). Compared to ISO maximal AC-bctivation was 66% (RA) and 57% (LV), maximal FC-increase was 85% (RA) and 49% (LV). In contrast to ISO TER-effects were mediated solely by ~-AR since they were antagonized only by the ~o-AR antagonist ~CI 118,551 (erythro-(±)-1-(7-methyIindan-4-ylbxy)-3-isopropylaminobutan-2-ol) but not by ~-AR antagonists. In vivo i.v. infusions of TER (dose-rang~ 25-300 ng/kg/min for ~and ISO (dose-range 5-70 ng/kg/min for lOmin) dosedependently increased HR and P -and shortened PEP and HRcorrected QSo-time. Pretreatme~S~f the vo]unteers with the ~-AR selective antagonist bisoprolol (10mg p.o. 2h before the lhfusions) caused significant larger shifts to the right of the dose-response curves to ISO than to TER: Ratio ED after Bisoprolol / ED after Placebo HR(EDsog) Psyst(ED10 %) PEP(ED25 %) QS2(EDT.5%) ISO 

We have initiated a clinical phase I study with dose escalation of CYC in combination with doxorubicin (DOX) in a patient with soft tissue sarcoma refractory to DOX which was administered in a dose of 80mg/sqm as 30min iv infusion either alone or after CYC in a loading dose of 2 and 4mg/kg/2h followed by 6 and 12mg/kg for 24h by continuous iv infusion, respectively. The addition of C¥C caused a tumor regression and significantly increased myelosuppression. CYC blood levels were in the range of 0.7-1.1 and 1.0-1.7 ~M, respectively. During the first 48h after DOX administration plasma AUCs as well as urinary excretion were increased by >40% for DOX and >250% for doxorubicinol. As with etoposide, CYC may significantly contribute to the toxic side effects of DOX and its metabolites by an inhibition of their biliary excretion. Thus, the clinical study should be continued with great caution. Furthermore, the amelioration of tumor response caused by the addition of CYC cannot be solely ascribed to a modulation of MDR but may be due to a higher retention of DOX and its metabolites. Opiods (e.g. sufentanil; suf) and benzodiazepines (e.g. midazolam; mid) are drugs of first choice to accomplish an effective analgosedation in patients at the intensive care unit. In 11 head trauma patients suf (25 -200 p.g/h) and mid were infused for 48 hours according to the desired pharmacodynamic response. Following the initial bolus of suf (2 lxg/kg), during the infusion of both agents and after the end of the suf infusion multiple blood and urine samples were collected. Concentrations of suf were determined by a specific and sensitive RIA. From the initial decline following the bolus and from the terminal slope after stopping the infusion, the distribution (21, 22) and elimination (2z) of suf were characterised by the corresponding half-lives (tv=). CL-values were calculated from the steady state plasma concentrations and the equation CLR=Ae/AUC , respectively. In the table the mean -SD and ranges of the most important pharmacokinetic parameters are summarized, Besides vital signs intracrau/al pressure (ICP) was monitored. [SH]LSD binding was specific, saturable, and depended ~ incubation time, protein concentration and previous handling of tissue, i.e., use of fresh or frozen tissue.

In c~ntrast, studies with [aH]ketanserin were highly non-specific.

Therefore, we reccmmerd [aH]LSD as ligand of choice to study platelet 5-HT~ receptor binding in humans.

Furthermore, repeated measurement of individual data over time shc~id be interpreted cautiously, especially when data from depressed patients are tinder ex~tion. Ketoprofen and ibuprofen exert their therapeutic effects via inhibition of prostanoid synthesis. The aim of the present study was to determine dose response-relationships between two oral doses of ketoprofen resp. ibuprufen and platelet aggregation and prostanoid formation in man. For this we designed a open four-way cross-over study. 12 healthy female volunteers received for two consecutive days, followed by a five day drug-free interval, either ketoprofen 3x25 mg/d (LDK) or ketoprofen 3x50 mg/d (HDK) or ibuprofen 3x200 mg/d (LDD or ibuprufen 3x400 mg/d (HDI). The response criterions, determined before and on both days of each treatment period were: platelet aggregation in response to collagen, platelet thromboxane-synthesis (as TXB2 in platelet rich plasma after aggregation, measured by RIA) and 24-h-urinary excretion of PGE-M (7~z-hydroxy-5,11-diketotetranor-prosta-1,16<1ioic acid, main urinary metabolite of PGE2, measured by gaschromatography/tendem mass spectrometry). In addition, concentrations of ketoprofen and ibuprofen were measured in blood plasma 4 h after the morning dose of each drug treatment. Plasma concentrations of ketoprofen rose 1.9 fold (day 1) resp. 2.9 fold (day 2) if doses were doubled, for ibuprofen the increase was 1.6 (day 1) resp. 1.7 (day 2). Platelet aggregation in response to 2 lag collagen was significantly reduced by LDK (-57.2%, p< 0.005), HDK (-75.2%, p< 0.005), LDI (-22.2%, n.s.) and HDI (-27.9%, p< 0.05) as compared to control. With exception of LDI versus HDI, the differences between the treatments were significant at p< 0.05. Between the two days of each treatment period no significant differences were seen. TXB2-synthesis was significantly decreased by LDK (-94.8%, p< 0.005), HDK (-98.9%, p< 0.005), LDI (-57.4%, p< 0.005) and HDI (-71.8%, p< 0.005) as compared to control. The differences between the treatments were significant at p< 0.05, between the two days of each treatment period no significant differences were seen. All four treatments reduced 24-h-urinary excretion of PGE-M significantly (p< 0.05) in the range of 60.0% (HDK) to 43.2% (LDIL There were no significant differences between treatments, only a non-significant trend could be observed, corresponding to the inhibition of platelet aggregation and TXB2-synthesis. Our data show that platelet aggregation and TXB2~synthesis were inhibited in a dose-related manner, whereas PGE-M excretion in urine showed no significant difference between the four drug schedules. Both ketoprofen dosages were more effective in reduction of platelet aggregation and TXB2-synthesls than ibuprofen in low or high dosage.

Institute of Clinical Pharmacology, Hannover Medical School, D-3000 Hannover PHARMACOKINETICS OF THE PROSTACYCLIN-ANALOGUE TAPROSTENE R. Terlmden, W. Lintz, J. Gerloff. . Taprostene (T) combines marked inhibition of platelet aggregation with weak vasodilation. Compared to native prostacyclin, its pharmacodynamie potency is only about 1/10, but its stability in vitro and in viva exceeds prostaeyelin by far. For the determination of T in biological matrix, an offline HPLC-separation, followed by capillary gas chromatography-negative-ion eheminal ionization mass spectrometry was developed. The limit of quantification is 50 pg/ml. The pharmacokinetle parameters were evaluated in a randomized 3-way cross-over study in 12 healthy male volunteers, receiving a 2 hour infusion with 1.5 #g/kg BW, 3 #g/kg BW, and 6 pg/kg BW. The therapeutic range is between 3 and 6 #g~kg BW. Blood was sampled for 48 h in heparlnized tubes, immediately eentrlfuged and the plasma alkafized by the addition of 50 #1 NaOH. Urine was collected for 72 hours atier start of the infusion, the speeimans were also alkalized. The pharmaeoldnetie parameters were calculated by iterative curve fitting assuming a 3-compaRment model using the TOPFIT-program.

Pharmacokinetic parameters of T (means (SD)) The distribution and elimination aIter i.v.-administration follows a 3-exponential fall of the plasma concentration curve. All 3 half-fives are of therapeutio relevance, since the partial AUC's contribute almost equally to the total AUC. Total and renal Clearance are constant in the dose range tested. The terminal biologieaI half-fifo of T is considerably longer than that of native prostaeyelln. In combination with a better stability in vitro T is the superior candidate for therapeutic use.

Gl~nentlml GmbH ForschtmgszenWam. Abt. Hmrampharmakologie trod Abt. Phalan~oldnetlk. Ziegle~tr.6. 5100 A~ah~

In active peptic ulcer disease omeprazole (OM) represents a drug of first choice for rapid pain relief and healing whereas CBS can effectively prevent ulcer relapse if Helicobacter pylori (Hp) is eradicated. Therefore it is very likely that both drugs will be given concomitantly and consequently drug interactions have to be taken into considerations especially since solubility and consequently absorption of Bi from CBS is dependent on the intragastric pH which will be elevated by omeprazole. Therefore we have tested in a placebo-controlled cross-over study in 6 healthy volunteers whether OM (40 rag/die for one week) affects the plasma concentration time profiles (AUC) and urinary excretion (Ae) of Bi following a single oral dose of 240 mg CBS. OM and Bi have been measured by HPLC and AAS respectively. In addition, intragastric pH-values were monitored for 8 h.

As expected, control pH-values (2.1) were significantly (p=0.0001) elevated to 4.7 by OM which was eliminated with a tV2 of 1.3 --. 0.Th (mean +--SD) and an oral CL of 387 -221 ml/min. The increase of intragstric pH was related to the AUC of OM (r=0.98; p=0.016). OM-induced changes in intragastric pH affected absorption of CBS since AUC (mg/l x rain) and Ae (mg/8h) were significantly (p=0.04) higher during OM treatment (10.3 + 9.5 and 1.9 resp.) if compared to placebo ( A multicanter study on the efficacy of Galanthamine hydrobromide in the symptomatic treatment of senile dementia of Alzheimer's type was used to get pharmacokinatic data on the population of 56 patients between 60 and 85 years. Patients received 20 -50 mg/die Galanthamine in 10 mg tablets for 10 weaks. Concentrations of Galanthamine in red cells 1 to 4 hours after the application have been measured by HPLC. Aktivities of acetylcholinesterase in the hemolysate during inhibition and after reactivation were obtained by radio-and photometric measurement of diluted and undiluted sampels respectively (1). After oral administration of 10 mg Galanthamine Cma x of 16 -211, median 61 ng/ml and Ima x of 12 -71, median 37 % have been found. These values for the population of elderly patients were in the same range as respective data from young healthy volunteers (2). From 10 to 20 mg dose and concentration were lineary related. Interindividual variances of Galanthamina concentrations in red cells ranged from 12 to342 ng/ml and red cell acetylcholinestarase activities varied between 1.1 and 6.9 U/ml. Inhibition was calculated and ranged 11 -73 percent. The relation between inhibition of acetylcholinesterase and efficacy of Galanthamine will be analysed at the end of the trial. Conclusions: Galanthamine concentrations and acetylcholinesterase activities may serve to establish patients compliance and to contribute to the differentiation between responders and nonresponders. 

When P. Martini developed his concept of a doubleblinded clinical trial, an informed consent of the patients was not in the focus of interest. Today patients have to be informed due to ethical and legal reasons. There is a trend towards a more and more detailled information in the practise of clinical trials.

Information of patients has been discussed in relation to doctor-patient-relationship and patients compliance. A possible influence of information to the result of the study has not yet been analysed, Not regarding ethical-legal aspects, it will be shown that detailed information may influence patient's expectations. The more detailled they get information, the more they will speculate about their treatment, Independently, whether the patients on an average guess the treatment right or wrong, there is a polarisation of results. We are validating a new multi-user computerized test battery developed by the IBIS research team of the Bochum University for use in clinical pharmacology. With this system psychometrical tests of basic brain functions are available, like tests of simple reaction time, complex reaction time, vigilance, concentration (simple calculation task), motoric coordination, memory (word pairs or figures) and abstract language-free logics (2 versions). We performed complete testing of over 150 naive healthy male or female volunteers aged 20-40 years. The scores of both language-free versions of the logics test are correlated (alternate form reliability r =0.77), depend upon level of education and are positively correlated (r =0.42) with the % correct responses in the concentration test, the complex reaction test and the memory test with figures but not with word pairs. There is a correlation (r=0.43) between the % correct responses in the complex reaction test and the memory test for figures, which are both based on figural elements. There were no correlations between the other tests of the test battery. Reaction times are normally distributed and increase with task complexity. In a study with 6 times repeated testing with 3-4 days interval of 50 young healthy volunteers most tests showed clear learning effects which reached a plateau after 5 to 6 test sessions. Test-retest reliability increased towards the end of the training period. The advantages of a computerized test battery are the objectivity of testing and the ease of data management. Validation by studies with reference CNS drugs is planned. In the past several dopamine autoreceptor agonists were developed in order to decrease an enhanced dopaminergic neurotransmission. Roxindol (5-Hydroxy-3-(4-phenyl-l,2,3,6tetrahydropyridil-(1 )-butyl)-indol)is a newly compound which has a high dopamine autoreceptor agonistic potency at D2 receptors. Presumably, pituitary D2 receptors operate at the same affinity state like autoreceptors in the CNS. Roxindol was applied to 6 healthy male volunteers receiving in randomized order at 9.00 h 1 mg Roxindol or placebo p.o.. Blood samples were drawn every 30 min from 8.00 h until 13.00 h. Roxindol had no effect upon the regular pattern of ACTH and cortisol secretion. 30 min after administration of Roxindol an enhanced release of growth hormone from 1.1 + 0.2 ng/mlto4.3 + 2.0 ng/mlat 11.00 h could be observed (p < 0.05). Area values differed significantly, too (Roxindol vs. placebo, x + SD, 18.1 ± 17.1 vs. 7.7 ± 8.0 ng x h/ml, p < 0.03). Starting with prolactin baseline concentrations at 9.00 h of 5.3 ± 1.4 ng/ml a pronounced reduction to 1.1 ± 0.4 ng/ml at 11.00 h was observed. From 10.30 h to 12.30 h significant differences between mean plasma prolactin concentrations (p < 0.01 ) emerged. Area values differed significantly, too (Roxindol vs. placebo, x ± SD, 14.8 ± 13.5 vs. 34.4 ± 18.1 ; ng x h/ml, p < 0.03). Roxindol has potent effects on the facilitation of growth hormone secretion and the inhibition of prolactin release. Both findings are related to a dopamine receptor agonistic effect. Roxindol is potentially useful in the treatment of endocrine diseases such as prolactinoma. Particle size, water solubility and stability at different pH have influence on the bioavailability of newer oral cephalosporins. This results from in vitro investigations and from pharmacokinetic studies.

In contrast to phenoxymethyl penicillin the bioavailability of cefixime from a syrup (oily basis) is only marginally less than from a dry suspension formulation (aqueous basis) because of the slow absorption process of cefixime from the GIT which allows equal good dissolution of cefixime particles from both formulations. -Simultaneously administered antacids or H:z-blockers have an unfavourable influence on the bioavailability of cefpodoximeproxetil in the fasting state but not on that of cefotiamhexetil. The reason is the better water solubility of cefotiamhexetil at higher pH up to 5 with respect to cefpodoximeproxetil. -Cephalosporin esters are stable at low pH, but isomerize at higher pH (7-8) to antibacterialinactive delta-2-cephalosporins. Cefotiamhexetil is the only cephalosporin ester with noticeable instability and tendency to isomerize at pH 5-6, as it occurs in the small intestine. Therefore, the inactive isomer of cefotiamhexetil has been detected in serum, but not of other oral cephalosporins.

Department of Pharmacology, University of Regensburg, Universit/itsstrasse 31, D-8400 Regensburg, Federal Republic of Germany In view of possible interactions between antidiabetic drugs and other medicaments prescribed simultaneously as well as possible influences of concomitant treatment on intermediary metabolism we evaluated appr. 400 prescriptions using from a panel of doctors treating patients with IDDM and NIDDM and one or more accompany diseases. A potential interaction could be recognised in 19 % of cases. 70 % of these involved possible interactions between concomitant other medication and insulin or oral antidiabetlc drugs, whereas 30 % of possible interactions were related, to interactions inside the group of concomitantly administrated drugs. Most interactions involving insulin or oral antl-diabetic drugs occured with antihypertenslve drugs (thiazide-diuretics, 8blockers).

Equally frequent were interactions between sulphonylureas and lipid lowering agents (esp. cloflbrate) and between insulin and salicylates taken together amount to another 30%. 7 % of the interactions were classified as severe, 77 % as moderatly severe and 6 % as having little or no clinical relevance.

According to the recommendation of the In contrast to the antihistaminic effect of Diphenhydramine (D), that of Terfenadine (T) develops in isolated smooth-muscle preparations with latency. It proceeds even after the drug's removal from the bathing-medium, and cannot be reversed by repeated washing. A high affinity binding-site aside from the H~-receptor has been presumed to explain this phenomenon. The aim of this study was to reveal the significance of this additional binding-site on the kinetics of T by functional experiments in isolated tissues. Furthermore physicochemical properties of T should be determined. In a cascade-set-up oxygenated Tyrode's solution perfused a guinea-pig heart suspended according to Langendorff before superfusing guinea-pig ileal strips. The heart simulated a presystemie extraction. Histamine-induced contractions of the ileal preparations served to quantify a compound's antihistaminic effect. D passed the coronary vascular bed almost completely. Even high concentrations of T did not impair the ileal response to histamine after one hour or more, indicating a strong tissue-accumulation of T. In experiments on 4~Ca-labelled phosphatidylserine-monolayers T was found to be very effective in displacing 45Ca from the membranes, the half-maximum effective concentration (ECs0): T = 5'10 "~ reel/l, Propanolol (P) = 2"10 -2 reel/1 and D = 5"104mol/1. Tensiometric determinations of the surface activity in aqueous solution showed T to be a potent surface active compound with a CMC (criticmicellar-coneentration) of about 2"10 "s mol/l. D-and P-HC1 revealed a CMC of 2"10 "l mol/l and 10 -1 mol/l, respectively. In conclusion, T is an amphiphilic compound with a high surface activity due to an extreme enrichment at interphases. These properties of T in particular may be responsible for its striking tissue accumulation and for its peculiar kinetics. Dienogest (17a-Cyanomethyl-lTS-hydroxy-4,9-estradien-3-one) is a synthetic progestagen demonstrating a strong effect in the MePhail-Clanberg test in the rabbit. It is a component of the preparation CertostaP for oral contraception in women. In comparison to its high progestaganic activity, receptor binding of Dienogest to the rabbit uterus eytusol receptor has been found to be rather low (Progesterone = 100; Dienogest = 13; Levonorgestrel = 390). Preliminary studies un pharmacokineties and biotransformation of Dianogest in rabbits could not explain the discrepancy between low receptor binding and high progestagenJc effect. With steroid hormones, the non-protein bound part in plasma is regarded as the biologically active form. We have determined the non-protein bound part of Dienogest in plasma of female rabbits, rats, B~gle dogs and female volunteers using the method of centrifugal ultrafiltration with centrifree ® micropartition systems, q'he same procedure was carried out with the progestagen Levonorgestrel. Dienogast was found to have a considerably higher portion of free, non-protein bound compound in plasma, running up to 10% of the total substance concentration, in comparison to Levonorgestrel. This large proportion of free vs. protein bound steroid will obviously contribute to the strong progestagenic effect of Dienogest in vivo.

Hans The pharmacokinatics of the histamine H2-receptor antagonist cimetidine includes discontinuous absorption processes resulting in double peak occurrence after peroral administration of a single dose in the fasted state, with a frequency of at least 50% in humans and 30% in dogs. The most probable explanations for this phenomenon appear to be site-specific absorption and/or variations in the fastedstate gastrointestinal motility. The aim of the present study was to examine the contribution of different pharmacokinetic processes and administration modes to the systemic availability in dogs. This includes administration site and pH dependence of the absorption rate, clearance variations as well as potential correlations of the obtained parameters with different phases of the intestinal fasted-state motility cycle. A pharmacokinetic model with two lag-times and two absorption rates was evaluated regarding its applicability for cimetidine. Cimetidine pharmacokinetics in the dog is described by a two-compartment-model with an average terminal half-life of 104 _+ 21 h. After p.o. administration a considerable intra-and interindividual variability was found. No significant clearance non-linearity was detected up to a 20 mg/kg dose. However, at low plasma concentrations the renal clearance was higher, which is explained by (saturable) active tubular secretion. In the oral studies the frequency of double peak occurrence was 35%, while it was 20% for duodenal infusions. This indicates that gastdc emptying and site-specific absorption are both contributing to this phenomenon. Duodenal infusion studies on the pH dependence of absorption indicate that its rate and extent are higher at pH 8 (bicavailability, F=0.92) and pH 4 (F=0.87) than at pH 6 (F=0.72). In general the mean absorption times after duodenal infusions were slightly shorter for administration in the active phase than in the quiescent phase (55.6 vs. 62.3 rain).

Applying the two-lag-time model to the absorbed fractions it was possible to characterize the discontinuous absorption profiles. A population pharmacokinetic model was used to quantify the magnitude and variability of the various processes in the collective. After subcutaneous administration of morphine-HCl (20 mg/kg b.w.) maximum levels of M, M3G and M6G amounted to 3563 ng/ml, 10975 ng/ml and 289 ng/ral, respectively. In animals chronically treated with M (20 mg/kg bid) trough concentrations of M, M3G and M6G were measured and found nearly unaltered in the course of a ten-days" treatment. After the prolonged treatment the trough concentrations of parent drag and metabolites were determined. The CSF/blood ratio for M and M3G amounted to 1.6 and 0.1. The concentration of M6G in CSF fell below the detection limit (5ng/ml); i.e. whereas trough concentrations of M in CSF were found higher than in plasma the concentration of both glucuronides in CSF was much lower than in plasma. M, M3G and M6G are eliminated from plasma with half lifes of 92, 116 and 100 min, respectively. The similarity of the elimination rates suggests that the formation of the metabolites determines the rate of their elimination.

In separate experiments the half life of elimination of intravenously applied M3G (25 mg/ kg b.w.) was determined with 32 rain. M6G was not available in sufficient amounts to run similar experiments. In summary, there is no indication from our experiments that glucuronides behave like lipophilic drugs or for an accumulation in CSF.

Abteilung Pharmakologie, Universit/it Kid, Hospitalstrage 4, 2300 Kiel

Scherer. C. R. 1), H. Brunner 1), U. A. Wachter 2) S.

Contmon methods to detect and determine xenobiotics such as phaz~/aceuticals, illegal drugs, etc. in an organism by means of urine or blood samples are hampered by the fact that these xenobiotics mostly are eliminated within a few days after intake. In our attempt to uncover doping in athletes, we have developed a new GC/MS method for the determination of androgens in human hair allowing the determination of these compounds even months after its use. In brief, a hair sample is washed in an ultrasonic bath, dried and ground down% in a ball mill under liquid nitrogen. The powder received is extracted three times with ethylacetate. The combined extracts are evaporated under nitrogen and dissolved in ethylacetate/cyclohexane The new intravenous anesthetic propofol (P) is used for induction and maintainance of general anesthesia. Only few informations are available about the binding of this phenolic compound to human blood proteins. Therefore the binding of P to fresh human serum (I-IS), human serum albumin (HSA) and hemoglobin (HH) was studied by means of equilibrium dialysis, which was performed in 1/15 M phosphate buffer (pH 7.4). P concentration was deterrmined by HPLC after pre-column extraction and consecutive fluorimetric detection. Using HS and P in a concentration range from 39 ng/ml -90 /zg/ml (12 steps) %-binding was extremly high (range: 96 -99 %) and nearly independent of the substrate concentration. Even at maximal P concentration (limited by the poor solubility of the phenolic compound in aqueous solutions) a saturation of binding sites was not observed. Also 1% HSA (P concentration range from 9.7 ng/rnl -32.4 #g/ml (12 steps) could not be saturated with P; atypically, the percentage of P bound to HSA increased with higher and decreased with lower substrate concentration: e.g. with 32.4, 1.25 and 0.0097 tzg/ml P binding amounted 72, 70 and 58 %, respectively. 1% HH (same P concentration range as used with HSA) showed a rather typical binding behavior for P; at least at high P concentrations a decrease of the fraction bound indicated a saturation of binding sites: e.g. with 32.4, 1.25 and 0.0097 #g/ml P binding amounted 57, 64 and 65 %, respectively. It is assumed that the interaction betweeen HSA and P induces a conformational alteration of the macromolecule whereby at higher concentrations binding capacity for this ligand is increased. Cycl0sporin has several side effects, the most important of which are h~pertensi0n and renal dysfunction. Calckm channel blockers have been reported to protect the kidney during cycl0sporin-induced nephr0t0xicity and they are potent drugs in the treatment of h~portersi0n. Thus calciun antagonists, such as nitrendipine, seem t0 be good ~h'ugs for treatment of hypertension in patients receiving cycl0sporin. We studied the effects of altrandipine on he~odynamics and on cycl0sp0rin pharmac0kinetics in 6 patients with terminal renal failure. Kinetic parameters were calculated after a single 0ral dose of 20 mg nitrendipine, a single 0ral dose of 120 mg cycl0sporin and after concomitant therapy. S~eady-state plasma levels were determined after a 6 days' ural treatment with 20 mg nitrnndipinn twice daily or 120 ~g ciclespurin twice daily or after concomitunt therapy.

The maxi~mm blood cycl0sporin level (c~ax) was 286 ng/~l und it occurred 2,4 h after the oral dose of 120 ~ cycl0sporin. The AUCsh was 530 ng*ml'l*min. During treatment with the c0~ination of cycl0sporin and nitrendipine: C~a x = 239 ng/ml, t~x = 1,6 h, AUCsh = 421 ng*ml-l*ml. The ~aximu~ plasma nitrendipine level (C~ax) was 1,54 ng/ml and it occurred 2,8 h after the 0ral dose of 20 mg of nitrendipinn. The AUCsh was 4,93 ng*ml-l*min. Durig treatment with the c0~binatian ¢~ax = 2,02 ng/ml, tma x = 1,6 b, AUCsh = 5,0 ng*ml'l*m~n. The c0nc0~itunt thermpy with cycl0spurin and nitrendipine shortened the tma x. For cycl0sporin this effect was significant. At the time of Cma x for cycl0sporin, the blood pressure is sigoificant higher than before treatment. The si~ultane0us administration of nitrendipine is able to compensate for this effect. Since 1961 alcuronium (ALC) is used as a long acting stabilising muscle relaxant. There is poor information about its pharmacokinetics and its concentration-effect kinetics in patients, so we determined the ALC plasma concentrations after a bolus injection (12 h), the urinary excretion (24 h) and the neuromuscular blocking effect during the anaesthesia (about 2-4 h). 5 patients undergoing oral surgery (sex:4 m,l f; age: 35-69 years; body weight: 65-84 kg, ASA status 1-2) participated in this study after giving informed consent. Anaesthesia was induced with thiopentone and sustained with midazolam and fentanyl. Neuromuscular monitoring was done at the m.adduct.policis stimulating the n.u]naris over the wrist with supramax.voltage in train-of-four technique. Muscle relaxation was started with the application of an iv bolus of ALC (0.25 mg/kg).

[ALC] was determined after deproteinisation with acetonitri] by HPLC (Spherisorb 5-CN~ uv detection at 294 am, elution consisting of acetonitri] (46%) and Na2SO 4 (0.05 Mol/l in 0,005 Mol/l H2SO4), int.stand. ]audanosine, detection limit 0.05 rag/l) .

[ALC] varied between 2 (at 5 ') and 0.I mg/l (at 12 h) and could be fitted by TOPFIT to a mammillary 3-compartment model (t/2 term : 7.5 (3-16)h;~Vdss: 28.8 (23-37) ~l, Vcent r :5.1 (1.5-13.5) 1; k'13:0.110 rain-±; k31:0.028 min-~; V :16.8°1; k14" 0.241 rain-l; k41:0.142 rain-l; V4:6.8 1. Urina-3 " ry recovery was 74 % of the dose. The neuromuscular blockade .w.as fitted to the Hill equation E(t)=Ema x * y(t) H / [ICs0 + y(t) M] attached to one of the peripheral compartment (named 4), which exhibited the characteristics of a shallow compartment. [ALC] in this compartment and the neuromuscular blockade was correlated exhibiting a S-shaped concentration effect curve with IC50 = 0.035 rag/1. This is in contrast to the measured [ALC] 6-Mercaptopurine (6MP) released from azathioprine (AZA) may be oxidized by xanthine-oxidase (XO) to thiouric acid (katabolic pathway) or enter nucleetide interconversion pathway via thioinosinic acid (anabolic pathway). Thiopurinemethyl-transferase (TPMT) can methylate 6-MP, its riboside and nucleotide to their corresponding derivatives (6MMP, 6MMPR, 6MMPN). This additional pathway is of interest, since Lennard et al. (Clin Pharmacol Ther 41,18-25; 1987) found a correlation between a genetically controlled lack of TPMT and bonemarrow toxicity in children after chronic 6MP treatment. In a previous study (Naunyn-Schmiedeberg's Arch Pharmacol 346,R41;1992) we observed that pretreatment of mice with allopurinol (ALLO) results in the expected increase of 6-MP concentrations in intestinal mucosa, plasma and liver, however in spleen and bone-marrow 6MP and 6TG concentrations were reduced. Since it is unknown, to what extent 6MP and snsbolie metabolites are methylated in tissues, we studied the pharmacokinetics of 6MMP, 6MMPR and 6MMPN in intestinal mucosa, liver, spleen, and bone-marrow of mice (35-40g) after a single oral dose of 50 mg/kg of AZA with and without ALLO (20 mg/kg) pretreetment.

Concentrations of 6MMP and 6MMPR were determined after extraction by HPLC. The extraction of an acid hydrolized sample yielded total 6MMP (consisting of 6MMP, 6MMPR and 6MMPN) from which the concentration of 6MMPN was calculated. 6MMP, 6MMPR and 6MMPN were observed in all tissues. Maximum concentrations of total 6MMP (i.e. after acid hydrolysis) were reached 4 h after dosing and were as high as concomitant 6MP-concentrations (liver 20, intestinal mucosa 20, spleen 3, bone-marrow 0.5 #mol/kg). 6MMP accounted for only 1%, 6MMPR for 40%, and 6MMPN for 60% of the total concentration. While pretreatment with ALLO had only a marginal influence on 6MMP concentrations in intestinal mucosa and liver, concentrations in spleen and bone-marrow were reduced. Conelnslons: Methylation predominantly takes place intracellularly at the riboside and nucleotide level of 6MP. Inhibiting the katabolic pathway (ALLO) will not increase the concentration of methylated metabolites, in fact may even decrease their concentration (spleen, bone-marrow). These results indicate that pretreatment with ALLO will hinder 6MP from entering the anabolic pathway. In the light of these results the influence of ALLO on the therapeutic and toxic effects of AZA will be reevaluated. Institut ffir Pharmakologie der Medizinischen Univ~rsit~t zaz Lfibeck, Ratzeburger Allee 160, D-2400 Ltabeck. The mucolytic agent carbocysteine (CMC, S-carboxymethyl-L-cysteine) has been used for many years as a probe for the xenobiotic sulfoxidation capacity of a given population (Mitchell SC, Waring RH [1989] , Pharmacol. Ther. 43: 237-249). From more detailed investigations, basically performed in our laboratories, metabolic formation and renal excretion of significant amounts of cysteinyl sulfoxide metabolites can now definitely be ruled out. Instead, thiodiglycolic acid (TDGA), its sulfoxide (TDGA-SO), and S-carboxymethylthio-L-cysteine (CMTC) have unequivocally been identified as major routes of elimination ( Introduction. In man, comedication of cyclosporine A (CsA) and the calcium channel blocker diltiazem resulted in pronounced pharmacokinetic interactions (Brockm611er et al., Eur J Clin Pharmacol 38: 237), which are likely to reflect interactions on cytochrome P-450 3A isoenzymes. In this study the effect of diltiazem on the ratio of CsA and its metabolites 1 and 17 in different organs of rats was studied. Methods. Male SD-rats were treated with CsA alone (30 mg/kg/d, n=8) or in combination with diltiazem (60 mg/kg/d, n=8) for three weeks. 24 h after the last application concentrations of CsA and its metabo]ites M1 (AM9) and M17 (AM1) were measured after extraction from the tissuehomogenates by gradient HPLC using 210 nm UV-detection. Results. The concentrations and the relations of the metabolites 1 and 17 to CsA are given as median and range (* p < 0.05, ** p < 0.01, U-test): Comedication with diltiazem did not result in significant changes of concentrations of CsA and its metabolites and ratios of tissue concentrations to blood concentrations did not change. However, ratios of CsA primary meta* bolites to parent compound were significantly increased during diltiazem treatment. The data confirm, that the CsA-diltiazem-interferences are mainly due to interactions in secondary CsA-metabolism, 

CsA [pg/mll M1 [pg/ml] M17 [

Clostridium botulinum C2 toxin and Clostridium perfringens iota toxin modify monomeric actin by ADP-ribosylation. Here we studied the influence of nucleotides on the ADP-ribosylation of cell membrane bound actin. ADP-ribosylation of endogenous actin was stimulated by ATP in membranes from human platelets, rat liver, hamster fat cells and guinea pig fat cells. In rat liver and hamster fat cell membranes stimulation of ADP-ribosylation was dependent on the type of nucleotide (ATP > ATP'~S > AMP-PNP > ADP, respectively GTP > GTP,~S > GDP). In contrast, ADP-ribosylation of monomeric muscle actin is stimulated by ADP and less by ATP (Aktories et aL J. Physiol., Paris, 1990, 84, 262-266) . Further, the observed effect was independent from ATP-hydrolysis, because it was observed in the presence of stable ATP-analogs.

Using immunoblot techniques we showed that treatment of rat liver membranes with nucleotides (ATP > ATP~,S > ADP > AMP-PNP > > GTP = GDP = GTP~,S) resulted in release of actin monomers which were then ADP-ribosylated by C2 or iota toxin. Besides synthetic substrates (arginine-methylester and agmatine), glutamine synthetase has been shown to be a substrate of the transferase. Here we report on the ADP-n'oosylation of a-skeletal muscle actin by ADP-rt A. Monomefic G-actin was modified by ADP-rt A in a NAD-dependent manner. The ADP-ribose protein bond formed by ADP-rt A showed the same sensitivity towards hydroxylamine as the ADP-n'oose actin linkage which was formed by iota toxin at arginine-177. However, pre-ADP-ribosylation of actin in tb,~presence of unlabelled NAD and iota toxin did not inldbit further [~P]ADP-ribosylation by ADP-rt A and vice versa. Furthermore, proteolytic peptide mapping of actin ADPribosylated either by ADP-rt A or by iota toxin revealed a difference in the labeled peptides formed, indicating different acceptor sites for the ADP-ribose. Whereas iota toxin modifies exclusively monomeric Gactin but not F-actin, F-actin served as a substrate for ADP-rt A. Polymerized actin incorporated one tool ADP-ribose per tool actin, whereas monomeric actin is modified by two tool ADP-ribose per mol actin.

The monomeric actin ATPase which is inhibited by iota toxin-induced ADP-ribosylation was not influenced by ADP-rt A. Whereas actin modified by iota toxin showed no polymerization, actin modified by ADP-rt A showed the same extent of polymerization as did control actin. However, the rate of polymerization was decreased compared to control actin.

The data indicate that actin is modified by a eukaryotic ADP-ribosyltransferase and that this modification causes functional consequences. The ADP-ribosylation catalyzed by ADP-rt A is clearly different from labelling by bacterial toxins. ADP-rt A may be a model transferase for studies on endogenous ADP-ribosylation of actin. Pit-1 is a cell-specific activator of growth hormone (GH) and prolactin (PRL) gene Iranscription. In addition, it was shown that this factor mediates TRH and cAMP regulation of PRL and hTSHI3 gene expression. The regulation of the l~-gene by these mediators is predominantly mediated by three Pit-I DNA-binding sites in the 5'-flanking DNA of the gene. To determine the role of these sites in TRH and cAMP action on the TSH8 gene, we mutated these sites either individually or collectively to sequences which should not bind Pit-1 (/vl mutations) or to the consensus Pit-1 binding sites of the GH and PRL genes (P mutations). M mutations at site 1 (hTSHI~-I, -119 to -104 bp), at site 3 (hTSHI3-3, -72 to -58 bp) or at all three sites reduced TRH and forskolin induction in transfected GH 3 cells by 70-80%. However, mutation at site 2 (hTSHB-2, -104 to -89 bp) resulted only in a 20% reduction in TRH or forskolin induction. P mutations also reduced TRH or forskolin induction by 30-40%. Since Pit-1 is phosphorylated in rat pituitary ceils following activation of either protein kinase A or C, we tested the effect of phosphorylation on Pit-1 DNA-binding in vitro. Non-phosphorylated Pit-1 bound less well to hTSHl~-sites than to a rGH site (rGH-1). In contrast to the ~3H-1 site, the sequences hTSHI~-I and hTSHI3-3 bound 3-8fold more phosphorylated than non-phosphorylated Pit-1.

Thus unlike the marked basal activation of GH and PRL gene transcription by Pit-1 in the somatotroph and lactoa'oph, the TSHI3 gene contains unique Pit-1 DNA-binding elements that may limit basal activation in the thyrotroph by poorly binding of non-phosphorylated Pit-l, while conferring hormonal regulation by increased binding of phosphorylated Pit-1. An approach to antineoplastic therapy not exploited yet to a greater extent is to target enzymes relevant for cAMP turnover. It has been shown that in tumor cells cAMP concentrations are reduced, as compared to normal tissue (Hickie et al, 1974, Biochem.Biophys.Res.Comm. 59, 167-173) . Moreover, inhibition of 3',5'-cyclic nucleotide phosphodiesterase (PDE) has been found to correlate with enhanced cAMP-levels and with the extent of growth inhibition in various cell lines (Drees, M.;PHD Thesis 1992, University of Kaiserslautern). We have found selective inhibition of the rolipram-sensitive isoenzyme PDE-IV to play a predominant role for growth inhibitory effects. A new selective PDE-IV inhibitor was found to achieve significant cell growth inhibition in a spectrum of human tumor xenograft lines. A human colorectal xenograft line (CXF 280) sensitive in the colony forming assay was grown as solid tumor in nude mice. Analysis of its PDE-isoenzyme pattern revealed that about 6 % of total cAMP PDE activity originate from PDE-I, 26 % from PDE-II and 68 % from PDE-IV. Since selective PDE-IV inhibition correlates with markedly enhanced intracellular cAMP levels and also with the extent of growth inhibition this isoenzyme appears to be a promising new target for antitumor therapy. In order to obtain further evidence for the involvement of protein kinases in the short-term ACTH-stimulated aldosterone synthesis in rat zona glomerulosa cells, the effects of three different compounds with protein kinasa inhibitory properties were investigated. Staurosporine, H-7 and trifluoperazine inhibited ACTH-stimulated aldosterone release in a dose-dependent manner. The corresponding IC50-values were 350 nM (staurosporine), 28 muM uM (H-7), and 30 muM (trifluoperazine). These values are in agreement with data reported from other in vitro systems. While the inhibitory effect of H-7 was reversible upon washing of the cells with inhibitor-free medium, the inhibition was maintained in cells treated with staurosporine or trifluoperazine. This difference is probably related to different membrane permeability of the compounds. No cytotoxic effects were observed at any concentration of the substances. In contrast to the stimulated production, basal release of aldosterone even at the highest drug concentrations tested was not completely inhibited. We thus conclude that protein kinases may play a crucial role in short-term ACTH-stimulated aldosterone production in rat glomerulosa cells. Hoyer I Functional responses to stimulation of rat 5-HT~c receptors expressed in A9 cells were studied using whole cell voltage clamp recording technique. Stimulation of 5-HT~c receptors with 5-HT {3x10 a mol/I) evoked stable outward currents in cells clamped at -50 mV. Pretreatment with the protein kinase C activator phorbol myristic acetate (PMA; 10 ~ tool/I, 60 min) reduced the 5-HT-induced current amplitude by 46 + 6 % of the control value (n = 12). Pretreatment (45 min) with the PKC antagonist calphostin C (10 ~ mol/I) reduced the PMA-induced desensitization from 46 + 6 to 8 + 5 % (n = 9). One h after incubation with PMA (10 s tool/I) injection of the serine proteinphosphatase calcineurin (0.1 unit/cell) increased the 5-HT-induced response up to 34 + 7 % (n = 10). In A9 cells which were incubated 24 h with the 5-HT~c receptor agonist mCPP (meta chlorophenyl-piparazine HCI; 10 s mol/I), 5-HT-induced responses were reduced by 35 + 5 % of the control value (n = 15). Injection of calcineurin (0.1 unit/cell) in mCPP treated cells enhanced the 5-HT-induced responses by 27 + 6 % (n = 11). These results suggest that in this system rat 5-HT~c receptors are desensitized after phosphorylation by PKC. This desensitization can be counteracted by calcineurin-induced dephosphorylation. In response to various agonists a number of platelet proteins becomes phosphorylated at tyrosine residues. The responsible kinases, however, still await identification. Platelets contain high levels of non-receptor protein tyrosine kinases of the src-family of which pp60 c'src is the most abundant, suggesting an important role in platelet physiology. All agonists tested, including thrombin, vasopressin, collagen, the calciumionophore A23187, PMA (phorbol 12myristate, 13-acetate), which indirectly or directly activate protein kinase C, increased overall kinase activity of antibody-immobilized pp60 c-src isolated from clarified platelet extracts. On the other hand, elevation of cyclic AMP directly by forskolin or indirectly by prostaglandin E 1 did not affect the activity of the enzyme. To substantiate the differences in enzyme activity we determined K~, and V_a~ values of resting and thrombinstimulated platelets. Thror~in incre'~s~ substrate affinity of pp60 c'src as indicated by a 2-to 3-fold decrease of the K m values for ATP and exogeneous protein substrates. V,,,~ x values were n~St altered. As a consequence of thro~']n-stimulation, pp60 c'src becomes translocated from the membrane fraction(s) to the cytoskeleton, suggesting new formation of substrate-enzyme complexes. Indeed, we found many of the major tyrosine-phosphorylated proteins (145, 130, 80, 60, 58, 54 and 38 kDa) to be cytoskeleton-associated in agonist-stimulated platelets. To rule out modifications of pp60 c'src as a probable cause of these e~nts, we analysed tryptic phosphoamino peptides of immunoprecipitated azP-labeled pp60 c'src of unstimulated and stimulated platelets. The agonists listed above induced an approximately 2-fold increase in pp60 c'src phosphorylation, mainly at Ser-12 resulting from protein kinase C activation. We suggest that phosphorylation at Set-12 increases the substrate affinity ~ src of pp60 -and may also be responsible for translocation of the enzyme closer to putative substrates at the cytoskeleton. opening of a receptor operated cation channel (ROC), stimulation of phospholipase C (PLC) and a subsequent release of calcium ions from intracellular stores via an 1,4,5-inositol trisphosphate (IPa) mediated pathway, inhibition of ade~nylyl cyclase (AC) by a Gi-protein dependent mechanism, adhesion, aggregation and degranulation of platelets. The aim of this study was to identify the ligand specifity of platelet ADP receptors by functional assays using intact platelets. We investigated the effect of various adenosine phosphates, adenosine phosphate derivatives and nucleotide diphosphates on platelet calcium influx, calcium mobilization, cAMP levels and platelet aggregation.

Our results indicate that the platelet purinergic receptors are very specific with respect to ADP. ATP and ATP-2¢-S which are potent activators of most purinergic receptors have only little effect on cAMP levels, no effect on calcium influx and mobilization and probably an inhibitory effect on aggregation. Adenosine diphosphates substituted in the phosphate moiety [ except the u,/3-methylene-derivative which has no effect on platelets ] activate calcium influx and mobilization and inhibit AC although less potently than ADP itself. Changes in the ribose-ring abolish the ADP effects on AC/G i protein, while calcium responses and reversible aggregation can stilI be induced by 2'-Desoxy-ADP but not by ADP-acyclic aldehyde and ADP-acyclic alcohol. Nucleotide diphosphates such as GDP, IDP or UDP do not affect calcium regulation, platelet aggregation or adenylyl cyclase. The results suggest that human platelets contain at least two classes of ADP-receptors which can be distinguished by their ligand binding properties.

Medizinische Universit~tsklinik, Klinische Forschergruppe, Josef-Schneider-Stral~e 2, W-8700 Wfirzburg, Federal Republic of Germany 196 SYNTHETIC LWOPEPTIDES ARE EFFECTIVE ACTIVATORS OF HUMAN PLATELETS Michaela Berg Synthetic lipopepfide analogues of the N-terminus of bacterial lipoprotein have been shown to induce activation of macrophagas, neutrophils and lymphocytes. We studied the effect of the lipopeptide, N-palmitoyl-S-[2,3 -bis(palmitoyloxy)-(2RS)-propyl]-(R)-cysteinyl-(S)-seryl-(S)-lysyl-(S)-lysyl-(S)-lysyl-(S)-lysine (Pam3Cys-Ser-(Lys)4), on several functions of human platelets. Pam3Cys-Ser-(Lys)4 led to the aggregation of platelets and induced the secretion of serotonin from preloaded platelets with an effectiveness similar to that of thrombin. These cellular effects of Pam3Cys-Ser-(Lys)4 were concentration-dependent being half-maximal at 5-10 ~tg/ml and maximal at 20-30 ~tg/ml. The lipopeptide isocarba-Pam3Cys-Ser-(Lys)4 was as potent as Pam3Cys-Ser-(Lys)4 , whereas Pam3Cys-Ser-Gly was without effect. Since activation of platelets by several agents including thrombin involves the phosphorylation of proteins on tyrosine residues, we tested the effect of Pam3Cys-Ser-(Lys)4 on tyrosine phosphorylation. The lipopeptide stimulated tyrosine phosphorylation of several proteins, including 40 kDa, 75/80 kDa and 95/105 kDa species, which were also tyrosine phosphorylated in response to thrombin. Stimulation of tyrosine phosphorylation by the lipopeptide was time-dependent and reached a maximum after 5-10 s (40 and 75/80 kDa proteins) and 90 s (95/105 kDa proteins). The results show that lipopeptides are strong activators of platelets. The potential pathophysiological signiftcance of this finding remains to be elucidated. We have studied the effect of modified lipoproteins on monocyte adhesion to human umbilical vein endothelial cells. Therefore, a high capacity adhesion assay system using BCECF-labeled monocytes was developed. TNFo~ increased human monocyte adhesion 7-fold after 6h and remained high for at least 18 h. The human monocytic THP-1 cell line or endothelial cells were treated with modified lipoproteins which are suggested to contribute to the development of atherosclerosis, probably via increasing monocyte adhesion. Neither LDL, nor Fe-or Cu-oxidized LDL and acetylated-LDL were able to increase monocyte or endothelial cell adhesiveness significantly. Furthermore we tested the hypothesis whether induction of monocyte differentiation with phorbol esters for 72 h may change their adhesiveness. A short-term 2-fold upregulation was observed during the first 6-12 h, but then declined back to starting level within the next 12 h and sustained for the next 60 h. Cholesterol loading of PMA-differentiated THP-I cells with acetylated-LDL was also ineffective to increase THP-1 adhesiveness. The obtained data reveal that in our system the adhesion of monocytes appears not to be stimulated by native or modified lipoproteins. Human promyelocytic leukemia cells (line HL-60) were induced by dimethylsulfoxide (DMSO) to differentiate to granulocytic cells with concomitant loss of proliferation capacity. In addition, increases of superoxide anion production and of acid phosphatase activity were observed. Furthermore, decreases in the expression of c-tnyc and c.rayb protooncogene mRNA were found by northern blot analysis. These results were compared to effects after the treatment with the anthracycline antitumor antibiotics doxorubicin and aclacinomyein A. For this comparison the cells were incubated with nearly equitoxic concentrations of the drugs. While aelacinomyein A also showed the ability to induce the differentiation into myeloid direction with all features seen with DMSO, there was no similar effect with doxorubiein. With aclaeinotuyein A decreases of e-rnyc and e-myb mRNA levels, comparable to that, seen after DMSO, in the intensity as well as in the time course, were observed. On the other hand, even after extremely toxic concentrations of doxorubiein, no changes in c-myc and c-myb expression were detected. These findings suggest that the modulation of the e-myc and c-tnyb oncogene expression by aclaeinomyein A is related to its DMSO-like differentiation inducing capacity, independent of general eytotoxieity, and may therefore be a characteristic effect for this way of differentiation. Similar studies will be undertaken with other antitumor agents; such studies may have therapeutical consequences. The immunosuppressive drugs cyclosporin A and FK 506 interfere with calciumdependent signal transduction pathways which regulate gene transcription in Tcells. The complex of immunophilins and cyclosporin A or FK 506 inhibits the calcium-binding phosphatase calcineurin. This inhibition is assumed to prevent nuclear translocation of a subunit of the transcription factor NF-AT and, thus, interieukin-2 gene transcription and T-cell activation. The effect of cyclosporin A and FK 506 on membrane depolarization-induced gene transcription in pancreatic islet cells was investigated. Previously it was shown that membrane depolarization and calcium influx activates CRE-mediated gene transcription in pancreatic islet cells via calcium/calmodulin-dependent protein kinase II phosphorylating and activating the CRE-binding transcription factor CREB. After transient transfection of a glucagon-reporter fusion gene into pancreatic islet cell lines, both cyclosporin A and FK 506 inhibited glucagon gene transcription induced by membrane depolarization and calcium influx, FK 506 being more potent than cyclosporin A (IC 50 of about 0.5 nM or 50 nM respectively). These drugs also inhibited gene transcription when an oligomerized glucagon CREreporter construct was used or after stimulation with 8-Br-cAMP. In accordance with the notion that CREB is the cognate transcription factor binding to the CRE, the transcription induced by the transactivating domain of CREB fused to the DNA-binding domain of GAL4 was inhibited by cyclosporin A and FK 506. Rapamycin, a FK 506 related compound, antagonized the inhibitory effect of FK 506 but not cyclosporin A, suggesting that FK 506 and cyclosporin A may act through complex formation with their distinct intracellular immunophilins. The inhibitory effects of cyclosporin A and FK 506 were prevented by overexpression of the calcium-binding phosphatase calcineurin. These results obtained in pancreatic islet cells suggest that CREB-and CRE-mediated gene transcription may depend on calcineurin phosphatase activity. The molecular mechanism of action of cyclosporin A and FK 506 underlying their pharmacological properties and toxicity may thus include the inhibition of CRE-mediated gene transcription. 

Male Wistar rats were injected intracerebroventdculady (i.c.v.) with ANG II (10 ng, n=10) or physiological saline (n=10) as control and sacdfied 15 min later. Brain areas known to obtain AT1-or AT2-receptors such as midbrain, brainstem, hypothalamus and caudate nucleus were removed. Total RNA was extracted and cDNA were reverse transcribed. The cDNA corresponding to c-los mRNA was then amplified via polymerase-chainreaction (PCR). PCR products were analysed with slot-blot and the bands measured densitometricelly. In the midbrain and in the brainstem, ANG II induced a significant increase of c-fos mRNA when compared with control. In the hypothalamus and in the caudate nucleus, no effect was observed. Glucose transporters (GLUT1-4) represent a family of homologous membrane proteins exhibiting a remarkable tissue specificity of expression. GLUT1 and GLUT4 bind the dlterpene forskolin with hi.gl~ affinity and can be labeled covalently with its photoreactive, iodinated derivative ~IAPS-forskolin; the binding is specifically intfibited by glucose and cytochalasin B. We have previously shown by tryptic cleavage experiments that the binding site of forskolin is located in the C-terminal third of the glucose transporter, probably in the proximity of the membrane spanning helix 9. In order to further localize the binding site, the tryptophane residues W388 and W412 in the glucose transporter GLUT1 were altered to leuclne by sitedirected mutagenesis. The mutant eDNA was ligated into an expression vector driven by the cytomegalo'drus promoter (pCMV), and introduced into COS-7 cells by the DEAE-dextran method. After 48 h, celis were homogenized and plasma membranes were isolated by differential centrlfugation. In these membranes, glucose transport activity was assayed after solubilization and reconstitution into lecithin liposeines. Binding of forskolln was assessed by photulysis of membranes in the presence of zzSiAPS-forskolin, electrophoretic separation, and autoradiography. Immunoblotting with anti-GLUT1 antiserum indicated that comparable numbers of glacose transporters were present in plasma membranes from cells transfected with wild-type GLUT1, GLUT1-L388 or GLUTl-IA12. Transfection of the wild-type GLUT1 gave rise to a three-fold increase in the reconstituted glucose transport activity recovered from plasma membranes. Transfection of GLUT1-L388 produced a 70% lower increase than that of the wild-type GLUT1, whereas transfection of GLUTI-IA12 failed to increase the reconstituted transport activity. Photolabeling of the GLUT1-L412 with iZalAPS-forskolln was not different from that of the wild-type GLUT1, whereas the GLUT1-L3gg incorporated 70% less photolabel than the wildtype GLUTI. Thus, transport activity and forskolin binding were similarly reduced (70%) in the mutant GLUT1-L388. In contrast, GLUT1-L412 exhibited normal forskolin binding in the absence of any detectable transport aetlvity. These data suggest a partial dissodation of the binding sites of forskolin and glucose in the GLUT1. Whereas both tryptopbanes W388 and W412 appear indispensable for the function of the transporter, only W388 is involved in the binding of the inhibitory llgand forskolin.

Institut fftr Pharmakologie und Toxikulogie der RWTH Aachen, Wendllngwcg 2, D-5100 Aachen, and Institut fiir Pharmakologie und Toxikologle der FU Berlin, Thielallee 69-73, D-1000 Berlin, FRG. It is generally believed that insulin receptors are derived from a single gene generating two alternatively spliced mRNA species. Employing a partial receptor agonlst, the covalently dlmerized insulin derivative B29,B29'-suberoyl-insulln, we have prevlously demonstrated a heterogeneity of the signal transduction by insulin receptors in two cell systems (Weiland et at., Prec. Natl. Acad. SCI. USA (1990 USA ( ) 87:1154 USA ( -1158 . Therefore, the present study was designed to further characterize a possible heterogenelty of insulin receptors in different rat tissues. Binding of zz~I-insulin to insulin receptors and its inhibition by unlabeled insulin or B29,B29'-suberoyl-insulin was assayed in plasma membranes from brain, spleen, adlpoeytes and liver. ICs0-values of B29,B29'-suberoyl-insulin were significantly different in all tissues investigated (brain < spleen < adlpoeytes < liver). In contrast, ICs0-values of insulin were identical with the exception of spleen membranes (spleen < brain = adlpocytes = liver). Furthermore, the IC5~ ratios (B29-dimer/insulln) were significantly different ranging from 0.68 (brain) to 12.8 (liver). The apparent molecular weights of the a-subanits of insulin receptors, as labeled with a photoreactive insulin derivative, were indistinguishable in adipoeytes, liver and spleen, but significantly lower in brain. In order to characterize the alternative spficing of the known insulin receptor gene, seven overlapping fragments of the ~-subuult were generated by PCR-amplifieation from eDNA, and were analyzed by digestion with restriction enzymes. Only the two known receptor isoforms (IRA and IRB) were detected in the tissues investigated (brain and spleen, IRA; liver, IRB; adlpoeytes, IRA + IRB). Thus, the tissue-specific expression of IRA and IRB was not correlated with the differences in the ICz0values and -ratios of insulin and B29,B29'-suberoyl-insulin. Moreover, in membranes from cells overexpressing IRA or IRB, the IC~o-ratios were identical. These data demonstrate a high degree of heterogeneity of insulin receptors in rat tissues. This heterogeneity is unrelated to alternative splicing of the known insulin receptor gene.

Institut fiir Pharmakologie und Toxikologle der RWTH Aachen, Wendlingweg 2, D-5100 Aachen, FRG. In recent studies with isolated rat hepatocytes we determined the kinetic characterizations of a renin-lrthiblting hydrophilic linear peptide EMD 56133. We found that the affinity to the "multispecific bile acid transporter" (MT)is very low in contrast to the hy.drophobic linear peptide EMD 51921which shows a rapid biliary elimination via MT. In inhibition studies we found that both peptides are not transported via the same carrier. We used the expression cloning strategy to isolate the carrier protein responsible for the transport of the hydrophilic p_eptide. Injection of total rat liver mRNA in Xenopus laevis oocytes results in a functional expression of a sodium-independent uptake of the hydropttilic peptide. This uptake could be inhibited by addition of tubocurarin, ouabain and urdabeled EMD, In contrast uptake was not inhibited by alanine, chelate and taurocholate. No increase of EMD uptake was measured after injection of the mRNA fraction which encodes for the hepatocellular bumetanide transport protein. The same could be foundafter injection of eDNA clone prLNaBA which s + code for the strictly Na -dependent taurocholate transport protein.

To enrich the mRNA species encoding the peptide transporter total rat liver mRNA was size fractionated on a linear sucrose gradient. Injection of size fractionated mRNA resulted in increase of EMD uptake in a single mRNA fraction.

Institut ftir Pharmakologie und Toxikologie der Justus-Liebig Universitiit, Frankfurterstr. 107, 6300 Giessen, FRG

Norharman (NH), the "simplest" B-carboline, has been found endogenously (Schoutan and Brulnvels, 1986 , Pharmacol. Biochem. Behav. 24: 1219 -1123 . It was stated that it could have a marker function in alcoholic patients, reflecting a long standing metabolic alteration leading to increased formation and/or reduced decomposition. Besides high affinity norharman binding sites, which have been discovered in rat brain (Pawllk and Rommelspacher, 1988, Eur. J. Pharmacol. 147: /_63-171; Pawlkk et al. 1990 , J. Chem. Neuroanatomy 3: 19-24), another class of binding sites in the liver has been found, which successfully could be solubilized from P2 membrane homogenate (0.25% w/v Triton X-100). Displacement studies with uulabled NH revealed an IC50-value of about 5 * 10-9:ram (n=3). The values for K D and Bma x, determined by scatchard plot analysis, were found to be 26-+8nM and 11-+3 pmol/mg protein respectively (n=14). No displacement of NFI binding could be detected using MAOinhibitors (pargyline, brofaromine, harmaline, harmine, Lilly51641; 0.2nM-0.6raM), benzodinzepine llgands like diazepam,/]-CCE and RO15-4513 (0.3nM-301~ r) or the dopamine D 2 ligand haloperidol (/nM-l~alVl). Displacemeut studies with NH derivatives having substituents at different positions, showed that it is important for the speclflty of NH binding that atoms no. 1, 8 and 9 remain free of any substitnant.

The study was supported by the DFG (AZ: HE 916/7-1). The correct characterization of enzyme-substrate or receptorligand interactions inmultiphasic systems, such as biological membranes in an aqueous environment or reversed micelles in a hydrophobic one, requires the quantification of substrate or ligand concentrations in the compartment which is relevant for direct substrate or liga~d access to the enzyme or receptor. Detailed investigations of the mutual relationship between ligand and inhibitor actions on membrane-associated enzymes, based on the analysis of the respective local concentrations of both the ligand and the inhibitor, are missing so far. A strategy is proposed here to derive local inhibition constants for competition of substituted imidazoles (type II ligands) and steroids (type I ligands) for binding to microsomal cytochrome P450c17 (CYPI7) by elimination of contaminating hydrophobic men~rane effects. This approach utilizes the concept that accumulation of small hydrophobic ligand molecules in biomembranes results in a linear dependence of apparent dissociation constants on membrane-lipid concentrations. Accumulation within membrane lipids (expressed as the partition coefficient Kp) rises in the order tinidazole (275) The minimum functional unit of sodium pumps is a heterodimer composed of a catalytic ~r subunit and a E, subunit. The a' subunit has three isoforms (~1, a2, a3), and the f& subunit has, at least, two isoforms (1~1 and f~2). While e isoforms show a remarkable degree of sequence conservation, the two 6 isoforms have only 40% identical amino acids. Using Xenopus oocytes as an expression system, we could recently demonstrate that the adhesion molecule on glia, AMOG, a 1~2 isoform of the mouse sodium pump, is as effective as the I~1 isoform in combining with ~1 subunits to yield a functional el/E,2 sodium pump (Schmalzing et al. JBC 267, 20212, 1992) . A major function of the ~ subunit is to support the correct folding the e subunit. Because of the low sequence identity of the i~ isoforms, the conformational constraints exerted on their assembly partners may be different. To probe conformation, we have determined ouabain binding which occurs with highest affinity to phosphorylated forms of the sodium pump. Four isozymes, al/1~1, e3/~1, el/E,2, and e3/~2 were expressed in oocytes by cRNA injection, cDNAs for el and e3 were from Torpedo californica and rat, respectively, cDNAs for r~l and E,2 from mouse.

Direct evidence for a physical interaction between the different combinations of a and I~ subunits was obtained by immunoprecipitation of a/l'$ complexes with monoclonal antibodies to the It, subunit. Scatchard analysis of the binding data indicates that the i~2 subunit confers about 60% higher affinity for ouabain on sodium pumps than the E,1 subunit. The difference in ouabain binding affinity might reflect a ~ isoform-specific modulation of the structure of the ~ subunit. Alternatively, the f~ subunit itself may constitute a minor portion of the ouabain binding site.

1Max-Planck-lnstitut for Biophysik, Frankfurt/Main, Germany 2Neurobiologie, ETH H6nggerberg, Z0rich, Switzerland. The kinetics of phenacetin-O-deethylation (POD) was investigated in a genetically engineered V79 cell line (XEMd-MZ) expressing rat CYP 1A2 in comparison with freshly isolated rat hepatocytes. The V79 cells XEMd-MZ were incubated in the presence of phenacetin at 9 concentrations ranging from 0.25 to 20 p.M in 2.5 ml DMEM culture medium for 24 hours. Hepatocytes were incubated as suspension cells in 50 ml round bottom flasks with 12 (two rats), 8 (one rat), and 6 (one rat) concentrations of phenacetin ranging from 0.1 to 50 I~M for 15, 30 and 60 minutes. Supernatant medium was kept at -20oC for later analysis. V79 cells expressing CYP 1A2 deethylated phenacetin, whereas parental V79 cells and the V79 dedved cell lines CYP 281 and 1A1 did not. The formation of paracetamol from phenacetin in the CYP 1 A2 expressing V79 cells was linear with time up to at least 4 hours and up to 6 x 105 cells. In isolated rat hepatocytes POD activity was linear with time up to 30 to 60 minutes and up to 2 x 106 cells per ml. The POD activity of XEMd-MZ cells and freshly isolated hepatocytes adhered to Michaelis-Menten kinetics. The apparent Km values were similiar in 4 series of experiments with the XEMd-MZ cells. In hepatocytes the high affinity Krn was slightly lower than in the XEMd-MZ cells. However, the apparent average Vmax found with the XEMd-MZ cells (14.90 pmol/min/106 cells) corresponded to the apparent Vmax for the high affinity site of the hepatocytes (18.12 pmol/min/106 cells). Eadie-Hofstee plots of the kinetics of phenacetin-O-deethylation yielded a biphasic kinetics in the case of hepatocytes. This indicated at least two active catalytic sites. In contrast, XEMd-MZ cells yielded a linear kinetics as they are defined for a cytochrome P450. Therefore, genetically engineered cell lines for expression of cytochromes P450 warrant promising opportunities as precise analytical tools in drug metabolism and toxicity studies. The diversity of G-protein isoforms of tx-, g-and ~,-subunits identified by molecular cloning or chromatographic pro~edures is rapidly growing. On the other hand, there is increasing evidence for specific subunit composition of G-proteins for transducing distinct receptor signals. Thus, we were interested in subunit composition of Gproteins. Therefore, in addition to polyclonal subtype-specific peptide antisera against various oz-and B-subunits, we developed a specific antibody against the C-terminal peptide of Y2 and ~3' This antibody (AS 292) was validated employing recombinant 'i'l-, Y2-and ~i x ' 3 subun ts e pressed in E. coli. Application of the antiserum to purified mammalian lI~(-complexes of various species and tissues showed that this antiserum detected three protein bands. The ~,-subunit of transducin (Y/l), purified from bovine retinae, which showed only one band at 7.8 kDa, was not recognized by AS 292. Comparison of the mobility of the three immunoreactive ~,-subunits detected in mammalian brain revealed that two l[-subunits had lower apparent molecular weight and one had a higher one relative to ~l-mobility. Furthermore, the dominant protein, which migrated closely to 1(1, appeared as doublet bands on SDS-PAGE. Based on the apparent molecular weight on SDS-PAGE, we suppose that these three proteins correspond to the Y2 and Y/3 gene products and to a peptide termed "1(5" on the basis of electrophoretical mobility 0/obishaw et al., 1989, L Biol. Chem., 264, 15758-15761) . In accordance with this finding, a known peptide sequence of "1[5" is of striking similarity to ~2-and l~3sequences. Examining gy/-complexes purified from various species, differences in the composition of the y-subunits detected by AS 292 were demonstrated. Fulther purification of bovine brain g](-complexes on octyl-Sepharose resulted in partial separation of y/-subunits but not of their tightly associated g-subunits. This study supports the concept of high diversity of y/-subunits. Institut fiir Pharmakologie der Freien Universit~t Berlin, Thielallee 69/73, D-1000 Berlin 33 Binding of [3H]GDP to transducin in bovine rod outer segment (ROS) membranes was measured in dim red light and bright white light as well as to purified G protein. In dim red light, binding of [3H]GDP (0.1 gM) was a monophasic and rapid reaction, reaching a plateau after 20 rain of incubation at 25°C. Illumination reduced the initial binding rate by about 60 %. Furthermore, binding equilibrium was observed only upon prolonged incubation. Saturation experiments in dim red light indicated the presence of one population of [3H]GDP binding sites (Bma x -400 pmol/mg protein; K d -100 nM). Upon illumination, the total number of binding sites was not altered, but 70-80 % of the binding sites exhibited a lowei apparent affinity (K d -600 nM). When [3H]GDP (0.1 [£V[) was prebound to ROS membranes in dim red light, illumination caused a rapid release of bound [SH]GDP (approx. 50 % within 10 rain). Most importantly, binding of [3H]GDP to both dark-adapted and illuminated ROS membranes was largely reduced upon pertussis toxin treatment. Furthermore, when illuminated membranes were treated with hydroxylamine, known to inactivate rhodopsin, these membranes exhibited similar [3H]GDP binding characteristics as darkadapted membranes. Finally, purified transducin did not bind [3H]GDP. It is concluded i) that dark-adapted "inactive" rhodopsin interacts with transducin and ii) that, in contrast to illuminated "active" rhodopsin, which causes a GDP-GTP exchange, "inactive" rhodopsin apparently induces a GDP-GDP exchange at transducin. By these characteristics, i) the noise of the signal transduction system is largely prevented and ii) upon illumination rhodopsin can immediately activate apparently prebound transducin. Molecular cloning studies have demonstrated the existence in mammals of five genes encoding distinct muscarinic acetylcholine receptors (mAChRs) which are heterogeneously expressed in the central nervous system and the target organs of the autonomic nervous system. In order to determine the factors responsible for regulation of mAChR expression, we have screened a mouse genomic library and isolated the gene encoding the mouse m4 mAChR with a 12.5 kb 5'-flanking region. The gene contains no introns in the coding sequence, and encodes a polypeptide of 479 amino acids with a high degree of amino acid identity to other vertebrate m4 mAChRs. To verify that this ganomic clone codes for a functional mAChR peptide, the Mm4 mAChR was subcloned into an eukaryotic expression vector and stably expressed in Chinese Hamster Ovary cells. Binding experiments demonstrated that the Mm4 mAChR is able to bind various muscarinic ligands with affinities typical for mAChRs (i.e., Quinuclidinyl bermilate: K D of 22 pM; atropine: K D of 0.5 nM; pirenzepine: K o of 116 riM). In addition, the muscarinic agonist, carbachol distinguishes high and low affinity binding sites ( One major substrate protein was phosphorylated with [3'-32p]GTP in membranes of human leukemia (HL-60) cells. The phosphoprotein comigrated with J~-subunits of heterotrimeric GTP-binding proteins (G proteins) in different gel systems. Upon solubilization of the phosphorylated membranes, the phosphoprotein could be immunoprecipitated by a G protein [$-subunit specific antiserum. The [~-subunit phosphorylation was transient and was found to be specific for GTP and its analogue, guanosine 5'-O-[3'-thio]triphosphate. When phosphorylated membranes were incubated with various nucleotides, the bound phosphate was specifically removed by GDP, suggesting that the phosphate can be retransferred onto GDP. Divalent cations, preferentially Mg 2+ and Mn 2+, were required for both phosphorylation and dephosphorylation. The phosphorylation was stable against treatment with NaOH, but sensitive to treatment with heat, HCI and hydroxylamine. Moreover, treatment of the membranes with the histidinemodifying agent, diethyl pyrocarbonate, resulted in loss in phosphate incorporation. It is concluded that G protein [~-subunits are involved in a guanine nucleotide specific enzymatic activity transferring the 3,-phosphate from GTP to GDP, most likely via a phosphorylated histidine intermediate. In membranes of human leukemia (HL-60) cells the ~-subunits of heterotrimeric GTP binding proteins (G proteins) were phosphorylated by GTP transiently. The covalently bound phosphate could be selectively removed by addition of GDP. When the phosphorylation was performed in the presence of the chemotactic peptide fMet-Leu-Phe (fMLP) two alterations in the time course of the reaction were observed. At very early time points (< 1 min), activated formyl peptide receptors increased the phosphorylation of G protein [~-subunits, while at longer incubation times (> 5 rain) the dephosphorylation was enhanced. In the presence of GDP (the phosphate acceptor substrate), the velocity and the extent of the dephosphorylation process was increased, both in the absence and presence of flVILP. In contrast, addition of either the metabolically stable GDP analogue, Homologous desensitization of fl2-adrenergic receptors ([3AR) is thought to be effected via phosphorylation of C-terminal domains of the receptor by the specific l$-adrenergic receptor Idnase (I3ARK) and by binding of ~arrestin to phosphorylated receptor. In order to study the proposed mechanism of [~AR-desensitization we purified the components involved in this process as recombinant proteins. For that purpose, human ~32AR, 13ARK and 13-arresfin were overexpressed in Spodoptera frugiperda cells (Sf9-cells) using the baculovirus expression system. In the case of I~2AR purification, the cells were harvested and the membrane fraction was solubilized with several detergents, e.g. digitonin, lauroyl sucrose and dodecyl-13-maltoside.The solubilized receptor was purified on alprenololsepharose. I~ARK and 13-arrestin were purified to apparent homogeneity by ion exchange chromatography. Purified ~2AR was reconstituted in lipid vesicles and phosphorylated up to 6 tool Pi/tool ~2AR by purified ~3ARK. The extent of receptor phosphorylation was markedly enhanced (up to 10 fold) by addition of G-proteins (such as Gs and GO) or by resolved ~/-subunits of G-proteins. These results indicate that 13Tsubunits serve as anchor for pARK and facilitate the translocation of the kinase to membrane-bound receptor. Gs was much more potent than the retinal G-protein transducin. In this context we also studied the interaction of the reconstituted components with phosducin, a protein which causes inhibition of GTPase activity of several G-proteins. The potential role of phosducin in the regulation of receptor desensitization will be discussed. We have recently purified phosducin, a phosphoprotein thought to be specifically localized in retina and pineal gland, from bovine brain. Its mRNA was identified in bovirm brain as well as in many other tissues by Northern blot hybridization and by reverse transcripfiordpolymeras~ chain reaction, showing that phosducin is a ubiquitous protein. The eDNA of phosducin was expres~d in bacteria and the recombinant protein purified. We had shown earlier that phosducin inhibits GTPase activity of the three different G-proteins Gi, Go and Os, whereas no inhibition was seen on GTPTS-binding.To study with which G-protein subunit phosducin mainly interacts, crosslinking studies were done with the water-soluble homobifunctional crosslinker, bis(sulfosuccinimidyl)suberate (BS3). "Fnese studies indicate that phosducin has a greater affinity to the [~'~-than to the ~x-subunit of Go. Overall effects in 15-adrenergic signal transduction were assessed with membranes of A431 cells, human epitheloid carcinoma cells. These membranes were stimulated at the receptor, G-protein and adenylyl cyclase level with isoproterenol, sodium fluoride and forskolin and adenylyl cyclase activity was measured with and without phosducin. In these assays phosducin showed a negative regulatory influence at the G-protein level, so that its net effect on l~-adrenergic receptor signal transduction is negative. All inhibitory effects of phosducin were markedly reduced following its phosphorylation by cyclic AMP-dependem protein kinase. For further characterization of the functional role of phosducin the cDNA of phosducin was cloned into an eukaryotic expression vector and phosducin overexpressing cell lines were established. The effects of phosducin on signal transduction at the cellular level were then assayed by measuring the ~receptor simulated adenylyl cyclase activity as well as l~adrenergic receptor desensitization. Recently we reported that the transcriptional activity of the Gin-2 gene is increased by 40% in ventricular nuclei of rats treated with a 96 hours infusion of isoprcnaliue (F.U. Miillcr et ak, Circ Res 1993, in press ). In order to study the time course of this regulation, ran-on transcription assays were performed in ventficular nuclei of rats treated for 12, 48 and 96 h respectively with isoprenaline (ISO, 2.4 ml/kg,d) or with 0.9% NaCI as control (CTR). To test whether this effect is B-adrenoceptor mediated we studied ventricular nuclei of rats treated with ISO in combination with propranolol (ISO+PROP, propranolol 9.9 rag/kgod). ISO for 12 h did not change Gin-2 geue transcriptional activity compared to CTR (ISO 12 h: 31+3 ppm, CTR 12 h: 39±4 ppm G proteins play a central role in various signal transduction cascades and have been implicated in the pathophysiology of human heart failure. However, former studies measuring G proteins/mRNAs used tissue homogenates. Thus, it is not known whether the different G protein c~-subunits are distributed uniformly in the heart or whether the observed changes in G protein expression indeed reflect changes in cardiomyocytes or in non-myocardial cells. To study the physiological distribution of G protein a-subunit mRNA in the heart, in-situhybridization was performed on cryosections (6 #m) of paraformaldehyd-fLxed adult rat hearts, using single stranded $35-or digoxigenin-labeled cRNA probes complementary to the 3'region of Gs~, Gia-2, Gic~-3 and Goes, dipping autoradiography or antibody detection. Specificity of the signals was controlled by parallel hybridization of brain sections, hybridizations using 35S-labeled cRNAs in sense orientation, pretreatment of slides with RNAse A and/or DNase 1 and by studying the distribution of t-myosin heavy chain mRNA for comparison. Results: 1. In the heart, the level of overall abundance is Gsc~> > Giet-2> Gic~-3 > Goa, which corresponds with northern blot data. 2. Labeling of the myocytes did not differ significantly between left or right ventricles, atria or septum. 3. The endothelial layers of the endocardium, epicardium and blood vessels, as well as connective tissue cells, were labeled more markedly than the myocardium. We conclude that allocation of autoradiographic image to the tissue structure by in-situ-hybridization provides important information about G protein localization in the heart, which may be crucial in understanding discrepant results of G protein measurements in the literature. Apparent differences in Gst~-, Gi~-2-, Gia-3-and Goes-mRNA levels between endocardial layer and myocardium are likely to influence measurements of G protein expression in heart tissue, e.g. in endomyocardial biopsies. 

In smooth muscle cells of different origin, prostaglandin (PG) E~ activates the E-type prostaglandin (EP) receptor subtypes EPI, EP2 and EP3. The PGE~ analogue, sulprostone, activates EP1 and EP3 receptors. The stable prostacyclin analogue, iloprost, activates prostacyclin (IP) receptors and EPI receptors. In human thrombocytes, PGE 2 and iloprost activate adenylyl cyclase, presumably through the heterotrimeric GTP-binding protein (G-protein) Gs. Sulprostone inhibits iloprost-induced cAMP formation. We used membranes from the human erythroleukemia cell line, HEL, as a megakaryocytic model to study the effects of PGs on the activation of G-proteins. lloprost, the selective IP receptor agonist, cicaprost, PGE 2 and sulprostone increased high-affinity GTPase activity. The increase in GTP hydrolysis induced by iloprost, cicaprost and PGE~ was only partially pertussis toxin-sensitive, and the one induced by sulprostone was almost completely pertussis toxin-sensitive. All PGs stimulated cholera toxin-catalyzed [~P]ADP-ribosylation of immunologically identified ~-subunits of G c proteins. In addition, they stimulated incorporation of the photoreactive GTP analogue, [e-~zP]GTP azidoanilide, into these proteins. Furthermore, iloprost, cicaprost and PGE~ stimulated incorporation of [:-~zP]GTP azidoanilide into esubunits of G^-proteins. We suggest that (1) PGEz activates Gcproteins v~a EPI and EP3 receptors and G~-proteins via EP2 receptor, (2) sulprostone activates Gi-proteins via EPI and EP3 receptors and that (3) In biochemical studies using membrane preparations or reconstituted systems it has been shown that G-protein coupled receptors can activate the respective G-protein in the absence of an agonist ("empty" receptor). The aim of our study was to search for functional evidence of G-protein (G~) activation by empty 6-adrenoceptors in intact cardiomyocytes isolated from guinea-pig ventricles and from human atrial and ventricular tissue. Ica was measured by the whole-cell voltage clamp technique and used as an indicator of B-adrenoceptor coupling to G s. The signal transduction pathway was amplified by the adenylyl cyclase activator forskolin in a concentration of 0.5 ~M which enhanced Ica two-fold. The 6-adrenoceptor blockers atenolo[ and propranolol were used to test functional activity of the empty receptor. In guinea-pig myocytes, atenolol reversibly reduced forskolin-enhanced Ica in a concentrationdependent manner: i0 nM for threshold, 50 nM for half-maximum and 300 nM for maximum inhibition (50% of stimulated Ica ) . Atenolol reduced Ica at all test potentials o~ the current-voltage @elationship. The reversal potential and steady-state inactivation were not affected, i0 ~M of atenolol were required for significant reduction of Ica in the absence of forskolin.

Propranolol also reduced forskolin-enhanced Ic~ in the absence of a S-adrenoceptor agonist. Similar results were obtained in human atrial and ventricular mocytes. We conclude that the empty B-adrenoceptor can stimulate Ica and therefore, agonist-free, G--protein cou~ed receptors can be active in intact cells. This phenomenon could play a role in the development of B-adrenoceptor hypersensitivity during therapy with S-adrenoceptor blockers.

Pharmakologisches Institut, Universit~t -GHS -Essen, Hufelandstra~e 55, D-4300 Essen i, FRG 220 LABELLING OF THE STIMULATORY G PROT~N GS¢ SY 2',3'-D~ALV~lv~GTP

We have probed the guanine nucleotide binding pocket of G protein --subunits by employing the 2',3'-dialdehyde analogue of GTP. The dbose ring of GTP was opened by oxydation with pedodate to yield oGTP. This analogue is capable of inhibiting the binding of [3sS]GPTyS to the purified recombinant stimulatory G Protein t~-subunit (rG,,) with an affinity comparable to GTP. Similarly, the association rate of [~t-nP]oGTP binding to rG,, proceeds with a rate identical to that obtained with GTP. In contrast to GTP, oGTP binds to rG,~ in a quasi-irreversible manner since dissociation can only be achieved by denaturation of the protein. In the presence of oGTP, rG,, does not activate adenylyl cyclase in the $49 eye-reconstitution assay; this inhibitory effect of oGTP cannot be overcome by subsequent addition of GTPTS or GTPTS +isoproterenol; oGTP, however, is not an inactive nucleotide, since, in $49 eye-membranes, adenylyl cyclase activation by rG,,R187E, a GTPase-deficient mutant of G,, is observed in the presence of oGTP. The rate of adenylyl cyclase deactivation (0.03 min 4) following stimulation by oGTP-liganded rG,,R187E corresponds to the mutant's catalytic rate of GTPase activity. Covalent incorporation of oGTP into rG,, is achieved by borhydrid reduction. Based on the primary sequence of G~,, 7 fragments of distinct size are predicted to occur upon cyanogen bromide-(CNBr)-cleavage. Covalent incorporation of oGTP into rG,, followed by CNBr-eleavage results in an altered peptide map as assessed by HPLC and eleetrophoresis on modified Sch~gger-Jagow gels. These results show that (i) the guanine nucleotide binding pocket of G protein ~t-subunits contain lysine residues which interact with oGTP via Schiff-base formation; (ii) based on its unique binding kinetics, oGTP can only support activation during a single turnover. These properties make oGTP a versatile tool for disrupting signalling pathways that are controled by GTP-binding proteins. 

Addition of the 2',3'-dialdehyd analogue of GTP, oGTP to various membrane preparations results in concentration-dependent inhibition of adenylyl cyclase activity. This effect persists even after removal of oGTP by centrifugation followed by repeated washing and can be attributed to irreversible binding of oGTP to the ~t-subunit of the stimulatory G Protein G,,. We have used human platelet membranes to study receptor-G protein and G protein-effector regulation following disruption of the endogenous pathway. Human platelets are endowed with Az-adenosine and prostacyclin receptors which stimulate adenylyl cyclase and only express the long form of G,, (G,~_0 as determined from immunoblots using a G,, specific antiserum. Hence, the physiologic activation of adenylyl cyclase by these receptors can only proceed through interaction with G~,. L. Following oGTP pretreatment, adenylyl cyclase activity can be restored by the addition of purified recombinant G~; both the short and long splice variants, rG,,.~ and rG~,,.L, are capable of stimulating the platelet adenylyl cyclase in a concentration-dependent manner. Whereas both rG~,,., and rG~,,_ L are virtually identical in the $49 cyc-reconstitution assay, rG~,_ L is about 10-fold more potent in platelet membranes. In contrast, the increment in adenylyl cyclase activity which is produced by stimulation of the platelet A2-adenosine receptor is comparable for both rG,~., and rG~,.~. These findings document that (i) oGTP can be used to delineate the subtype specificity in G protein-controled pathways; (ii) the A~-adenosine receptor couples with similar efficacies to G~.~ and G,~.L. (iii) The platelet adenylyl cyclase appears unique among the known types of adenylyl cyclase as it discriminates between two forms of G~,, thus pointing to a functional role of alternative splicing. Neuropeptide Y, angiotensin II and noradrenaline are potent intrarenal vasoconstricting neurotransmitters and hormones. We have compared their ability to stimulate the phospholipases C and A~ in rat renal slices. Phospholipase C activation was determined as formation of [~H] inositol phosphates in [~H] inositollabelled slices during a 45 min incubation in the presence of i0 mM LicI to block inositol phosphate degradation. Phospholipase As activation was determined as formation of prostaglandin E= during a 30 min incubation and assessed by a radioimmunoassay. Noradrenaline concentration-dependently (ECso 3 ~M) stimulated inositol phosphate formation via activation of both ~-and ~-adrenoceptors but i00 ~M noradrenaline did not enhance prostaglandin E= generation. Angiotensin II (i ~M) did not stimulate inositol phosphate formation but enhanced that of prostaglandin E=; this was not blocked by 1 ~M of the ATe-receptor selective antagonist losartan. Neuropeptide Y (i ~M) did not significantly affect basal and did not potentiate noradrenaline-stimulated inositol phosphate formation. In contrast neuropeptide Y enhanced prostaglandin E= generation although somewhat less effectively than angiotensin II. We conclude that the intrarenal vasoconstrictors neuropeptide Y, angiotensin II and noradrenaline activate different signaling pathways. Moreover, activation of phospholipase A2 by angiotensin II and neuropeptide Y does not appear to occur secondarily to phospholipase C activation. Rat renal cortex contains ~i~-and ~-adrenoceptors in approximately similar density. We have compared their coupling to inositol phosphate (IP) formation in renal slices labelled with [~H]inositol using column chromatographic detection.

In [~H] prazesin competition binding studies methoxamine was 46-fold ~-selective whereas noradrenaline (NA) did not exhibit detectable selectivity for known ~-adrenoceptor subtypes. Despite the ~A-selectivity of methoxamine, the ~-selective antagonists 5-methylurapidil and (+)-niguldipine inhibited i00 ~M methoxamine-stimulated IP formation by less than 50% with high affinity whereas inactivation of ~-adrenoceptots by chloroethylclonidine (CEC) treatment (i0 ~M, 30 min, 37 ° ) reduced it by 54%. Similar data were obtained with noradrenaline. It has been proposed that renal ~-adrenoceptor-stimulated IP formation occurs secondary to Ca ~+ influx (J. Biol. Chem. 265: 17601, 1990 ). However, chelation of extracellular Ca =~ by 5 mM EGTA did not inhibit noradrenalinestimulated IP formation in CEC-treated renal slices. Forskolin has been reported to inhibited noradrenaline-stimulated IP formation in rat kidney (Eur. J. Pharmacol. 148:441, 1988) . We found that inhibition of noradrenaline-stimulated IP formation by CECtreatment and 20 ~M forskolin was additive indicating that forskolin primarily inhibits the ~-adrenoceptot response. We conclude that rat renal ~-and ~adrenoceptors both couple to IP formation but their coupling efficiency may be different. Adrenaline and NPY mobilize Ca ~÷ from intracellular stores in HEL cells in an apparently inositol phosphate independent manner but this Ca 2÷ mobilization abates rapidly (Am. J. Phyisol. 255:E880, 1988; J. Biol. Chem. 264:4986, 1989) . Cross-desensitization experiments with adrenaline, moxonidine and NP¥ revealed that desensitization occurs with phenylethylamine and imidazoline agonists at ~-adrenoceptors and involves homologous and heterologous components.

Although protein kinase C activation by phorbol ester rapidly desensitized adrenaline-and NPY-mediated Ca 2÷ increases, the protein kinase C inhibitors staurosporine (i ~M), calphostin C (i ~M) or H8 ((N-[2-(methylamino)-ethyl]-5-isoquinolinesulfonamide 2HCl, i00 ~M) had only weak and inconsistent effects. In contrast, Ca ~÷ mobilization was enhanced by a 24 h treatment with the tyrosine kinase inhibitor herbimycin (I BM). In Ca =÷ containing buffer, the Ca ~÷ store modulators ryanodine (i0 ~M) and thapsigargin (50 riM) enhanced hormone-stimulated Ca =~ elevations. In Ca=÷-free buffer (5 mM EGTA added) thapsigargin partially inhibited adrenaline-and NPYstimulated Ca ~* increases. We conclude that multiple protein kinases may be activated following stimulation of HEL cell ~=A-adrenoceptors and NPY receptors but that protein kinase C activation does not carry a dominant part in the rapid desensitization. Ca 2÷ mobilization appears to occur from a thapsigarginsensitive store via a ryanodine-rather than inositol-l,4,5-trisphosphate-sensitive channel. Analogues of S-prenylated cysteine like N-acetyl-S-trans,trans-farnesyl-Lcysteine (L-AFC) have previously been shown to inhibit the carboxyl methylation of proteins carrying a C-terminal S-prenylated cysteine residue and to block the endotoxin-activated serum-, but not the phorbol esterelicited chemotaetic response of mouse macrophages. As the ~/subunits of heterotrimeric signai-transducing guanine nucleotide-binding proteins (Gproteins) are posttranslationally modified at their C-termini by methyl esterification, we have determined the effect of L-AFC on formyl peptide and complement C5a receptor-mediated G-protein activation using myeloid differentiated HL-60 cells as a model system. While investigating the effects of treating intact HL-60 cells with L-AFC, we found that the compound interfered with receptor-mediated G-protein activation even when added to membrane preparations. L-AFC inhibited basal and receptor-stimulated binding of guanosine 5'-O-(3-thiotriphosphate) (GTP[S]) to and hydrolysis of GTP by HL-60 cell membranes. Receptorstimulated GTP[S] binding and GTP hydrolysis were more sensitive to L-AFC inhibition than basal G-protein functions. Inhibition of receptormediated G-protein activation was also observed for N-acetyl-S-all-transgeranylgeranyl-L-cysteine, S-trans,trans-famesyl-3-thiopropionic acid, but not for N-acetyl-S-trans-geranyl-L-cysteine, N-acetyl-L-cysteine, or the methyl ester of L-AFC, suggesting that the isoprenyl moiety (C~5 or C2o ) and the carboxyl group, but not the peptide bond of the isoprenyl cysteine compounds were required for inhibition. Most interestingly, we found that exogeneous S-adenosyl-L-methionine was apparently not required for and S-adenosyl-L-homoeysteine did not attenuate the inhibitory action of L-AFC. Taken together, our results suggest that isoprenyl cysteine compounds inhibit the formyl peptide and complement C5a receptormediated activation of G-proteins in granulocyte plasma membranes, but indicate that this inhibition is unlikely to be due to inhibition of protein carboxyl methylation. a16, a member of the tXq subfamily of G-protein u subunits, was recently identified in human hematopoietic ceils. The expression of al6 decreases markedly during hematopoietic cell differentiation, raising the distinct possibility that this a subunit is involved in regulation of cell growth and/or differentiation. In order to elucidate the function of this novel c~ subunit, we cloned and mutagenized its cDNA to obtain a constitutively active tx subunit (txlt-R186C). Cultured mammalian (COS-l) cells were transfected with both wild-type (WT) and mutant cDNAs. Expression was confirmed by immunoblotting using a rabbit antiserum raised against the C-terminal decapeptide of tx16. Cells transfected with the cDNAs of C~q-WT and the constitutively activated mutant aQ-R183C were analyzed ftr comparison. In agreement with earlier results, 6xpression of aQ-R183C led to a = 3.5-fold increase in inositol trisphosphate formation and expression of t~-WT was without effect, The constitutively activated mutant a16-R186~ caused a two-fold increase in the formation of inositol trisphosphate in intact COS-1 cells, while the wild-type t~16 subunit had no effect. These data suggest that tx16 , similar to other members of the family, is involved in coupling cell surface receptors to stimulation ~ phospholipase C. To further analyze the molecular mechanisms by which txl6 is coupled to stimulation of phospholi_pase C we expressed a~6-WT and t~lt-R186C cDNAs in Escherichia coli usin.g the T7 RNA polymerase/promoter system. The recombinant proteins were purified from soluble fractions of bacterial lysates and reconstituted with phospholipase C-t1 and phospholipase C-t2, which were transiently expressed in COS-1 cells. It is expected that this approach will allow to analyze in detail the molecular mechanisms by which activated tx~6 stimulates inositol phosphate formation in human hematopoietic cells. We have recently reported that human HL-60 granulocytes contain a phosphoinositide-specific phospholipase C (PLC) that is stimulated by free ~3,-subunits of heterotrimefic signal-transducing guanine nucleotidebinding proteins (G-proteins) and have subsequently identified this PLC as PLC~2. Here, we have addressed the questions (i) whether the receptormediated activation of the retinal G-protein transducin results in stimulation of PLC/~ and (ii) whether this stimulation proceeds via f3~/subunits. To this end, we have reconstituted the light receptor rhodopsin, the purified a and fl~ subunits of the retinal G-protein transducin, and recombinant PLC/~z and examined the effects of receptor activation on the hydrolysis of phosphatidylinositol 4,5-bisphosphate supplied as an exogeneous PLC substrate. G-protein-depleted retinal rod outer segments (ROS) and lysates of COS-1 cells transfected with the eDNA of PLCfl~ were used as sources of light-activated rhodopsin and PLC/3z, respectively. As reported before, PLC~ 2 was markedly stimulated by the free /~'t, but not by heterotrimeric transducin (at.Gt~,'/~3't). Addition of ROS and the poorly hydrolyzable GTP analogue guanosine 5'-0-(3thiotriphosphate) to PLC/~2 and a t GD~"/~Tt led to a marked stimulation of inositol phosphate formation by PLCfl2. This stimulation was markedly attenuated upon chemical deactivation of rhodopsin with hydroxylamine and was not observed when adenosine 5'-O-(3-thiotriphosphate) (ATP[S]) rather than GTP[S] was present in the incubation medium. ROS activation of PLCfl2 in the presence of GTP[S] closely correlated with ROS stimulation of high affinity binding of GTP[S] to transducin and was prevented by pretreatment of transducin with pertussis toxin and NAD +. Taken together, our results suggest that activation of PLCfl2 by G-protein ~3,-subunits may be an important mechanism by which receptor-mediated activation of G-proteins leads to enhanced inositol phosphate formation. We have recently reported that phospholipase C-B2 (PLC/~2), a novel member of the family of phosphoinositide-specific phospholipases C, is markedly stimulated by G-protein ~'~ subunits. The eDNA of PLC~2 has been cloned from a human HL-60 leukemia cell eDNA library and encodes a protein consisting of 1181 residues. PLC/~2 is homologous to other PLC isozymes within two domains designated X (R253-L501) and Y (E541-M797). The unique mode of regulation of PLC/~ 2 prompted us to investigate the structure-function-relationships of this enzyme by sitedirected mutagenesis of its eDNA. The mutant PLC~ 2 polypeptides were transiently expressed in cultured mammalian (COS-l) cells. Detergent lysates of transfected cells were immunochemically analyzed for PLC/~ 2 expression and assayed for PLC activity using exogeneous phospholipid vesicles containing phosphatidylinositol 4,5-bisphosphate as substrate: (i) A chimeric PLC/~ consisting of the N-terminal 676 residues of PLC~2 and the C-terminal 842 residues of PLC~ 1 was expressed at levels comparable to the wild-type PLCs, but showed no catalytic activity. (ii) Two N-terminally truncated forms of PLCB 2 lacking the first 86 and 235 residues, respectively, were not expressed in COS-1 cells. (iii) A C-terminally truncated form of PLCfl21acking the last 176 amino acids was catalytically active and stimulated byboth Ca 2+ and B~" subunits. (iv) Finally, a deletion mutant of PLC/~ 2 lacking most of the C-terminus distal to domain Y (PLCB2A(F819-El166)) was indistinguishable from the wild-type enzyme in terms of the level of expression in COS-1 cells and the degree of stimulation by Ca 2+ and/~3' subunits. Our results suggest that the Y domain of PLCB~. is involved in the catalytic functions of the enzyme, that the Nterminus is required for PLCB2 expression, and that the C-terminal portion of PLCB2 is required neither for phosphoinositide hydrolysis nor for B'~ subunit stimulation. Thus, the binding sites for Ca r.+, the phospholipid substrate, and G-protein/~, subunits reside in the N-terminal two thirds of PLCB 2. The 813'1 and Bly2 dimers of heterotrimeric signal-transducing guanine nucleotide-binding proteins (G-proteins) were overexpressed in baculovirus-infected insect cells using the baculovirus replacement vector pAcUW51. The B and 3' subunit cDNAs were inserted into regions of this vector controlled by the polyhedrin and the pl0 promoter, respectively. Both promoters are active during the very/ate phase of infection. The B7 complexes were overexpressed by infecting Spodoptera frugiperda (Sf9) insect cells with the recombinant baculoviruses, detergent-solubilized from particulate fractions of the infected cells, and purified to near homogeneity. Recombinant BlYx and B13'~ dimers carrying a mutation known to block y subunit isoprenylation and membrane attachment of r7 dimer (Bxy~C71S and Bx3,2C68S) were produced using the same protocol, except that these By dimers were purified from soluble fractions of infected cells and that the purification was performed in the absence of detergents. Both wild-type and mutant Yx and q,2 subunits formed tight complexes with the 81 subunit as evidenced by co-sedimentation upon sucrose density gradient centfifugation and segration as B't dimers upon purification. The wild-type, but not the mutant ~, subunits were modified by isoprenylation, as shown by the incorporation of [3H]mevalonate into the former, but not into the latter pelypeptides. The ability of the different wild-type and mutant By subunits to interact with G-protein c, subunits and effectors was analyzed by examining their effects on the pertussis toxin-mediated [nP]ADP-ribosylation of a subunits and on the activity of phosphoinositide-specific phospholipase C-B2. It is expected that this approach will allow to precisely map the structural elements of B and 7 subunits required for the interaction with each other and with other transmembrane signalling components. In the bovine adrenal gland, neuropeptide Y (NPY) is co-stored with catecholamines. Activation of nicotinic acetylcholine receptors (nAChR) is the stimulus for the secretion of catecholamines and NPY. Exogenously applied NPY depresses the nicotine-induced hormone secretion (Higuchi et al., J. Pharmacol. Exp. Then, 244:468, 1988) . Nicotine-evoked release is associated with the activation of inward depolarizing cationic currents and a concomitant rise in internal calcium levels. In whole-cell patch clamp experiments, NPY (0.2, 2, 20 ~tM) reduced in a concentration dependent manner inward currents through nAChR-cham~els. Voltage-dependent Ca 2+-, Na +-, and K÷-charmels were not affected. Peptide YY (PYY), a 70% sequence homologue of NPY, as well as C-terminal peptide fragments , NPY(18-36)] were also active. The rank order Of potency was NPY, NPY(16-36), NPY(18-36) > PYY > [Leu 3I , Pro34]NpY, compatible with the existence of NPY-receptors of the Y3 subtype on bovine chromaffin cells (Michel, Trends Pharmacol. Sci., 12:389, 1991) . In the cell-attached mode of the patch clamp teclmique, it is possible to record nAChR activity under conditions, where the charmels are isolated from the bath-milieu outside the patch pipette. NPY applied to the bath decreased the open probability of nAChR-channels, indicating that diffusible second messengers must be involved in the transduction between Y3receptors and nAChR-channels. Accordingly, extracellular pertussis toxin pretreatment (500 ng/ml) and intracellular GDP-B-S (200 p.M), both known to inactivate G-proteins, abolished the intfibitory action of NPY on nicotinic currents. As well, including adenosine 3',5'-cyclic monophosphate (cAMP, I00 ~tM) in the patch pipette prevenled the inhibitory action of NPY. q~ae same was true for membrane permeant cAMP analogues (8-Br cAMP, dibutyryl cAMP, 10 and 100 ~tM, respectively) and the phosphodiesterase inhibitor 3-isobutyl-l-methyl-xanthine (IBMX; 100 p.M). In contrast, inhibitors of the cyclic AMP dependent protein kinase A, synthetic peptide fragments of the endogenous sceletal muscle protein kinase inhibitor [PKI(5-24), PKI(14-24) amide, 1 and 10 I.tM, respectively; Cheng et al., J. Biol. Chem., 261:989, 1986 ] markedly increased the NPY-induced inhibition of nicotinic currents. It is suggested that the co-hormone NPY inhibits catecholamine secretion by interfering with the cAMP-dependent phosphorylation of nACtLR-channels, thereby decreasing their reactivity upon cholinergic stimulation. Circulating human lymphocytes or neutrophils are often nsed to investigate biochemical mechanisms underlying psychiatric diseases. In these ceils, like in neurons, intact Ca 2+homeostasis and Ca2+-mobilisation processes are of great importance for cell function and activation. Since intracelhilar Calcium [Ca2+]cmobilization and inositolphosphate (IP) formation are part of the same signal transducfion pathway, we investigated in parallel experiments N-formyl-methionyl-leucyl-phenylalanin (fMLP)-indueed increases in inositolphosphates and [Ca2+]i in human lymphooytes and neutrophils from healthy donors. Additionally, [Ca2+] i after stimulation with phytohemagglminine (PHA) and anti-CD3 was determined. For Ca2+-measurements, ceils were incubated with fura-2aeetoxymethylester (3#tool/l) and stimulated with fMLP (l#mol/l). Fluorescence was measured using a SLM 4800 spectrofluorometer, [Ca:~+]l was calculated due to Orynkiewiecz et al (1985) . For evaluation of inositolphosphate-tunaover, cells were incubated with [3H]-inositol (l#Ci/106 cells) for 24h. In the presence of lithium (lOmmol/l) ceils were stimulated by the addition of fMLP for different times. IPs were separated by anion-exchange chromatography. In neutrophils maximal [Ca2+]i is obtained after -8s, which is 5-10 fold over basal level, t 2+ whereas in lymphocy es, maximal [Ca ] i is only 150% of basal which is achieved after 30-45s. After 300s in both cell types, [Ca2+]i is only slightly increased over basal level (120%). Comparison of IP formation after stimulation with fMLP revealed less pronounced differences. Basal IP 1 levels in nantrophils are lower than in lymphocytes and maximal increase in IP 1 formation is higher in neutrophils. However, these differences are not statistically significant. PHA-induced [Ca2+]t in lypmphocytes is significantly higher than anti-CD3-mediated increase in [Ca2+]i, the smallest increase is induced by fMLP. Determination of cell composition by FACS analysis revealed no individual correlation between response and amount of T-cells. Stimulation of neutrophils and lymphocytes with the same agonist results in second and third messenger formation which differs in kinetic and maximal increase. These differences must be taken into account when human blood cells are used to investigate changes in Ca2+-homeostasis during psychiatric diseases.(e.g. Dubovsky et al., Am.LPsychiat. 149:118-120, 1992 In many cell types, hydrolysis of phosphatidylcholine (PC) is enhanced by receptor-mediated phospholipase D (PLD) activation.

Evidence is emerging from experiments using non-hydrolysable GTP analogues that G proteins control PLD activity (Billah and Anthes, Biochem. J. 269: 281, 1990) . In isolated rat atria, we studied the effect of sodium fluoride (F-) and of magnesium chloride (Mg 2+) on PLD activity; both F-and Mg 2+ have been widely used to study G protein activation. PC as the substrate of PLD was labeled with [3H]-myristic acid. The enzyme catalyzes the formation of choline, labeled phosphatidic acid (PtdOH) and, in the presence of 2 % ethanol, phosphatidylethanol (PETH).

It was found that F-(0.3-10 mM) as well as Mg 2+ (I-50 mM) enhanced the formation of PtdOH and PETH (up to 5-fold) in a concentration-dependent manner. Mg2+-free incubation significantly lowered the resting activity of PLD. The phorbol ester PDB (0.02-0.6 ~M), in contrast to 4~-phorbol-13e-acetate (PAc), slightly reduced the F--evoked formations of inositol phosphates and PtdOH, but enhanced PLD activity as reflected by rises in PETH and choline.

In addition, 0.2 ~M PDB potentiated the effect of F-on PLD activity. It is concluded that PLD activity in heart tissue is directly controlled by a G protein and a protein kinase C ~hrough independent pathways.

Supported by the Deutsche Forschungsgemeinschaft and the German-Israeli-Foundation (GIF) Pharmakologisches Institut der Universit~t, Obere Zahlbacher Str. 67, W-6500 Mainz, Germany PHOSPHATIDIC ACID PHOSPHATASE FROM RAT BRAIN A. H6er, W. I_ache, E. Oberdisse* The cleavage of phosphatidylcholine (PtdCh) by phospholipase D (PLD) to phosphatidic acid (PtdA) and choline has been shown to be receptor-regulated in several tissues. From PtdA diacylglycerol (DAG) is subsequently released by a phosphatidic acid phosphatase (PAP). DAG is known as an intracellular second messenger activating protein kinase C. Therefore, the PAP enzyme may play an important roIe in this signal transduction mechanism. There is only limited evidence that the PAP activity is regulated via membrane receptors. We have investigated some properties of the PAP activity from rat brain membranes, using 32p-PtdA as exogenous substrate which was synthesized from DAG and [y_32p] ATP. The divalent cations Mg 2+ and Ca 2+ were not essential for the enzyme. It is probably a Mg2+-independent PAP which has also been described in rat liver membranes, whereas the membr'aneous PAP from HL 60-cells is Mg2+-dependeut. ZnC12 (100 ~tM) inhibited the rat brain PAP activity completely. The dependence on pH showed a broad optimum between 6 and 7. NaF, which is known to inhibit PAP, showed a half-maximal inhibition of the enzyme at 3 mM NaF. Propranolol has been reported to inhibited PAP activity in whole cell preparations. In contrast to these results we found that propranolol activated or inhibited the enzyme depending on the concentration of propranolol. We also tested the effects of several detergents on PAP activity. Triton X-100, cholate and CHAPS increased the enzyme activity; the highest activation was found at concentrations of 3, 10 and 30 mM, respectively, whereas SDS was inhibitory at concentrations higher than I0 ~tM. The K m value of the enzyme was estimated to be about 150-200 ~tM. These data show that neuronal tissue contains a PAP activity with a substrate affinity comparable to other substrate-specific lipid-hydrolysing enzymes as phospholipase C. ] p), to G proteins and its regulation by fMet-Leu-Phe (fMLP) was studied in membranes of differentiated HL-60 cells. At concentrations exhibiting similar functional activities (inhibition of high-affinity agonist binding to formyl peptide receptors), similar amounts of GTP?S (3 nM) and Gpp[NH]p (30 nM) were bound. For optimal fMLP-stimulated binding of either GTP analogue, both GDP and Mg 2÷ were required with ECso values of ~0.15 and 0.5 ~tM (GDP) and ~20 ~tM and ~100 ~tM (Mg2÷) for Gpp [NH] p and GTPyS binding, respectively. Maximal fMLP-stimulated binding of GTPyS was about 2-fold higher than that of Gpp[NH]p. fMLP-stimulated binding of GTPyS was also observed with Gpp[NH]p instead of GDP, but the extent was reduced to the level of fMLP-stimulated Gpp[NH]p binding. On the other hand, in the presence of both GDP and GTP~S, fMLP reduced binding of Gpp [NH] p. Further differences between the two GTP analogues were observed by measuring the initial rates of fMLP-stimulated binding. In the presence of GDP, fMLP-stimulated binding of GTPyS exhibited a rather long linear increase, whereas fMLP-stimulated binding of Gpp[NH]p rapidly reached equilibrium, which kinetic behavior was also observed upon replacement of GDP by Gpp[NH]p in the GTPyS binding reaction. The data thus provide evidence for an essential advantage of GTPyS over Gpp[NH]p in receptor-stimulated binding to G proteins and suggest that this "additional" reaction, which apparently requires relatively high concentrations of both GDP and mM Mg ~÷, involves a transphosphorylation process which is possible with GTPyS but not with Gpp[NH]p.

Institut for Pharmakologie, Universit~it Essen, Hufelandstr. 55, D-4300 Essen, FRG

In human embryonic kidney ceils (HEK), stably expressing the m3 muscarinic acetylcholine receptor, the muscarinic agonist carbachol stimulated phospholipase D (PLD). This receptor-mediated stimulation of PLD could be mimicked by A1F~" suggesting that G-protein(s) are involved in the regulation of PLD activity. In the present study we have investigated the effects of guanine nucleotides and divalent cations on PLD activity in permeabilized HEK ceils. PLD activity was assayed in [aH]oleic acid prelabelled cells that were permeabilized with digitonin (8 p.M) by measuring ~ the formation of H-labelled phosphatidylethanol (PEt). The GTP analogue guanosine 5'-O-[~/-thio]trisphosphate (GTP~S) caused a concentration-dependent increase in PEt production (ECs0= 7.5 ~tM; maximal increase of about 100% at 100 ~tM GTP~S). Both basal and GTPyS-stimulated PLD activities were inhibited by GDP and guanosine 5'-O-[B-thio]diphosphate (GDPBS; IC~0 values 0.8-3 mM). Mg ~+ ions concentration-dependently enhanced basal and GTP'/S-stimulated PLD activities. Free Ca z÷ was not required for PLD activity, however higher Ca ~+ concentrations Co-0.1 ~tM) inhibited both basal and GTPTS-stimulated PLD activities. Only in the absence of exogenous ATP, which by itself did not modulate PLD activity, adenosine 5'-O-[y-thio]trisphosphate (100 ~tM) increased PEt production by 70% suggesting an involvement of a nucleoside diphosphokinase in PLD activation by the ATP analogue. In conclusion, the stimulation of PLD activity by GTP~S and its inhibition by GDP and GDPBS supports the hypothesis that G-proteins are involved in the regulation of PLD activity. Elevated free Ca ~+ concentration inhibited PLD, which may indicate an inhibitory regulation of PLD by intracellular Ca z+ in this cell type. 

Antigen-induced cross/inking of IgE bound to its high affinity receptor (FceRI) at the surface of basophils or mast cells triggers a number of biochemical events culminating in the release of several inflammatory mediators. Crosslinking of FceRI stimulates tyrosine phosphorylation of several proteins, including two of 42 and 44 kDa. Proteins with identical molecular masses were tyrosine phosphorylated in response to diverse growth promoting agents in other cell types and have been identified as mitogen-activated protein (MAP) kinases. We therefore investigated the possible involvement of MAP kinase in the signalling induced by different secretagognes in rat basophilic leukemia ceils (RBL). Crosslinking of FceRI led to the stimulation of MAP kinase activity, which co-purified on Mono Q chromatography with 42/44 kDa proteins that were tyrosine phosphorylated in response to FceRI crosslinking and which reacted with anti-(MAP kinase) antibodies. MAP kinase activity was also stimulated by the calcium ionophore A23187, a potent receptor-independent secretagogne. Similarly, in RBL cells which express the G-protein coupled muscarinic receptor ml, carbachol increased MAP kinase activity and induced secretion. FceRI crosslinldng-, A23187-and carbachol-induced MAP kinase activation was rapid and reached a maximum after 1-2 rain. The dosedependence of secretagogue-induced MAP kinase activation correlated with that of increases in serotonin release from activated cells. Downregulation or inhibition of protein kinase C as well as incubation of cells with the tyrosine kinase inhibitor genistein markedly inhibited MAP kinase activation in parallel with serotonin release. Taken together, these findings functionally link MAP kinase activation to signalling events leading to _mediator release in RBL cells. The chemotactic peptide, N-formyl-L-leucyl-L-phenylalanine (fMLP), increases the cytosolic free calcium concentration ([CaZ+]~) in human neutrophils with subsequent activation of superoxide (02-) production. To investigate further the role of increases in [Ca2+]~ in fMLP-induced O~production we stud'ed the effects of maitotoxin, a marine toxin from Garnbierdiscus toxicus that activates non-selective cation channels (NSCs) and phosphoinositide hydrolysis, and thapsigargin, a Ca2+-ATPase inhibitor that activates NSCs without activation of phosphoinositide turnover, on [Ca2+]~ in human neutrophils and dibutyryl cAMPdifferentiated HL-60 cells, and 02-production in human neutrophils. Both maitotoxin (50 ng/ml) and thapsigargin (100 riM) produced significant increases in [Ca2+]i in human neutrophils and differentiated HL-60 cells.

The increase in [Ca~+]i due to maitotoxin was fully dependent on extracellular Ca :+, and was substantially inhibited by econazole (10 #M) and 1-{.g-[3-(4-methoxyphenyl) propoxyl]-4-methoxyphenylethyl}-lI-Iimidazole hydrochloride (SK&F 96365; 30 #M),. blockers of NSCs. In contrast the increase in [Ca~+]~ due to thapsigargln was due to both release of Ca 2+ from intracellular stores and entry of external Ca 2+ and was poorly inhibited by econazole and SK&F 96365. Thapsigargin stimulated Mn ~+ entry into differentiated HL-60 cells, but maitotoxin did not. By themselves neither maitotoxin nor thapsigargin stimulated O~-production in human neutrophils. However, they both potentiated the O~-production induced by fMLP. Furthermore, they were both able to prime 02production: in the presence of either maitotoxin or thapsigargin a subthreshold concentration of fMLP was able to stimulate 02-production. These effects of both maitotoxin and thapsigargin were substantially inhibited by econazole. Taken together the above results suggest that in human myeloid cells maitotoxin and thapsigargin activate different NSCs. Furthermore activation of NSCs is important, but not sufficient, for stimulation of O 2" production. So far three splice variants of the aa-gene coding for two ao proteins have been identified by molecular clSning. A third form of G, ("Ga-~") is detectable in mammalian membranes, using subtyt~-spe~£fic antibodies (Spicher et al., FEBS Lett. 307, (215) (216) (217) (218) . In order to get deeper insight into the heterogeneity of G O proteins, we purified them from mammalian brains. Employing a four-step purification protocol resulted in resolution of three ao-SUbtypes. According to the predicted pI values, "~,,~" eluted after a,,9 and a,,1 from a Mono Q column. SurprisingTy, when supplem~e3ted witfi"tl~-complexes, only ~ and ao2 were substrates for pertussis toxin (gl'X)-mediated ADPYn'bosylation whereas "a^o" was not. However, this was not due to an artificial modification during the purification procedures as evaluated by using specific C-terminal antibodies and ADP-fibosylation of membrane-bound proteins with subsequent immunoprecipitation. Since the pl value of "anq" was more acidic compared to ~ol and ang, we speculated whe~er this form represented an endogenously ,gDP-fibosylated G o. However, neither specific anti-ADP-fibose-antibodies nor treatment with HgCI~ supported this assumption. Reconstituion experiments with all thre~ a.o-subtypes revealed different pharmacological properties on calcium cliannel activity in cultured PTX-pretreated human S/-I-SY5Y cells.

Following stimulation with dopamine, somatostatin, the tz-selective agonist [D-Ala2,N-Me-Phe4,GlyS-ol]-enkephalin (DAGO) and the gselective agonist [D-Pen2,PenS]-enkephalin (DPDPE), only un~ mediated inhibitory effects on calcium channel activity. In contra~'ff,, upon carbachol stimulation, both c¢_~ and ao2 were able to transduce inhibition of calcium channel activit) ~. However, following stimulation with various .... agonists neither reconstituted "an,, v3 ",, influenced calcium channel actiwty nor d~d GTP1lS-preactwated Uo3 .

From these experiments we ~onclude that the novel form "t~o3" may be a third gene product. Experiments are underway to eluci-date the peptide sequence. Subtype-specific immunoprecipitation of G-protein ct-subunits photolabeled in the absence or presence of agonists revealed profound differences between p.-and ~5-opioid receptors in coupling to G i-and Go-type G-proteins. Activated ~-opioid receptors preferentially coupled to Gil, whereas activated ~t-opioid receptors more effectively coupled to Gi3. Counting of radioactivity incorporated into the immunoprecipitates of membrane proteins photolabeled in the presence of DPDPE and DAGO showed that DPDPE stimulated photolabeling of Ctil and cti3 by 90% and 28%, whereas DAGO stimulated the incorporation of the GTP analogue into (:til and cti3 by 21% and 177%, respectively.

The relative stimulation of photolabeling of ot.il by DPDPE and of ~i3 by DAGO reached a maximum aider 90 sec and thereatter constantly declined. Both agonist stimulated receptors activated Gi2 to the same extent. Immunoprecipitation experiments with Sol-and ctocommo nantisera suggest that either receptor also differentially coupled to the G osubtypes, Gol and Go2. In conclusion, la-and 6-opioid receptors appear to discriminate between PTX-sensitive G-proteins and lead to activation of distinct G-protein subtypes. . Coincubation of the cells with cycloheximide (100 #g/ml) prevents G protein up-regulation, suggesting that this effect is due to de novo protein synthesis. Reconstitution of $49 lymphoma cyc membranes with cholate extracts verifies that G s from morphine pretreated cells is fully active. On the other hand, studies with Gpp(NH)p, which directly activates inhibitory G proteins at low nanomolar concentrations, indicate that G i displays reduced efficacy to control adenylate cyclase upon chronic morphine exposure. Although both high-and low-affinity #opioids act at the same receptor population and induce a similar degree of tolerance as assessed by the desensitization of adenylate cyclase regulation in SH-SY5Y cells, our data suggest that they induce distinct biochemical changes responsible for the generation of tolerance. While the high-affinity ligands seem to uncouple receptors from their subsequent signal transduction systems, the lowaffinity #-agonists cause qualitative and quantitative adaptations at the level of G proteins. The concomitant regulation of both inhibitory and stimulatory acting G proteins documents their mutual functional interdependence. Neuronal cells and endothelial cells contain Ca2÷/calmodulin-regulated isoforms I and III of nitric oxide synthase (NOS), respectively. So far, these isoforms have been considered "constitutive" enzymes. Other cells such as macrophages can be induced with lipopolysaccharide (LPS) and cytokines (e.g. interferon-~) to express Ca2+-independent isoform II of NOS. In previous studies we found that the low Ca2÷-sensitive NOS activity that is sometimes seen in RAW 264.7 macrophages disappeared when the cells were exposed to LPS and interferon-~. This prompted us to investigate whether the expression of Ca2+/calmodulin-dependent isoforms of NOS is regulated by cytokines. Our cell line of bovine aortic endothelial ceils (BAEC) expresses marked Ca2÷/calmodulin dependent NOS activity (isoform II]), and additional NOS activity cannot be induced with LPS and a variety of cytokines tested. Interestingly, the Ca2*-dependent NO synthesis of intact BAEC treated with tumor necrosis factor (TNF-a, 1000U/ml for 24h) decreased by 30-40% when compared with vehicle-treated control cells. Also, the specific NOS activity of homogenates from TNF-ct-treated BAEC was reduced by about 30%. No Ca2÷-independent (induced) NOS activity was detected under these conditions. Western blot analyses were performed using a specific monoclonal antibody (H32) to endothelial NOS. Identical amounts of cell homogenate from control BAEC or TNF-a-treated BAEC were loaded to run the sodium dodecyl sulfate-polyacrylamide gel electrophoresis, but the specific 135 kDa band recognized by antibody H32 was much weaker in the lane containing the material from the TNF-a-treated BAEC. Northern blot analyses were performed using a radiolabeled 1575 bp bovine endothelial NOS eDNA fragment as a probe, mRNA was extracted from control-and TNF-a-treated BAEC. In both cases, the blot showed hybridization to a 4.8 kb mRNA. However, less mRNA was found to be transcribed in the TNF-a-treated BAEC. Reverse transcriptase polymerase chain reaction using primers based on sequences common to all three known isoforms of NOS detected amplified transcript from control BAEC, but virtually no transc~fpt from TNF-cc-treated BAEC. We conclude that cytokines can downregulate the expression of Ca~+-dependent isoforms of NOS (at least isoform III), and therefore those isoforms can no longer be considered "constitutive" enzymes. Here it is demonstrated that endothelium-dependent relaxations of isolated coronary arteries are considerably more sensitive to inhibition by MB than relaxations induced by direct activators of sGC. Similar data where obtained in the presence of superoxide dismutase (SOD), indicating that the diverse potencies of MB were not due to superoxide-induced inactivation of NO. Subsequent experiments revealed that MB more potently affected NO synthase (NOS) than sGC. Conversion of L-arginine to L-citrulline was inhibited by the dye in a concentration-dependent manner (IC50 = 5.3 ~tM), and concentrationresponse curves were only marginally shifted rightwards (ICs0 = 9.2 ~M) in the presence of SOD. Reactions of MB with the NOS cofactors NADPH, tetrahydrobiopterin, FAD or FMN, assayed as MB-dependent reduction of cytochrome c, were found to be too slow to account for the decreased enzyme activities seen in the presence of the dye. Purified sGC was about one order of magnitude less sensitive towards inhibition by MB than NOS. When the purified enzyme was maximally stimulated with 0.1 mM S-nitrosoglutathione, cGMP formation was reduced by 50 % at -60 ~M MB. Moreover, the effects of MB on sGC were uncomplete (maximal inhibition of 70% at 1 mM MB). These data demonstrate that MB is only a poor inhibitor of sGC and, in the presence of SOD, acts primarily as a direct inhibitor of NOS. According to recent data, NOS is a hemoprotein, so that MB may exert its effects by oxidation of ferrous heme iron, which seems to be involved in the NOS-catalyzed oxidation of L-arginine. Nitric oxide (NO) is synthesized from L-arginine by cytochrome P450like haemoproteins, one of which is calcium/calmodulin-dependent and predominantly localized in neuronal and endothelial cells. The subcellular localization of this enzyme in endothelial cells, however, is still a matter of debate. Therefore, we compared the subcellular distribution of NO synthase (NOS) in bovine (BAEC) and porcine aortic endothelial cells (PAEC) and in endothelial cells cultured from human umbilical cord veins (HUVEC) by (i) monitoring the NG-nitro -L-arginine-sensitive conversion of L-[3H]arginine to L-[3H]citrulline and (ii) by Westernblot analysis with a specific monoclonal antibody raised against NOS from BAEC. In native and cultured BAEC and in cultured HUVEC, NOS activity was largely confined to the microsomal fraction (80-100%), whereas enzyme activity was equally distributed among the cytoso/and membrane fraction in native and cultured PAEC. Total NOS activity was approximately 10 times higher in native PAEC than in native BAEC or cultured BAEC, PAEC or HUVEC. Moreover, in all endothelial cell types tested, specific NOS activity in the microsomal fraction was consistently higher (3-6 times) as compared to the cytosol. The difference in immunoreactivity between the cytosolic and microsomal NOS from PAEC appeared to be even greater (10-15 times), possibly reflecting a difference in conformation between these two enzymes. This notion was supported by the finding that SKF 525A, a cytochrome P450 inhibitor which can also inhibit NOS, was approximately 10 times less effective in blocking NOS activity in the cytosol of cultured PAEC as compared to the membrane fraction. Further subcellular fractionation of native PAEC and cultured BAEC revealed that the majority of the microsomal NOS activity and content was associated with the endoplasmic reticulum and not with the plasma membrane. The functional implications of these findings remain to be elucidated. However, they argue against the concept of a shear stressdependent NO release from endothelial cells triggered by changes in the fluidity of the plasma membrane, thereby affecting NOS activity. Previous studies provided evidence for a contribution of intmcellular depletion in free thiols to the development of tolerance occuring during repeated treatment with higher doses of organic nitrates. Therefore, the content of free thiols (-SH) and of disulphides (S-S-) was measured with a histochemical method in intimal and medial smooth muscles of porcine coronary arteries. After equilibration in oxygenated Krebs-Henseleit solution (37°C, pH 7,4), ring segments with and without endotbelium were treated for 30 min with either vehicle or 100gM of isosorbide-5-mononitrate (IS-5-MN), glyceryl trinitrate (GTN), S-nitroso-D,L-n-acetylpenicillamin (SNAP) or N-(3-Nitmtapivaloyl)-lcysteineethylester (SPM 3672). Paraffin sections of the arteries fixed in formol or bouin solution were treated according to a published method (Sippel, TO 1978, Histochem. J. 10:585-595) and detection of-SH and S-S-was performed by measuring the extinction at 520 nm with a microscope photometer. In the presence of endothelium the content of-SH in controls (n=8) gave an extinction of 0,127±0,013 in the intima, 0,120~-0,010 in the media and 0,124±0,008 in the whole smooth muscles of the vascular wall indicating a homogenous distribution. The corresponding values for S-S-~vere 0,684~4),084, 0,535±0,120 and 0,612±0,073. Removal of endothelium induced a reduction in S-S-in the intima to 63,2a:1,9% and increase in -SH in the media to 143,9±9,1%. Treatment with all nitrates resulted in a similar reduction in -SH in the media and intima of intact and denuded artery segments to 50-75% of the control values. In contrast, S-S-content was less affected but also reduced by IS-5-MN, SNAP and SPM 3672 in intact and by GTN, SNAP and SPM 3672 in denuded arteries. Investigation of vasodilating effects of the nitrates used in the same artery type revealed a development of tolerance only for GTN and IS-5-MN as indicated by the ratios of halfmaximal effective concentrations before and after tolerance induction amounting 14,4+1,7 and 5,4_.+1,3, respectively. SPM 3672 exhibited no cross-toleranee to GTN. Preincubation with captopril induced a significant (p<0,05) reduction in the degree of tolerance development to GTN (6,4+1,1, p<0,05). However, captopril shifted dose-response curves for GTN in non-tolerant vessels to the right and in tolerant vessels to the left and both differences did not achieve statistical significance. Thus, we conclude that nitrate tolerance is associated with but problably not caused by intraeellular thiol depletion. Nitric oxide synthase (NOS) can be induced in several ceils by bacterial lipopolysaccharides (LPS), cytokines and a pleiora of other agents and conditions that exert an oxidative stress. NOS induction requires gene transcription and protein biosynthesis. The factors controlling transcription of the NOS gene are unknown. In many cells one of the nuclear transcription factors -kappa B (NF-kB) -integrates the intracellular signals elicited by oxidative stress by enhancing transcription of certain genes encoding signaling and defense proteins. NF-kB is present in the cytosol in an inactive form that is activated by removal of an inhibitor, IkB. Activation is specifically prevented by thiol compounds and even more potently by iron chelators, such as diethyldithiocarbamate (DETC). To clarify whether NF-kB is involved in NOS gene transcription we studied the effect of DETC on induction of NOS activity by LPS in murine bone marrowderived macrophages (MQ) cultured in L-cell-conditioned medium. Nitrite release (detected by the griess reaction) and NO formation (detected by electron spin resonance (ESR) spectroscopy) were taken as parameters of NOS induction. DE-I'C, which did not directly affect NOS activity in cell homogenates, inhibited induction of NOS (IC50 0.5 mM), if DETC was present during exposure of MQ to LPS. Inhibition was accompanied by the trapping of intracellular free nonheme iron by DETC, as detected by ESR spectroscopy. Addition of heme to MQ did not prevent the effect of DETC, indicating that a lack of heme, an essential cofactor of NOS, was not responsible for the absence of NOS activity. We conclude that NF-kB is involved in the signal transduction pathway that leads to induction of NOS in MQ. Cyclopiazonic acid (CPA), a mycotoxin from Aspergillus and Penicillium, has been described to selectively inhibit the Ca2+-ATPase of the sarcoplasmic reticulum (SR) in smooth muscle (Uyama et al., Br J Pharmacol 106: 208-214, 1992) . Since nitrovasodilators have been suggested to relax vascular smooth muscle via stimulation of the SR-Ca 2 +-ATPase resulting in enhanced Ca2+-sequestration (Lincoln and Cornwall, Blood Vessels 28: 12%137, 1991), we investigated the effect of CPA on coneentrationvasorelaxant response curves (CVRCs) of two nitrovasodilators, sodium nitroprusside (SNP) and isosorbide dinitrate (ISDN), in the organ bath. Helical s~ps of rabbit aortae were incubated with 100 #tool CPA for 30-50 rain and then stimulated with 10 -7 mol/l noradrenaline (NA) or with 80 mmol/l KCI. Upon stabilization of tension, CVRCs were recorded by cumulative addition of the vasodilators. In preparations contracted by NA, CPA shifted the CVRCs of SNP and ISDN to the right; the calculated EC50-values (tool/l, x 10 -8) without / with CPA were 1.46 / 7.4 for SNP and 6.96 / 38.6 for ISDN. In contrast, in the K+-depolarized preparations, the CRVCs of the nhrovasodilators were shifted to the left; EC50values (mol/1, x 10 -7) without / with CPA were 29.2 / 2.26 for SNP, and 87.6 / 3.16 for ISDN. Also, the contractile response to 500/zmol/1 CPA alone could be totally antagonized by a single dose of SNP (3x10 -6 mol/l). -While the nature of the rightward-shift of CVRCs by CPA in NA-stimulated tissues remains obscure, the latter data suggest that activation of the SR-Ca2+-ATPase is not a major process in the chain of events leading to vasorelaxation by nitrovasodilators. (1982) 1187). The aim of the present study was to examine the endothelium dependence of the eateeholamine-indueed relaxation of rat cerebral arteries. Furthermore we attempted to classify the responsible aminergie receptor. Cumulative or singledose response curves on isolated rat basilar arteries with a diameter between 300 and 500 I.tm were performed with NE, isoproterenol (ISO) and terbutaline after preconstrietion of the vessels with 5-HT 3x10 -7 M or K ÷ 40 raM. The maximal relaxing effect after preconstriction with K ÷ 40 mM amounted to 68%, n=23, 40%, n=6 and 13°/0, n=12, respectively. In basilar arteries the presence of nitro-Larginine (L-NA) lxl0 "4 M abolished the relaxing effect of carbachol (CCh) and markedly reduced the relaxation induced by NE and ISO. For example, after preconstriction with K ÷ 40 mM the amount of relaxation by NE lxl0 -~ M before and after application of L-NA was 53,0±3,5% and 8,2±3,2%, respectively (n=5, SEM). Equivalent results were obtained after perfusion of the vessel with Triton X 1:1000 for about 20 see. Also in the medial, the anterior and the posterior cerebral artery the relaxation caused by NE and ISO was reduced by L-NA. The relaxing effects of NA, CCh and glyceryl trinitrate (GTN) were dependent on the concentration of K*. Though the force of constriction produced by K* 40, 80 and 120 mM was comparable, the relaxation induced was stronger at lower Kt concentrations, whereas the relaxation by papaverine proved to be independent on the K÷-cortcentrafion. Methoxamine and the seleetiw c*2-agonist BHT933 had no relaxing effect. The effect caused by NE was inhibited by propranolol, atenolol and CGP20712A (J~l-selective antagonist). But neither the dependence on the antagonist concentration nor the shape of the dose response curves suggest a competitive inhibition. Micromolar concentrations of ICI 118551 (J3z-selective antagonist) did not considerably affect the NE-indueed vasorelaxation. In summary, the NE-induced relaxation is sensible to L-NA or to perfusion with Triton X. Furthermore the relaxing effects of NE, CCh and GTN are dependent on the K÷-concentration. These results suggest that the effect is mainly mediated by endothelium-derived NO. With the commonly used pharmacological tools, the aminergie receptor enuld not be unambiguously identified. The proteolytic enzyme thrombin triggers several receptor-mediated cell effects. Recently a functional thrombin receptor was cloned and characterrized as a G-protein-coupled receptor.

Thrombin cleaves a peptide of this receptor creating a new N-terminus that acts as tethered ligand for the receptor. A thrombin receptor activating peptide (TRAP, 14 As) was identified which mimics thrombin effects on platelets and other cells endowed with thrombin receptors. We studied the contractile and endothelium-dependent relaxant effect of TRAP in comparison to thrombin in ring segments of porcine pulmonary arteries. The integrity of endothelium was assessed by the relaxant effect of bradykinin (i0 nmol/l). In PGF2~ (3 ~mol/l)-precontracted arteries with intact endothelium TRAP at 0.i -i0 ~mol/l caused a transient relaxation which was absent after mechanical removal of endothelium. The EC50 value (0.75 ~mol/l) was about three orders of magnitude higher than that of thrombin. After the relaxant response to TRAP the subsequent addition of thrombin showed a weaker relaxant effect compared to the control, but bradykinin-induced relaxation was unchanged. In deendothelialized vessels TRAP (0.i -i0 ~mol/l) exerted contractile responses. The TRAP effect was reversible. In comparison to thrombin the effective concentrations of TRAP were about three orders of magnitude higher. Hirudin, heparin and e-NAPAP did not block the TRAP effect. These findings suggest that TRAP directly activates thrombin receptors in endothelial and smooth muscle cells. Bradykinin is known to relax coronary smooth muscle by stimulation of endothelial ceils, resulting in an enhanced formation of EDRF as well as in endothelium dependent hyperpolarization (EDHF). Both mechanisms are described to be due to autacoid-induced increases in endothelium free Ca2+ concentration. Since the regulatory mechanisms of EDHF are still obscure, we investigated the differences between the regulation of EDRF and EDHF formation. Methods: Endothelial free intracellular Ca2+ was measured by the fura-2 technique. Endothelium dependent relaxation was determined in isolated coronary strips by isotonic registration. Formation of EDRF was measured as increases in cGMP in coronary arteries. Results: In the presence of the bradykinin antagonist HO 140 (D-Arg-[Hyp 3, Thi 5, D-Tic 7, Oic8]-bradykinin, both components of endothelium mediated relaxation (i.e. EDRF, EDHF) were diminished, while lmM N~L-Nitro-Arginin (L-NNA) only prevented the formation of EDRF. The K÷ channel blocker tetrabutylammonium (TBA, 3 mM), which attenuated bradykinin-induced Ca2+ entry into endothelium cells, almost completely diminished EDHF-mediated relaxation, while L-NNA-sensitive relaxation was weakly attenuated. In agreement with these results was the finding that TBA only slightly weakened increases in cGMP induced by 100 nM bradykinin (305+97.6 vs. 420-2-46.5 pmol/g w.w. under control conditions). Combination of TBA and L-NNA completely prevented all kinds of endothelium dependent relaxations (and increases in cGMP) induced by bradykinin. Conclusion: We suggest that bradykinin-induced formation of EDRF and EDHF are closely related to activation of endothelial B2 receptors. Activation of endothelial K+ channels, which plays a minor role in the EDRF formation, is essentially important for endothelium-derived hyperpolarization of coronary smooth muscle.

Institut f'tir Pharmakologie und Toxikologie, Universitat Graz, Austria. Supported by the Fends zur Ftrderung der wissenschaftlichen Forschung (P 8581).

Introduetlon: Since a variety of sutstances such as presser substances, thrombin and cytokines was shown to increase endothelin (ET) secretion in cultured endothelial ceils, several intracellular messengers may be involved in the regulation of basal and stimulated ET secretion. Activation of protein kinase C (PKC) was found to be stimulatory whereas nitric oxide (NO) and/or cGMP may have inhibitory effects. The role of changes in intracellular free Ca2+-concentration is unclear since both stimulatory and inhibitory effects have been reported. Methods: Endothelial cells were obtained from porcine aortae by enzymic digestion, cultivated to confluence in OPTI-MEM medium (0.8 mM Ca 2+, 10 % foetal calf serum [FCS]), and used after 1 or 2 passages. Cells were incubated with the different drugs for 5 hr in culture medium without FCS, and ET released was determined by RIA. Results: Basal ET release increased linearly with time (1-5 hr). The PKC activator phorbol 12-myristate 13-acetate (PMA; 400 nM) stimulated ET secretion 1.5-fold. In the presence of the PKC inhibitor H9 (N-(2aminoethyl)-5-isoquinoline-sulfonamide) PMA-stimulated ET secretion was completely and basal release partially (minus 60 %) inhibited. The nitric oxide synthase inhibitor N~0-Nitro-L-arginine (L-NNA) (100 gM) increased and the NO donor sodium nitroprusside (1 mM) inhibited PMAstimulated ET secretion in most preparations. The calcium ionophore A23187 (lktM) inhibited basal and PMA-stimulated ET release (minus 70 %). L-NNA had no appreciable influence on this inhibition. In the absence of extracellular Ca 2+, basal and PMA-stimulated ET secretion was -20 % lower than in its presence, and the inhibitory effect of A23187 was partially reversed. Conclusion: These results indicate that, in euhured PAEC, the activation of PKC results in a considerable increase in ET secretion which is largely independent of extracellular Ca 2+. Increasing the cytosolic Ca 2+concentration by A23187 depresses ET secretion through a mechanism which is largely independent of cGMP. Myocardial contractility is depressed in endotoxic shock. Plasma levels of cytokines, e.g. IllB or TNFc~, are elevated in patients with endotoxic shock. In smooth muscle cells and cardiomyocytes cytokines can induce the formation of nitric oxide (NO) which activates guanylyl cyclase. The mechanism of the depression in contractility (e.g. by vasodilatation of smooth muscles, direct effects on cardiomyocytes) has not been firmly established. Therefore, we investigated the effects of 1116 and TNFc~ on contractile response (extent of cell shortening in % of cell length) and cGMP content (pmol/mg protein) in isolated ventricular cardiomyocytes from guinea pigs. Furthermore, the influence of the NO-synthase inhibitor N -nitro-L-arginine (L-NNA) on these effects were studied. Incubation of the cardiomyocytes with Illfl (25 /uj/ml) or TNFot (100 ng/ml) for 18 h led to a decrease in contractile response to 44 % or 43 % of control value and to an increase in cGMP content to 178% or 161% of control value (p<0.05, n = 7 -12). In the presence of the NO-synthase inhibitor L-NNA (0.1 retool/l, 18.5 h) the contractility-decreasing and cGMP-elevating effects of I11B and TNF, were abolished.

It is concluded that the depression in myocardial contractility in endotoxic shock can be explained in part by a direct negative inotropic effect of cytokines on cardiomyocytes. This effect is mediated by NO leading to an increase in cGMP content. Upon stimulation of human platelet aggregation induced by different agonists a permanent increase of cGMP was measured. This cGMP increase was generally believed to be a negative feedback mechanism terminating early events of activating signal transduction. We were interested in this cGMP increase associated with the platelet activation. Surprisingly our data proved an efflux of cGMP from the activated thrombocytes. Discriminating between intracellular and extracellular cGMP we were able to demonstrate that only cGMP in the extracellular space increases, while the level of intraplatelet cGMP actually decreases. Therefore, the overall cGMP accumulation within the first minutes after the agonist addition results from extrusion of cGMP, which thereby escapes hydrolysis by intracellular phosphodiesterases. In contrast, upon direct activation of soluble guanylyl cyclase by nitrovasodilators, like sodium nitroprusside, increased cGMP remains mainly inside the cells. Thus the efflux of cGMP from platelets seems somehow to be associated only with the platelet activation. Additonal experiments showed that an elevation of intracellular calcium and activation of protein kinase C are likely to be involved in promoting cGMP ' efflux. Our results are discussed in opposite to the general hypothesis that cGMP increase during platelet aggregation represents a negative feedback mechanism limiting activation. According to our data an apparent cGMP increase actually results from extrusion of cGMP rather than soluble guanylyl cyclase activation. This cGMP efflux provides in addition to phosphodiesterase cleavage a mechanism lowering intraplatelet cGMP thus favouring the aggregatory response. The present study was undertaken to characterize the inhibitory effects of some sydnonimines on cytosolic ionized calcium [Ca2+]i, aggregation and on adhesion of stimulated human platelets. In addition, the ability of the cyclic GMP-analog 8-PCPT-cGMP to inhibit increases in [Ca2+]i-levels and aggregation was measured. All investigations were performed with gelfiltered, plasma-free platelets. The 3-morpho}inosydnonimine (SIN-I) inhibited platelet aggregation stimulated by either ADP (4-8 pM in the presence of fibrinogen), collagen (2-101~g/ml) or thrombin (0.02-0.05 IU/ml) with half-maximal concentrations (.IC50) between 70 and 200nM, respectively. The agonist-evoked [Ca2+]i-increases in fura-2loaded platelets were inhibited by SIN-I less sensitively than aggregation (IC50 between 500 and 11300 nM). Adhesion on subendothelial extracellular matrix was measured with [2-3HI-adenine labelled platelets stimulated by 8 ~JM ADP in the absence of fibrinogen. The platelet preparation was free of any aggregatory response but about 20% of the platelets adhered to the matrix. SIN-I dose-dependently inhibited adhesion (IC50 approximately 300 nM). Furthermore, 8-PCPT-cGMP inhibited platelet aggregation stimulated by either ADP, thrombin or collagen with IC50 between 50 and 200 pM. The cGMP-analog inhibited agonist-evoked increases in [Ca2+]i in similar ranges. The more sensitive inhibition of platelet aggregation by SIN-I compared with inhibition of [Ca2+]i-increases suggests that some of the effects of nitric oxide (NO) on platelets might be independent of cytosolic ionized calcium. The difference in sensitivity was not observed by directly applying the cGMP-analog. Therefore, we conclude that some antiplatelet effects of NO-releasing agents are not cGMP-mediated. In vivo as well as in vitro veins are more sensitive to organic nitrates than arteries. Since the oxygen stress in the arterial system is higher than in the venous system, we investigated in isolated, corresponding A. and V.fem. from rabbits wether a change in oxygen tension might alter the sensitivity of the vessels to organic nitrates.

Helically cut vessels were mounted isotonically in Krebs-Henseleit (KH) solution with a preload of 0.25 and 0.7 g for the V.fem. and the A.fem..

The KH-solution was either equilibrated with 95~ O2+5% COz (normoxia) or 95% N2+5% COz (hypoxia).

The vessels were maximally depolarized with KCl (120 mmol/l for V. and 60 mmol/l for A.). During normoxia the EC50 for GTN-induced relaxation of the A.fem. was 2 orders of magnitude smaller than that of the V.fem..

Hypoxia was followed by a slight relaxation of the A.fem. and a more pronounced dilation of the V.fem.. Cumulative concentration response curves were unchanged in V.fem..

However, in A.fem. the response to the lower GTN-concentrations was enhanced, thereby decreasing the EC-50 from Ixl0 -4 mol/l (normox.) to ixl0 -6 mol/l (hypox.).

There was no longer any difference observed in the dose response curves of the A. and V.fem.. Similar hypersensitivity during hypoxia could be induced in coronary arteries of the pig.

Neither the application of SOD during normoxia nor the inhibition of Cyt-p450 dependent metabolism of GTN by cimetidine during hypoxia did alter the response of arteries to GTN. Thus, it is tempting to speculate, that the GTN-induced arterial relaxation might become an important factor in ischemic tissue. Institut fur Pharmakologie Universit~t zu K61n Gleueler Str.24 L-arginine (L-ARG) is the physiological precursor of endothelium-derived nitric oxide (NO) which plays a role in the regulation of blood pressure. Molsidomine (MOL) spontaneously releases NO and thereby induces vasorelaxation. Analogues of L-ARG like L-nitro-arginine-methylester (L-NAME), on the other hand, increase blood pressure by inhibiting the enzyme NO-synthase. As L-ARG is also involved in other metabolic processes, we investigated the effects of chronic administration of L-ARG (2 g/100 ml, N=8), L-NAME (5 mg/100 ml, N=8) or MOL (3 mg/100 ml, N=8) in drinking water vs. controls (N=8) over a period of 5 months in Munich Wistar Fr6mmter (MWF) rats. In monthly intervals, the animals were weighed, and 24h-drinking water uptake and urine excretion were determined. Systolic blood pressure (BP) was recorded monthly by tail plethysmography. The heart weights were determined at the animal's death or at the end of the experimental period. Mean basal BP was 143,5 -+ 2,2 mm Hg. It was unaffected by L-ARG or MOL, but continuously and significantly increased during L-NAME: 2 rats treated with L-NAME died in the course of the experiment, one of which had excessive BP values. L-NAME and MOL treatment resulted in significantly lower mean body weights as compared to control or L-ARG, although daily food intake was similar. In L-ARG-treated rats, drinking water and urine volumes were significantly higher at all times than in the other groups. Consistent with their higher blood pressure, L-NAME-treated MWF had the highest relative heart weights. We conclude that long-term oral administration of L-ARG in MWF does not influence BP or growth, whereas L-NAME inhibits growth and induces hypertension and cardiac hypertrophy which may be deleterious. The increased water uptake and urine excretion in L-ARG-treated rats was probably due to an osmotic effect of L-ARG and not mediated by NO, as water and urine volumes in MOL-treated rats did not differ from controls. Nitric oxide (NO) and nitric oxide generating agents like sodium nitroprusside (SNP) or 3-morpholinosydnonimine (SIN-l) stimulate the mono-ADP-ribosylation of a cytosolic 39 kDa protein in various tissues. This effect is independent of the known action of NO on guanylyl cyclase and cGMP. We purified the protein from human platelet cytosol and after N-terminal sequence analysis identified this protein as the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH'). Nitric oxide, SNP and SIN-1 stimulate the auto-ADP-ribosylation of GAPDH in a time and concentration dependent manner. Furthermore, ADP-ribosylation of GAPDH is associated with inhibition of enzyme activity. Basal as well as stimulated ADP-ribose transfer is inhibited by the SH-alkylating agent Nethylmaleimide and the radiolabel of modified GAPDH is removed by HgCI 2 treatment, suggesting an ADP-ribose-cysteine bond to be present in the ADP-ribosylated form of GAPDH. Bovine serum albumine with a Snitrosylated SH-group can serve as a model of ADP-ribose transfer from NAD + under certain experimental conditions. Further experiments point to NAD + as the necessary cofactor for the ADP-ribosylation reaction, excluding the possibility of a reaction sequence involving a NADglycohydrolase followed by nonenzymatic ADP-ribose transfer to GAPDH. There was no radioactivity associated with GAPDH, when [carbonyl-14C]NAD+ is used, suggesting that the ADP-ribose moiety of NAD + is transferred to the protein. Further studies demonstrate the requirement of a + reactive SH-group besides a NAD binding site as a prerequisite for a NOstimulated ADP-ribosylation reaction. Soluble guanyly| cyclase (sGC) is a heterodimer, and we recently cloned and sequenced both enzyme subunits (¢ and 8). Ferroprotoporphyrin IX (heme a) represents a prosthetic group of sGC, and it has frequently been speculated that this heme group may mediate the activation of soluble guanylyl cyclase by NO. We investigated this issue using site-directed mutagenesis to exchange conserved amino acids thought to be involved in heme protein interactions. The proximal histidines of all known heme and myoglobin chain types show a leucin in the -4 position, and two such motifs (LXXXH) were found to be present in sGC. The respective histidine residues (106 and 442 on the B-and ¢-subunits, respectively) were changed to phenylalanin, and five other histidines were mutated to serve as controls. In the cytosol of COS-7 cells, which had been used for the expression of cDNA clones, bahai activities of the wild type were 127 + 29 pmol cGMP x mg "~ x min "I (mean ± SEM; n = 9). All mutants were still catalytically active, although basal activities were decreased to variable extends (10 to 64 pmol x mg "1 x min'l). Mutation of His-f06 resulted in a complete loss of enzyme stimulation by sodium nitroprusside (SNP), whereas all other mutants were still sensitive to NO-containing compounds. These results demonstrate that His-106 of the Bl-SUbunit is essential for NO-induced stimulation of sGC. Thls histidine residue may either participate in heme binding or mediate NO-heme-induced changes in protein conformation leading to enhanced rates of cGMP formation. The functional domains of Ict cGK, the catalytic domain (residues 335-670) and the regulatory domain (residues 1-352) were cloned using PCR techniques. The catalytic subunit was cloned in pmal C-2 vector and expressed in E. coli as a fusion protein with maltose-binding protein (MBP). The regulatory subunit was cloned in puc18 and expressed in E. coli. The 77 kD catalytic subunit-MBP fusion protein was purified from j'M101 cells using amylose-affinity chromatography. Although the fusion protein was soluble no catalytic activity could be detected. Proteolytic cleavage of the MBP fragment with Factor Xa yielded in a spontaneously aggregating catalytic subunit. The 38 kD regulatory subunit was purified from JM101 and E222 cells using DEAE-ion exchange chromatography and cAMP-resin affinity chromatography. The protein is proteolytic sensitiv and readily cleaved at residue 77. The two cGMP binding sites A and B are functional but differ in their kinetic constants from the wild type protein. The dissociation constants are shifted 10-fold for both sites (Kt) site A = 100 nM and Kt) site B = 800 nM) and off-rate sudies revealed a normal fast dissociation site for site A (Kof f = 3.7 rain -1) but also a fast site for site B (Koff = 0.33 rain-l). The results show that dissecting the cGK in its functional domains gives a functional regulatory subunit and a non-functional catalytic subunit. The latter finding agrees with the observation that the native enzyme cannot be expressed catalytically active in E. coll. Relaxation of smooth muscle is thought to be mediated by cGK. This enzyme consists of two identical subunits of 75kD each and two isoforms (1~ and If~) can be isolated. The cDNA of the enzyme was cloned from bovine tracheal smooth muscle (1) and expressed in eukaryotic cell lines as a functional protein (2). To study whether cGK can also be expressed in E, coil, we used the "T7 expression system" (3). The cDNA of cGK type le was cloned into the bacterial expression vector pET3a (3) and transformed in E~ coil strain RL21 (DE3) (3). Analysis of cell protein by SDS-PAGE and western-blotting showed, that after induction with IPTG a 75kD protein was efficiently accumulated (up to 30% of total protein) and could be stained with antibodies specific for cGK. However, most of the protein was in an insoluble form and no phosphotransferase activity could be detected. These "inclusion bodies" could be solubilized with urea and D'IT under alkaline conditions but could not be renatured to an active enzyme. In order to prevent aggregation of cGK we constructed two further expression vectors. First, the 22aa leader peptide of the pectat lyase gen (pelB) from Erwinia carotoyora was fused to the N-terminus of cGK and the fusion protein expressed using the same T7 system as mentioned above. The leader peptide could direct the export of cGK into the periplasmatic space of E.coli, where productive folding of cGK could take place. Second, the cGK gene was cloned into the prokaryotic shuttle plasmid pWLQ1, which allows expression under control of the tac-promotor in E.coli and the gram positiv Corynebacter,ium glutamicurn. Neither the fusion protein with the pelB leader se~luence nor the pWLQl-construct led to the production of soluble and active enzyme. These results suggest that the cGK can be expressed in E_=. coil only in an insoluble und inactive form, In contrast to the situation in eukaryotic cells, bacteria may lack systems for posttranslational modification and/or correct protein folding I"chaperones ") that are crucial for the activation and productive folding of cGK. Two degenerate oligodeoxy-nucleotide primers designed after highly conserved peptide sequences in cyclic nucleotide-gated cation channels of retinal and olfactory neurons were used to PCR amplify homologous sequences from rabbit heart, smooth muscle and brain. Screening of a rabbit aorta cDNA library with the 32p-labeled PCR products as probes yielded a full length cDNA, cGA1, consisting of 5449 bp. The open reading frame encodes for a protein of 732 amino acids. The calculated Mr is 84,043. The predicted amino acid sequence is 93 % homologous to the bovine cAMP-gated cation channel of the olfactory system and 57 % homologous to the bovine retinal cGMP-gated cation channel. Injection of cRNA containing the complete coding region of cGA1 into Xenopus oocytes induces the formation of a cGMP/cAMP-gated cation channel with the basic properties found in other cylic nucleotide-gated cation channels. Our data indicate that cyclic-gated cation channels are not specific for retinal and olfactory neurons but are also expressed in other tissues like brain, heart and aorta. At present the physiological role of this class of ion channels in these tissues is not known. 

Two types of G-protein-coupled receptors with high affinity for adenosine can be distinguished by binding studies (Aj and A~). Recently, another adenosine-binding protein of unknown function with lower affinity for adenosine was purified by Hutchinson and Fox (JBC 1989; 264:19889-19903) from human placental membranes and named adenotin. Now, we purified adenotin from bovine heart membranes and determined protein-and binding-data. The purification includes the following steps: i. Extraction from membranes by low concentrations Stiroulation of Al-adenosine receptors by agonists leads to the activation of different second messenger pathways. It is not known whether these pathways are triggered separately by different Al-adenosine receptor subtypes or whether a single receptor is responsible for the various effects. To clarify this question we stably expressed in CHO cells the cDNA of a rat brain Al-adenosine receptor cloned from a hippocaropal cDNA library. These CHO cells do not contain endogenous Al-adenosine receptors. Transfection gave rise to several clones expressing up to 1000 frool of Al-adenosine receptors per rog roerobrane protein, These receptors had the expected binding properties. In these cell lines, we investigated the second roessenger effects caused by stimulation of the Al-adenosine receptor. First, the adenylyl cyclase activity was strongly (>70%) inhibited in an Al-adenosine receptor specific roanner (Cyelohexyladenosine, CHA > R-PhenyMsopropyladenosine, R-PIA > 5'-N-Ethylcarboxyamidoadenosine, NECA). Second, the inositol trisphosphate production was weaidy (+40%) enhanced in a concentration-dependent roanner via this receptor. Thus, our resuhs show that a single Al-receptor type is capable of triggering at least two different effector pathways.

Laboratory of Molecular Biology, University of Munich, Max-Planck-Institute of Biochemistry, 8033 Marfinsried, FRG

One of the main physiological actions of adenosine in the central nervous system, the inhibition of neurotransmitter release, is mediated by pertussis-toxin sensitive G proteins, but can be regulated by adenosine independently of cyclic AMP levels. The central depressant effects of adenosine have been linked to a variety of effector systems. In contrast, activation of G proteins by adenosine A 1 receptors can be moni- The fall in glomerular filtration rate and the inhibition of renin release by adenosine is thought to be mediated by adenosine Al-receptors. In rats on low sodium diet renal response to adenosine is potentiated, whereas rats on high sodium diet show a reduced response to adenosine (Osswald et al, Pfluegers Arch 357:323, 1975) . The present experiments were carried out to determine whether the different Na diets induce changes in Al-raceptor affinity or density. Male Sprague-Dawley rats (200-280 g) received either a low Na diet (0.02 %), standard diet plus 1% NaC1 as drinking water, or high Na diet (1.5 %) plus tap water for four weeks. Rats were sacrificed, exsanguinated and kidneys removed quickly. Glomeruli were isolated and purified up to 95 % by a sieving method. Membranes from glomeruli were prepared according to Lohse et al (Naun Schm Arch Pharmacol 326:69, 1984) . Adenosine Al-receptors were characterized in binding studies using 125J-HPIA or 3H-DPCPX. Before binding was carried out (250 ,al Tris-Mg-boffer, pH 7.4, 100 #g protein) membranes were treated with adenosine deaminase (2.4 u/ml) to eliminate endogenous adenosine. Unspecific binding was determined in the presence of 10 "~ M theophylline (J-HPIA) and 10 "a M cold DPCPX (3H-DPCPX). Specific binding was saturable and reversible for both ligands. Binding curves were fitted by nonlinear regression (Hill equation Whereas the A I receptor-mediated' inhibitory effect of adenosine on the release of noradrenaline (NA) in the CNS is well established, a modulatory role of presynaptic A 2 receptors has not been observed so far. Recently, we found that the release of NA evoked by superfusion of hippocampal tissue with 3,4-diaminopyridine (3,4-DAP) was much more sensitive to manipulations of intracellular cAMP than that evoked by electrical fields (Lauth et al., NSAPh 346:R21, 1992) . Therefore, we now studied the question, whether stimulation of A 2 receptors, which are positively linked to adenylate cyclase, affects the NA release evoked by 3,4-DAP (200~M, 1 min) in rabbit hippocampus slices. In this model A 1 receptor a~onists (L-phenylisopropyladenosine, PIA~ 2-chloro-N -cyclopentyladenosine, CCPA) acted inhibitory, whereas preferential or selective A2-agonists (N-ethylcarboxamidoadenosine, NECA; 5'-[Ncyclopropyl]carboxamidoadenosine, CPCA; 2-p-(2carboxyethyl)phenylethylamino-5'-N-ethylcarboxamido adenosine, CGS-21680) showed inhibitory or no effects. In the presence of the A 1 receptor antagonist 8-oyclopentyl-l,3-dipropylxanthine (DPCPX, I~M) the inhibitory effects of A1-agonists were abolished.

In contrast, the A2-re~eptor agonists now exhibited significant facilitatory effects, which could be antagonized by the additional presence of the A 2 antagonist 3,7-dimethyl-l-propargylxanthine (DMPX, 10~). From these findings we conclude that in addition to inhibitory A~ also facilitatory ~2 receptors may exist on no~adrenergic nerves in the rabbit hippocampus. Adenosine binding proteins in the cytosol could explain the relatively high adenosine tissue content under nol"moxie conditions. Hershfield and Kredich (Science 202:757, 1978) reported that SAH-hydrolase from placenta can bind adenosine (kn ~ 0,5/~M). Thus SAH-hydrolase could function as an "deep" compartment for adenosine preventing its deamination. In the present study we isolated SAH-hydrolase from bovine kidneys to describe its adenosine binding characteristics.

Pharmakologisches Institut der Eberhard-Karls Universit~t Tiibingen, and CAA (5'-(N-cyclopropyl)carboxamidoadenosine) 20 out of the 45 synthesized purine analogs inhibited H202-formation in a dose dependent manner. Structure response relationships demonstrate that derivatives of N 6substituted andl2',3'-didesoxyadenosine, six-ring xanthines and C -substituted isoquinolines were able to decrease H202-production.

Intitute of Pharmacology, University of DUsseldorf, Moorenstr. 5, W-4000 DUsseldorf, Federal Republic of Germany

Adenosine receptors have been originally distinguished as Al-receptors that inhibit and A2-receptor that activate adenylate cyclase. According to the structure-activity-relationship of various adenosine analogs, the A2-receptor could be further subdivided into high-affinity A2a-receptors (adenylate cyclase independent) and low-affinity A2b-receptors coupled to adenylate eyclase. We investigated the effects of the A2-receptor agonlst CGS 21680 

Hypoxia represents a stimulus for endothelial cell proliferation. Since hypoxia is also associated with a pronounced release of adenosine, we have tested the ability of endothelial cells to respond to typical adenosine receptor agonists. Primary cultures of human umbilical vein endothelial cells were obtained by collagenase dispersion and maintained in medium 199 containing 20% fetal calf serum and endothelial ceil growth supplement. Cell proliferation was assayed with [3H]thymidine incorporation. Addition of adenosine deaminase resulted in inhibition of basal [3H]thymidine incorporation. Likewise, xanthine amine congener (XAC), an adenosine receptor antagonist, produced a concentration-dependent decrease in endothelial cell proliferation with a half-maximum effect at about 10 nM. These observations imply participation of endogenous adenosine in FCS -stimulated endothelial growth. Inhibition of basal [3It]thymidine incorporation was much more pronounced by pertussin toxin-treatment suggesting the involvement of Gcdependent pathways in controlling endothelial cell growth. In the presence of adenosine deaminase, the adenosine receptor agonist N-ethylylcarboxamid-adenosine (NECA) stimulates endothelial growth with a half-maximum effect at about I0 nM. However, both the Rand S-isomer of N6-phenylisopropyladenosine (PIA) were much less effective than NECA. Our observations suggest that adenosine receptors are capable of promoting endothelial cell growth. We propose that locally released adenosine is a candidate for mediating neovascularisation induced by hypoxia and other forms of tissue damage. However, the receptor subtype involved remains unclear at present. PMN are important inflammatory cells involved in acute tissue injury. Tins study investigates whether agonist-induced (FMLP) PMN stimulation may be synergistically antagonized by PGE 1 (PG) and SIN-I, a NO-donor. PG (10 nM-10 tzM) iNaibited dose-dependently FMLP-induced B-glucuroindase (B-G) and oxygen radical (02") release. SIN-1 (100 ~M) alone was ineffective, but potentiated PG-induced inhibition. Isobutytmethylxantinne (IBMX; 0.5 raM) potentiated the inhiintory effect of PG. The combination SIN-1 (100 I-~M) plus IBMX (0.5 raM) did not further iathibit ll-G release, but almost abolished 0 2" generation. Langendorff hearts of rabbits were perfused at coustant volume and subjected to 2 h of low-flow isehaemia followed by 30 rain of reperfusion (MI +RP). Nitric oxide (NO) was measured by the oxyhaemoglobin-tealmJque in the coronary effluent. Basal NO release was significantly cl{mini~hed in early reperfasion (15 rain), but enhanced in a later phase (30 rain). The total amount of NO formation by bradykkfin (0.05 ~M, infused for 3 min) was significantly reduced after MI+RP. Both effects were completely prevented by iloprost treatment (30 nM Iloprost also significantly redtlced the rise in left ventricular enddiastolic pressure during MI+RP and preserved creatine kinase specific activity in the left ventricle. Iloprost also prevented cardioceronary injury induced by indomethacin treatment during MI + PP. The results indicate that both endogenous and exogenous PGI 2 in addition to a cardioprotective action also protect the coronary endothelium inehiding preservation of endothelial NO formation.

In response to various stimuli activated platelets produce the procoagulant and vasoconstrictant prostanoid thromboxane A 2. Thromboxane is a strong platelet agonist itself and its synthesis in stimulated platelets thus provides a selfamplificating feed forward mechanism favouring adhesion, aggregation and secretion of procoagualant granulae content. Release of arachidonie acid from membrane lipids by activation of a calcium dependent phospholipase A 2 is generally accepted to be the rate limiting step in thromboxane synthesis. The intracellular free calcium concentration is therefore believed to be the single regulatory element in platelet araehidonic acid metabolism. Increasing the intracellular free calcium concentration by means of thapsigargin, a specific inhibitor of endoplasmatic calcium-ATPase, indeed leads to substantial thromboxane synthesis and platelet activation. However, in the presence of the thromboxane receptor antagonist BM 13177 (Sulotroban) thapsigargin still stimulates liberation of arachidonic acid but thromboxane synthesis is dramatically decreased. The free arachidonie acid is almost exclusively converted into biological inactive 12-I-[ETE via 12-1ipoxygenase pathway. Receptor occupation seems thus to be an equally important prerequisite for thromboxane synthesis as is intracellular calcium increase for the release of the substrate arachidonic acid. The nature of the coupling between receptor and prostaglandin-endoperoxide (PGH2)-synthase is so far unknown, possible mechanisms may involve a change in compartimentation of liberated arachidonate and/or PGHrsynthase. In previous studies we have demonstrated that centaur activation of the intrinsic coagulation cascade in whole human blood in vitro is accompanied by genermtlus of 5-1ipoxygenaue-darived cysteinyl-leukotrlanes (iT). By supplementation experiments it was shown that the ~-lipaxygenase pathway activated resides within peripheral mosucytes IPM) but sot iu pelpuorphunuclear leukuo~tes, Indeed, PN (5 or 15x105/mlJ incubated in recalcified platelet-rich (CaP'-POP) or platelat-poor plasma generated immunoreantive LT in a time-dependent fashion. When PM 5x105/ul were iunubatnd for 60 uin in Ca~'-PBP in the presence of lyuine analogues known to inhibit binding of plasuln(ogen) to monoophio plasain[agenJ blndiug sites, u ooncsntration-dnpendent decrease in cysteinyl-LT formation was observed: W~-acetyl-L-lysl~e, ICe! 2,23 mM; 6-aulnohexanoic acid~ ICe! 604 pH; tranexanic acid, IOs~ 55 ~M, Furthermore, plasmln inhlbitors such as aprotiuiu and uz-antiplaumin were similarly fouhd to inhibit cysteinyl-LT generation by FN in Cu~+-PRP (ICs~ Hlg KIU/ml and 506 aN, respectivelyJ, In contrast, plusninoHen activators such au urekinase and streptoklnaue up to 10 and 100 U/nl, respneblvely, led ho oeacsutration-dependent stimulation of the cyuhninyl-~f production by PN in Cat~-PHP, ~one of the inhibitory compounds used had any such effect when P~ lncubabad iu Hank's balanoed salt solution [NBHH)were stiuuluted with ionopbore A 2HIB? (I ~NJ, On the other hand, neither urokiuase nor utreptokinaue had any direct stimulaLor~ effect on Lhe 5-1ipoxygeaaee noLivitp of P~ incubated in HBSH, Incubation of 15slO s PN/ml in NBHH for 60 min did not induce auy delectable LTB5 production, However, addition of exogenous glu-plasuiu (0,000~ -1,40TA U/ul) Lriggered release of LTUs with u uanimuu (58g ! 60 pg/ml) at 0,175 U/ml of gluplasuin, In this uettiug~ the utinulatory effect of glu-plas~in 0.I75 U/ml was antagonized by 6-aninobeuanoic acid (lOam 79 pN) and tranexuulc acid (ICes ~,~ pW), On Lhe other hand, plauuisogeu containing Lhe lyuine binding sites of plaumis but its catalytic cenLer in a oryptlc form only, was unable to stimulate the ~TN~ production in FN incubated la NHSH, These results show that i) oonLact activation-mediated utlmulaLion of the 5-1ipexygenmse pathway in PE proceeds via plasuin, li) lyulne binding uite-nediated mssociaLioa of plasnin with the uonuuytie membrane is required, ilil the oatal~tic center o~ the plmumln molecule is of absolute importance for the stbmulatory effect observed, Plasmin, u hitherto unknown stimulus of Lhe Human neutrophils have an important function in the physiology and pathophysiology of inflammatory and ischemic processes and may contribute to severe tissue damage, e.g. in myocardial infarction. Regulation of neutrophil activation by hormones, intercellular messengers or drugs is therefore very important. Several studies demonstrated that activation of neutrophils can be inhibited by cAMP-or cGMP-eievating vasodilators, but the underlying molecular mechanisms of this inhibition is not clear. Our laboratory has characterized the role of vasodilator-stimulated protein phosphorylation including the 46/50 kDa protein VASP (Vasodilator-stimulated phosphoprotein) in the inhibition of human platelets. In the present study we investigated whether a comparable inhibitory intracellular signal transduction system is present in human neutrophils. Our data show that 1) human neutrophils contain high levels of the protein VASP, 2) VASP is rapidly phosphorylated in neutrophils in response to cAMP-or cGMP-elevating vasodilators, and 3) phosphorylation of VASP correlates with the vasodilatorevoked inhibition of the fMLP-induced calcium response in neutrophils. We conclude that VASP phosphorylation may be an important component of the intracellular mechanisms of vasodilator action in neutrophils. We hope that further elucidation of inhibitory signal transduction pathways in neutrophils contributes to a better understanding and development of new therapeutical approaches for inflammatory or ischemic diseases. Cyclosporine A (CSA) is used as immunosupressive drug to prevent transplant rejection. A major unwanted effect of cyclosporine is the often observed nephrotoxicity, the mechanism of which is not yet clear, The prostaglandin synthesis of the kidney is modulated by CSA: thromboxane synthesis is enhanced while the synthesis of the vasodUatory PGE 2 is reduced. The aim of the present study was to investigate the mechanism of the inhibitory effect mediated by CSA. Rat renal mesangial cells were stimulated with interleukin 1 (11-1) to induce prostanoid synthesis. U-1 enhanced the specific activity of the cyclooxygenase (COX), determined in the microsomal fraction of these cells, in a concentration dependent manner. Recently, it was shown that COX activity consists of two isoforms, which differ significantly with respect to their mRNA, protein structure and functional role, We could now show that I1-1 had no effect on COX 1 mRNA or protein expression, which is constitutive in mesangial cells, but significantly enhanced the COX 2 mRNA and protein mass (determined by Northern and Western blot techniques), resulting in enhanced PGE2 synthesis. CSA suppressed the II-l-induced PGE 2 release and COX activity. It had no effect on the constitutively expressed COX 1, but inhibited the I1-1 -induced mRNA expression and the corresponding COX 2 protein, These data show that I1-1 selectively induces the COX 2 Isoform in mesangial cells and that CSA interferes with this induction resulting in a reduction of the enhanced synthesis of PGE 2 in these cells, The effect of various ionophores (e.g. nigericin and monensin) on the inositol i,4,5-trisphosphate (IP3)-sensitive calcium store in RI~mSF cells suggests that a low potassium but high sodium concentration is maintained in this calcium storing organelle. To investigate whether this is due to the presence of intracellular sodium potassium pumps in the membrane of this erganelle, we examined the effect of the lipophilic glycoside digitoxigenin on ionophore-induced calcium release from this IP3-sensitive store in saponin-permeabilized cells. In fact, nigericin-induced calcium release was decreased or abolished by the glycoside, indicating that the outside to inside potassium gradient of the calcium store had been diminished. Furthermore, digitoxigenin inhibited calcium uptake into the ~P3-sensitive store when sodium was present in the incubation medium. Sodium itself, in the absence of digitoxigenin, decreased calcium uptake by the IP3-sensitive store. Finally, intracellular nonmitochondrial calcium sequestration in permeabilized RINm5F cells was inhibited completely by thapsigargin, the specific inhibitor of sarcoplasmic/endoplasmic reticulum calcium pumps. In conclusion, these results are taken as evidence for coexistence of a sodium potassium pump, sodium calcium exchange and a thapsigargin-sensitive calcium pump in the membrane of the IP3-sensitive calcium store of RINm5F cells. Ruthenium red is a widely used research too] in biochemistry and pharmacology. It inhibits not only mitochondrial Ca ~* uptake with a high potency, but also Ca z÷ release from the sarcoplasmic reticulum and antagonizes the effects of Capsaicin by blocking neuronal ion transport. The structural features which confer these properties are unknown. Using the mitochondrial Ca ~÷ transport as a model system we compared the effects of ruthenium red which is a trinuclear oxo-bridged ammine complex of ruthenium to those of other metal ammine complexes, Mononuclear ruthenium(III) and ruthenium(I~) hexsmmine complexes proved to be inhibiters of mitochondrial Ca z÷ uptake (IC~0s of 5,2 ± 0.4 ~M and 10.9 ± 0.4 pM as compared to 0.19 ± 0.02 ~M for ruthenium red). Cobalt(III) hexammine was equally potent, indicating that the presence of ruthenium as the central atom was not obligate for the inhibitory activity. The maximal effect of all these compounds was an inhibition of the initial velocity of Ca z* uptake by around 90%. The tetrammine complex of copper(II) was a less potent and less efficient inhibitor than the above hexammine complexes (ICs0 83 ± 9 DM, maximal inhibition 64 ± 4%), indicating that the number of ammine ligands correlates with the potency of these compounds. Apparently, ruthenium red interferes with cellular ion transport by virtue of its polycationic nature, which is determined not only by the charge of the central atoms but also by the partial charges of its 14 ammine ligands.

Institut fQr Pharmah01ogie und T0xik010gie, Un~versit~t GOttingen, D-3400 G6ttingen

Frijus-Plessen, G. Jahn*, H. Foth, E. Jurkiewicz*, K. Hirsch-Ernst, and G. Hunsmann* The antiviral efficiency of 3'-azido-3'-deoxythymidine (AZT) is frequently decreased during long term usage which might be caused by a shift of the kinetics of AZT uptake and/or conversion to the active nucleotide AZT triphosphate. In this study, we compared the uptake of AZT into human peripheral lymphocytes and either uninfected or HIV-1-IIIB-infected Jurkat cells, a permanent human T-cell-line. The cells were incubated at a eytokdt of about 1% for 4 h with 100 and 450 ~M AZT. The substrate concentrations in the extracellular medium were not substantially lowered throughout the incubation. At 100 gM AZT, the intracellular level of the unchanged nucleoside was threefold higher in lymphocytes than in Jurkat ceils (2.2 vs. 0.8 nmol/106 cells), whereas at 450 I.tM AZT the intracellular concentration approached almost the same level in both cell types (4,8 vs. 5.3 nmol/106 cells). The values decreased slightly (20 %) between 60 and 240 rain of incubation indicating that the inside-outside balance of AZT has been changed because of elimination by metabolism or decrease of the inside-directed AZT flux. HIV-l-IIIB-infected Jurkat-cells showed slightly lower intracellular AZT concentrations (10 %) than uninfected ceils. The inhibition of a multiple drug transporter-mediated efflux of drugs by 10 I-tM verapamil had no effect on the uptake of AZT in Jurkat cells and marginally decreased the AZT levels in lymphocytes (4 %). However, 10 ~tM forskolin induced a substantial diminuation of intracellular AZTlevels in both cell types. These results indicate, that a mdr-mediated efflux is not likely for AZT. But a forskolin-sensitive process must be assumed to contribute to the kinetics of AZT uptake, effiux and/or metabolism. Indirect imrnunofluorescence experiments using antisera directed against the nineteen amino acid N-terminal fragment of Porin 31HL as well as experiments with monoclonal antibodies, indicated the presence of porin in the plasma membrane of endothelial cells cultured from human umbilical vein and pig aorta. In order to reveal a possible functional role of this protein we studied the effects of the antisera on cytoplasmic free Ca 2+ and on intracellular cGMP levels which reflect NO syntase activity, i.e, a physiologically relevant function of endothelial cells. The tested antisera by themselves barely affected intracellular Ca 2+ and cGMP levels, but clearly enhanced histamine-as well as bradykinininduced rises in intracellular free Ca 2+ and cGMP levels. Since we have recently characterized an endothelial LC-type CIchannel which shares several biophysical properties with porins, we have tested whether the antibodies directed against porin do affect the function of endothelial LC-type CI-channels. Two monoclonal antibodies against Porin 31HL and one antiserum were tested in patch clamp experiments. Both monoclonal antibodies and the antiserum blocked endothelial CI-channels. Our results demonstrate the existence of porin in the membrane of vascular endothelial cells and suggest a functional role in agonistinduced activation of the cells. This protein may be identical with the endothelial LC-type CI-channel. Bile acids undergo an enterohepatie circulation involving specific Na+-dependent bile acid transport systems in the sinusoidal membrane of hepatocytes and brush border membrane of ileal enterocytes. This organotropism of bile acids suggests to use them as delivery systems to achieve a specific targeting of drugs to the liver and the ileum and to improve the enteral absorption of poorly absor'bable drugs and peptides. Therefore, hybride molecules composed of a drug and modified bile acid molecules were synthesized. The drugs chlorambueil, various HMG-CoA-reductase inhibitors and peptides were eovalenfly linked to 7c*, 12c~-dihydroxy-3B-(o~aminoalkyloxy)-513-cholan-24-oic acids. Transport studies with isolated hepatocytes and ileal brush border membrane vesicles and photoaffinity labeling revealed a specific interaction of these drug-bile acid conjugates with the Na+.-dependent bile acid transport systems in contrast to the parent compounds which did not interfere with the bile acid transporters. In liver perfusion experiments the bile acid character of these molecules was demonstrated: 1. The affinity of drugs to the hepatic and ileal bile acid transporters was increased by at least one order of magnitude by coupling to bile acids. 2. 2"he renal secretion of a drug could be changed into a hepa~biliary one by attachment to bile acids. 3. The intestinal absorption and secretion of peptides into bile could significantly be increased by coupling to bile acids. 4. The intracellular release of an active drug could be influenced by the side chain of the bile acid moiety in the drug-bile acid conjugates as well as by the linker structure. In conclusion, these studies show that modified bile acid molecules can be used as shuttles to achieve an organotropic delivery of peptides and drugs to tissues possessing bile acid transport systems. In epithelial cells the plasma membrane is differentiated into two structurally and functionally distinct domains, the apical membrane, which faces the external milieu, and the basolateral membrane, which contacts neighboring cells and the blood supply. We have identified an 80kD glycoprotein complex as a marker protein for constitutive exocytosis at the apical surface in the canine kidney epithelial cell line MDCK. The isolation and characterization of the gp80 cDNA revealed that the protein is the cartine homolog of the rat protein TRPM-2 (Testosteron-repressed message-2), the human proteins apolipoprotein J and complement lysis inhibitor (CLI) and the bovine protein GP III which has been isolated from the chromaffme granules of the adrenal medulla.

These results indicate that the protein is secreted constitutively at the apical surface in kidney cells, constitutively at the basolateral surface in hepatocytes and via the regulated pathway in neuronal cells. We plan to characterize by in vitro mutagenesis, gene transfer and expression of the cDNA the structures in the protein that determine its complex intracellular transport. Most importantly we want to analyse if a single slructure determines its routing in epithelial and in neuronal cells or whether there are separate signals for each pathway. Protein disulfide isomerase (PDI) was considered to be involved in the hepatic uptake of certain organic anions. With polyclonal antibodies against PDI and a 60 kDa bumetanide binding protein from rat liver plasma membranes this hypothesis was investigated. For production of polyclonai antibodies against the PDI, a 324 bp fragment of the corresponding cDNA was cloned in a bacterial expression system. The resulting beta-gaiactosidase-PDI-fusion protein was then used for immunization (anti-PDI-Ab). The 60 kDa bumetanide binding protein used for immunization (anti-Bum-Ab 60) was isolated by preparative 2D-gel electrophoresis and electroelution With these antibodies the respective proteins were localized in rat liver plasma membranes and in different organs as shown by two dimensional gel electrophoresis and cell immunoflourescence. Protein disulfide isomerase was identified by the anti-PDI-Ab monospecifically as a 55 kDa protein (pI = 4.5) without any crossreaction with the other antibodies. Crossreactivity was observed for a 60 kDa bumetanide binding protein (pI = 7.1) with antibody anti-Bum-Ab 60 and with an antibody against P450 UT-H (P450 IID1). It is concluded that hepatic membrane binding proteins for bumetanide may have common epitopes with certain P450 proteins but not with the protein disulfide isomerase. we ' could demonstrate that not oral salt loading, but feeding dominantly influences renal dopamine (DA) excretion. In order to investigate, whether sympathetic nerve activity mediates the renal response to feeding, we studied urinary DA excretion after bilateral renal denervation (DNX) in fed and fasted animals.

Male Sprague-Dawley rats (200-300 g) were anesthetized with thiopental. After an abdominal midline incision kidneys were denervated by cutting visible nerve fibers and by painting the renal artery with 10% phenol ethanolic solution (DNX, n=8). Sham operated animals served as controls (CON, n=6). On days 5 and 7 after surgery, rats were kept in metabolic cages for 24h. Using a cross over design, one half of the groups was kept with access to chow (FED) on day 5 whereas chow was withheld on day 7 (fasted, FST). The other half of the groups was treated in the reversed order. Success of the DNX procedure was confirmed by measurement of norepinephrine tissue content in both kidneys (NE-kdy) at the end of the experiments. In 24h urine, catecholamines were determined by HPLC, creatinine and sodium photometrically. See The data show that the feeding induced, more than 2-fold increase in urinary DA excretion is not influenced by denervation of the kidneys. We conclude that the renal response to feeding is independent of the efferent sympathetic nerve activity. ACE-inhibitors can reduce proteinuria in diabetic nephropathy as shown in many but not in all clinical trials. To assess the influence of the duration of diabetes mellitus (DM) on the antiproteinuric effect of ACEinhibitors we studied diabetic rats in two experimental series of early (1-3 weeks after STZ injection) and late (9-15 weeks after STZ) stage of diabetic nephropathy. Sprague Dawley rats (body weight 240-260g) were made diabetic by an intraperitoneal injection of STZ (60 mg/kg). Animals which developed blood glucose levels above 300 mg/dl within four days were used for the experiment. Rats were housed in metabolic cages for 24h once a week. Urine was collected for 24 hours and analyzed for protein, creatinine and sodium. The ACE-inhibitor Quinapril (QUI) was administered via a gastric tube at a dose of 20 mg/kg during the first 3 weeks after STZ. Compared to controls (vehicle treated diabetic rats) QUI did not reduce protein or creatinine excretion in the early stage of DM. In the late stage of experimental DM, however, QUI (60 rag/1 added to the drinking water, approximately 25 mg/kg per day) reduced the pathological ele9ated proteinuria by 27 % (n= 11). Also in this experimental group QUI did not change creatinine excretion. Our data show show that QUI can reduce proteinuria in normotensive rats only in the late phase of DM. The involved mechanisms of action are unclear at present and require further investigations. [333] [334] [335] [336] [337] [338] 1988) suggested that one reason for the increased GFR in DM is a decreased activity of the tubuloglomerular feedback (TGF). We therefore examined in the present study the effect of dipyridamole (DIP) which is known to enhance TGF activity via an adenosine-mediated mechanism, on kidney function in non-fasted diabetic male Sprague-Dawley rats. One day after STZ (60 mg/kg i.p.) daily treatment with DIP (1 g/kg via a gastric tube) was started in one group (STZ+DIP) and with vehicle only in another group (STZ). Non-diabetic animals, also receiving vehicle, served as controls (CON). After 3 weeks the rats were anesthetized with thiopental (80 mg/kg) and the left kidney was prepared for micropuncntre experiments. We assessed TGF activity as the difference between stop-flow-pressures (ASFP) in the early proximal tubule at 0 and 50 nl/min orthograde perfusion of Henle's loop from endproximal. The In the present study we analyzed whether inhibitors of the endothelium-dependent relaxation can affect the renal response to exogenous and endogenous ADO in rats.

Methods: Male Sprague Dawley rats (250-350 g) were anesthetized with 100 mg/kg i.p. thiopantal. The left kidney was exposed by flank incision and a electromagnetic flow probe was fitted around the renal artery to monitor continooualy renal blood flow (RBF) using the Carolina flowmeter. Exogenous ADO was administered as single injections via a catheter placed into the thoracic aorta. Renal response to endogenous ADO was assessed by measuring RBF after releasing a 30 sec renal artery occlusion. The resulting postocclusive vasoconstriction has been shown to be an ADO-mediated phenomenon [Osswald et al. (1977) Angiotensin II (ANG II) and hs respective binding sites have been localized in circumventricular structures such as the subfomical organ (SFO), areas known to play an important role in central volume and electrolyte regulation. Furthermore, there is increasing evidence for the presence of osmoreceptors in these structures although these have not yet been exactly localized. Intracerebroventricular (i.c.v.) injections of hypertonic saline and ANG II induce similar effects with respect to water intake, thirst and renal sodium excretion. Previously, we have demonstrated that the natriuretic response after i.c.v, injection of hypertonic saline is partly mediated by a central angiotensinergic mechanism.

In the present study we tested the hypothesis whether the centrally induced natriuresis after hypertonic saline involves the SFO as suggested by the occurence of osmosensitive cells as well as a high concentration of ANG II receptors in this brain area. All experiments were performed in conscious Wistar rats. Urine was collected via a chronically implanted ureter catheter. I.c.v. injection of hypertonic saline (0,3 M; 5gl; n:6) induced an 84% increase of renal sodium excretion (144,4 + 22,5gM/60 re_in before; 268,9 ± 64,5gM/60 min after p<0,05) without significant changes in urinary flow or blood pressure. After injection of the specific AT 1 receptor antagonist, Losartan, into the SFO (5gg/200 nl ; n:8) the natriuresis after hypertonic saline i.c.v, was completely abolished without significant changes in urinary flow and blood pressure. Our results suggest that the centrally induced natriuresis after hypertonic saline is mediated by an angiotansinergic mechanism involving AT 1 receptors in the SFO. On isolated $2 segments of nonperfused proximal tubules of the rabbit kidney the effects of proteinkinase C activating and inactive phorbolesters or proteinkinase C inhibitors like staurosporine on the TEA (tetraethylammonium) uptake were tested. With the experimental conditions chosen the cellular accumulation of TEA reflects the TEA transport across the contraluminal membrane. The phorbolesters were added to the bath solution at 5 different concentrations. The TEA concentration in the incubation solution amounted to 1,375 E-07 tool/1. Phorbolesters significantly influencend the renal organic cation transport. A. time-and dose dependent effect was observed. The phorbolesters tested had a maximum effect at a concentration of 1E-07 mol/1. The highest influence had phorbol-12-myristate 13-acetate (PMA) whitch induced a 7 fold higher TEAaccumulation compared to solvent. The inactive metabolite 4 a phorbol-12, 13 didecanoate (PDD) lead only to a slightly higher accumulation compared to solvent. These effects were abolished by staurosporine. Staurosporin by its own at concentrations <E-08 mol/1 diminished the cellular TEA-uptake, at higher concentrations no effect was measured. A longtime stimulation of proteinkinase C lead to a reduction of cellular uptake. The effect of the stimulated proteinldnase C may be due to a phosphorylation of membrane proteins leading to a stimulation of the basolateral cation transport. Longtime exposure to phorbolesters may lead to a proteolytic degradation of the stimulated proteinkinase C. Inhibition of ANP clearance receptor mediated metabolism of ANP in vivo could be a valuable therapeutic tool for regulating endogenous levels of A/~P. We characterized AP 539 (l-(2-Naphtoxyacetyl)piperidin-4-carbonyl-N -(L)-arginyl-N -(L)-isoleucinyl-N -(L)-aspart~l-N~-(L)-argin~l-N~-(L)-isoleuciny~amid) by ANP C ~eceptor expressing CHO-cells and ANP A receptor expressing COS 7 cells as a specific ANP C receptor ligand. In anaesthetized rats AP 539 (i0, 30 and i00 ~g/ kg/min for 1 h) slightly depresses systolic blood pressure and causes a dose-dependent increase in diuresis, natriuresis, urinary cGMP excretion and urinary sodium/potassium ratio. ANP and cyclic GMP plasma levels were also elevated under infusion of AP 539. BACKGROUND: Measuring Na+K+-ATPase inhibition has become one of the standard methods for the study of endogenous digitalis like factors and certain factors involved in volume homeostasis, blood pressure regulation and several diseases, e.g. essential hypertension. The aim of the present study was to elucidate the effect of storage and processing + + conditions of plasma or serum upon levels of endogenous Na K -ATPase inhibitors. METHODS: Samples of 1) human serum and plasma, or 2) Wistar rat plasma, deprotainized plasma and plasma passed through a C-18 bonded reverse phase column (to free it from ouabain-like substances) were incubated at either 4°C or 25°C. Sample fractions were removed at definite time intervals and stored at -80°C. All samples were extracted on C-18 bonded disposable reverse phase columns. Substances retained were eluted with methanol, evaporated to dryness and reconstituted in Tds buffer. Ouabain-sensitive inhibition of rat brain and rat kidney Na+K+-ATPase was determined in a colodmetric assay using a modification of the Fiske & Subbarow method adapted to the microplate format. RESULTS: We found that Na+K+-ATPase inhibitor levels were stable only in deproteinized plasma. All other groups exhibited inhibition increasing from 0-20% to stable maximum levels between 90% and 100%. Maximum inhibitory activity in samples incubated at 4°C was observed after 3 days but after only 2 hours when incubated at 25°C. Identical results were obtained in pooled and individual samples. Sedal dilutions of extracts revealed dose-response curves steeper than those of cardiac glycosides. DISCUSSION: Our study shows that withdrawal, processing and storage conditions are critical for the reproducibility of Na+K+-ATPase inhibitor determinations.

Thus, deproteinized plasma should be prepared immediately at 4°C and stored at -20°C or colder. However, the existence of a plasmatic precursor of a Na+K+-ATPase inhibitor that is released in vitro cannot be excluded and will require further investigations.

Heidelberg, FRG (BS,MR,RR); Ciba Geigy Ltd., Basle, Switzerland (AD).

The ryanodine receptor is the main calcium release structure in skeletal and cardiac sarcoplasmic reticulum. In both tissues, phosphorylation of the ryanodine receptor has been proposed to be involved in the regulation of calcium release. The ryanodine receptor-calcium release channel complex is also known to be modulated by adenine nucleotides in a calcium-dependent manner. We have examined the ability of the purified cardiac ryanodine receptor to serve as a substrate for phosphorylation by purified PK-A, PK-G, and PK-CaM. A high stoichimetry of phosphate incorporation was observed for PK-CaM (938+48 pmol Pi/mg purified channel protein), whereas the level of phosphorylation was considerably lower with PK-A and PK-G (264+114 and 96+6 pmol/mg, respectively). This implies that PK-CaM is the physiologically relevant protein kinase that is capable of phosphorylating the channel protein at a minimum stoichiometry of 2 mol Pi/mol of tetramer. Additional evidence in support of this hypothesis is provided by the observation, that PK-CaM activity comigrates with the ryanodine receptor over several steps of purification. In addition, the role of adenine nucleotide binding on the ryanodine receptor was studied, using the 2',3'-dialdehyde ATP analog (oATP); oATP induces a calcium release from passively loaded sarcoplasmic reficulum vesicles. This calcium release can be resolved in an initial fast phase with half-lives in the 100-200 millisecond range and a second slow phase (tl/2---1-3 see). The oATP induced calcium release is specifically inhibited by ruthenium red and millimolar Mg ~+ concentrations. Furthermore, we show that [¢t-3~P]oATP is covalently incorporated in the puttied ryanodine receptor. These results unequivocally demonstrate an adenine nucleotide binding site on the ryanodine receptor and make its identification possible. Chinese hamster ovary (CHO) cells were stably transfected with the cloned smooth muscle ~ subunit of the L-type calcium channel in the absence (CHO~) or presence (CHO~B3) of the putative smooth muscle B-subunit. CHO~ and CHO~B3 cells expressed L-type whole-cell Ba 2+ currents. Coexpression of the ~ and B subunit increased significantly the dihydropyridine (DHP) binding sites from 141 to 223 fmol/mg and the maximal inward current from -13.3 to -25.2 ~A/cm 2. The ~ subunit apparently increased the time to peak and shifted the half-maximal activation from +9.5 to +0.7 mV, whereas half-maximal steady-state inactivation was similar in both cell lines. Ba 2+ currents of CHO~ and CHO~6 cells were inhibited by various DHP's including PN200-110 and the phenylalkylamine devapamil (D888). At a holding potential (HP) of -80 mV D888 induced a tonic block with an IC50 of 10~M in CHO~ and CHO~B3 cells. The D888 block was voltage-dependent. At HP -40 mV the ~7~ of D888 was 0.1~M. In the presence of 10~M the half-maximal steady-state inactivation was shifted from +1.5 to -68 mV in CHO~ cells but only from -2.2 to -17 mV in CHO~B3 cells. These results show that the coexpression of the smooth muscle ~I subunit with its putative B subunit effects th~ number of DHP binding sites and conducting channels and reduces the affinity for D888 presumably by changing the inactivation kinetics of the channel. In electrically excitable ceils of vertebrates voltage-dependent Ca2+-channels (VDCCs) modulate physiological processes like excitation-contraction (ECC) coupling, neurotransmitter-release and transcription. The ion conducting subunits (at-subunits) of different types of vertebrate VDCCs have been cloned and s~ueneed and their biophysical and pharmacological properties are currently investigated in heterologous expression systems. VDCCs have also been identified in invertebrate tissues where they seem to play an important role for ECC by controlling neurotri~n, smitter (e.g. glutamate) release in the neuromuscular Junction and CaZ'r-influx into muscle cells. The structure of invertebrate Ca2+-channels is unknown and none of their subunits have yet been cloned or expressed. Here we report for the first time the cloning and characterization of an invertebrate Otl-SUbunit from Musca domestica larvae. We isolated a eDNA clone encoding a 1646 amino acid polypeptide (Mdl~l; calculated Mr: 189,308), which shares sequence homology with 1,4-dihydropyridine (DHP)-sensitive Ltype (51 -55 %) and DHP-insensitive Otl-SUbuhits (35 -37 %). Mdlot 1 mRNA and the polypeptide itself are almost exclusively expressed in the larval body wall musculature as shown by whole-mount in situ hybridization, semiquantitative polvmerase chain reaction and immunostalning with a sequencedirected antibody. This ~l-subunit forms functional voltage dependent Ca 2+channels after heterologoffs expressiQn of Mdlt~ 1 cRNA in Xenopus ooeytes.

Reversible binr~ng studies with (-)-[3H]desmethoxyverapamil and photoaffmity labelling with [ H]azidopamil direefly,~llowed to identify a 133 kDapo.lypeptide that brads phenylalkylamine CaZ~--antagonists with a CaZ'r-charmel characteristic 1~.harmacologieal profile. However, Mdlt~t is distin~et from this 133 kDa polypeptide as demonstrated by immunoblotting anal photoattlnit)' labelling. Our data there£o.re provide biochemical evidence for the existence ot at least two sp~i.es of CaZr-channels in Musca domestica larvae. As voltage dependent CaZ'r-ehannels in the insect mnseulature seem to play a central role for ECC, the Mdlal-SUbanit may represent a potential molecular target for a novel generation of insecticides. Two fluorescent, high affinity ligands for L-type Ca2+-channels that are selective for the 1,4 dihydropyridine binding domain (4,4-Difluoro-5,7-dimethyl-4-bora-3a,4a-diaza)-3-(s-indacene)propionic acid, coupled to (+)-and (-)-l,4-dihydro-2,6dimethyl-4-(2-trifluoromethylphenyl)-3,5-pyridinecarboxylic acid 2-(aminoethyl) ethyl ester hydrochioride, (+) and (-)DMBODIPY-DHP) and the phenylalkyl-amine binding domain (5-{3-[3-(4,4-Difluoro-5,7-dimethyl-3a,4a-diaza-4-bora-indacene-3yl)propionarmdo]phenethyl-N-methylamino } -2-isopropyl-2 -(3,4,5 -trimethoxy-phenyl)-valeronitfile, racemic (R,S) DMBODIPY-PAA), respectively, serve as acceptors (emission wavelength 517 nm) in fluorescent resonance energy transfer (FRET) when tryptophan residues of the pore forming Otl-subuint of L-type Ca2+-charmels are excited at 285 nm and the ligands are specifically bound. FRET therefore allows to study Ca2+-channel kinetics , equilibrium binding studies or to investigate the interaction of drugs that interact in a heterotropic allosterie manner without seperation of bound and free ligand. The application of this novel method for rapid drug screening, kinetic properties of the Ca2+-charmels and its stability will be demonstrated.

The experiment was performed with Ca 2+channels (383 pmol/mg protein) at 22°C. The initial concentration (Lo) of the ligand (-)DMBOD~PY-DHP was 17 nM and of the protein 9.2 ~tg/ml (Ro = 3.3 nM).

[LR],at equ,ilibrinm = 1.3 nM; k+l = 0.0149 rtM -t rain "t. Dissociation of the ligand was initiated by 1.5 ~tM (+).[sradipine. Best fit parameters for the dissociation kinetics were: k 1 = 0.437 rain "1 (34%); k_ 2 = 0.084 rain "1 (6ff~). 

The role of endothelium for the regulation of vascular tone by production of vasorelaxing and contracting factors is well known. EDRF is released continuously under basal conditions or can be stimulated by different endothelial agonists. The influence of calcium channel blockers on the calcium dependent formation of EDRF is controversly discussed. In the present study we investigated the vascular activity of different C 2 substituted 1,4-dihydropyridines (DHP) on isolated porcine basilary an~ coronary arteries in the presence and absence o f intact endothelium. Isolated ring segments of the arteries bathed in oxygenated Krebs-Henseleit-solution were precontracted with PGF2a. Endothelial function was checked by substance P induced endotheli~.fia dependent relaxation (+~n, do). Endothelium was mechanically damaged or blocked by L-N nitro arginine, a selective inhibitor of the EDRF-synthesis (-Endo). After a second precontraction with PGF2a, nitrendipine (NTD) and its 2-phthalinudomethyl (I 1990 The data show that, inhibition of EDRF-synthesis leads to a profoundly loss in vascular activity of the DHP studied. The degree of reduction varies with the structure of the DHP and is tissue dependent. This indicates that (1) the loss of basally released EDRF may not exclusively be responsible for the decreased vasorelaxing effect and (2) The effects of the new calcium-antagonist CRE 319 against isehemia and reperfusion-indueed arrhythmias and ereatine kinase (CPK) release were investigated using anaesthetized rats (60 mg/kg, pentobarbitone i.p.) with transient (7 min) eoronasy artery occlusion. The hemodynamie parameters such as heart rate (HR), systolic and diastolic blood pressure (BPs,BPd), pressure rate product (PRP)were measured. The test drugs or vehicle were administered Lv. 10 mi~ prior to occlusion (bolus). Each group consisted of 16 rats. CRE 319 in two doses (2.5 and 10 I~g/kg) completely prevents mortality (Mortality Index, MI reduction from 50 to 0%). Moreover, CRE 319 in the dose of 5 lag/kg reduced signifieantiy MI from 50 to 6%, whereas Nifedipine(NIF) and Nimodipine (NIM) in the same dose abolished MI to 18 and 25%, respectively. Reperfusion-indueed ventricular fibrillations (VF) were observed in 100% of control, whereas such episodes occured in 0, 13 and 12% of the rats treated with three doses of CRE 319, respectively. In eompaxison, NIF reduced VF-incidenee to 30% and NIM to 33%. CRE 319 in all doses markedly reduced also the incidence of ventriculas taehycardia (VT) as follows: from 100 to 25, 53, 62%, respectively. In contrary, NIF and NIM do not prevent VT-ineidanee at all. CRE 319 in all doses significantly reduced V'F duration, ff it occured (VF reduction from 28.8 to 0,2,3.2s, respectively). In comparison, NIF reduced VF duration to 4.0 and NIM to 5.8 s. CRE 319 significantly reduced also VT-duration in two doses (2.5 and 10 ~tg/kg)-reduetion from 42.6 to 10.5 and 20.3 s, respectively, whereas only NIF reduced VT-duration to 24.8 s.CRE 319 in the lowest dose significantly diminished CPK activity in plasma (from 20.5 to 12.9 U/gProt.) and only NIF exerted similar effect (13.6 U/gProt.).

In conclusion, CRE 319 is a very effective drug in protecting the animals against reperfusion-indueed arrhythmias, more powerful than Nifedipine and Nimodipine, which also exert eytoproteetive properties. The discouraging outcome of clinical intervention trials with phosphodiesterase(PDE)inhibitors (e.g. promise-study) frequently is ascribed to the proarrhythmic and energy-"wasting" effects of PDEI|l-inhibition. In order to quantify the functio'aal and energetical consequences of inotropic stimulation with the PDE-inhibitor Milrinone (Mil) and ~he Ca-channeI-agonist Bay K8644 (Bay), beth compounds were compared in chronically failing (LAD-ligature, 8wk, n=16) and sham operated (n=18) rat hearts. These were orthogrede[y peffused with K.H.-solution (32C, preload 4mmHg, afterload 66mmHg, Ca 0.9raM, 4Hz). The infarct size was around 30-50% of the LV circtmaference. In-rive, LVEDP was the most prominentty changed hemodynamic parameter (5 to 13 mmHg), in the isolated hearts, Bay increased LV power output (P) concentration-dependently (l-30nlvi) by up to 150mJ/min in sham and t 2{) red/rain in failing hearts. The EC50 appeared unchanged at about 5nM. Mil inreased P by 150m J/rain in sham but only 80 mJ/min in LAD-ligated hearts (.EC50: 40uM, unchanged). At power increments up to about 100m J/rain the oxygen consumption under Mi[ stimulation was neariy twice as high as under Bay stimulation in sham as well as in LAD-ligated hearts. Thus, Mil significantly lost inotropic efficacy in chronically failing rat hearts. This was not observed with Bay. A di~wn-regulation of the cAMP-generating system appears responsible for the loss of efficacy of PI) E-inhibition since a decrease in B-receptor density as well as the inotropic efficacy of l soprenaline could recently been demonstrated in the same animal medel. The increase in oxygen consumption during equieffective stimulation was sigqaificantly more pronounced with Mi[ than with Bay. The efficiency of heart work under Mil stimulation was comparable to that with Isoprene[inc. Bay stimulation was energetically comparable to a Ca-elevation in the perfusate (data not shown). Thus, in chronically failing rat hearts, under conditions of constant afterload and constant heart-rate, direct Ca-elevation by a Ca-channel agonist appears superior t~ cAMP-elevation by PDEinhibition with regard to inotrcpic efficacy as well as energetics. 

We investigated the effects of constituents of clove oil, eugenol (1), aceteugeno] (2), /~-caryophyllene (3) and caryophyllene oxide (4), on the catecholamine-dependent strong rested-state contraction of guinea-pig papillary muscles. This contraction was reduced by all four compounds with the respective EC50 values (/~mol//): 183, 300, 270, 29.

Since, in the absence of POTENTIAL ( calcium inward current was concemration-dependently and reversibly reduced by both substances.

(4) was about six times more potent than (1), the respective EC50 values being 32 and 200,umol/1. The effect was due to a diminution of ICa at all membrane potentials tested (see Fig.) without a shift of the voltage-current relationship. Thus, the constituents of dove oil, eugenol and caryophyllene oxide must be considered as calcium current inhibitors. 

By simultaneous registration of the appropriate parameters dose response curves were established for the inotropic, chronotropic, dromotropic and coronary dilating effects of these dihydropyridines. All substances examined induced a concentration-dependent reduction of the contractile force (CF). NTD 1 proved to be the strongest negative inotropic calcium antagonist (EC50cF = 5.3 .10 -8 molll) followed by NITR (ECSOcF = 1.6 • 10 -7 tool/I), NTD 4 (ECSOcF = 9.5-10 -7 tool/I) and NTD 3 (ECSOcF = 8.3. 10 -6 tool/I) in descending order. The corrected coronary flow (CFI) was increased by all dihydropyridines in a concentration-dependant manner, and again NTD 1 turned out to be the most effective substance (EC50cFI = 5.0 .10 -9 molll) followed by NITR (EC50CFI = 2.7.10 -8 mol/I), NTD 4 (EC50cR = 4.8.10 -8 mol/I) and NTD 3 (EC50CFI = 3.7.10 -7 molll) in descending order. The heart rate (HR) was decreased up to a maximum of -35 % by NTD 1 and NITR and the negative chronotropic effect of NTD 4 and NTD 3 was even less pronounced (NTD 4:-25 %, NTD 3:-18 %). Regarding the AV conduction time (AV) NTD 3 and NTD 4 showed no significant effect at all, whereas the negative dromotropic effect of NTD 1 and NITR reached significance under the influence of the respective maximum concentration (NTD 1:+20 %, NITR: +14 %). The vascular selectivity, expressed by the EC50c#EC50cR ratio, was most distinct by NTD 3 (22.7) and NTD 4 (19.8) followed by NTD 1 (10.5) and NITR (5.9).

MODULATION OF INTRACELLULAR CALCIUM BY EXTRA-CELLULAR MAGNESIUM W. Vierling + , A. Stampfl ++ As shown earlier, magnesium exerts a negative inotropic effect in heart ventricular muscle and reduces the calcium inward current (Dichtl and Vierling, Eur. J. Pharmacol., 204, 243, 1991) . Accordingly, we found that Mg 2+ diminishes the stimulation-dependent transient increase in intracellular free calcium concentration. The ratio of the fluorescence at 340 and 380 nm wavelength (R340/380) was determined as a measure of [Ca2+] i in encymatically isolated, fura-2loaded single myocytes from adult guinea pigs. The maximum R340/380 during the transient was diminished from 0.9 to 0.76 after elevation of [Mg2+]o from 1.2 to 9.6 mmol/1. In contrast, the diastolic [Ca2+]i did not change after magnesium elevation. Also omission of Mg 2+ had no detectable influence on resting [Ca2+] i at normal (140 mmol/1) sodium concentration. However, there was a strong increase of R340/380 (from 0.46 to 0.9) during rest in magnesium-free solution after sodium withdrawal even at a very low [Ca2+]o (50/~mol/1). Readdition of 1.2 mmol/1 Mg 2+ led to a remarkable decrease of R340/380 (to 0.52). Control experiments in Mg2+-containing solution showed only a slight increase in resting [Ca 2 +]i after Na + withdrawal. It is concluded that, in heart ventricular muscle, magnesium strongly influences [Ca2+] i by reducing the stimulation-dependent calcium influx into cell but probably also by activating cellular calcium efflux, particularly at an impaired sodium-dependent calcium elimination. 

The antiischemic effects of the phenylalkylamine Ca2+ antagonist emopamil in experimental brain isehemia prompted us to investigate the pharmacological and biochemical properties of a high affinity, cation-sensitive binding site for (-)-[3H]-emopamll in guinea-pig brain and liver (Zech et al., Eur. J. Pharmacol. 208:119-130, 1990) . Several drugs with reported antiischemic effects in heart or brain (e.g. amiodarone, trifluperazine and opipramol) inhibited (-)-[3H]emopamil binding with high affinity (IC5o values < 30 nM). Photoaffinity labeling with the arylazide phenylalkylamine [3H]azidopamil allowed us to identify this binding site on a 22 kDa polypeptide (designated emopamil binding polypeptide, EBP) in the endoplasmie reticulum of liver, kidney, lung and adrenal gland. The digitonin-solubilized 22 kDa polypeptide fully retained its pharmacological properties and co-migrated with (-)-[3H]-emopamil binding activity upon sucrose density gradient centrifugation and gel filtration. From the sedimentation coefficient (% 20 = 12.0) and Stokes radius (6.0 run) a molecular mass for the protein detergan[ complex of > 280 kDa and for the protein moiety of > 84 kDa was calculated. Therefore the EBP must be associated with itself or other proteins in an oligomeric complex. Partial purification of the solubilized EBP was achieved by ion exchange chromatography, sucrose density gradient centrifugation and denaturating gel filtration.

[3H]Azidopamil also specifically photoaffinity labeled a 27 kDa polypeptide, which, in contrast to the EBP, bound sigma "receptor" ligands (haloperidol, 1,3ditolylguanidine) with nanomolar affinity. The 27 kDa polypeptide and the EBP share several biochemical and pharmacological properties (similar molecular weight, tissue and subcellular distribution, photoaffinity labeling by [3H]azidopamil, high affinity for the antidepressant opipramol). We therefore propose that the EBP represents a novel sigma "receptor" subtype, which could play an important pathophysiological role in the development of ischemic tissue damage. Effects of Dotarizine 0.3 mg/kg vs. saline (n=10 each) i.v. on cerebral blood flow (CBF) were tested at various arterial carbon dioxide tensions (PaCOz) in lightly anaesthetized rats. CBF was recorded as the dorsal sagittal sinus puncture outflow per unit time. In parallel, arterial and sagittal sinus blood gas and acid-base status, and cardiovascular, respiratory and electrocorticogram (ECoG) functions were monitored. In saline controls (CON), a 2.09 fold PaCOz increase raised CBF by a factor of 1.87 ±0.71 (n=lO) at a mean arterial blood pressure (MAP) of 93.6 ±31.8 mmHg and an ECoG spiking rate of 172.2 ± ll0.4/min, yielding CBF= 1.98-PaCO2+24.5. Five min after Dotarizine, a 2.24 fold increase in PaCO2 raised CBF by the factor 2.31 at MAP 70 ±28.5 mmHg and ECoG spiking of 176.8 ±85.9/min, giving CBF=2.36-PaCO2+4.32. Subsequent stepwise PaCO~ increase led to control rat CBFco~=2.32-PaC02 + 85.4 (r=0.5439; 54 measuring points at any MAP; equal slope found by othersl). Dotarizine led to the linear regression CBF~o~:5,22.PaCO~-83.5 (r=0,668; 33 points at any MAP). Effects on cerebral arteriovenous difference in 02 content (avD02), cerebral metabolic rate of 02 (CMR-02) and ECoG are described and discussed. Dotarizine raised the cerebral perfusion autoregulatory response to PaCO2 changes without depressing ECoG activity; it prevented the hypercapnia-induced reduction of CMR-O2 found in controls. Dotarizine (0.05 mg/kg'min; 30 min) infusion to rats under normocapnic conditions produced weak peripheral vasodilation (hardly affecting MAP) and clear-cut cerebrovascular dilation (CBF increase) concomitant with CMR-Oz increase z. 

T. Ginap,Y, vanVelthoven,~, ~heremet, D,d, DosleyandT.d. Feuers~in

In recent years the existence of four types of voltage-sensitlve calcium channels (V$CCs), designated L, N, P and T, hove been defined which differ in their conductances, rates of activation and inactivation, and pharmacology. Phenylalkylamines (e.g. verapamil), benzothiazepinas (e.g. diltiazem) and 1,4-dihydropyridinas (DHPs) (e.g. nifedipine) are the most we11-known groups which hove been found to interact at pharmacologically distinct sites on the L-YSCC. Despite the presence of a high density of binding sites for L-VSCC ligands in mammalian neuronal tissue, the role of L-Y$CCs with respect to nerve terminal function is unclear. Feuerstein et al. ( 1991, Eur. J. Pharmacol. 198 : 37-42) have shown that a reduction in depolarization intensity (from 26.2 to 17,2 mM K + applied for B rain) or the (~-conotoxin GVIA-mediated block of N-V$COs (which predominantly subserve neurotransmitter release from nerve terminals) enabled L-V$CC activators to enhance serotonin release from rabbit hippocampal slices. Both the non-dihydropyridine FPL 64176 ((2,5-dimethyl-4-carbomethyl-3-(2-benzyl)benzoyI) pyrol) and BAY K 8644 produced a similar concentrationdependent enhancement of 15 mM K+-ovoked [:~H]-NE release from human and rat neocorticel slices. At I pM a 70-80% increase above control values was observed. The L-VSCC blocker isradipine (ISR) ( O, I pM) completely antagonized the effact of both FPL 64176 and BAY K 8644, implying a direct interaction with neuronal L-Y$CCs. At a stimulus of 25 mM K + BAY K 8644 ( I IxM) did not significantly enhance [3H]-NE release from rat neocortical slices. At the same concentration of K +, FPL 64176 ( I pM) was also almost ineffective in enhancing [3H]-NE release. Overall, the effects in human and in rat tissue were very similar. FPL 64176 was a very weak inhibitor of [3H]-ISR binding to rat neocortical membranes with an ICso of 1411M. In contrast, Bay K 8644 was at least 300 times more potent in this regard with an ICSO of 44nM. These findings indicate that FPL 64176, a novel non-dihydropyridine L-VSCC activator being structurally unrelated to DHPs, may enhance depolarizationevoked [3H]-NE release like BAY K 8,644. FPL 64176 appears not to interact directly with the DHP binding site of L-Y$COs and can serve as an additional pharmacologic teal to elucidete the function of neuronal L-Y$CCs. The frequency-dependence of the class III action of dofetilide (D) has not been characterized in detail. In particular, it has not been established wether D exhibits reverse frequency-dependence that is a decreasing class III action with increasing stimulation rate. Hence, the frequency-dependence of the effects of D on QTinterval and refractory period (RP) were examined in isolated electrically stimulated guinea pig hearts (Langendorff preparation, 35sC) , and its influence on the antifibrillatory efficacy of D was evaluated by the ventricular fibrillation threshold (VFT). At a concentration of 0.1 pM D increases QT from 276+7 to 333+8 ms and RP from 238+9 to 318+14 ms at a stimulation rate of 60 bpm (p<0.05), At a stimulation rate of 240 bpm, QT as well as RP are no longer significantly prolonged (QT: 188+4 vs. 195+7 ms; RP: 171 4-7 vs. 169_+15 ms), However, this reverse frequency-dependence (RFD) occurs with slow offset kinetics corresponding to an offset rate of 0.0931 _+0.007 per beat. According to the RDF of its class III effects, D increases the VFT significantly from 70+3 to 90+3 mA (p<O.05) during stimulation with 60 bpm, but has no effect on the VFT at a stimulation rate of 240 bpm (36+2 vs. 35+3 mA). However, an increase of the VFT can be also observed upon a shortlasting increase in stimulation rate. Thus, when the VFT is determined at a stimulation rate of 60 bpm following the induction of up to 12 extrasystoles at the closest coupling intervals possible, D causes an increase of the VFT from 47-+4 to 63_+7 mA (p<O.05) despite its RFD. Hence, the slow offset-kinetics stabilizes the class III effect of D during several beats even at a high stimulation rate. Conclusion: The present study demonstrates that the RFD of a class III antiarrhythmic agent can be overcome in part by slow offset kinetics. This results in antifibrillatory efficacy even in the presence of salvos of very closely coupled extrasystoles. Potassium channel openers are believed to exert smooth muscle relaxant activity by hyperpolarizing the cell membrane potential which in turn shuts off voltage dependent calcium influx. Up to now a causal relationship between hyperpolarisation and changes of contractility remains elusive. 1,6-dihydro-l-[2-(3-pyridinylcarbonylamino)ethyl]-6 -oxo-3 -pyridine carboxylic acid methylester ($1818) and its N-methylamide derivative (S 3935) are novel agents found to exert potent hyperpolarisation in smooth muscle cells of guinea pig pulmonary artery. The conventional glass microelectrode method was used. The shifts of membrane potential were 14 mV at 10 -7 M and 29 mV at 10 -6 M for S 1818, and 26 mV 10 -7 M and 28 mV at 10 -6 M for S 3935. The known potassium channel openers cromakalim and rilmakalim (HOE 234) provoked corresponding shifts of the resting membrane potential at the concentration of 5,10 -7 M by 15 mV and 18 mV, respectively. In all experiments glibenclamide (10 -7 -10-6 M) antagonized the respective hyperpolarisations suggesting that S 1818 and S 3935 are potent potassium channel openers. In guinea pig pulmonary artery contracted by 25 mM KCI cromakalim and rilmakalim exhibited powerful antispasmogenic effects. Surprisingly, the novel agents S 1818 and S 3935 completely failed to antagonize contractions up to a concentration of 10 -5 M.

In guinea pig ileum contracted by various agents such as angiotensin II, acetylcholine, BaCI2, histamine, serotonine, and prostaglandin F2~, cromakalim and rilmakalim exhibited relaxing effects whereas $1818 and $3935 again turned out to be ineffective up to the highest concentration tested (10 -6 M). These results suggest that hyperpolarisation of the cell resting membrane potential does not necessarily influence contractility of smooth muscle cells. The new diazabiclononan derivative bertosamil strongly delayed repolarization of action potentials in rat ventricufar muscle: 10pmol/l caused a prolongation of action potential phase 1 to 810% of control measured at a repolarization level of 0mV. In order to clarify the underlying mechanism effects of bertosamil (0.1-10gmol/l) on repolarizing transmembrane currents in rat ventricular myocytes were analyzed. Single ventricular cells were obtained by a standard enzymatic procedure. Ionic currents were measured with the whole cell voltage clamp technique. The transient outward current (ito) was activated at positive test potentials from conditioning potentials of -80 to -120mV. The amplitude of the peak outward current was decreased in a concentration dependent manner. The threshold concentration was 0.1~mol/l and half maximal ito-inhibition occurred at 3gmol/l bertosamil (n=5). Besides the decrease of ito-amplitude bertosamil accelerated the time dependent inactivation of ito: 3pmol/l bertosamil reduced the inactivation time constant to 17ms (control 43ms), corresponding to a decrease of the current-time integral to 19% of control (integration period 0-50ms). The voltage dependent inactivation of ito was not affected by bertosamil. The delayed outward rectifier (iK) , measured at the end of the one second test pulse was only slightly affected by bertosamil: it decreased to 77% of control at 3pmol/l (n=4). The inwardly rectifying K-current (iKl) and the calcium current (isi) were moderately decreased: 3@mol/l bertosamil reduced iKl to 84% (n=5) and isi to 85% of control (n=3), respectively. In conclusion, the strong inhibition of repolarizing K-currents by bertosamil is suggested to be the predominant mechanism which leads to prolongation of action potentials in rat ventricular cells. Preconditioning (PRECON), which consists of a brief period of ischemia and reperfusion, reduces infarct size subsequent to a longer lasting ischemia by increasing myocardial tolerance to injury and cell death. The contribution of KATp-Ch to this cardioprotective action was investigated in anesthetized open-chest pigs using the KATp-Ch opener bimakalim (BIM) and the KATp-Ch blocker glybenclamide (GLY). All groups of animals were subjected to lh occlusion of a distal branch of the LAD coronary artery (TOC) followed by 2h reperfusion. Controls (CO) were studied without additional interventions. PRECON (10 min coronary artery occlusion + 15 min reperfusion) or PRECON subsequent to GLY infusion (1 mg/kg/10 rain, iv) was performed just prior to TOC in two series of experiments. BIM (3 # g/kg bolus + 0.1 #g/kg/min, iv) was infused either for 15 min just before TOC (BIM A), or for 75 rain starting during TOC 15 rain prior to reperfusion (BIM B). The area at risk (AAR) was determined using fluorescein dye injected into the left atrium after reocclusion of the coronary artery and expressed as percent of left ventricle (LV). The infarct size (IS) was stained by p-NBT and indicated as percent of area at risk (AAR Whole-cell patch-clamp recordings were performed on single RINm5F cells at about 35 ° C. The patch pipette was filled with KCl-solution (in namol/l: 140 KCI, 10 NaC1, 1.1 MgC12, 0.5 EGTA, 1 Mg-ATP, 10 HEPES, pH=7.2, adjusted with KOH) and the bath contained NaC1solution (in mmol/l: 140 NaC1, 4.7 KC1, 1.1 MgC12, 2 CaCI2, 10 HEPES, pH=7.4, adjusted with NaOH). The cell resting potential was variable between -20 and -60 mV (mean -45 + 1, n = 122) and frequently depolarizing spikes were observed. Addition of 100 #reel/1 diazoxide to the bath hyperpolarized the potential to -70+1 mV (n=141) and abolished all potential fluctuations. The increase in cell resting potential was associated by an increase in whole-cell current both in inward and outward direction. In order to stabilize the cell potential, all experiments were performed in the presence of 100 tzmol/1 diazoxide. Addition of glibenclamide depolarized the cell potential dose dependently (IC50=9 nmol/1, Hill-coefficient 0.8). Even at maximal dose (100 /zmol/1) cell potential was not completely inhibited (-21+4 mV, n=5). The effects of glibenelamide were poorly reversible. Similar results were obtained with glimepiride (Hoe 490, IC50 = 14 nmol/1, Hill-coefficient 0.7). Hoe 511 (6-chloro-3a-hydroxy-3 -methyl-2 -methylimino-indano < 2,1B >thiazolidin-hydrochlorid, which was shown to have anorectic effects in experimental animal, caused a dose dependent depolarization of cell potential (IC50=37 #reel/l, Hill-coefficient 1.4) and a decrease of whole-cell current. With 1 mmol/1 of Hoe 511 the cell potential was completely and reversibly inhibited. It can be concluded that the cell resting potential in RINm5F cells is only partially determined by glibenclamide-sensitive K+(ATI~) channels. Hoe 511 blocks this channel type as well as other K + channels, which participate in generating the cell potential. Rilmakalim (HOE 234), a potassium channel opener, is effective against contractile response to asthma mediators in guinea pig airways and acts on acute existing bronchospasms. It has been shown that bradykinin (BK), neurokinin A (NKA), and PGE 2 are potent stimulators of luminal chloride secretion in tracheal epithelium probably by increasing production of intracellular cAMP. Moreover, increased epithelial chloride secretion is an early event of the response to specific allergen challenge in the airway mucosa. The aim of this study was to investigate if the BK-, NKA-, or PGE2-induced chloride secretion can be influenced by the potassium channel openers rilmakalim and cromakalim. For this purpose the membranous portions of the dog trachea were dissected and mounted in the Ussing chamber. Short-circuit current (ISC), a measure of the luminal chloride secretion, and membrane conductance were measured. Rilmakalim reduced the basal epithelial chloride conductance, whereas cromakalim was a potent stimulator of epithelial chloride secretion when applied to the mucosal side of the tracheal tissue. Preincubation of tracheal tissues with rilmakalim (10 -6 -10 -4 mol/I, mucosally) dosedependently decreased the short-circuit current after addition of BK (10 -6 mol/I), NKA (10 -7 mol/I) and PGE 2 (10 -6 mol/I), the IC50 values being about 5x10 -5 mol/I). Cromakalim was as effective as rilmakalim against PGE 2 stimulation but was ineffective in the presence of BK and NKA. Rilmakalim is significantly less effective when applied to the serosal side of the tracheal tissue.

In conclusion, the potassium channel opener rilmakalim reduced the basal chloride conductance and significantly decreased BK-, NKA-, and PGE2-induced chloride secretion in canine tracheal epithelium. The predominant site of action of rilmakalim is the luminal side of the tracheal tissue, indicating its interference with chloride channels. In previous studies with inside-out membrane patches exposed to nucleotide-free solutions, saturating concentrations of sulfonylureas did not cause complete block of the ATP-dependent K + channel (KnT1,-channel) of pancreatic B-cells. We therefore examined the regulation of the effectiveness of sulfonylureas in inside-out patches of naouse pancreatic B-cells. Channel inhibition by saturating concentrations of tolbutamide (1-5 mmol/1) remained incomplete even in the presence of submaximally inhibitory concentrations of ATP (20 ~mol/l), adenylyl-i~midodiphosphate (AMP-PNP, 30 ~mol/1) , adenosine-5'-O-(2-thiodiphosphate) (ADP-I~-S, 50 ~tmol/l) or ADP (100 ~tmol/1) at the cytoplasmic face of the membrane. The inhibitory effects of tolbutamide were reversible both in the presence and in the complete absence of Mg 2+ ions. The findings suggest that sulfonylureas are partial agonists the mechanism of action of which does not involve alteration of protein phosphorylation of the KnTt,-channel and / or regulatory proteins. Oncogenesis in vivo and transformation in vitro induced by Rous sarcoma virus (RSV) is strictly linked to the expression of the v-src oncogene. The v-~rc oncogene product pp60 v-src is the prototype of a family of nonreceptor tyrosine-specific protein kinases. Transformation mediated by pp60 v-~re is associated with an increase in tyrosine phasphorylation of a large number of cellular proteins. Yet, only few of these proteins have been identified with respect to their physinlogical function, and very few of them seem to be directly involved in the transformation process. However, there is some evidence that those substrates of pp60 v-~rc that might be critical for transformation are located at the plasma membrane. Using the whole-cell patch-clamp technique, we found that the K ÷ current propotties of normal and of RSV-transformed chicken embryo fibroblasts (CEFs) show dramatic differences. In nontransformed CEFs the main membrane current is a delayed outward K ÷ current which is tetraethylam-..........

momum-sens~t~ve, but 4-arnmopynduae-ms_ens~tive. This K current ~s almost independent of the intracellular Ca z÷ concentration and becomes completely inactivated at positive membrane potentials. In contrast, in transformed CEFs this K ÷ current changes into a ~aninactivating current thai; now strongly depends on the intracelluler CS z÷ concentration. This CaZ÷-dependent K ÷ current is blocked by the scorpion toxin charybdotoxin (ICSO: 19 nM)~ whereas the K ÷ current of normal CEFs is insensitive to charybdotoxin (up to 300 raM). The same K ÷ currant alterations were found after microinjection of purified, enzymatically active pp60 v'are into normal CEFs, but not al~er infection of CEFs with the Rous-associated virus RAVS, which lacks the v-~r¢ ancogene. Our results surest that the ancogene product ppS0 v-arc modulates existing K ÷ channel proteins~ leadin~ to profound electrophysiolngical and phm'macological alterations of K ÷ currant properties in RSV-transformed CEFs. Furthermore, our experiments identify for the first time K ÷ channels as possible substrates of the oncogene product ppS0 v-arc. Frequency-dependent depressant effects of a drug on slow channels in the AV node are important in its efficacy against supraventdcular tachycardias. The purpose of this study was to compare the use-dependent effects of both drugs on AV nodal conduction in isolated guinea pig hearts perfused by the method of Langendorff. Adenosine 3 /zM and verapamil 0.01 #M caused a comparable prolongation of the atrioventricular conduction time and reduction of sinus rate. The time dependence of drug induced changes in atrioventricular conduction was characterized after abruptly changing the atrial pacing rate. The basic cycle length was abruptly shortened from 240 ms to 180 ms. The resulting time constant for adenosine ('c=467_.+187 beats; ~+S.D.) was significantly (9<0.05) longer than that for verapamil (x=264_ 121 beats).

In conclusion adenosine had more pronounced frequency-dependent effects on atdoventricular conduction than the calcium channel blocker verapamil. This may result in a higher clinical efficacy of adenosine in supraventricular tachycardias where the AV node forms a part of the reentrant circuit. The effect of carbachol on force of contraction (contraction frequency 0.5Hz) and on resting intracellular Na ÷ activity (ai,a) of guinea-pig atrial trabeculae was investigated before and after pretreatment of the animals with pertussis toxin (PTX 120 Bg/kg i.v.; 24 hr). In the absence of FTX, IBM carbachol decreased atrial force of contraction (F¢) to 6.8±2.3% of the predrng value (n=22). This effect was not significantly changed by increasing the carbachol concentration to 300~M. After pretreatment with PTX, carbachol produced a concentration dependent increase in Fc which was significant at 10BM and maximal at 100BM (138.3 ± II.3% of control; n=4). In 9 atrial trabeculae, resting membrane potential (Em) and aisa as determined before and in the presence of 100BM carbachol by means of conventional and ion-sensitive microelectrodes was -77.8±0.4mV and -81.7 ± 0.SmV (P<0.05) and 8.3±0.2 and 8.5±0.4mM, respectively. In 4 PTX-pretreated preparations, E® (-79.2±0.6mV) did not hyperpolarize upon addition of 100BM carbachol (-79.5±0.5mV) but aisa rose significantly and reversibly from 8.1±0.4 to 9.0 ± 0.2mM. After impairment of the Na ~ pump activity of PTX-pre treated trabecnlae with 10,M dihydroouabain (DBO), 100,M carbachol increased aisa by 1.6±0.5~ (n=4) as compared with 0.gmM in the absence of DHO. In 5 atrial trabeculae which were not pretreated with PTX, 30~M DMO increased a*sa from 8.6±0.6 to 10.2±0.SmM and depolarized E® from -78.6±0.9 to -77.1± 0.4mV. Additional superfusion of the preparations with 300~M carbachol produced no significant change in ats. but hyperpolarized Em significantly to -82.5±0.3mV (P<0.05). In contrast to ventricle (Korth & KOhlkamp, Pflfigers Arch 403:266,1985) , PTX-catalyzed ADP-ribosylation of Gi/Go inhibitory G proteins is essential for the carbachol induced increase in a~,~ in atrial cells. The rise in a~,a is probably responsible for the positive inotropic effect of carbachol in PTX-pretreated atria. Endothelin (ET) is not only a very potent vasoconstrictor agent but has also been shown to induce positive inotropic effects in the heart of several species, including humans. In this study we investigated to which signal-transduction pathway(s) ET couples in the human heart. For this purpose we determined in right atrial appendages from patients without apparent heart failure undergoing coronary artery bypass grafting the effects of ET on adenylyl cyclase (AC) and inositol phosphate ~IP) ~eneration. In right atrial slices ET-I (i0-~-i0 -~ mol/l) concentration-dependently stimulated IPgeneration; the pEC50-value was about 7.6. ET-3 was approximately 30 times less potent than ET-I indicating that the ET-receptor involved is of the ET Asubtype. In right atrial membranes ET-I and ET-3 did not affect basal AC-activity. On the other han~, both ET-I (io-ll-10-9mol/l) and ET-3 (10 -9-10-Vmol/l) concentration-dependently inhibited iO~mol/l isoprenaline-and l~mol/l forskolinstimulated AC-activity; ET-I (pECs~-value = 9.5) was approximately I00 times more p6tent than ET-3 (pECso-value = 7.25) indicating that the ET-induced AC-in~ibition is also mediated by an ETa-receptor. We conclude that in the human right atrium ETA-receptors exist that couple to two different signaltransduction-pathways:

in low concentrations to inhibition of AC-activity and in (50-100 times) higher concentrations to stimulation of IP-generation. 

In connection with a programme on assessing the responsiveness of hypertrophied hearts to adrenoceptor stimulation, we investigated the functional responses of isolated hypertrophied hearts 'taken from spontaneously hypertensive rats (SHR) and from rats with induced aortic stenosis (ASR) of 18-20 and 32-34 weeks of age to the c~ 1adrenoceptor agonists cirazoline, methoxamine and phenylephrine and compared them with those from age matched Wistar Kyoto rats (WKY) and "sham" operated controls. The influence of cardiac hypertrophy on al-adrenoceprer quantity was investigated by means of radioligand binding experiments. The ASR showed a pronounced degree of cardiac hypertrophy, which increased progressively with age. The inotropic responses to the a 1-adrenoceptor agonists cirazoline (Ernax 15 ± 2.1% and 80 + 8.1% increase, respectively) and methoxamine (Ema x 25 ± 3.6% and 75 ± 7.4% increase, respectively) were strongly reduced in ASR-hearts compared to control hearts. The inotropic response to phenylephrine, however, remained unchanged in ASR of 18-20 weeks of age. In ASR-hearts of 32-34 weeks of age, however, the response to phenylephrine also proved impaired. The al-adrenoceptor density was reduced in hearts taken from ASR of 18-20 weeks of age when compared to hearts taken from control animals (Brnax 138 ± 8.0 and 154 ± 4.2 fmol/mg protein, respectively) and diminished even further in ASR of 32-34 weeks of age (Bma x 105 ± 6.4 and 139 ± 4.4 fmol/mg protein, respectively). From these results it may be concluded that advanced cardiac hypertrophy is associated with a reduction of al-adrenoceptors and a loss of inotropic responsiveness to al-adrenoceptor stimulation.

Dept. Pharmacotherapy, University of Amsterdam, Academical Medical Centre, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands.

U. Jahnel ~, H. Nawrath ~, J. Rupp ~, and R. Ochi 2 5-Hydroxytryptamine 10 i~mol/I (5-HT) exerted a positive inotropic effect associated with an increase in the Ca 2+ current (Ica) in the human right atrium. For detailed analysis, L-type Ca 2+ channel currents were recorded from cell-attached patches using 100 mmol/I Ba 2+ as charge carrier. Upon repetitive depolarizations from -80 to 0 mV at 2 Hz, Ca 2+ channel activity was identified by the appearance of burstlike inwardly directed currents; unitary conductance of the Ca 2+ channel amounted to 25.8 pS. Typically, Ca 2÷ channel openings during some depolarizing steps were followed by variably long periods of quiescence (blank sweeps). During superfusion with 5-HT, ensemble averaged (mean) current was enhanced to 160% of control. The increase in mean current was related to an increase in the channel availability, defined as the ratio of current containing sweeps to the total number of depolarizations. Unitary conductance, open probability within a current-containing sweep, mean open and mean shut times remained unaffected by 5-HT. These results indicate that the 5-HTincrease in macroscopic Ica in the human atrium reflects an enhanced activity of Ca 2÷ channels which may involve a cAMPdependent phosphorylation of the channel protein thereby causing a shift from an unavailable to an available state of the Ca 2÷ channel. In the failing human heart the 1]-adrenergic receptor system is markedly desensitized, as evidenced by a raassively decreased positive inotropic response to fi-adrenergic receptor agonists, while contractile responses to forskolin and calcium are largely preserved. This decreased responsiveness has been attributed to a reduction in the number mainly of the 1]1receptor, and a decrease in function of both the El-and the 1]2-receptor subtypes. In order to understand the molecular mechanisms of these changes, the expression of the following components of the 1]-adrenergic receptor system was assessed in healthy and failing human left ventricular tissue at the protein and the mRNA level: 131-and 1]2-receptors, 1]-adrenergic receptor kinase (I~ARK) -1 and -2, and i-arrestin-1 and -2. Receptor levels were measured by radiologand binding, 1lARK levels by determination of enzymatic activity using rhodopsin as substrate, and fi-arrestin by Western blot analysis as well as by determining inhibition of the signalling function of purified reconstituted 1]-adrenergic receptors. mRNA levels for all proteins, as well as for 1]-actin and glyceraldehyde-3phosphate dehydrogenase as controls, were measured by reverse transcription/polymerase chain reactions. Ill-Receptor levels were decreased by 50%, whereas 1]2-receptor levels were unchanged in failing hearts; these values were exactly paralleled by the values for the corresponding mRNAs. 1lARK activity was increased about 2-fold, which can be traced back to a selective 3-fold increase in lARK-1 mRNA. Expression of i-arrestin-1 and -2, which is high in human heart, was the same in control and failing hearts both at the protein and the mRNA levels. Thus, a reduction in Ill-receptor mRNA and an increase in lARK-1 mRNA are probably responsible for decreased ill-receptor numbers and increased 1lARK activity, respectively. The latter probably results in enhanced receptor phosphorylation and, hence, decreased receptor function. These data show that the decreased function of the B-adrenergic receptor system in failing hearts can be traced back to specific alterations in the respective mRNA levels. Myocardial contractility is depressed in endotoxic shock Plasma levels of cytokines, e.g. IllB or TNFo~, are elevated in patients with endotoxic shock. In smooth muscle cells and cardiomyocytes cytokines can induce the formation of nitric oxide (NO) which activates guanylyl cyclase. The mechanism of the depression in contractility (e.g. by vasodilatation of smooth muscles, direct effects on cardiomyocytes) has not been firmly established. Therefore, we investigated the effects of I116 and TNFc~ on contractile response (extent of cell shortening in % of cell length) and cGMP content (pmol/mg protein) in isolated ventricular cardiomyocytes from guinea pigs. Furthermore, the influence of the NO-synthase inhibitor N -nitro-L-arginine (L-NNA) on these effects were studied. Incubation of the cardiomyocytes with Illfl (25 /~g/ml) or TNF~x (100 ng/ml) for 18 h led to a decrease in contractile response to 44 % or 43 % of control value and to an increase in cGMP content to 178% or 161% of control value (p<0.05, n = 7 -12). In the presence of the NO-synthase inhibitor L-NNA (0.1 retool/l, 18.5 h) the contractility-decreasing and cGMP-elevating effects of I11B and TNFt~ were abolished.

It is concluded that the depression in myocardial contractility in endotoxic shock can be explained in part by a direct negative inotropic effect of cytokines on cardiomyoc~es. This effect is mediated by NO leading to an increase m cGMP content. Upon stimulation of human platelet aggregation induced by different agonists a permanent increase of cGMP was measured. This cGMP increase was generally believed to be a negative feedback mechanism terminating early events of activating signal transduction. We were interested in this cGMP increase associated with the platelet activation. Surprisingly our data proved an efflux of cGMP from the activated thrombocytes. Discriminating between intracellular and extracellular cGMP we were able to demonstrate that only cGMP in the extracellular space increases, while the level of intraplatelet cGMP actually decreases. Therefore, the overall cGMP accumulation within the first minutes after the agonist addition results from extrusion of cGMP, which thereby escapes hydrolysis by intracellular phosphodiesterases. In contrast, upon direct activation of soluble guanylyl cyclase by nitrovasodilators, like sodium nitroprusside, increased cGMP remains mainly inside the cells. Thus the efflux of eGMP from platelets seems somehow to be associated only with the platelet activation. Additonal experiments showed that an elevation of intracellular calcium and activation of protein kinase C are likely to be involved in promoting cGMP efflux. Our results are discussed in opposite to the general hypothesis that eGMP increase during platelet aggregation represents a negative feedback mechanism limiting activation. According to our data an apparent cGMP increase actually results from extrusion of cGMP rather than soluble guanylyl cyclase activation. This cGMP effiux provides in addition to phosphodiesterase cleavage a mechanism lowering intraplatelet cGMP thus favouring the aggregatory response. The present study was undertaken to characterize the inhibitory effects of some sydnonimines on cytosolic ionized calcium [Ca2+]i, aggregation and on adhesion of stimulated human platelets. In addition, the ability of the cyclic GMP-analog 8-PCPT-cGMP to inhibit increases in [Ca2+]i-levels and aggregation was measured. All investigations were performed wit.h ge~filtered, plasma-free platelets. The 3-morpholinosydnonimine (SIN-I) inhibited platelet aggregation stimulated by either ADP (4-8 pM in the presence of fibrinogen), collagen (2-101~g/ml) or thrombin (0.02-0.05 IU/ml) with half-maximal concentrations (IC50) between 70 and 200nM, respectively. The agonist-evoked [Ca2+]i-increases in fura-2loaded platelets were inhibited by SIN-I less sensitively than aggregation (IC50 between 500 and 1000 nM). Adhesion on subendothelial extracellular matrix was measured with [2-3HI-adenine labelled platelets stimulated by 8 I~M ADP in the absence of fibrinogen. The p~atetet preparation was free of any aggregatory response but about 20% of the platelets adhered to the matrix. SIN-I dose-dependently inhibited adhesion (IC50 approximately 300 nM). Furthermore, 8-PCPT-cGMP inhibited platelet a&~regation stimulated by either ADP, thrombin or collagen with IC50 between 50 and 200 pM. The cGMP-analog inhibited agonist-evoked increases in [Ca2+]i in similar ranges. The more sensitive inhibition of platelet aggregation by SIN-I compared with inhibition of [Ca2+]i-increases suggests that some of the effects of nitric oxide (NO) on platelets might be independent of cytosolic ionized calcium. The difference in sensitivity was not observed by directly applying the cGMP-analog. Therefore, we conclude that some antiplatelet effects of NO-releasing agents are not cGMP-mediated. Halothane (Halo) has been reported to sensitize the myocardium towards the effects of exogenous catecholamines in patients and laboratory animals. This study was aimed at investigating the catecholamine-sensitizing effects of Halo as well as the underlying cellular mechanisms in human myocardium. The positive inotropic effects (PIE} of isoprenaline (ISO) and Ca 2+ were studied in isolated human papillary muscle strips (1 Hz, 37°C) with and without exposure to Halo (2%). The effect of Halo on adenylate cyclase (AC) was investigated in human myocardial membranes and membranes of wild type and Gse-deficient (cyc') S49 cells. Halo augmented the PIE of ISO (PIEmax 8.1 +0.8 mN, EC50:0.027 pM, n=8) compared to control conditions (PIEmax 3.7 +0.2 mN, ECs0:0.049 pM, n=9). The potency and effectiveness of the PIE of Ca 2 + was unchanged. Halo (1%) increased basal as well as ISO-and guanylylimidodiphosphate (Gpp(NH)p)-stimulated AC in human myocardial membranes (p<0.05). Treatment of membranes with pertussis toxin increased AC by 40 % and abolished the effect of Halo. Halo had no effect on forskolin-stimulated AC or on basal, ISO-or Gpp(NH)pstimulated AC in the presence of 5 mM MnCI 2. The same results, i.e. a pertussis toxin-sensitive increase of AC-stimulation by Halo, were obtained in $49 cyc" or wild type cell membranes. The effects of Halo were not different between native $49 cyc" or cyc" membranes reconstituted with recombinant rGse expressed in E. coll. The data suggest that Halo stimulates AC and sensitizes AC following stimulation by I&-adrenoceptor agonists and guanine nucleotides due to an impairment of Gie-function. This mechanism may play a role in the Halosensitization of myocardial AC towards catecholamines. We have prepared a panel of monoclonal antibodies (mAb) to the myocardial Na-Ca exchanger expressed in Sf9 insect cells [Li et ai.,(1992) J Biol Chem 267:7828] to analyze the im~unoreactive domains and the topological organization of this membrane protein, mAbs reacting strongly with western blots of the recombinant protein were selected to screen an expression sublibrary containing exchanger cDNA fragments. These fragments coded for partial sequences of the exchanger protein and were expressed as fusion proteins with B-galactosidase in E.coli. Positive clones, containing specific linear mAb binding sites 15-155 amino acids in length, were identified for 4 different mAbs. All epitopes were localized to the long hydrophilic loop connecting transmembrane segments 5 and 6. The immunodominant region, that was targeted by 3 out of 4 mAbs, is a highly charged domain in the carboxyterminal half of the loop. Binding studies with a 3H-labeled high affinity mAb showed that the immunodominant region is located on the intracellular surface of the membrane. The same mAb was used to determine the density of exchanger molecules in membranes from different cell t~pes. Newborn rat heart cells contain about 6 x i0 exchanger molecules/cell. Exchanger levels and transport activities in individual cell lines appeared to be well correlated. Cell. Cardiol. 22: 439-452, 1990; B. Wohlfart, Acta Physiol. Scand. 106: 395-409, 1979) . If these changes reflect intrinsic properties of heart muscle, they should also be present in single myocytes. We have therefore investigated PEP and PRP in cells isolated by enzymatic dlssociation.

Contractions were induced by electric field stimulation (I Hz; 30°C) and measured as cell shortening by means of a photodiode array. Test intervals were interpolated between trains of 20 regular pulses. In control cells mechanical restitution of contractility could be described by a biexponential function with time constants of 0.18±0.03 s and 25.8± 4.8 s. The time constant of the monoexponential fitting of the data in ventricular strips was 3.44 s (n=10). In myocytes, maximum PRP occurred after 15 to 30 s of rest and, when normalized to steady-state, amounted to 3.52±0.54 (n=10). This value compares to 2. 45±0.26 (n=ll) in ventricular strips at 35°C. RF as calculated from the decay of PRP was 0.83±0.16 in single cells and 0.85±0.01 in ventricular strips. Caffeine (3 mM) reduced maximum PRP to 1.34±0.17 at 30°C in cells and completely suppressed PRP in ventricles. Our results show that the adaptive changes of contractility in response to transient perturbation of stimulation are also present in myocytes. Therefore we conclude that they are due to the same basic mechanisms.

Institut Elevation of force to < 1.2-fold was considered "inactive". Inhibition of (3H)ouabain binding to guinea-pig cardiac membranes (IC50) was used to assess their interaction with (Na++K+)-ATPase.

R2: Cardiac excitability could be increased or decreased under the influence of different antiarrhythmic drugs. One determinant of excitability is the threshold voltage (Vth). When it is rendered to abnormal, an increased propability of arrhythmias is to expect. The influence of low therapeutic (10pM), therapeutic (30pM) and toxic (100pM) concentrations of lidocaine (L) and of low therapeutic (3FM), therapeutic (10pM) and toxic (100pM) concentrations of (+)-sotalol (S) on cardiac stimulation threshold (ST) were studied, to evaluate, whether they might produce conditions of an increased or decreased excitability of the myocardium. Experiments were performed in isolated Langendorff perfused guinea pig hearts. At fourty points, on the epicardial surface, eight in the horizontal direction in angles of 45 degrees around the heart, beginning at the intraventricular septum and fife in the vertical direction, in regular distances from the basis to the apex, ST was determined at a stimulation of 300 beats/rain in a control group (n=18) and under the influence of different concentrations of L (n=10) and S (n=10). During control conditions, ST ranged around the heart between 0.8+0.1mV and 1.6±0.2mV (~±SEM). There was a rather homogeneous distribution of ST, that no value of one point was significantly different to this of the contiguous. Recently, hypertonic solution has been shown to be an effective resuscitation solution for the treatment of hemorrhagic hypotension. The purpose of this study was to investigate if an increase of sodium concentration caused by the use of hypertonic solutions has direct effects on the conduction system and on contractility of the heart. Therefore the effects of sodium concentrations from 90 to 170 mM under normal perfusion and from 120 to 170 mM during global ischemia, respectively, have been investigated in isolated guinea pig hearts perfused by the method of Langendorff.

Conduction intervals were not altered by stepwise increase of sodium concentration from 90 to 170 raM, Under normal conditions the left ventricular pressure (LVP) showed a significant decrease to 775:2% (x+S.E.; p<0.01) if the sodium concentration was below 110 raM. An increase of the concentration above the normal value of 130 mM did not influence the LVP. Also during ischemia the increase of the sodium concentration from 120 mM up to 170 mM caused no changes of the LVP-depression induced by global ischemia. While ~,arying sodium concentration neither under normal conditions nor during global ischemia changes in atrioventricular and intraventricular conduction could be observed. Therefore, in conclusion, possible changes of the sodium concentration caused by the use of hypertonic solutions may not influence cardiac contractility and electrophysiological parameters directly. An important parameter to determine cardiac excitability is the threshold voltage (Vth). Its changing to abnormal could produce conditions for an arrhythmogenic potency. We studied whether low therapeutic (0.1pM), therapeutic (lpM) and toxic (3pM) concentrations of propafenone (P) or low therapeutic (lpM), therapeutic (l.5pM) or toxic (3pM) concentrations of flecainlde (F) alter the matrix of stimulation threshold (ST) that conditions for the appearance of reentrant or automatic arrhythmias could be produced. Experiments were performed in isolated Langendodf pedused guinea pig hearts. At four(y points, on the epicardial surface, eight in the horizontal direction in angles of 45 degrees around the heart, beginning at the intraventricular septum and fife in the vertical direction, in regular distances from the basis of the apex, ST was determined at a stimulation rate of 300 beats/min in a control group (n=18) and under the influence of different concentrations of P (n=10) and F (n=10). Within the control group, there was a rather homogeneous distribution of ST, values ranged between 0.8+0.1mV and 1.6+0.2mV (~'+SEM) and from one point to the contiguous, there was rio significantly different value. P at a concentration of 0.1pM induced a mostly significant reduction of ST (0.5+0.06mV to 0.84+0.5mV) against control, whereas under 1 arid 3pM ST was significantly elevated (values ranged from 1.12±0.1mV to 2.33±0.3mV at lpM and between 1.1±0.1mV and 2.3±0.1mV at 3pM). The highest inhomogenity of distribution of ST was present under the influence of lpM of P. In 17,5% a significantly different value of ST from one point to the contiguous was present. ST remained within control range under the influence of lpM and 1.5pM of F (values from 0.81±0.08mV to 1.39±0.12mV at lpM and between 0.66+0.04mV and 1.06±0.08mV). Only at 3pM F, ST was significantly elevated (1.56±0.17mV to 2.59±0.28mV). The higher the drug concentration, the more homogeneous the matdx of ST. At lpM F, 10% of signiticandy different values of ST from one to the nd~hbeured point were present and at 3pM there were 5%. The results of the present study show, that the two class Ic antiarrhythmic agents differently affect cardiac ST especially in therapeutic concentrations. The increased inhomogenity within the matrix of ST under the influence of P might bear an increased proarrhythmic risk. The olten discribed proarrhythmic risk of a therapy with F seems not likely to be to explain by changes of ST. Propafenone, a class lc sodium channel blocker additionally provides a Badrenoceptor blocking activity, a Ca ++ antagonism and is able to prolong the action potential duration. The aim of this study was to evaluate the dose response relationship of propafenone on the cardiac parameter and conduction system. Spontaneously beating isolated guinea pig lmarts were attached to a modified Langendorff perfusion system. After an equilibration period of 30 rain the control values of the All nodal-(AV), His bundle-(HV), intraventricular conduction time (QRS) and the repolarization interval (QT interval) and the sinus rate were measured, followed by 60 min of perfusion with 0.1, 0.5, 1.0 & 5.0 ~tM propafenone (n=6 for each concentration). AH, HV, QRS, QT and sinus rate were registrated after 10, 30 and 60 rain respectively.

Propafenone had almost no effect on the AH conduction interval. The HV interval was significantly prolonged (p<0.05 ANOVA) by 5,0 and 1.0 yam propafenone. The conduction velocity in the ventricular myocardium was significantly reduced after 10 mm by 5.0 pM propafenone. During the observation period there ~vas no change in the frequency dependent QT-time when 0.5 and 0.1 laM propafenone were admimstered. Higher concentrations of propafenone resulted in a significant increase of QT interval aRer 30 min of perfusion. The spontaneous sinus rate was reduced in a dose deoendent manner. Maximal reduction of sinus rate of 57.8 ± 4.4% (mean ± S.E) could be observed when 5.0 I~M propafenone was used.

With the exception of the AH interval propafenone showed distinct dose response relationships. A steady-state was reached after 10 min for the HV and the QRS interval prolongation and after 30 mth for QT interval prolongation and sinus rate reduction, Therefor we conclude that stable and reproducable results on the conduction and pacemaker system in isolated hearts can only be obtained after more than 30 rain of perfusion with propafenone, Zentrale Versuchstieranlage, Karl-Franzens-University G-raz, Austria * Dep. of Internal Medicine, Karl-Franzens-Uinversity Graz, Austria ** Dep. of Zoology, Karl-Franzens-University Graz, Austria SDZ 211-939~A NEW CARDIOTONIC Na+CURRENT ACTIVATOR G. Scholt),sik, R. Salzmann, U. Quast, R. Berthold, H. Ott, A. Schaad*, E. Wettwer* Synthetic compounds which prolong cardiac Na + currents and increase force of contraction have recently been discovered, of which DPI 201-106 was the first and structural prototype (Scholtysik et al. N.S. Arch. Pharmacol.329: 316, 1985) . Several follow up compounds were described as SDZ 210-921 (Scholtysik, J. Cardiovasc. Pharmacol. 14: 29, 1989 ), BDF 9148 (Armah et al. Br. J. Pharmacol. 100:321 P, 1990 ) or PD 122860 (Haleen et al. JPET 250: 22, 1989 . We now introduce a new derivative, SDZ 211-939, (-)-(S)-6-Amino-a-[(4-diphenylmethyl-l-piperazinyl)methyl]-9H-purine-9-ethanol. SDZ 211-939 was investigated for cardiotonic activity in-vivo and in-vitro and for cardiac electrophysiological effects in-vitro. In anaesthetized rats, SDZ 211-939 at single doses of 0.03 to 0.3 mg/kg i.v. proved to be an effective cardiotonic combined with retardation of repolarisation in the ECG. In Langendorff-perfused hearts from rats and rabbits, SDZ 211-939 at concentrations between 10 -8 and 10 -6 mol/I increased force of contraction, increased (rabbit) or decreased heart rate and prolonged repolarisation in the ECG (rat). These effects were reversible on washout. In guinea-pig left driven atria SDZ 211-939 prolonged the functional refractory period, increased force of contraction (bell shaped concentration-response curve) and decreased beating rate in the right atria (3 x 10 -8 to 3 x 10 -6 mol/I). Potassium-depolarisation (22 raM) of left atria abolished the effects. Transmembrane action potentials of guinea-pig and dog papillary muscles were prolonged by SDZ 211-939 (3 x 10 -9 to 10 -6 mol/I). Voltage clamp experiments in guinea-pig ventricular myocytes revealed that SDZ 211-939 (3x10 7 to 10 -5 mol/I) inhibited Na+-current inactivation leading to a persisting Na + inward current which could be blocked by q-IX (10 -5 mol/I). The effects were reversible upon wash out. Interestingly, replacement by iodine of NH 2 in position 6 of the purine nucleus preserved the inotrcpic activity in guinea-pig left atria, suggesting the possibility of synthetizing a high affinity radiolabel for the cardiac Na + channel. In conclusion: SDZ 211-939 may serve as a very useful tool for the investigation of mammalian cardiac Na + current regulation. The effect of emetine dihydrochloride on the maximum upstroke velocity (qmax) of the action potential in papillary muscles and Purkinje fibres was studied by the conventional intracellular microelectrode technique. Macroscopic and single channel current (INa) were recorded with the patch-clamp technique using the cell-attached mode. qmax was significantly reduced by emetine dihydrochloride in papillary muscles (EC50 = 11.8 pmol/I, n = 5-7, P < 0.001) and in Purkinje fibres (EC50 = 12.9 pmol/I, n = 5-6, P < 0.001) without changing the resting or maximum diastolic potential. In ventricular myocytes at a concentration of 30 pmol/I emetine dihydrochlodde (n = 12) caused a reduction of the open state probability by increasing the sweeps without activity. This increase was threefold, while single channel conductance remained unchanged. After a washout period of 15 minutes the effect was not reversible (n = 2). The shape of the steady state current-voltage relationship was not significantly changed (30 IJmol/I, n = 3). At a stimulation rate of 1, 2 and 3 Hz peak INa was reduced in a frequency-dependent manner (30 I~mol/I, n = 2). The mean current per record was decreased by emetine dihydrochlodde to the same extent at all measured holdir~g potentials at a stimulation rate of 2 Hz (30 pmol/l: E h = -120 mV, 67.5 + 6.0 %, n = 5; E h = -100 mV, 58.5 + 5.2 %, n = 5; E h .= -90 mV, 67.4 ~ 5.1%, n = 5). 1985) . A few reports suggest a role for mexiletine hydrochloride in treating torsade de pointes (Campbell, N Engl J Med 316:29-34, 1987) . Therefore, the action of mexiletine hydrochloride on disopyramideinduced EADS and EES was studied in spontaneously beating Purkinje fibres by means of the intracellular microelectrode technique. In order to increase the probability of induction of EADS and EES by disopyramide, the potassium concentration of the bathing solution was reduced to 1,35 mmol/l. In five controls, this change did not elicit EADS and EES. After a control period of 30 min, the preparations were exposed to disopyramide at a concentration of 10, and 20 or 30 I~mol/I, for 60 min each. As soon as EADS and EES appeared, mexiletine (2,5 and 5,0 I~mol/I) was added to the bathing solution for 60 rain.

Mexiletine (2,5 pmol/I) completely abolished EADS and EES in one out of five, and at the higher concentration (5 pmol/I), in one out of four experiments. On return to disopyramide-containing, mexiletine hydrochloride-free bathing solution, suppression of EADS and EES was completely reversible. In conclusion mexiletine was capable of stopping EADS and EES completely in a small fraction of experiments only, and less effective in this respect than potassium channel openers like pinacidil (3, 30, and 100 pmol/I) (Brosch and Heistracher: Naunyn-Schmiedeberg's Arch Pharmacol 345: R76, 1992) . Pharmacol 346: R 24, 1992) . In the present experiments we analysed the influence of NDM on the cardiac action potential. Guinea-pig atria and papillary muscles were kept at 32 ° C in Krebs-I-Ienseleit's solution (2.5 retool/1 Ca 2+) gassed with carbogen. Standard microelectrode techniques were used to measure the action potential. At a pacing rate of 1 Hz (1 ms square pulses, 1 ms duration, 1.5 ~imes threshold voltage), NDM (100, 300 and 1000 #M) reduced the action potential amplitude of atria and papillary muscles while the membrane resting potential remained unchanged. The maximum depolarization velocity was decreased concentration-dependently from 222.5 + 10.4 V/s to 155.0 + 10.2 V/s in papillary muscles and from 187.5 + 4.6 V/s to 106.0 + 5.8 V/s in atria. The recovery of the sodium current from inactivation was slightly delayed in both tissues. In papillary muscles, the action potential duration was reduced from 204.2 _+ 4.4 ms to 111.8 + 3.7 ms and the functional refractory period (determined with twin pulses of equal strength) decreased from 222.5 __. 4.2 ms to 145.0 + 6.3 ms. In contrast, prolongations were seen in atria. The action potential duration increased from 64.5 + 1.8 ms to 92.8 + 4.8 ms while the functional refractory period was lengthened from 93.3 + 1.7 ms to 500.0 +_ 18.0 ms. In papillary muscles depolarized by 25.9 retool/1 KCI and stimulated at a rate of 0.1 Hz, NDM (300 ~tM) reduced (in the presence of 1.25x10 -6 mol/l orciprenaline) the upstroke velocity, the amplitude and the duration of calcium-mediated slow response action potentials. We conclude that NDM reduces the inward sodium and calcium current in all cardiac tissues while the repolarizing potassium currents are influenced differently in the atrium and the ventricle. Recently, cationic amphiphilic model compounds have been described (Borchardt et al., Biochem. Pharmacol. 42, $61-$65, 1991) in which the ability to perturb phospholipid membranes is not paralleled by hydrophoblcity (logP'). These compounds were applied to check whether their potency to block cardiac sodium channels correlates with their ability to lower the phase transition temperature ( A T) in liposomes of dipalmitoylphosphatidic acid (DPPA). The action on sodium channel function was determined in guinea-pig cardiac preparations: The threshold of alternating current (50 Hz) to induce arrhythmia (ACT) and the threshold voltage of rectangular pulses (3Hz) to elicit regular contractions (RPT) were measured in isolated left atria; the maximum rate of depolarization (dV/dt) was measured in papillary muscles (0.1 Hz). The drug concentrat%a~s at which excitability was reduced to 50% of tlie contol value served as a measure of potency. We investigated the effect of the g-adrenergic agonist isoproterenol (ISO) on the class I properties of the antiarrhythmic drug propafenone (1 rag/l) at the intact papillary muscle from rat hearts. The potential-dependent steady-state availability (hoe) of the fast cardiac sodium current (Iiq~+) was investigated with the loosepatch-clamp technique, which is suitable to clamp this current on intact muscle preparations even at physiological sodium concentrations. The use-dependent depression of INn+ was evaluated at a stimulation rate of 10 s -1 from a holding potential of -100 inV. We found no alteration of hoo by propafenone alone. The addition of ISO (0.1/zM) showed no augmentation of the maximal available sodium current. In contrast we have shown previously that ISO [Kirstein M(1991) 1.53 (p<0.001) ). Both substances fiR)-and (S)-Diprafenone) exerted considerable frequency dependent blockage of the cardiac sodium channel: Recovery from inactivation is described by two exponential functions. The fraction of the slow exponential is considerably increased by both enantiomers. From our dam we conclude that the action of (R)-and (S)-Diprafenone on the cardiac sodium channel is partially stereoselective, but both cnantiomers are active as class Ic antiarrhythmic agents. Thus it is possible by separation of (R,S)-Diprafenone in its pure enantiomers to separate the B-blocking action ((S)-form) from the sodium channel blocking action ((R)-and (S)-form). R56865"* is a synthetic drug displaying digitalis-antagonistic, cardioprotective and cerebral antiischaemic actions. Previous studies have revealed a blocking effect at different cardiac and neuronal ion channels. We have analyzed its action at the voltage-dependent sodium channel of embryonic rat ventricular myocardial cells.

Sodium current (I~,) was recorded from single cultured cells using the whole-cell configuration of the patch-clamp technique (-20 °C) at reduced extracellular Na (70 mM). Holding potential was -100 mV; depolarizing pulses to -30 or -20 mV (20 ms) were applied at 0.2 Hz or varying intervals. R56865 (1-30 ,uM;n>3) was applied extracellularly via the bath solution or a microsuperfusion system. R56865 reduced In, concentration-dependently with an IC50 of 2/~M without obvious changes of Irma time course. The effect was fully reversible within 3 rain of continuous drug-free superfusion. An increase in pulsing frequency from 0.2 to 1.6 Hz in the presence of 3 ,uM R56865 increased Ir~, block from 50 to 83 %. I~a recovered monoexponentially with a time constant of 2.2 s in the absence of pulses. During series of 10 pulses applied at a constant frequency of 1.1 Hz and separated by 10-s intervals allowing for recovery, Irma showed use-dependent block by 3 ,aM R56865, and INn of the 10th pulse was reduced as a function of pulse duration, by 34 % at 3 ms, 46 % at 10 ms, 59 % at 50 ms, and 60 % at 150 ms. In the presence of 100 ,uM vemtridine, R56865 decreased the time constant of the veratridine-induced post-repolarization Na tail current from 200 to 134 ms at 3 ,aM and from 182 to 68 ms at 10 #M.

In conclusion, R56865 is a reversible blocker of the cardiac sodium channel which apparently binds to its open and inactivated conformations while dissociating when the channel is in the closed resting state. Isolated perfused rabbit hearts were treated with increasing concentrations of norepinephrine (NE, 0.01, 0.1, 0.5 pM) either alone or in presence of propranolol (P, 0.i pM). For analysis of the epicardial activation and repolarization process an ep~cardial mapping (256 unipolar leads) was performed. For each electrode the activation and repolarization time was determined. From these data the "breakthrough-points" (BTP) of epicardial activation were determined. At each electrode an activation vector (VEC) was calculated giving direction and velocity of the local excitation wave. The beat similarity of various heart heats (under NE) compared to control was evaluated by determination of the percentage of identical BTP and of similar VEC (deviation ~ 5°). Moreover at each electrode the local activation recovery interval (ARI) and its standard deviation (of 256 leads, dispersion, DISP) were determined. NE alone (0.01, 0.1, 0.5 pM) led to an increase in left ventricular pressure, heart rate and DISP with concomittant frequency dependent reduction in ARI, and to changes in the epicardial activation pattern (reduction in BTP, VEC). We found that in the presence of P (0.I DM) NE prolonged ARI and reduced ARI-dispersion. This effect was not due to changes in heart rate. The disturbing effects on the activation pattern were diminished. From these results we conclude, that NE prolongs the relative action potential duration probably via stimulation of a-adrenoceptor and enhances cellular coupling. Thus, the antiarrhythmic properties of propranolol may at least in parts be due to an unmasking of class Ill like NE-effects. ( Antiarrhythmic compounds with class III action prolong action potential duration (APD) mainly by affecting K ~ channels. We have studied the frequency-dependence of the effects of such agents, i.e. E-4031 (l-[2-(6-methyl-2-pyridyl)-ethyl]-4-[4~methylsulphonyl-amino-benzoyl]piperidine), dofetilide, almokalant and tedisamil, in guinea-pig papillary muscle in order to detect possible use-dependence of this block. Action potentials were recorded with conventional microelectrode technique at the stimulation frequency of 1 Hz. Frequency-dependence of APD (90% of repolarisation) was measured after 2 min at 0.2, 0.5, i, and 2 Hz preceded by 90 s of rest. The muscles were exposed to increasing concentrations of the compounds. All drugs prolonged APD in a concentration-dependent manner, however, the dependence of increase in APD on stimulation frequency was different. With E-4031 (I ~M), maximum lncrease in APD was 81±10 ms at 0.5 Hz, the effect became smaller both at higher and lower frequencies. With tedisamil (3 ~M) the largest increase was 35±6 ms at 0.2 Hz, higher frequencies produced less increase (25±6 ms at 2 Hz). With dofetilide (i0 ~M) and almokalant (i ~M), the maximum increases were 56±4 ms and 47±5 ms, respectively. The latter 2 compounds showed only little rate-dependent action. Control APD also varied with frequency, however, when analysing time-matched controls, we found a change in APD with time, which amounted to -11±6 ms at 0.2 Hz, -8±6 ms at 0.5 Hz, -7±2 ms at 1 Hz, and +3±5 ms at 2 Hz. Correcting for these spontaneous changes, the APD prolonging effects of almokalant and dofetilide also became freguency dependent. In conclusion, the blocking mechanlsm of the agents studied may show some use.-dependence, but measurements of membrane currents are required to analyse this effect in more detail.

Institut, Universit~t -GHS -Essen, Hufelandstra8e 55, D-4300 Essen i, Germany

In various heart tissues, a transient outward current (Ito) is activated by depolarizing voltage clamp steps from -50 mV up to +40 mV. At +40 mV, Ito peaks after about 3 ms in ventricular myocytes from rat hearts; the time course of inactivation can be appropriately described by two time constants, "~fast = 14 ms and ~s~ow = 220 ms. Verapamil, nifedipine and quinidine in concentrations between 1 and 30 pmol/I accelerated the inactivation of Ito, whereas peak Ito was less affected. In contrast, 4-aminopyridine (4-AP) 1 mmol/I preferentially depressed peak Ito, whereas the time course of inactivation was rather slowed down. The efficacy of 4-AP was diminished at short and enhanced at long pulse intervals (reverse use-dependence). The time course of drug action at +40 mV was calculated by the fractional changes of Ito. Quinidine, nifedipine and verapamil produced a concentrationdependent block of Ito increasing during the depolarizing voltage clamp step. Conversely, the block by 4-AP of Ito decreased during the step. We conclude from our experiments that verapamil, nifedipine and quinidine bind to the Ito channel in the open state at positive potentials. In contrast, 4-AP obviously binds to the channel in the closed state at negative potentials and relief of block is observed at positive potentials. The two ~odium channel modifiers DPI 201-106 (4-[3'-(4"-benzhydryl-l"piperazinyl)-2'-hydroxy-propoxy]-iH-indole-2-carbonitrile; DPI) and BDF 9148 (4-[3'-(l"-benzhydrylazetidine-3"oxy)-2'-hydroxy-pro~oxy]-iH-indole-2-carbonitrile; BDF) produce similar positive inotropic effects in guinea-pig papillary muscle, but prolongation in action potential duration (APD) induced by BDF is significantly less than by DPI. Since both agents impalr ingctivation of sodium current, this difference in APD prolongation could be due to additional effects of either compound on potassium currents. To test this hypothesis, we have used the whole-cell voltage clamp technique in enz~matically dissociated ventricular myocytes of gulnea-pig hearts and measured the effects of DPI and BDF on the inward rectifier (I~1) and the two components of the delayed recti-fi~ (IKs and IKr ) . IKI was studi6d-using long lasting (~6 s) ramp puIses from +40 to -i00 mV. After 8 mln of exposure, DPI (3 ~M) reduced I~i to 65.7±8.2% of control at -iOO mV and to 69.6~f8.3% at -50 mV, whereas BDF (i ~M) did not significantly affect LK~ I s was studied with test pulses from -40 to .~v m~. DPI reduced IKs to 44.9±19.6% of pre-drug controls, which was m6~e than the spontaneous decline to 68.1±13.0% in time matched control cells. BDF did not significantly affect IKs. The rapid component IKr was measured as the tail current after a repolarizing step from +50 to -40 mV. DPI reduced the tail current to 19.2±1.5%, again BDF had no effect. Our results show, that DPI reduced IKI and both components of the delayed rectifier ~Ks and IKr , whereas no effects of BDF on potassium currehts could be detected. These additional effects of DPI are sufficient to explain the more pronounced prolongation in APD observed with this agent than with BDF.

Institut, Universit~t -GHS -Essen, HufelandstraBe 55, D-4300 Essen i, FRG 

T. SchrSder, J.P. Hering, G. Hellige

The echo signal intensity from blood can be increased by central venous injection of solutions containing air bubbles. To examine not only the right but also the left side of the heart the transpulmonary echo contrast agent SH U 508 (Schering AG, Berlin, FRG) was developed. SH U 508 is a suspension of 2 pm monosaccharide micro particles (galactosel.

To experimentally investigate the effect on cardiac function the agent was directly injected into the coronary arteries in 6 anaesthetised sheep. SH U 508 was injected intracoronary amounts of 2 ml (n = 12), 4 ml (n = 10) and 8 ml (n = 9) at a concentration of ,~00 mg/ml.

The An important feature of the non-failing heart is the ability to keep cardiac output (CO) at a constant level despite varying afterloads. In the failing heart increases in afterload cannot be counter-balanced and lead to a decrease in cardiac output.

To investigate whether treatment with positive inotropic drugs is efficient in restoring this regulatory mechanism in stunned myocardium we tested the following compounds: digitoxin, adrenaline, enoximone, adibendan, and R 80122, a highly selective phosphodiesterase III inhibitor.

In the guinea-pig heart-lung preparation ventricular function was assessed while ai~erload was increased from 60 to 100 cm H~O within 4 rain. Then hearts were submitted to 15 min of global normothermic ischemia. After 15 min of reperfusion at 60 cm H~O drug or solvent was added to the reservoir and allowed to act for 15 min (digitoxin: 20 rain) before a second ventricular function curve was established.

In stunned hearts treated with solvent only the enhancement of afterload was not accompanied by an increase in +LVdP/dt~ as observed before ischemia. In these hearts CO at 100 cm H20 was reduced to 36% of the value reached at 60 cm H~O. After treatment increases in +LVdP/dt~ were observed in the following order: adrenaline (3 x 10 -7 reel/l) > R 80122 (1 x 10 .6 reel/l) > adibendan (1 x 10 -~ reel/l) > enoximone (1 x llY' reel/l) = digitoxin (1 x 10 -~ reel/l). In contrast the rank order in improving CO was: R 813122 > adrenaline > enoximone > adibendan = digitoxin. These results suggest that positive inotropic agents can fully or partially restore CO regulation in stunned myocardium. However, drug effects on CO cannot be predicted from the respective effects on +LVdP/dt,~. R 80122 exerted a more pronounced influence on CO than on +LYdP/dt~ and may thus improve hemodynamics in failing hearts with less energy consumption as compared to adrenaline.

Janssen Res. Found., Raiffeisenstr. 8, W-4040 Neuss 21, FRG; "current address: Schwarz Pharma AG, W-4019 Monheim, FRG The antifibrillatoric effÉcacy of an antiarrhythmic agent is conventionally evaluated by determination of the ventricular fibrillation threshold (VFT). Because of some serious problems in methodology the VFT does not seem to be appropriate for determination of antifibrillatoric properties under ischemic or reperfusion conditions. The aim. of this study was to set up an in-vitro model that allows determination of the antifibrillatoric property of an antiarrhythmic agent using sustained ventricular fibrillation during reperfusion.

Isolated guinea pig hearts were perfused according to the method of Langendorff. By redncing the coronary flow to 10% of the control value of 6 ml/min the hearts were exposed to global ischemia. Simultaneously the left ventricle was paced at twice the diastolic threshold. The basic cycle length was 300 ms and slightly longer than the spontaneous beat-to-beat interval. Ventricular fibrillation O/F) could first be observed after 9.6 • 0.6 rain (mean ± sam) of global ischemia. A first series of experiments was able to show that VF was stable for at least 90 min under global ischemic conditioms. VF was defined by an irregular electric activity on the epicardiac surface ECG and by a complete loss of ventricular contraction force development.

In addition the dependency of the VF stability on the length of ischemia was evaluated. Therefor in three experimental groups coronary flow was abruptly increased to the pre-ischemic value (reperfusion) after 15, 30 and 60 min of global ischeima respectively. After short-term ischemia 60% (n=10) of the initiated VF were stable over a 120 min period after the onset of reperfusion. Long-lasting global ischemia for 60 rain also failed to provide enough stability of VF, only 7 out of 20 experiments (35%) showed sustained VF over the whole 120 rain period, whereas after 30 min of global ischemia VF was stable in all experiments (n=30).

In conclusion the characteristics of this model for sustained VF during reperfusion are consisted with reported data (ref to Manning & Hearse 1984 J.Moll.Cell Cardiol. 16, 497-518) . Therefor this in-vitro model provides an easy to use approach for screening the antifibrillatoric efficacy of an antiarrhythmic agent. Oxygen-derived free radicals have been suggested to contribute to tissue injury associated with myocardial ischemia (HI) or reperfusion. However, the origin of these free radicals is still unclear. We investigated the role of catecholamines as a source of oxygen free radicals during HI by examining whether the effects of superoxide dismutase (SOD) on MI are altered by depleting catecholamine stores. We used isolated electrically driven rabbit hearts (Langendorff, constant pressure: 70 cm H20, Tyrode solution, Ca *÷ 1.8 mmol/l, 180 min-t). MI was induced by left coronary artery branch occlusion (CAO) and quantitated from epicardial NADH-fluorescence photography. Catecholamine stores were depleted by pretreating rabbits with reserpine (7.0 mg/kg i.p. 24h before preparation). SOD (48 U/ml) did not significantly influence coronary flow or left ventricular pressure (p>0.05). SOD significantly diminished HI after repetitive CAO (-40%)(p<0.05). In hearts with depleted catecholamine stores, MI was not significantly affected by SOD (p>0.05). Accelerating the pacing-rate after CAO from 180 min -I to 300 min -i , a procedure stimulating noradrenaline-overflow, was accompanied by a significant enlargement of MI in control hearts (+50%)(p<0.05). Pacing-rate dependent growth of HI could be completely prevented either by SOD-treatment of isolated hearts or by reserpine-pretreatment of the rabbits (p>0.05). Conclusion: Oxygen-derived free radicals considerably contribute to tissue injury during HI and they are derived mainly from noradrenaline in isolated rabbit hearts .  THE INFLUENCE OF NIFEDIPINE ON ISOLATED HEARTS TAKEN  FROM DIABETIC AND/OR HYPERTENSIVE RATS. A.J, Pijl, M,G.C. Hendriks, K.L. Ksm, M. Pfaffendorf and P.A.van Zwieten

We investigated the impact of diabetes mellitus and hypertension on the isolated rat working heart preparation. Spontaneously hypertensive rats (SHR) and normotensive Wistar Kyoto (WKY) rats of 12 weeks of age with streptozotocin induced diabetes were used. The hearts were per/used with Tyrode solution of 37°C (calcium-and glucose concentrations of 2.23 and 11.1 mmol/I, respectively) at a physiological pH. They were paced at a frequency of 5 Hz (pulse width: 5 ms, voltage: 10% above treshold, 2-4V). The preload and the afterload were fixed at 15 cm H~O and 70 cm H20, respectively In this preparation the following parameters were determined: LVP (left ventricular pressure), +dP/dtm., (maxima( rate of pressure increase), -dP/dtr,=x (maximal rate of pressure decrease), AO (aortic output), CF (coronary flow) and the CO (cardiac output). In separate experiments we corrected every individual group for the glucose concentration and the afterload. For the hearts from hypertensive rats the afterload was increased to 110 cm H20, and for those of diabetic rats the glucose concentration was elevated to 30.0 mmol/l. It appeared that as a result of this "physiological" correction the groups could not be compared with each other anymore. For every individual group of hearts a dose response curve was constructed for nifedipine. This was done under standard conditions and under more adjusted conditions. In conclusion: 1) The rat working heart shows a worse performance in vitro under "hypertensive" and/or "diabetic" circumstances when compared to the non-pathologic situation. 2) In order to make an appropriate comparison for the 4 groups it is preferable to treat them all under standard conditions instead of adjusting the glucose concentration and/or afterload for every individual group. 3) There are no significant differences in sensitivity for nifedipine, as reflec- In human idiopathic dilated cardiomyopathy (IDC) a decreased positive inotropic effect of cAMP-dependent agents is associated with sympathoadrenergic activation, downregulation of l~-adrenoceptors and an increase in inhibitory G protein t~-subunits (Gin). We measured these parameters in bovine hereditary cardiomyopathy (bCMP) to study whether similar mechanisms may underly the contractile dysfunction in this form of heart failure. Catecholamine tissue concentration were determined with HPLC and electrochemical detection. l~-adrenoceptor (I~-AR) and c~-adrenoceptor densities (~-AR) were measured in membrane fractions with radioligand binding using 125I-CYP and 3H-prazosine, respectively, and Giu in crude homogenates by pertussis toxin catalysed ADP-ribosylation in right and left atria and ventricles (n=8 each). Freshly obtained hearts from the slaughterhouse served as controls (CTR; n=8). Results: In all 4 chambers of hearts with bCMP tissue concentrations of noradrenaline and adrenaline were decreased by 36-69 %, respectively, whereas dopamine concentrations were increased by 105-218%. In CTR, I~-AR and ct-AR were about 50% lower in both atria than in both ventricles (28/54 [6-AR], 5/12 lt~-AR] fmol/mg protein). In bCMP, I~-AR was markedly decreased in all chambers by about 90% whereas ~-AR remained unchange~L Pertussis toxin catalysed ADP-ribosylat~on and SDS-PAGE revealed 3 bands in the 40 kDa region in atrial tissue and only 2 in ventricular tissue of CTR and bCMP. Total PTX-substrates were increased in all 4 chambers in bCMP by 34-90%. This increase was almost exclusively due to an increase in the predominant, lowest band, amounting to 61-160%. We conclude that changes in myocardial catecholamine levels and the B-adrenoceptor-G-proteinadenylyl cyclase pathway in bCMP closely resemble changes in IDC. This suggests that bCMP, at least in this respect, may be a suitable model for human IDC. The aim of this study was to investigate the cardioprotective effect of ACE-Inhibitors in rats suffering from heart failure after myocardial infarction. In normotensive Wistar rats either sham surgery or myocardial infarction (MI) by ligation of the left coronary artery were performed. Six weeks after this procedure rats were treated with either placebo (H20 , lml/kg/die, p.o.) or the ACE-inhibitor moexipdl (2-[2-[ [1-(Ethoxycarbonyl)-3phenylpropyl]amino]-I-oxopropyl]-6,7-dimethoxy-1,2,3,4tetrahydroisoquionoline-3-carboxylic acid, hydrochlodde) (MO, 10mg/kg/die, p.o.) for another six weeks. After the treatment pedod catheters were implanted chronically in the femoral artery and vein as well as in the left ventricle of the heart. On the following day artedal blood pressure, heart rate, end-diastolic left ventdcular pressure (EDP) and myocardial contractility (dp/dtmax) were recorded in conscious animals dudng control conditions (0.9% NaCI, lml/h i.v.) and dudng increased aftedoad (methoxamine (c¢-[1-Aminoethyl]-2,5-dimethoxybenzylalcoholhydrochlodde), lmg/ml/h i.v.). Rats were then sacrificed and the hearts were examined for heart weight and infarct size. The increase in arterial blood pressure as well as the reduction in heart rate during methoxamine were not different among groups. Regardless of therapy rats with MI showed greater increases in EDP (Sham: 4,0+1,3; MI-H20: 15,2+1,7"; MI-MO: 18,1:P.2,3" [mmHg]) and smaller increases in dp/dtma x (Sham: 1,2+0,1; MI-H20: 0,5±0,3*; MI-MO: 0,5:t:0,2" [mHg/s]) when compared to sham operated rats dudng increased aftedoad. However, moexipril led to a significant reduction of total heart weight compared with placebo treated rats with MI (Sham: 1,3:t:0,04; MI-H20: 1.5±0,09"; MI-MO: 1,2+0.08 [g]). We conclude that the cardioprotective effect of ACE-inhibitors applied dudng established chronic heart failure after myocardial infarction is rather due to a regression of myocardial hypertrophy than to improved mechanical properties of the heart. The bovine cardiomyopathy (bCMP) probably is a genetically determined dilated cardiomyopathy (DCM) which occurs in the breed Simmental x Red Holstein cattle and leads to severe heart failnre (HF) (Tontis et al., Schweiz Arch Tierheilkd 132, 105-116, 1990 ). In HF an increased activity of the sympathetic nervous system (SNS) can be regarded as a compensatory mechanism (Francis, Am J Cardiol 66, 33D-39D, 1990) . In order to study the role of SNS in bCMP we determined the myocardial content ofnoradrenaline (NA), adrenaline (A) and dopamine (D) by HPLC with electrochemical detection in right and left atria and ventricles from hearts with bCMP (n=8) and in bovine control hearts (CTR; n=8). In comparison, human failing hearts with I)CM (n=2-9) or ischemic cardiomyopathy (ICM; n=3-7) and human non/ailing hearts (NF; n=4-9) were studied. In both, bovine CTR and human NF we found higher concentrations of the three cateeholamines (pg/mg wet weight) in the atria (NA: 2261+_108 and 1308+157; A: 34+-3 and 23+_6; D: 253+-41 and 597+-182, respectively) than in the ventricles (NA: 1007+_78 and 801+_99; A: 27+_2 and 15+_4.; D: 66+11 and 303+_80, respectively). In bCMP NA and A decreased significantly (p<0.05) to 50-74% and 31-40% of CTR, respectively, whereas D increased significantly to 205-318% of CTR. In human DCM and ICM we observed a significant depletion of NA, A and D to 45-72%, 33-83 % and 3-6% of NF, respectively, We conclude that the high adrenergic drive in severe HF leads to a myocardial catecholamine depletion independent of the underlying cause and of regional factors. The augmentation of D in bCMP is probably due to a shift in the rate-limiting step of cardiac norepinephrine synthesis from tyrosine-hydroxylation to dopamine-hydroxylation and does not necessarily imply pathophysiological differences between human and bovine cardiomyopathy. The existence of hereditary heart failure in cattle was reported repeatedly and occurred in many different countries (see Tontis et al. Schweiz. Arch. Tierheilk. 132: 105, 1990; KOmper and Bahnemann, Tier~irztl. Prax. 20: 254, 1992) . In Switzerland the majority of cases affected crossbred cattle of the breeding Simmental x Red Holstein, suggesting genetic influences on the pathogenesis of the disease. The animals suffer from venous congestion and congestive edema of the submandibuiar area, the brisket and the ventral abdomen. Pathological findings such as cardiomegaly with ventricular and atrial dilatation and hypertrophy as well as histologically veryfied myocardial fibrosis confirm the diagnosis "bovine cardiomyopathy" (bCMP).

Since life threatening heart failure in humans is a problem of which pathophysiological questions are still open, the aim of the present experiments was to investigate, whether bCMP could serve as a model to gain further insight into the illness.

Ventricular strips obtained after slaughtering were suspended for electrical stimulation in-vitro from clinically healthy controls (n = 15) and from cattles with final stage bCMP (n=12). Muscles from bCMP showed impared relaxation and decreased responsiveness to isoprenaline and phgnylephrine. The maximum inotropic effect of exogenous Ca++ (7 x 10 "~ incl./I) was not different. Transmembrane action potentials showed reduced Vmax and durations in bCMP when compared to healthy hearts.

In addition, cattles with pathologically confirmed diagnosis of bCMP had significantly higher plasmaconcentrations of noradrenaline as compared with healthy animals (preliminary result of the measurements in a small number of 5 animals in each group).

These pharmacological chara~eristics of the heart muscle in bCMP and increased noradrenaline plasma levels suggest similarities to human heart failure and may encourage for further pathophysiological comparison. We studied the effect of simultaneous hypertension and increasing age on the c/1-adrenOceptOr .agonist phenylephrine (PhE) and the endothelium-dependent relaxation provoked by methacholine (MCh), in isolated aortic ring preparations taken from: normotensive Wistar-Kyoto (WKY) rats and spontaneously hypertensive rats (SHR) of 18 and 28 weeks of age, and from 18 weeks old spontaneously hypertensive stroke-prone rats (SHRSP). The aortic ring preparations of SHRSP proved more reactive to PhE compared to those obtained from the age-matched WKY and SHR. A significant age-dependent decrease in maximal response for adrenergic stimulation was observed for the preparations taken from 18 and 28 weeks old WKY and SHR. In preparations of 18 weeks old SHRSP the pD 2 of MCh significantly decreased compared to those taken from the 18 weeks old WKY and SHR. Ageing (18 to 28 weeks) caused no significant changes in pD2 to MCh in the rat aorta. In conclusion, there is an age-associated decrease in response to u~-adrenoceptor stimulation and a hypertension-induced decrease in sensitivity to endothelium-dependent relaxation. The smooth muscle relaxing effects of NO am confirmed by numerous studies. Vasoralaxant effects of orally administered inorganic NaNO 2 were investigated in a randomized study with SHR in order to test the hypothesis that the higher endogenous production of nitrite (NO2") from alimentary nitrate (NOs") may be one of the reasons why a lower blood pressure (BP) is found in vegetarians. Groups of 8SHR (BP, M_+SEM: 189_+3mmHg) were given drinking water ad libitum containing 4 different doses of NaNO2: 0, 25, 50 and 100 mMol. Sodium content of all groups was adjusted up to 100 mMol by supplemenation with NaHCO s. Indirect BP measurements at the end of the 8week period indicated dose-dependent hypotensive effects: 257-+6 (0 mMol NaNO 2 resp. 100 mMol NaHCO 3 = controls), 236-+8, 213-+3, rasp. 155+13 mmHg. Analysis of covariance (covarlable = body weight) revealed significant differences of treatment after 4 weeks (p<0.05). In addition effects of a 20% food-restriction were investigated (N =32), but no further hypotensive effects were found. Plasma-NO~ correlated with NO2"-content of drinking water (r = 0,609; p < 0.00 t ); Plasma-NO2" concentrations of treated animals ranged from 4.1 to 137.8pMol (HPLC). A follow-up study with 4 male SHR (baseline-BP = 186+8 mmHg) administered 3 weeks drinking water containing 100 mMol NaNO 2 indicated reversibility of the BP Iowedng effect: 2 days after switching over to 100mMOlNaHCO 3 BP increased from 160-+13 up to 261 -+19 mmHg, In orientating studias on 4volunteers plasma-NO~" (basal level, M-+SD: 30.6_+10.3/JMol) reached a max. level of 433.5-+183.3pM01 3hrs. after the consumption of nitrate-containing vegetables. In one subject Plasma-NO 2" increased from non detectable (<5/JMOl) to 15.2 pMol at 23 hrs.

Results of the present studies strengthen the hypothesis that subchronically administered nitrate -after endogenous conversion to nitrite -might have similar hypotenelve effects in vegetarians.

Department of Biol. Chemistry and Nutrition Science, Section: Pharmacology and Toxicology of Nutrition, University of Hohenheim, Fruwirthstr. 14, W-7000 Stuttgart 70, Federal Republic of Germany 369

To investigate the possible function of endothelin peptides in vivo we have established transgenic rats overexpressing their genes. In one series of experiments we chose the human endothelin-2 gene to generate transgenic animals. A genomic fragment including 2 kilobases (kb) of the promoter region and 1,3 kb of the 3'-flanking region has been injected into pronuclei of fertilized rat oocytes which were implanted into the oviduct of pseudopregrtant female rats, The presence of the transgene was checked using southern blot analysis of tail DNA from the offspring (n = 13). Two transgene postive founder animals could be identified which were used for breeding and further analyses were carried out in the F1 generation. Measurement of endothelin-2 in the plasma revealed significantly (p < 0.0001) elevated endothelin concentrations (2.46 +/-0.3 pM/ml) in transgenic rats as compared to control animals (0.52 +/-0.25 pM/ml). Expression analysis for human endothelin-2 mRNA showed high concentrations in the kidney, which is also a major site of its expression in humans, but specific mRNA signals were also detected in adrenal, lung, gut and brain. Measurement of cardiovascular parameters revealed that blood pressure and heart rate were not different in the transgenic animals. The present concept holds that endothelin acts predominantly on the autoerine/paracrine level. Our data are in support of this and show that a chronic increase of endothelin-2 in the plasma alone does not lead to changes in basic cardiovascular. . parameters, such as blood pressure. These transgenic arumals provide an important model for future studies into mechanisms of action and in vivo functions of endothelin peptides. Feeding New Zealand White rabbits (1.8 -2.0 kg) a cholesterol-enriched diet (0.5 % for 3 months, 3-% for 3_ month; CHOL) resulted in characteristic alterations in the coronary circulation without visible plaque formation. Langendorff-perfused hearts (40 ml/min) of CHOL-rabblts exhibited a significantly reduced decrease of coronary perfusion pressure (CPP) after short-term infusion (3 rain) of the endotheliumdependent vasodilators bradykinin (0.05 ~,molA), substance P (0.05 p, mol/1), and carbamoylcholiue (0.1 ~mol]l). This impairment was not due to a diminished release of the endogenous mediators endothelinm-derived NO (oxyhaemoglobin-technique) and prostacyclin (6-oxo-PGF 1 RIA) C~ -.

The effect of cholesterol on the endothelium-dependent relaxation was efficiently suppressed by oral nitrendlpine (CHOL+NIT) (20 mg/kg'd, withdrawn 72 h before the acute experiment). The The data are mean ± SEM of n experiments in each group ) p < 0.05 vs. CHOL Basal mediator-release was at comparable levels in all groups. Serum-cholesterol was not changed by nitrendipine in any group. The data suggest that clinical treatment with nltrendipine results in a marked preservation of endothelial function and endothelium-derived mediator release in hypercholesterolaemia by mechanisms which are independent of preventing plaqueformation.

Iustitut f~r Pharmakologie, Heinrich-Heine-Universitfit Dtisseldorf, Moorenstr. 5, W-4000 Diisseldorf, FRG Substitution of estrogens in postmenopausal women reduces the risk of cardiovascular diseases. Mechanisms of this protective effect of estrogens are still under discussion. We studied the acute effects of 176-estradiol (Schering AG) on tone of human internal mammary and coronary arteries in vitro. Arteries were obtained freshly during bypass surgery or heart transplantation. Ring segments were suspended in organ baths and preconstricted with prostaglandin F2a (0.3 -3 #M). TO avoid tachyphylaxis, 176-estradiol was added only in one concentration (0.03 nM -3 p.M) for each ring preparation, Within 30-40 minutes, 17i3-estradiol produced a significant relaxation of about -67 + 14% (% preconstriction; mean _+ SD) in mammary arteries, and of about -82 4-21% in coronary arteries; solvent controls (0.2% ethanol) showed only small relaxant effects (between -3 to maximal -25%). In mammary arteries with mechanically removed endothelium, 176-estradiol produced similar relaxtions (-74 + 12%) as compared to intact vessels. Mammary arteries from male patients responded somewhat less (-64 + 14%, n=16) as corresponding arteries from females (-75 4-12%, n=4). It is concluded that 176-estradiol induces endotheliumindependent relaxations in human mammary and coronary arteries in vitro, This effect might be more pronounced in arteries from female donors as compared to males. This direct vasodi!atory effect may contribute to the cardiovascular benefits of estrogens in postmenopausal women. The human skin is an area of interest for the evaluation of vasoactivity of local anaesthetics. But, as the normal perfusion of the skin is rather low, it is difficult to quantify precisely especially vasoconstrictive properties of local anaesthetics. The purpose of this study was to investigate, whether flux enhancement by neurogenic inflammation provides a condition that allows the evaluation of the vasoactivity of local anaesthetics. Test areas were the dorsal aspects of the forearms in 24 volunteers. Skin blood flow was measured with a computercontrolled, motor-driven two channel laser Doppler flowmeter at ten measurement points over a distance of 18 mm. Neurogenic inflammation was elicited by 1% histamine prick at the most distal measurement point; 100 ul of the test solution were intracutaneously injected into the flare of the neurogenic inflammation. Lidocaine 0:5% and mepivacaine 0.25% were investigated plain and with addition of adrenalin (50 ug/mt) and POR 8 (0.1 U/ml). Saline served as control. Plain solutions of lidocaine and mepivacaine exerted a mild vasoconstrictive effect on enhanced flow after histamine prick in comparison to saline. Addition of vasoconstrictors significantly reduced enhanced flow. Injection of saline produced further flux enhancement. The results of this study are in accordance with results obtained by other means. We therefore assume that the flare of neurogenic inflammation produces a valid condition for the evaluation of the vasoactivity of local anaesthetics. Flezelastine, l(2H)-phthalazinone,4-((4-fluorophenyl)methyl) -2-(hexahydro-l-(2-phenylethyl)-IH-azepin-4-yl)-, monohydrochloride, was examined with respect to its electrophysiological effects on rabbit ventricular and atrial myocardium. Conduction time was prolonged in a dose-dependent (2-20 #mol/l) manner by 18 to 62%. This effect was more pronounced in the right ventricular free wall than in the left auricle. The effective refractory period of the left atrial myocardium was lengthened by 36% (iHz), but there wasn't any influence on the papillary muscle. Heart rate was reduced by 12% at most. Contractility of left auricle and papillary muscle decreased dose-dependently from 16 to 41%.

In papillary muscle flezelastine (I #mol/l in Tyrode solution) caused an increase in action potential duration of 37% (33-41%) accompanied by a slight reduction of the maximal rate of rise (after transient increase) by 48%. The reduction of ~ is frequency dependent with a rate constant of 0.095 at 1.~x increasing to 0.110 at higher concentration (3 #mol/l). In Purkinje fibers there is a strong reduction of the notch -pointing to an effect on the if.

These results suggest that beside its antiasthmatic/ antiallergic properties flezelastine also exerts antiarrhythmic effects on different tissue levels which are similar to class I antiarrhythmic compounds. 

Disseminated intravascular coagulation (DIC) is characterized as systemic conversion of fibrinogen to fibrin and activation of platelets finally leading to microthrombosis. DIC occurs as complication of a wide variety of diseases. DIC-related microthrombi, detected as increased deposition of 125J-labelled human fibrin, could be induced by intravenous infusion of E. coli lipopolysaccharide (endotoxin, 25 mg/kg per 4 hr) in the liver and the kidneys of anesthetized rats. Endotoxin infusion was associated with marked consumption of fibfinogen and thrombocytopenia. Treatment with 1 -20 /~g/kg'min saruplase (r-scu-PA), that was infused concomitantly with endotoxin, dose-dependently and significantly reduced endotoxin-induced microthrombosis in the liver and the kidne s b 85 res . ±773 i ± 1065 + 1 171 ± 1 171 ±576" * p < 0.05 vs untreated endotoxia controls When saruplase (20/zg/kg-min) was administered only during the last two hours of endotoxin infusion, liver microthrombosis was still significantly dissolved by 69 %, whereas renal microthrombosis was insignificantly reduced by 34 %. There were almost identical positions of the dose-response curves of saruplase for the reduction of endotoxininduced microthrombosis and for the shortening of the euglobulin clot lysis time in normal rats as a measure of its fibrinolytic activity. Saruplase (20/zg/kg'min) did not modify thrombocytopenia (-79 ___ 5 % vs -63 _+ 6 %) and hypofibfinogenemia (-59 ___ 6 % vs -55 + 8 %) in endotoxemic rats. Saruplase per se did not affect plasma fibrinogen levels (187 _+ 24 mg/dl vs 193 _+ 43 mg/dl). Thus, in a fibrin-selective dose range saruplase is able to dissolve microthrombosis associated with DIC in endotoxemic rats. Griinenthal GmbH, Zieglerstr. 6, W-5100 Aachen The proteinase thrombin induces the formation of intravasal thrombi by the convertion of fibrinogen to fibrin. Additionally, thrombin may promote thrombus formation by its contractile action on vascular smooth muscle. We investigated the intracellular messengers which are involved in this thrombin (3U/ml) effect. Experiments were carried out on endothelium denuded isolated ring segments of porcine pulmonary arteries. The initial phase of thrombin-induced contraction was accompanied by a rapid temporary increase of intracellular inositol-l,4,5-triphosphate (IP3). The following sustained tonic phase of contraction was inhibited in part by ve~pamil (10~M) which blocks voltage dependent Ca~-channels, and by staurosporine (0.1~M), an inhibitor of protein kinase C (PKC). Staurosporine at the same concentration was found to inhibit the contraction induced by phorbol dibutyrate (50rLM-I~M) which activates PKC directly.

Verapamil had no inhibitory effect. We conclude that the interaction of thrombin with its receptor on vascular smooth muscle cells is followed by the activation of phospholipase C. Two second messengers are generated: IP which mobilizes Ca 2+ from intracellular stores an~ diacylglycerol which activates PKC. In our experiments activated PKC seems to increase the Ca 2+sensitivity of myosin light chain phosphorylat~gn rather than the activity of voltage dependent Ca ~channels.

The reduced contractile response to thrombin in Ca2+-free medium was still followed by an increase in IP 3 and inhibited by staurosporine. 

In anaesthetized dogs an electromagnetic flow probe and distally to it a concentric perspe× constrictor narrowing the diameter of the vessel to about 20 % of normal were placed around the left circumflex coronary artery, the latter leading to cyclic coronary flow reductions (CFRs, Felts et al. (1976) , Circulation 54:365). CFRs occurred with a frequency of 5 to 8 per 30 rain and were stable over 2 hours In controls. In the treatment groups after 1 h of CFRs either verapamil (V: 0.2; 0,5 mg/kg), trandolapril (r: 0,05; 0.1; 0.5; 1.0 mg/kg), or their combination (V 0.05 + T 0.02; V 0.1 + T 0.05; V 0.2 + T 0.1 mg/kg) were injected i.v. over 2 rain. The number of CFRs occuring in the 2rid hour were compared to the one before drug administration. Maximal doses of V reduced CFRs by 29 % and of T by 74 % during the second 30 min period after administration. As shown in the figure combining the two inactive doses of 0.1 mg/kg V and 0.05 mg/kg T already led to a reduction of CFRs of 70 %, i.e, the maximum achievable with higher doses of T alone. Heart rate, blood pressure and LVdP/dtm= , were lowered significantly by this combination. In a second experiment in anaesthetized dogs the effect on coronary blood flow (CBF) was measured. T alone ( [196] [197] 1991) . Chemical denervation (by 6-OHDA) delayed the begin of hypertension caused by DPSPX, shifting it from days 3/4 to days 6/7. Blood pressure values plateaued by the end of the second week and were in average 20 mm Hg lower than in animals which received only DPSPX.

Our results suggest an involvement of the sympathetic system in the genesis and maintenance of high blood pressure in this model.

However, other mechanisms are certainly involved and therefore we tried to evaluate the effects of ACE inhibition on this hypertensive model, systolic and diastolic blood pressure being measured in conscious restrained male Wistar rats with a tail cuff.

The rats treated with captopril received the drug in their drinking water (100 mg kg -I day -1) from day -I to day

28.

On day 9' ~onstant infusions of either DPSPX 90 ~g kg-h-" or saline were commenced using Alzet osmotic minipumps (i.p.).

On day 28, the rats were killed by decapitation, fragments of the left ventricle, mesenteric, renal and caudal arteries were collected and processed for morphological studies. Captopril prevented the development of hypertention in DPSPX treated rats and was without effect on blood pressure in control animals.

Concomitantly, the hypertrophic and hyperplastic changes induced in arteries by DPSPX were prevented by captopril. We have described a rat model which responds to repetitive episodic hypoxia mimicring sleep apnea (3-5 % lowest ambient oxygen, every 30s for 7 hrs / day for 35 days) with a chronic increase in arterial blood pressure. Denervation of the peripheral chemoreceptors prevents the increase in blood pressure. The purpose of the current study was to determine wether the peripheral sympathetic nervous system is instrumental in producing persistent blood pressure elevation in this model. Chemical sympathetic denervation was achieved and maintained by 3 intraperitonenl injections of 100 mg/kg of 6-OH dopamine on days 2, 4, and 28 of a 49 day experiment in two groups of male Wistar rats (320-410 grns). One 6-OH dopamine treated group (DA/H; n = 8) was subjected to intermittent hypoxia (for 40 days, begin at day 8) and the other 6-OH dopamine injected group remained unhandled in their usual cages (DA/C; n = 11). A third group was injected with placebo only and subjected to the same hypoxia (H; n = 13) while a fourth remained unhandled for 40 days (C; n = 13). Measurement of catecholamines in cardiac muscle homogenate using HPLC continued denervation in 6-OH dopamine animals. There were no significant differences among the 4 groups at baseline.

A Mean BP (%) -1,7 +8,5* -4,7 -9,1 LV/TBW (rag/g) 1,71 1,91' 1,72

1,95" A Hematokrit (%) -0,8 +6,1' -2,9 +2,7 A Hemoglobin (gin%) +0,1 +1 -1,7 +0,6 A = difference between baseline and end of the study; *significant, p < 0,05 The left ventricle-septum (LV)/total body weight (TBV) ratio was -like in all other animals with more than 35 days treatment with intermittent hypoxia -higher in the hypoxia treated groups (H, DA/I-I) at the end of the study. The lack of blood pressure increase in group DA/H -especially in comparision with group DMCshows that an intact peripheral sympathetic nervous system is a prerequisite for the persistent increase in blood pressure in response intermittent hypoxia. In this study we investigated the potency of the new ACE inhibitor moexipril (2-((1ethoxycarbonyl)-3 -phenylpropyl)amino-1-oxopropyl)-6,7-dimethoxy-1,2,3,4 tetrahydroisochinolin-3-earbons~ure and its active diacid, moexiprilat, in vitro and in vivo in comparison with enalapril. In vitro, moexiprilat exhibited a fourfold higher inhibitory potency than enalaprilat against plasma ACE (ICs0 1.5riM vs. 6 aM) and against purified ACE from rabbit lung (IC50 2.1 aM vs 7.5 nM). The prodrug moexipril proved to be three times more potent than enalapril in inhibiting purified ACE from rabbit lung (IC50 2.7 bt M vs. 7.5 IsM), whereas in plasma, where a partial enzymatic activation to the active diacid occurs, both substances are equipotent (IC50 0.18 I~M vs. 0.21 paM), The in vivo activity of both ACE inhibitors was determined in spontaneously hypertensive rats (SHR), which were treated with doses of 10 mg/kg per day by gavage for 4 weeks. The plasma parameters of the renin-angiotensin-system (ACE, angiotensinogen, renin and angiotensin I (ANG I)) were determined 2.5 h after gavage, before and 3, 14 and 28 days after the start of treatment. Both ACE inhibitors lead to a comparable decrease of blood pressure over 4 weeks and inhibited plasma ACE (76-85%) and reduced plasma angiotensinogen (27%) to a similar extent. In contrast, treatment with enalapril resulted in a more pronounced increase in plasma renin (15 vs. 8-fold compared to control) and in plasma ANG I (12.5-fold vs. 4.7-fold compared to control) compared to moexipril. In a further series of experiments SHR were treated with moexipril and e~,lapril for 4 weeks as described above. Blood pressure was recorded via the intraartzrial catheter, and the in rive inhibition of ACE was determined by iv injections of ANG I (I00 rig) and bradykinin (10, 100, 300 and 1000 ng). The presser responses to ANG I after enalapril treatment were inhibited to a greater extent (89%) than after treatment with moexipril (85%). However, both ACE irdaibitors showed a similar dose-dependent potentiation of the depressor responses to bradyldnin. Our results demonstrate that the active diacid form of moexipril is a highly potent ACE inhibitor. Orally applied as a prodrug, meexipril shows an ACE-inhibiting effect and antihyperteasive action comparable to enalapfil. We investigated receptor binding characteristics of e~-and r&-adrenoceptors as well as muscarinic receptors in left ventricular tissue of rat hearts. We used spontaneously hypertensive rats (SHR) and normotensive Wistar Kyoto rats (WKY) as controls. Rats of both groups were made diabetic by a single injection of streptozotocin (60 mg/kg). After 8 weeks binding assays were performed in membrane preparations of cardiac left ventricular tissue from age matched rats by means of [~H]-N-methylscopolamine for muscarinic, [~H]-prazosin for o~-adrenoceptor and (-)-[~%-iodocyanopindolol for ~-adrenoceptor binding. The binding characteristics (Bin,× and K~) are given in table 1. indicates a significant difference from the corresp0ndin 9 diabetic group (P < 0.05)

The presented data indicate that the number of muscarinic receptors is reduced only in SHR compared to the diabetic SHR, wheras o~adrenoceptors are reduced only in WKY compared to diabetic WKY. The reduction of F~-adrenoceptors is consistent for normo-and hypertensive rats compared to the corresponding diabetic group. Interestingly, no difference was found between normoglycaemic WKY and SHR. Diabetes mellitus is associated with changes in the structure and functional characteristics of the vasculature. We investigated the effects of various vasoconstrictors on small arteries of spontaneously hypertensive rats (SHR) with simultaneous diabetes. SHR and normotensive Wistar Kyoto control rats (WK¥) as welt as animals from both groups which experienced an 8 week period of severe diabetes after a single injection of streptozotocin (60 mg/kg) were used. Mean arterial pressures (MAP in mmHg) of both non-diabetic (n~30, 171 ±4.0) and diabetic (160+4.7) SHR were significantly higher compared to both non-diabetic (102±2.5) and diabetic (108±4.4) WKY, Mesenteric arteries were isolated and the preparations were normalized to their individual optimal lumen diameter in an isometric wire myograph. In this respect no difference was found between the four groups (E5-240 pro). The potency of the e~adrenoceptor agonist methoxamine was not different in the four groups (-log ECso 5.86 ±0.O7) but it was significantly more effective in diabetic SHR (Emo X in mmHg: 269+ 7.8) compared to diabetic WKY (200±12.1) and control WKY (232±12.3). For calcium ions, in the presence of 120 mM KCI, -log ECso-Values of preparations from diabetic SHR (3.5B-+0.06) were significantly different compared to those from non-diabetic SHR (3.81 ±0,06) and the latter differed significantly from preparations from nomdiabetic WKY (3.62±0.03). The Emo . of calcium-induced contractions was significantly higher in control SHR (2.78+3.5) compared to control WKY (236_+13.4). For KCI, in the presence of CaCI 2 1.8 raM, the concentration-response relationships were identical in all four groups I-log ECso (M): 1.46_+0.02) except for the Em,x found in preparations of diabetic SHR (181 _+18.6) and diabetic WK¥ (167 _+10.9) compared to non-diabetic SHR (231 +-. 13.7). It is concluded that, in contrast to the well known hypertensioninduced enhancement of the maximal contractile response of small resistance arteries to various stimuli, the diabetic state per se has no consistent influence on the responsiveness of these preparations. Little is known concerning the possible role of endothelium-dependent hyperpolarization in hypertension. For these reasons we have investigated the effects of compounds that interfere with the cellular potassium homeostasis on the (acetyl-r~) metacholine (MCh)-induced endothelium-dependent vasodilation in the perfused vascular bed preparation, obtained from spontaneously hypertensive rats (SHR) and age-matched normotensive Wistar Kyoto (WKY) rats, respectively. Since the maximal relaxation (Ema x, SHR: 87+2.4% and WKY: 85+3.1%, respectively) and pD 2 (SHR; 7.36+0.03M and WKY; 7.39 ± 0.06M, respectively) for the concentration-response curves were the same in both types of preparations there does not seem to exist any endothelial dysfunction in mesenteric resistance arteries of SHR. Pretreatment with 4-aminopyridine (4-AP) (0.5-1.0raM) decreased the pD 2 concentration dependently (SHR: 0.5mM; pD 2 =6.98±0.03M and 1.0mM; PD2=6.78±0.04M, repectively, versus WKY: 0.5mM; pD 2 = 6.92 + 0.02 M and 1.0mM; pD 2 = 6.37 ± 0.03M, respectively). Also tetraethylammonium (TEA) (1.0-10raM) shifted the concentrationresponse curve for MCh-induced vasodilation to the right (SHR: 1.0raM; PD2=6.82±0.04M and 10raM; pD2=6.49±0.03M, repectively, versus WKY: 1.0mM; pD2=6.65±0.05M and 10mM; PD2=6.09±0.03M, repectively). Both types of potassium channel blockers did not affect the maximal vasodilation. Inhibition of Na +/K+-ATPase with dihydro-ouabain (0,1-1.0mM) decreased the maximal vasodilator effect of MCh (SHR: O.lmM; Ernax=41-+9.5% and 1.0mM; Emax =18 + 8.1%, repectively, versus WKY: 0.5mM; Emax=76±3.7% and 1.0mM; Emax=43±7.9%, respectively) without altering the pD 2. Pretreatment with 20 mM K + containing physiological salt solution decreased both the maximal vasodilator effect and the sensitivity to MCh-induced vasodilation (SHR; PD2=6.69+0.18M, Emax=53+7.3% and WKY; pD2=6.83+0.O9M, Emax=61 +4.1%, respectively). The role of ATP-sensitive potassium channels may be ruled out since glibenclamide (0.1-1.0,uM) was without effect on the MCh-induced vasodilation. Differences observed between WKY and SHR preparations for TEA, 4-AP and dihydro-ouabain do not seem to influence the functional reponse induced by MCh.

Dept. of Pharmacotherapy, University of Amsterdam, Meibergdreef 15, NL-1105 AZ Amsterdam, The Netherlands. Angiotensin II (ANG II) has been reported to be a growth promoting factor for vascular smooth muscle cells and fibroblasts. Little is known about its influence on growth of endothelial cells (EC). We investigated the effect of angiotensin peptides on endothelial cells in vitro, using microvascular EC in primary culture, isolated from hearts of spontaneously hypertensive rats (S/dR) and normotensive Wistar Kyoto rats (WKY). Using a non-radioactive cell proliferation assay, we found that ANG II and its metabolite ANG III (ANG 2-8) at concentrations of 10"6M to 10 "11M significantly inhibited serum-stimulated endothelial proliferation of EC originated from SHR, but not from WKY. This effect could not be prevented by the specific ANG II AT1 receptor antagonist losartan, or by the ANG II AT2 receptor ligand CGP 42112A. On the other hand, losartan and CGP 42112A had an inhibitory effect on proliferation of serum-stimulated cells by themselves. The ANG fragments ANG 4-8 and ANG 1-7 as well as the vasoactive Gprotein coupled peptide hormones bradykinin and endothelin-1 did not significantly influence proliferation of coronary EC. Our results reveal an inhibitory effect of angiotensin peptides on endothelial cell proliferation that does not seem to be mediated by selective AT1 or AT2 ANG II receptors. Angiotensin receptors distinct from AT1/AT2 or possible interactions between AT1 and AT2 receptors with respect to endothelial proliferation may be involved. Molecular cloning of angiotensin II(Ang II) subtype 1(AT1) receptors from several tissues has revealed a structure similar to those of the G-protein-coupled receptors. Our previous work has shown that the stimulating influence of Ang II on spontaneous phasic contractile force of the rat portal vein is mediated by AT 1 receptors. In the present study we investigated the effects of cholera toxin(CT), dibutyryl cyclic AMP(db-cAMP) and pertussis toxin(PT) on Ang II-enhanced phasic contractions in the rat portal vein. Incubation of isolated rat portal vein preparations with CT(0.1, 0.5, 2 and 10 pg/ml for 4 h) and rib-cAMP(0.1, 0.5 and 1 mM for 1 h) decreased or abolished(high concentrations) the spontaneous phasic contractions. Ang II-enhanced contractions in the rat portal vein were concentration-dependently inhibited by CT and db-cAMP.

CT is able to AbP-ribosylate the G-subunit of stimulatory G protein(G,), thus causing the stimulation of adenylate cyclase and a rise in cAMP formation. The similar inhibitory effect of db-cAMP on Ang It-enhanced contraction may suggest that the inhibition of Ang II effects caused by CT might result from its ability to increase cAMP production rather than a blockade of an Ang II receptor signal transduction pathway. By contrast, pretreatment of the rat portal vein with 1 pg/ml PT for 4 h affected neither the spontaneous activity nor the myotropic response to Ang It, suggesting the absence of the involvement of inhibitory G protein (G=) in Ang II receptor signal transduction pathway.

In conclusion, the stimulating effect of Ang II in the rat portal vein may not be mediated directly by a Gi or G, coupled to Ang II receptors. It seems likely that the effect of CT on Ang II-induced contraction is due to its ability to increase cAMP production through Gs. Previous findings that hypertension travels with the kidney in renal transplantation studies between genetically hypertensive and normotensive rats suggest a major role for renal mechanisms in the pathogenesis of pdmary hypertension. However, the mechanisms underlying posttransplantation hypertension in recipients of a renal graft from geneticeliy hypertensive donors are currently not well understood. To investigate whether renal angiotensin II (ANG II) or argininevasopressin (AVP) receptors may be involved, we transplanted kidneys from 13 adult male stroke-prone spontaneously hypertensive rats (BHRSP) and 13 normotensive Wistar-Kyoto rats (WKY) to 26 bilaterally nephrectomized adult male FI hybdds bred from SHRSP and WKY parents. Four weeks after transplantation, renal grafts were removed from the recipients and processed for binding studies of ANG II receptors in intact isolated glomeruli and of AVP receptors in tubular membrane preparations. Nontrensplanted kidneys from age-and sexmatched SHRSP (n=10) and WKY (n=10) served as controls. At graft removal, systolic blood pressure was 146-3:11 mmHg in recipients of a WKY kidney and 194+7 mmHg in recipients of an SHRSP kidney (p<0.01). Binding charectedstics of ANG II and AVP receptors were as follows: There were no significant differences between control and grafted kidneys from either strain with respect to AVP receptor binding and the affinity of ANG II receptors. Maximum binding capacity for ANG II was increased in the glomeruli of grafted versus control kidneys from both strains. Since this effect occurred independently of the genetic background of the donors and the actual blood pressure of the recipients, it probably does not contribute to posttransplantation hypertension. Our data suggest that mechanisms related to ANG II receptors may padicipate in repair and adaptation processes of renal grafts after transplantation. We previously demonstrated that angiotensin II (Ang II) induces hypertensive, positive inotropic and chronotropic effects in the pithed rat preparation. In the present study we investigated the effects of several nonpeptide Ang II receptor antagonists: Iosartan (ATe), 1-[(4-amino-3-methylphenyl)methyl]-5-(diphenylacetyl)-4,5,6,7-tetrahydro-lH-imidazol [4,5-C] pyridine-6-carboxylic acid (PD123177, AT2), 4'-[(2-n-butyl-6-cyclohexylaminocarbonylamino-benzimidazole-l-yl)-methyl] biphenyl-2-carboxylic acid (BIBS 39, AT1 and AT2) and 2-n-butyl-l-[4-(6-carboxy-2,5dichlorbenzoylamino)-benzyl]-6-N-(methyl-aminocarbonyl)-npentylamino-benzimidazole (BIBS 222, AT1 and AT2), on cardiovascular responses to Ang II in the pithed rat. Male Wistar rats (320-350 g) were pithed and the diastolic blood pressure (DBP), left ventricular pressure (LVP), dP/dtm,× and heart rate (HR) were measured. Ang II (0.01-300 nmol/kg i.v,) induced dose-dependent increases in all of these parameters. Losartan (3 and 10 mg/kg i.v.) caused dose-dependent parallel rightward shifts ef the dose-response curves of Ang II, without altering the maximal responses. PD123177 (100 mg/kg, i.v.) did not significantly influence the responses to Ang II. BIBS 39 (3 and 10 mg/kg, i.v.) caused parallel rightward shifts of the dose-response curves of Ang II but depressed the maximal dP/dtm,x and HR. BIBS 222 (3 mg/kg, i.v.) produced a similar pattern of antagonism as BIBS 39.

The results indicate that in the pithed rat Ang II induces hypertensive, positive inotropic and chronotropic responses which are mediated by ATe-but not AT2-receptors. BIBS 39 and BIBS 222 are also potent antagonists of Ang II. However, these compounds influence the cardiac responses to Ang II in a manner different from that of a pure ATe-antagonist, suggesting a possible role of AT2-receptors in the heart.

Dept. Pharmacotherapy, Academic Medical Center, University of Amsterdam, Meibergdreef 15, NL-1105 AZ Amsterdam, The Netherlands

Meel, W.Wienen, M.Entzeroth, N.Hauel, M.Panzenbeck* and R.Winquist* BIBR 277, 4"-[(1,4"-dimethyl-2"-propyl[2,6"-bi-lH-benzimidazol]l'-yl) -methyl]-[1,1"-biphenyl]-2-carboxylic acid is a novel potent nonpeptlde AT1 receptor antagonist. I n t r a v e n o u s a d m i n i s t r a t i o n of 0.1-1 m g / k g of BIBR 277 caused dose-dependent shifts to the right of t h e dose-presser response curve of AII in pithed rats with a depression of the m a x i m a l reponse. After Lv. a d m i n i s t r a t i o n of 0.1 mg/kg of BIBR 277 to a n a e s t h e t i z e d rats, AII-induced presser responses were blocked for more t h a n 2 hours. Blockade was still about 50% after 2 hours. Oral a d m i n i s t r a t i o n of 1 and 3 mg/kg of BIBR 277 once daily produced dose-dependent reductions in m e a n arterial blood p r e s s u r e of conscious renovascular (2K 1C) h y p e r t e n s i v e rats. A d m i n i s t r a t i o n via the drinking water (0.3 mg/kg/day a n d I mg/kg/day) over 1 week per dose effectively lowered m e a n arterial blood p r e s s u r e of renovascular hypertensive rats. BIBR 277 was also effective in primates. Blood p r e s s u r e of saltdepleted cynomolgus m o n k e y s was lowered after oral a d m i n i s t r a t i o n of 0.3, 3 and 10 mg/kg. I n conclusion, t h e novel AII-antagonist BIBR 277 is a potent, orally active antihypertensive compound with a long d u r a t i o n of action.

A . P h a r m a Research, Dr. Karl T h o m a e GmbH, 7950 Biberach, G e r m a n y * Dept.Pharmacology, Boehringer Ingelheim P h a r m a c e u t i c a l s Inc., Ridgefield, CT, USA. In the present study we report on the biochemical properties of a novel angiotensin II (AII) receptor antagonist, -:l'-yl) methyl]-[1,1'-biphenyl]-2-carboxylic acid}. BIBR 277 inhibited 125I-AII binding to AT 1 sites in rat lung membranes with a K i of 3.7 + 0.7 nM (n=3) and thereby was 6-and 3-fold, respectively, more potent than the ATl-selective antagonist losartan (K i 23.7 + 2.5 nM) or its metabolite EXP 3174 (2-n-butyl-4-chloro-l-[(2'-(1H-tetrazole-5-yl)biphenyl-4-yl)methyl]imidazol-5-carboxylic acid (K i 10.4 + 0.9nM). PD 123.177, 1-[(4-amino-3-methylphenyl)methyl]-5-(diphenylacetyl)-3,5,6,7-tetrahydro-:lH-imidazol-[4,5-c]pyridine-6-carboxylic acid, selective for the AT 2 subtype, was inactive up to 10 ,uM. No displacement of the radioligand from AT 2 sites in rat adrenals by BIBR277 was observed in sub-micromolar concentrations (K i > 10.000 nM). In autoradiographic studies, using rat kidney and adrenal slices, BIBR 277 (1/~M) inhibited 125I-sarcosinel,isoleucine8-AII binding to AT 1 sites present in the glomemli, renal medulla and adrenal zona glomemlosa while binding to AT 2 sites present in the adrenal medulla was not affected. In additional binding experiments it could be demonstrated that BIBR 277 (< 1 /~M) did not interfere with other receptor systems relevant for cardiovascular function. In addition, BIBR 277 (10/zM) did not interact with other components of the human renin-angiotensin system such as serum angioteasin I-converting enzyme or plasma renin activity.

In conclusion, BIBR 277 selectively and potently interacts with angiotensin II receptors of the AT 1 subtype. 1"-biphenyl]-2-carboxylic acid} is a novel nonpeptide angiotensin II (AID receptor antagonist. Here we demonstrate the in vitro potency and selectivity in functional studies in the rabbit aorta. The concentration-contractile response curve (CCRC) to AII was shifted to the right in the presence of 10, 100, and 1,000 nM BIBR 277, respectively. However, the Hill slopes (nil) of the curves were significantly decreased at all concentrations (control: 2.12+0.08 vs. 1.76+0.22, 1.37+0.08, and 1.66+0.09 in the presence of BIBR 277; p<0.05). In addition the maximum contractile force was also significantly decreased to 45+3, 52+3, and 41+3% of the control, respectively. The calculated dissociation constant, KB, for BIBR 277 was 3.3+0.9"10-10M. Similar results were obtained for the CCRC to angiotensin III, yielding a K B of 1.5+0.6"10-10M. BIBR 277 did not exhibit agonistic activity up to 100 /~M in this preparation. The interaction of BIBR 277 with AII receptors was shown to be reversible since ~.' the observed insurmountable antagonism was concentrationdependently reversed in the presence of the surmountable antagonist losartan, demonstrating the specific interaction with the AII receptor; b: complete removal of BIBR 277 from the tissue baths during several washout-cycles partially restored the contractile response to AII; and e: a slow off-rate from AT 1 receptors could be demonstrated in binding studies. In the presence of 10 k~M BIBR 277, the CCRC to norepinephrine or KC1 were not influenced nor was endotheliumdependent relaxation inhibited in precontracted rabbit aortic rings. In conclusion, BIBR 277 is a potent and selective antagonist for the AII receptor in functional studies in vitro. The transgenic rat TGR(mREN2)27 harbouring the Ren-2 gene of the mouse is a new animal model for.hypertension, in these rats, the components of the circulating remn-angiotensin system (RAS) are suppressed, whereas renin gene expression and activity is increased in various tissues particularly in the adrenal gland. We treated these animals with the newly developed specific angiotensin II receptor AT 1 antagonist BIBR277 {4'-[(1,4'-dimethyl-2'-propyl[2,6'-bi-lHbenzimidazol]-1 '-yl) methyl]-[ 1,1 '-biphenyl]-2-carboxylic acid} applying three doses, 0.1, 1 and 3 mg/kg body weight, for 9 weeks. The two higher doses normalized blood pressure after one week of treatment and, after a period of four weeks, also increased plasma renin and angiotensin II concentration, while the lowest dose showed no significant effects on these parameters. The genes for renin and angiotensinogen are differentially regulated by the angiotensin II receptor blockade. The level of angiotensinogen mRNA is increased in liver, kidney, and heart of treated animals, while in aorta the amount is reduced. In the kidney, the expression of the endogenous renin gene was increased accompanied by a raise in renin concentration, while the level of transgenic Ren-2 mRNA was unaffected. In parallel, angiotensin II levels were decreased in the kidney. Furthermore, the severe glomerulosclerosis and proteinuria observed in untreated TGR(mREN2)27 was ameliorated even with the lowest dose of BIBR277 lacking any effect on blood pressure. In addition, the ratio heart weight to body weight was significantly reduced by the lowest dose indicating a decrease in hypertrophy compared to controls. These results together with a high sensitivity to converting enzyme inhibition show that blood pressure in TGR(mREN2)27 is angiotensin II dependent. In addition, the observed pathology of the kidney and heart seems to be caused not secondary to the hypertension but independently by an effect on the oca R A S n these organs. Thus, TGR(mREN2)27 represent an important new model for investigations of the role of tissue RAS in cardiovascular disorders. Interference with the enzyme peptide cascade of the renin-angiotensin system results in blood pressure reduction. Besides angiotensin II (ANG II) receptor antagonists and ACE inhibitors renin inhibitors are considered to be potential antihypertensive agents. The blood pressure lowering effects of the new renin inhibitor S 2864 (N-[N-(3-(4-Amino-1 -piperidinyl-earbonyl) -2 (R)-benzylpropionyl) -L-histidinyl]-(2S,3R,4S)-1-cyclohexyl-3,4-dihydroxy-6(2-pyfidyl)-2-hexylamide-acetate) were tested in anesthetized, sodium depleted rhesus monkeys following intraduodenal and intravenous administration. In vitro, S 2864 inhibited the activity of purified human plasma renin with an IC50 of 3.8 x 10 "10 tool/1 and did not affect other human aspartyl proteases. In rhesus monkeys S 2864 (2 mg/kg) given i.d. decreased mean arterial blood pressure significantly by 27 % from 84 + 8 to 62 + 6 mmHg for 90 rain. A marginal heart rate decrease (duration 45 rain) followed by a small increase after 120 rain was observed. Cumulative i.v. administration of S 2864 in doses of 1, 10 and 30 ~tg/kg significantly decreased systemic blood pressure, dP/dtma x and cardiac output while heart rate was not changed. Plasma ANG II levels as well as renin activity were dose dependently reduced after 10, 30 and 60 rain. In conclusion, S 2864 is a potent and specific renin inhibitor, able to reduce blood pressure in anesthetized sodium depleted rhesus monkeys following intraduodenal administration. In order to determine the relative contribution of the paracellular and transcallular pathways to the barrier properties of gastrointestinal (GI) epithelia, we set up a modified Ussing chamber model. Electrical short-circuit currants (Isc), transmembranous resistances (Rt) as well as [3H]-mannitol flux rotes (MFR) were measured as parameters for trans-and paracellular permeability. We intend to use this model for studying putative pathophysiological and pharmacological influences on barder propedies of the GI epithelium, which might be important for the understanding and treatment of several disease states like peptic ulcera and inflammatory bowel disease. To validate our model, we investigated inhibitors and enhancers of active transport mechanisms, "barder breakers", and compounds interacting with the cytoskeleton. Rat stomachs were separated from their musculeds propria in ice-cold preparation and mounted into the Ussing chambers which were filled with HEPES buffered solution and kept at 37°C. Electrical potential between apical and basolateral side of the chamber was clamped to zero potential and Isc measured using an automatic voltage clamp system. FI t was calculated from the deflections in Isc produced by 5 mV pulses (0.02 Hz, 0.5 s) passed across the preparation. MFR was measured by adding [~H]-mannitol (6.5 p.Ci) to the apical side and taking samples from the basolateral side. In control preparations we observed a slight but consistent reduction of R t and a small increase in MFR over 6 h. Isc showed marked variations between prep~arations, and tended to decrease with time. The calcium ionophore A 23187 (5.10 "~ M) increased Is~, but had no effect on R t and MFR. Ouabaln, in contrast, significantly reduced Is¢ without any effect on R t and MFR. The calcium channel blocker verapamil (10 "4 M) and antimycin A (10 p.g/ml), an uncoupler of oxidative phosphorylation, significantly reduced R t and slightly increased MFR, without affecting Isc. Cytochalasin D (10 ~tg/ml) which specifically changes tight junction ion permeability, markedly reduced R t which was parallel~l by an increase in MFR. Apical treatment of the epithelium with the "barder breaker" ethanol (15 % v/v) caused a significant reduction of R t and an increase of MFR, very similar to cytochalasin D.

The isolated native rat stomach epithelial preparation is highly suitable for studying barrier properties of GI epithelia. Ethanol seems to primarily increase the permeability via the paracellular pathway, thus breaking the epithelial barrier for hydrophilic compounds which after penetration into the mucosa might lead to subsequent damage. It has been reported for high doses of omeprazole, 138 and 430 mg/kg, to interfere with the peripheral conversion of thyroxine (T4) to triiodothyronine (T3) [Scand. J. Gastroenterol. 1985, 20 ISuppl. i08), 53-69]. We studied the effect of omeprazole, lansoprazole, saviprazole, SK&F 65601(2-[[(3-chlero-4-morpholino-2-pyridyl)methyl] sulfinyl]-5-methoxy-benzimidazole) and E-3810 (2-[[4-(3-methoxypropoxy)-3-methylpyridine-2-yl]methylsulfinyl]benzimidazole) on total serum T4 and T3 as well as on thyroid weights after treatment for 14 days with doses of 50, iS0 and 450 mg/kg. The selected doses represent supramaximal doses with respect to their antisecretory effects in rats (EDs0 in pylorus-ligated rats: 1-5 mg/kg). Omeprazole and SK&F 65601 caused slightly elevated serum T4 and T3 without showing dose-dependencies and no changes of the thyroid weights. The treatment period of 14 days was too short to expect the reported decrease in total serum T3 for omeprazole. E-3810 caused significant elevated serum T4 and T3 with an inverse dose-dependency without change of the thyroid weight. Saviprazole and lansoprazole dose-dependently decreased total serum T4 without affecting serum T3 levels. Although both compounds showed the same degrees of serum T4 lowering, there was a significant weight increase of the thyroid gland in the highest dose for saviprazole only. Lansoprazole showed the tendency of a similar increase in thyroid weight after prolongation of the treatment period up to 6 weeks at the highest dose tested. Further studies for saviprazole to elucidate the mechanism of decrease in serum T4 revealed that the compound does not interfere with the hormone synthesis, storage and release at the level of the thyroid gland. The decrease in serum T4 was caused by an increased hepatobiliary elemination of T4 due to hepatic induction of T4-eleminating enzymes. The increase in thyroid weight induced by 450 mg/kg saviprazole in rats could be prevented by the substitution of T4 dose-dependently. (control) and presence of antimycotic agents (i-1990 ~g/ml) by measuring histamine-stimulated [x~C]aminopyrine uptake (AU) as an indicator for cellular acid production, and trypan blue uptake (TB) and release of lactate dehydrogenase (LDH), acid phosphatase (AP) and cytochrome c oxidase (CC), as indicators of cellular integrity, during a 1 h incubation at 37~C. Substance concentrations ,resulting. in a 50% inhibition of AU or a 50% increase In TB, LDH, AP or CC were calculated and grouped into 5 classes (~g/ml): 4) 0-i0, 3) 11-30, 2) 31-100, i) 101-300, 0) >300. Overall toxicity (OT) was defined as the sum of the calculated class-numbers. Indometacin, known to damage w t! gastric mucous cells in vivo , was used for comparison. Results: According to increasing OT values the toxlclt~ rank-order was: griseofulvin OT:0 (AU:0, TB:0, LDH:0, AP:0, CC:0), flucytosine 0 (0,0,0,0,0), fluconazole 2 (0,0,i,i,0), nystatin ii (1,3,3,2,2), amphotericin B ii (2,2,2,3,2), indometacin 13 (2,3,2,3,3).

Summary: Use of isolated gastric parietal cells allowed to detect marked differences in the cytotoxic potential of different antimycotic agents. Amphotericin B and nystatin, the most noxious compounds, exerted cytotoxic effects in concentrations which may be reached in the gastric mucosa after oral application.

This may contribute to the gastrointestinal side effects observed during amphotericin B and nystatin therapy. The gastric H+,K+-ATPase is the pump responsible for secretion of HCI by the stomach. Acid pump inhibitors of the benzimidazole class have shown superiority to H 2 receptor antagonists in therapy of peptic ulcer diseases. Their mechanism of action involves accumulation of the compounds in the acid space of secretory canaliculus of the parietal cell. This is followed by conversion of the original compound to a reactive sulfenamide, that is a permanent cation. The sulfenamide is able to react with available cysteine SH-groups of proteins that are present in the canalicular membrane bordering the acid space forming disulfide derivatives. In the pre~ent study it was possible to define the cysteines of the H , K+-ATPase that react with the sulfenamides by using radiolabelled compound and acid transporting vesicles isolated from hog gastric mucosa. This is followed by tryptic fragmentation of the labelled enzyme and N terminal sequencing of the membrane peptides after SDS tricine gradient gel separation. Pantoprazole

(5-difluoro-methoxy-2-(3,4-dimethoxy-2-pyridinyl-methyl)-sulfinyl)-iH-benzimidazole), was shown to label only cysteine813 and cysteine822 in the extracytoplasmic loop between the fifth and sixth membrane spanning segments of the alpha subunit of the pump. In the sixth membrane spanning segment conformational change was detected since the preferred tryptic cleavage site prior to the fifth membrane spanning segment moved from arg775-Ieu776 to lys783-ser784 and, for enzyme with both cysteines labelled, to lys791-asn792. The specificity of reaction suggests that the protonated form of pantoprazole is able to access this region of the enzyme prior to the formation of the sulfenamide. The selectivity of pantoprazole for acid pump inhibition could be shown. "in vltro", gastric mucous cells were isolated enzymatically and enriched by counterflow centrifugation. Cell vitality was estimated in the absence (control) and presence of antimycotic agents (i-i000 ~g/ml) by measuring [3H]L-leucine incorporation (LI) into acid-precipitable material as an indicator for cellular protein synthesis, and trypan blue uptake (TB) and release of lactate dehydrogenase (LDH), acid phosphatase (AP) and cytochrome c oxidase (CC), as indicators of cellular integrity, during a 1 h incubation at 37°C. Substance concentratlons resulting in a 50% inhibition of LI or a 50% increase in TB, LDH, AP or CC were calculated and grouped into 5 classes (~g/ml): 4) 0-i0, 3) 11-30, 2) 31-100, i) 101-300, 0) >300. Overall toxicity (OT) was defined as the sum of the calculated class-numbers. Indometacin, known to damage gastric mucous cells "in vivo", was used for comparison. Results: According to increasing OT values the toxlclt'-~-~'yrank-order was: griseofulvin OT:0 (LI:0, TB:0, LDH:0, AP:0, CC:0), flucytosine 1 (0,0,0,i,0), fluconazole 1 (0,0,0, 0,i), indometacin 5 (i,i,i,i,i,i), nystatin 6 (i, 1,1,2,1), amphotericin B 20 (4,4,4,4,4) summary: Use of isolated gastric mucous cells allowed to detect marked differences in the cytotoxic potential of different antimycotic agents. Amphotericin B, the most noxious compound, exerted cytotoxic effects in therapeutic serum level concentrations. Since gastric concentrations after oral application may be even considerably higher these effects may contribute to the marked gastrointestinal side effects observed during both l.v. and oral amphotericin B therapy. Oral administration of ethanol (75 %, v/v, lml bolus) to fasted rats produced gastric hemorrhagic lesions in the mucosa which could be inhibited by octreotide at doses of 0.01-10.0 ng/kg given subcutaneously (s.c.). The dose-response relationship was bell shaped with maximum inhibition occurring at the extremely low dose of 0.1 ng/kg at which there was no effect on gastric acid secretion. The low dose needed for cytoprotection contrasts with the dose of octreotide required to inhibit growth hormone (GH) secretion where the IDs0 was 0.13 .u.g/kg (0.103 -0.159). The somatostatin analogue, SDZ 217-717 (5-(I)-citrullinoctreotide) neither prevented mucosal lesions at doses up to 10 ng/kg s.c. nor inhibited GH release at doses up to 100 p.g/kg (s.c.). In order to elucidate further the cytoprotective mechanism of octreotide we investigated its effect on ethanol-disturbed microcirculation of the rat mucosa using an in vivo microscopy system. The mucosal microcirculation was continuously observed with a video camera attached to a microscope and the experiments were analysed by an image analysing system. After pretreatment of the rats with octreotide for 15 rain, FITC-dextran (fluoresceine-isothiocyanate dextran) was given i.v. and 1 ml of ethanol (50%) or physiological saline was applied topically for 5 min to the mucosa. The microcirculation was observed for another 60 rain. Ethanol induced stasis of the microcirculation and a fall in mucosal blood flow within seconds. Stasis could be observed first in the collecting venules followed by the capillaries. Octreotide prevented the stasis dose dependently and significantly reversed the fall in blood flow with the same bell shaped doseresponse relationship. The ethanol-induced permeability changes in the microvessels of the mucosa monitored by extravasation of FITC from the capillaries were also reduced by octreotide. SDZ 217-717 had no cytoprotective effect. These results suggest that octreotide maintains the cellular integrity of the gastric mucosa by a mechanism other than the inhibition of gastric acid secretion. This effect is a feature of the endocrinologically active somatostatin analogue as shown by the ineffectiveness of 217-717. Octreotide appears to inhibit ethanol induced stasis in the microcirculation and extravasation from the capillaries. 1Sandoz-Pharma Ltd., CH-4002-Basel, 2University of Frankfurt, D-6000-Frankfurt.

Th. Buhl, A.W, Dunbar Somatostatin (SRIF) and several of its analogues are potent inhthitors of gastric acid and pepsinogen secretion after i.v. or s.c. administration. After oral administration, however, these oligopeptides are almost completely inactive, probably due to rapid degradation and poor absorption in the GI tract. Here we report for the fnst lime that a specially t ~ designed SRIF analogue EK+)Maltose-(D)Phe-Cys-Phe-(D)Trp-Lys-Thr-Cys-Thr-ul acetate (SDZ CO 611), potently inhibits gastric acid and pepsinogen secretion after oral administration in the dog. Acid and pepsinogen secretion was measured in gasUic fistula dogs after stimulation with bethanecbol 60 p.g/kg s.c.. An aqueous solution of the drug or control was applied orally 1 h before bethanechol stimulation. Gastric acid secretion was quantified by measuring total gastric acid output in 15 rain intervals. For the determination of pepslnogun secretion, stomachs were continuously perfused with hydrochloric acid (0.01 N, 2 ml/min) to keep gasUic pH < 4 during the whole experiment. This procedure had no effect on basal and stimulated pepsinogen secretion and allowed us to quantify pepsinogen secretion independently from simultaneous effects on acid secretion (Buhl T and Dunbar AW, Naunyn Schmiedeberg's Arch Pharmacol (1992) 345:R89) . SDZ CO 611 dose-dependently inhibited gastric acid secretion with an El)so of 1.5 ~tg/kg. The effect was maximal with 10 I~g/kg where we observed an almost total inhibition. In experiments with acid perfusion, SDZ CO 611 also inhibited pepsinogen secretion in a dose-dependent fashion, with maximal inhibition oeeuring at 10 p.g/kg. The El)so for the inhibition of pepsinogen secretion was 2.4 p.g/kg. The I~ antagonist cimetidine also completely inhibited gastric acid secretion at 30 mg/kg, but failed to inhibit pepsinogan secretion. Thus, SDZ CO 611 potently inhibits both acid and pepsinogen secretion with almost identical EDso-valnes. As pepsinogen content of gastric juice is reduced even in an acidic gastric environment, this effect cannot be due to its simultaneous inhibition of acid secretion, bat seems to be mediated by a direct and independent effect on gastric chief cells. As both acid and pepsin ate important in the pathogenesis of peptic ulcer disease (Soll AH, N Engl J Med (1990) 322:909.916 ), this appears to be an interesting profile for a novel antiulcer drug. Corlclu$ioll: Our data clearly demonstrate that SDZ CO 611 is an orally active somatnstatin analogue which potendy and efficiently inhibits gastric acid and pepsinogen secretion in gastric fistula dogs. 

Peristaltic contractions of the gut are mediated by cholinergic and tachykininergic neuroeffector transmission. The type of tachyldnin receptors involved in peristalsis was studied by use of NK-1 and NK-2 receptorselective tachykinin antagonists. Two methods for studying peristaltic motor activity in segments of the guinea-pig small intestine were used. The ascending enteric reflex (AER) contraction of the circular muscle was evoked by inflation of an intraluminal balloon and recorded orally from the distension site. Aborally When administered to rats either p.o. (60 min) or s.c. (30 min) before the intragastdc administration of 1 ml of 100% ethanol for 1 h, CO protected the gastric mucosa from ethanol-induced lesions. The inhibition was dose-dependent with a maximum effect obtained at 1 ng/kg s.c. (10 ng/kg p.o.). Quite surprisingly, doses above 100 ng/kg were not effective and thus the dose-response curve was bellshaped, In addition, gastric lesions induced by administration of indomethacin (30 mg/kg p.o. for 4 h) to rats were also inhibited in the same dose range (max. effect at 1 ng/kg s.c.). The dose-response curve for the inhibition of indomethacin-induced lesions was even more complex, since doses between 100 ng/kg and 10 ~tg/kg were ineffective but high doses (0.1-1 mg/kg) caused a protection which was complete at 1 mg/kg p.o.

Besides its cytoprotective effects, CO also caused an inhibition of basal gastric acid secretion in rats equipped with chronic gastric fistula after p,o and s.c. administration. The doses needed for this effect, however, were considerably higher and an EDs0 of 3.4 pg/kg s.c. was determined representing half maximal inhibition of acid output. CO also stimulated gastric motility in conscious rats, determined by measuring the emptying of small lead glass beads into the small intestine under Xray control. These effects occured at doses between 0.1 and 10 pg/kg s.c.

This demonstrates that the cytoprotective effects of CO at low doses cannot be explained by effects of CO on gastric acid secretion or gastdc emptying since those effects require much higher doses. On the other hand, the additional protective effect against indomethacinqnduced lesions obtained at high doses, seem to be due to the inhibito~ effect of CO on gastric acid secretion. In conclusion., CO is a true cytoprotective agent which inhibits gastric lesions at extremely low doses. Its mechanism of action remains unclear, Preclininal Research, SANDOZ Pharma Ltd., CH-4002 Basel, Switzerland 404

Schroeder P*, Elsenhans B ÷, Forth W ÷, SchUmann K + In Fe-deficient rats intestinal Fe transfer is increased only in the proximal small intestine. Such longitudinal differences can either he due to an intrinsic programme which is specific for the longitudinal location or to different luminal conditions. When ileal segments are transposed to the proximal jejunum they gain the ability to adapt their Fe transfer to the demand, though the transfer capacity of the original jejunum is not reached completely.

The present study investigates to what extent the adaptation of Fe-transfer to the demand is preserved when a proximal jejunal segment is excluded from its intestinal continuity. Segments from 20 male Sprague Dawley rats (15 cm, starting from the flex. duodeno-jejunalis) were transformed into blind loops with a distal stoma in the ventral abdominal wall. 20 rats were sham-operated.

Fe deficiency was introduced in half of the animals by feeding of an Fe-deficient diet during rapid growth. After 4 weeks the intestinal transfer of SSFe was investigated in vitro.

Water and glucose transfer showed no significant differences between any of the investigated segments, indicating unimpaired vitality in vitro. The enterocytes in the blind loops maintained their ability to adapt their Fe-transfer capacity to the demand. The mucosal surface of the blind loops, however, had decreased to approx. 50% of that of sham operated segments. Correspondingly, intestinal Fe transfer as related to the segments' length had decreased to approx. 60% of that found in Fe-deficient and Fe-adequate sham operated controls, respectively.

Consequently, there was no difference in the serosal to mucosal transfer-ratio for ~Fe. Thus, the lack of passing ingesta was no triggering factor to impaire the adaptative capacity for S~Fe transfer.

* Allg. 23 Kiel, Permeability of intestinal epithelia to small solutes is dominated by properties of the paracelhilar pathway; it rises upon tissue damage. In duodenal mueosa, permeability is controlled by prostaglandins and HCO 3-at the level of the occluding junctions (Macherey et aL, Am. J. Physiol., in press). Here we have verified a similar regulation in the ileum, a tissue prone to damage by cyclooxygenase inhibitors. As an indicator of permeability, electrical conductance (Ca) of isolated guinea-pig ileal mucosa was studied in Ussing-type chambers. Freshly mounted preparations showed a spontaneous rise of Ca that was particularly fast in nominally HCO 3--free solution~ These increases (e. g. from ~30 to ~60 mS/cm 2 during 80 rain) ceased and reverted towards initial values aider restoring, at constant pH, 20 mmol/l HCO 3-to both sides of the tissue. Addition to the serosal bath was more effective than luminal addition. The Ca drop required Na + and was antagonized by furosemide (10 -3 mol/l), both in the sernsal bath, suggesting cell uptake of riCO 3-at the basolateral membrane by a Na+-dependent transporter. The cyclooxygenase inhibitor, meclofenamate (3x10 -5 mold, added to both sides in the presence of riCO3-) accelerated the spontaneous rise of Ca with respect to control tissues. In both groups, PGE 2 (10 -6 mold, serosal side) reverted the increases; this effect was more marked in the presence of meclofenamate, so that Ca values were nearly equalized in the two groups. Effects on Ca were dissociated, in time course and, partly, direction, from changes in shortcircuit current (transient stimulation byHCO 3-and PGE 2, inhibition by furosemide). As this current measures anion secretion, the sustained changes in Ca cannot be secondary to cell swelling that might reduce the width of the intercellular space. Rather, in analogy to duodenal mucosa (Macherey and Petersen, Gastroenterology 97:1448, 1989 ) availability of prostaglandins may allow HCO 3-to lower junctional permeability. In vitro incubated, ileal mucosa seems to become deficient in endogenous prostaglandins, consistent with exogenous PGE 2 being able to stabilize Ca, provided HCO 3-is present at the serosal side. Human gallbladder has the ability both to absorb and secrete fluid. We have studied ion transport in preparatious from cholecystectomized patients. Tissues were partly stripped of the muscle layers for in vitro determination of short-circuit-current (I~), transepithelial voltage 0/ms), tissue conductance (Ca), and unidirectional ion fluxes (ji), in absorptive (Jrns) and secretory (Jsm) direction, using voltage clamp, pH-stat and tracer flux methods. ~ 30% of the tissues showed values of Yms and Ise close to zero and did not respond to secretagogues (=S; V/P, 3x10 -7 mol/l or 8-Br-cAMP, 10 -3 mol/l). Such tissues were discarded. In a group of 8 responsive preparations, Vms ~ 1.4 mV (lumen negative); Ise ~ 2.1 lamol/cm2h; Ca ~. 41.5 mS/cm :z. Ise and Vms were abolished by ouabain (3x10 -5 mold, serosal side) and roughly doubled by S. Similar though smaller stimulatory effects were obtained with prostaglandins E 1 or E 2. The CI" channel blocker, 5-nitro-2-(3-phenylpropylamino-) benzoic acid (NPPB; 3x10 -5 mol/l, luminal bath) abolished stimulation by S and also decreased spontaneous Vms and Is~. Ineffective was the diuretic amfloride at a concentration sufficient to block epithelial Na + channels (10 -4 mold, luminal side). Spontaneous and stimulated Is~ required the presence of either CI" or HCO 3" in the bath; for maximal stimulation, both anions were needed. S reduced j~o3 by ~ 0.2 (n=12) and raised j~o~ by ~, 0.4 pmol/cm2h (n=lS); the latter effect was antagonized by NPPB. Untreated tissues absorbed CI" at a rate of ~, 5 lamol/cm2h that was reduced by S to ~-1.5 lanol/cm2h (n=6) due to changes in J~. Net Na + transport was lacking and so were effects of S on jNa. We conclude that gallbladders from cholecystitic patients are either not engaged in electrolyte transport (damaged or fibrotie tissues) or secrete anions by electrogenic mechanisms that may be further stimulated by S. Consistent with animal models, stimulation may depend on induction of apical membrane anion channels. Intestinal epithelia from immunized animals respond to antigens with electrogenic secretion of anions, the thiving force for volume secretion, i. e, diarrhea. In this report, we show for the first time that airway exposure to an allergen can be sufficient to condition colonic mucosa for such a response. Guinea-pigs were sensitized by exposing the airways to a nebulized 2% ovalbumin solution for 15 min three times (day 1, 3, and 7). 14-28 days after the last treatment, animals were killed and the distal eolun was removed. The mucosa was isolated, mounted in Ussing-type chambers and continuously short-circuited for determination of tissue conductance (Ca), transpithelial voltage (Pd), and short-circuit-current (Ise). Serosal addition of ovalbumin (10-6 mold) caused a transient increase in Pa, Ca, and Is~, the latter, in isolated mucosa, peaking up to 6 lamol/cm2h within 1.7 rain. In a full thickness preparation, the effects were rather sustained though less dramatic. Ovalbumin was ineffective in non-sensitized animals. Pre-incubation with either the H, receptor antagonist, ceterizine, or the cyclooxygenase inhibitor, meclofenamate (both at 3xi0-5 mold in the serosal bath) abolished the response. The typical stimulation oflse by PGE, (10-6 mold) was largely enhanced in the presence of ovalbumin even when ovalbumin-induced Ise had already ceased. Histamine evoked a concentration-dependent (3x10-6-3x10-5 mold, serosul bath) elevation in ]se in sensitized and non-sensitized tissues. As with ovalbumin, this stimulation was blocked by meclofenamate or ceterizine. Isc effects were also suppressed by indomethacin (10-5 mold), furosemide (10-3 mol/l), trifluoperazine (3x 10-5 mold, all in the serosal bath), and reducing bath C1-to 5 mmol/l. PGEz, at a per se ineffective eoncentration (10-7 mold), enhanced the effect of subsequently added histamine (3x10-5 mold). Effects of histamine were not influenced by tetrodotoxin (10-6 mol/l) or the H2 receptor antagonist, cimetidine (10-5 mold, both in the serosal bath). We eonehide that ovalbumin, after airway sensitization, causes electrogenic secretion of CI-in an action mediated by histamine and subsequent formation of prostaglandins. Sensitization may also boost efficiency of secretagogues. the fractions of phenols or alkaloids respectively. Following a standard operation procedure (SOP), which is in use for many years in our laboratory and has shown good and reproducible results, the rat liver was perfused with a Krebs-Henseleit-buffer containing glucose and 0,5% bovine serum albumin. As vitality parameters of the live~, 02-consum ption and K+-release as well as the concentrations of the enzymes GOT and OPT were determined continuously. After a short time of equilibration the substances were applied to the buffer over a period of 30 min.. The bile fluid was collected in 10-min. fractions and weighed. The concentrations of bile acids were measured and the bile-acidproductions were calculated.

The ethanolic full extracts of Chelidonium majus L. induced a marked increase of bile flow, whereas neither the fractions rich of phenolics but without alkaloids nor the fractions rich of alkaloids buf free of phenolics induced a similar pronounced bile flow increase. The bile-acid-concentration is decreased during the bile flow increase, whereas the bile-acid-production is not altered. The glycosylated fraction of human serum albumin (HSA), reaching 0.4-2.6% in healthy subjects, is increased up to 2-3 fold in diabetic patients. Glycosylation, as welt as the presence of free fatty acids (FFA), separately slow down binding of the site II marker ligand dansylsarcosine (DS) to HSA. We examined for a possible effect of FFA additional to gtycosylation on binding of DS to HSA. The kinetics of the fluorescence shift following addition of DS to six different concentrations between 5.5 x 10"6 -1 x 10-4 of 7% glycosylated HSA, 25% glycosylated HSA and 25% glycosylated HSA with FFA were measured using the stopped-flow technique. A two-step association model was assumed, the second step being rate-limiting. Glycosylation of FFA free native HSA (7% glyc.) decreased the associtiation rate constant k2, which was (mean :1: SD) 686 + 61 s-1 for native HSA, and 385 :t: 10 s "1 for 25% gtycosylated HSA, respectively. The additional effects of FFA on binding of DS to 25% glycosylated HSA are described in the table: Interferon--is widely used an adjuvant in the chemotherapy of human neoplasms; long term interferon therapy is frequently associated with side effects which affect the thyroid gland such as hypo-and hyperthyroidism as well as thyroiditis. Previous studies in man have implicaW.xt that theses alterations in thyroid function result from indirect effects on the immune system and pituitary secretion. In the present study, we have tested the ability of interferon-a to exert direct effects on primary cultures of human thyroid epithelial cells: (i) Interferon-~ inhibits thyroid cell proliferation as determined by [aI-I]thymidine incorporation in a concentration-dependent manner with a half-maximal effect at -lng/rnl (50 pM). Inhibition of cell growth is observed in cells derived from normal thyroid as well as neoplastic tissue (autonomous and non-secreting adenoma, follicular and papillar carcinoma). This effect appears in part additive to that of suramin which is capable of antagonizing various mitogenic stimuli (EGF, TSH, serum factors). (ii) Over a similar concentration range, interferon-a suppresses thyroglobulin release by thyroid epithelial cells. (iii) Interferon-~x stimulates expression of major histocompatibility class (MHC) I but not MHC II genes as determined by fluroescence-activated cell sorter analysis after staining with the monoclonal antibodies A051 and Anti-HLA-DR, respectively. In contrast, incubation of thyroid epithelial cells with interferon--/ results in the enhancext expression of both MHC I and MHC II genes. These findings confirm that interferon-~t directly affects thyroid function and may explain the side effects of this cytokine observed in man. In addition, the antiproliferative action of interferon-~ and suramin may be exploited in the treatment of patients with advanced thyroid cancer for whom no therapeutic alternative exists. 

delays the inactivation of Na ÷ -channels in several tissues. Na* -channels are known to be present in islets of Langerhans, their role in the glucose-mediated insulin release is not yet established. To further evaluate this role in the stimulus-secretion coupling of the insulin-secretory mechanism, we studied the effect of BDF 91 48 on isolated mouse toancreatic islets. BDF 91 48 inhibited the glucose induced (16.7 mM) insulin release in a concentration dependent way, the effect being most pronounced at 1 O0 uM (361.3 ± 33.1 vs 142.7 ± 14.3 ~tU/5 islets, p<O.05). BDF 9148 (10 ~M) also inhibited insulin release evoked by tolbutamide (1 O0 I~g/ml) or KCI (20 raM) significantly but was ineffective on insulin release induced by low glucose concentrations (3 to 11.1 raM). BDF 9148 (10 IIM) neither affected a e Rb ÷ efflux nor 4~ Ca=* uptake induced by glucose, tolbutamide or KCI. ==Na-uptake was significantly enhanced in the absence of glucose though no effect was seen at stimulatory glucose concentrations. Electrical activity of mouse islets remained unchanged in the presence of 10 t~M BDF 9148 both, at 8 or 16 mM glucose. Our data show that in the mouse islet BDF 9148 inhibits glucose-stimulated insulin release. Since at stimulatory glucose concentrations no effect on = = Na* or Ca = * uptake or electrical activity is seen this suggests that BDF 9148 acts at a site following the uptake of Ca =* and Possibly indelaendent from the Na-channel activation. Early stages of uncomplicated diabetes are described to be associated with increased blood flow and reduced peripheral resistance. The mechanisms underlying this reduction of vascular tone is still unknown. We therefore investigated the effects of a prolonged exposure to pathophysiological Dglucose concentrations on agonist-induced EDRF formation in cultured porcine aortic endothelial cells. EDRF formation is strictly due to increases in intracellular free Ca2+ concentration ([Ca2+]i). Methods: [Ca2+] i was measured using the fura-2 technique. EDRF formation was determined as increases in endothelial cGMP levels, which is well established to serve as a biological marker for EDRF formation. Transport of D-glucose into endothelial cells was measured by 3H-D-glucose uptake. Results: Prolonged exposure (24 h) to pathophysiological amounts of Dglucose concentration-dependently amplified bradykinin-, ADP-, and ATPinduced Ca 2+ mobilization, while basal [Ca2+]i remained unchanged. This amplification was abolished in the presence of cytochalasin B, which was proved to be a suitable blocker of endothelial D-glucose uptake. L-glucose as well as D-mannitol failed to mimic the effect of high D-glucose, indicating that D-glucose-mediated amplification on agonist-induced Ca2+ mobilization represents a specific effect of D-glucose. In agreement to its effect on [Ca2+]i , preincubation with pathophysiological D-glucose concentrations amplified EDRF formation, induced by bradykinin, ADP and ATP, while basal EDRF formation remained unaffected. In contrast, high D-glucose failed to modulate EDRF formation, induced by the Ca2+ ionophore A 23 187, which is due to an unphysiological elevation of endothelial [Ca2+]i. Conclusion: We suggest, that prolonged exposure to pathophysiological D-glucose concentrations is responsible for increased EDRF formation, which in turn is strictly due to an enhanced Ca2+ response to autacoids, like bradykinin, ADP and ATP. This effect, which depends on D-glucose uptake into endothelial cells, may explain reduced vascular tone and peripheral resistance in early stages of diabetes mellitus. Trp shifts the dose-response curves for glucose-induced insulin re/ease two to three fold to the left (p<0.001), i.e. results in a sensitizing effect of isolated islets to glucose that is comparable to the insulinotropie effect of C~C7~. Especially above mentioned compounds which possess electron-drawing groups were effective. Other synthesized compounds with electron-donating groups had no effect. In contrast to CCK a the insulinotropic CCK. compounds were selective for the endocrine pancreas: they ha~ no or negl/glble agonistic or antagonistic effect on either the contraction of the guinea pig ileum or on amylase release of the exocrine pancreas or on the feeding behaviour of mice being supplied with either compound by an implantable Ahet ® pump for five days. It is concluded that by these highly selective compounds (peripheral) CCK-A receptor subtypes can be discriminated for the first time; the B-cell CCK-A receptors are different from those in smooth muscle and those for regulating appetite. The object of our investigations was the question whether a chemotherapy with Ciprofloxacin (CI) and Ofloxacin (O) had a stimulating or inhibitory effect on the unspecific cellular immun response in Balb-c mice. By means of an ex vivo model at which the investigated quinolones were administered orally in vivo, the phagocytic capacity of isolated peritoneal (PE) macrophages was examined in vitro, being based on the photometric measurement of hemoglobin after the phagocytosis of opsonised sheep red blood ceils (SRBC). Furthermore the PE-exsudates were analyzed (amount of PE-cells was determined, as well as differentiation according to percental compositien). The oral administration of the quinolones consisted of doses of a 100 or 200 mg/kg/d CI, as well as 50 or 100 mg/kg/d O, over different periods of time (1, 3 and 14 days). One control group was given a solution of 0.9% NaC1. The other group received a dose of 20 mg/kg/d p.o. of Cyclophosphamide (CY) which proved to decrease the phagocytic capacity of PE maerophages reliably and significantly~ A decrease of the phagocytic capacity of PE macrophages, depending on the administered doses, but not on the different periods of time could be shown for CI. 100 mg/kg/d CI caused only a minor decrease of the phagocytic capacity in comparison with the NaC1 control group, whereas the group receiving 200 mg/kg/d showed a distinct suppressive effect being compatible to the results received after CY administration. A decrease of phagocytic capacity also applied for the administration of O. In contrast, however, to CI, this depended on the doses as well as on the duration ef administration. In comparison to the CY group, a single administration of O showed minor decrease of phagocytic activity, however the administration of O over longer periods of time showed a stronger suppression of the phagocytic activity. In each case the decrease in phagocytic activity correlated with the administered doses. Thromboxane (TX) B2 and prostaglandin (PG) F2~ synthesis in clotting whole blood and collagen-induced platelet aggregation ex vivo were determined before as well as at various time points up to 72 h after intravenous injection of 600 mg azapropazone and, 10 or more days later, intramuscular injection of 30 mg ketorolac-trometamol in 6 healthy volunteers. For determination of TXB2 and PGF2~ synthesis 2 ml whole blood were incubated at 37°C for 1 h. After centdfugation at 2500 x g at 4°C for 15 min supernatants 'were analysed radioimmunologically. Platelet-rich plasma was obtained by centrifugation of 30 ml blood anticoagulated by sodium citrate at 150 x g at 20°C for 10 min. After dilution with platelet-poor plasma 1 ml aliquots containing 2 x 108 platelets were incubated at 37°C. Aggregation was induced by addition of 1-5 ~tg collagen /ml plasma. Azapropazone significantly inhibited TXB2 synthesis up to 6 h after injection. Maximal inhibition was obtained-5 min after injection, when TXB 2 synthesis was reduced to 38+4% of the basal value (357+87 ng/ml serum). Ketorolac-trometamol inhibited TXB 2 synthesis up to 24 h almost completely and thus to a significantly greater extent than azapropazone. Similarly, PGF2esynthesis was reduced by ketorolac-trometamol more effectively than by azapropazone indicating that both compounds act by inhibition of fatty acid cyclooxygenase. While platelet aggregation induced by 1 ~g collagen/ml was inhibited by azapropazene 1 h, but not 24 h after injection, ketorolac-trometamol caused a markedly greater inhibition of aggregation which lasted up to 24 h, even if higher concentrations of collagen were used. Our re-su~ts show that azapropazone given in a dose with an analgesic and antiphlogistic potency comparable to that of the ketorolac dose used affects platelet functions only moderately. This might be an advantage with regard to its use as post-operative analgesic. Proliferation and differentiation of T-lymphocytes are regulated by numerous extracellular factors as growth factors, hormones, antigens, interleukines, mitogens. The biochemical mechanisms that govern the responses of T-cells to these signals are now being elucidated. Increases in protein tyrosine phosphorylation have been found to be early events in the activation of resting T-cells.

In T-lymphocytes three members of the src-family of non-receptor protein tyrosine kinases (PTK). have been identified, p56 lck, p59fY n and p561Y n. It has been shown that p56 lck is associated with CDA or CDu receptors, p59fY n is associated to a subunit of the T-cell receptor CD~. The p~'otein to which p561Y n is associated is still unknown. The increase of tyrosine phosphorylation in proteins might be a consequence of the activation of one or more of those PTKs and/or of a tmnslocation of the enzyme(s) into other cell comparlments, as it has been shown for pp60 c-src in platelets. To answer these questions we have studied PTKs of the src-family in resting and activated human T-lymphocytes. We have isolated the distinct cell fractions and measured the enzyme activity using an antibody-immobilizedenzyme assay in the diffemt cell fractions and determinated the Vmax and Km for the enzyme present in these fractions. The PTK activity measured using a anti-p56 lck antibody was found in the membrane fraction from resting T-cells, but when the cells were stimulated with phytohemagglutinin (PHA), the activity was translocated into the cy~osolic fraction. This translocation was not observed when the T-cells were stimulated by an anti-CD 4 antibody. That means that the stimulation mechanisms are different for PHA and anfi-CD 4 antibodies. On the other hand, the enzyme from the membrane fraction in resting cells shows the same Vma x but two fold higher Km compared to the enzyme from the cytosolic fraction, indicating that there is no biochemical activation of the translocated enzyme after stimulation with PHA. The increase in tymsine phosphorylation of proteins during T-cells activation may be caused by a rearrangement of enzymesubstrate relationships due to enzyme translocafion. We tested the influence of the 3 substances on 5-1ipoxygenase using glycogen-elicited rat peritoneal neutrophil granulocytes, on cyclooxygenase and 12-1ipoxygenase using human platelets and the antioxidative activity using a cell-free peroxidation assay of arachidonic acid and iron/ascorbate. Leukotriene B., 5hydroxyeicosatetraenoic acid (5-HETE; 5-1ipoxygenase), ~ 12hydroxyheptadecatrienoic acid (12-HHT; cyclooxygenase), 12hydroxyeicosatetraenoic acid (12-HETE; 12-1ipoxygenase) and the arachidonic acid metabolites from peroxidation were separated by rp-HPLC and quantified by UV-detectiom Results:

• EC~0-activities bisabolole b-oxide A b-oxide B 5-1ipoxygenase ! 30pM 170pM no effect ,~yclooxygenase, !80pM no effect no effect , , 12-1tpoxygenase no effect no effect no effect ,,, antioxidation no effect :no effect ~no effect

Bisabolole inhibits 5-1ipoxygenase and cyclooxygenase without affecting 12-1ipoxygenase. Bisabololoxide A inhibited only 5lipoxygenase but not the other enzymes tested. Bisabololoxide B showed no effect in our test systems. The lag of effect on the peroxidation system up to the concentration of 2001abl shows, that the inhibitory actions of bisabolole and bisabololoxide A are not due to antioxidative properties. Different glucocorticoids (GC) applied intravenously, dermally or in vitro exert only small differences in the ability to raise RM 3/1 macrophages (Me) from human peripheral monocytes (Int Arch Allergy Appl Immunol 91:175,1990 , 94:278,1991 . The present experiments were designed to study the efficacy of oral application of GC in this model. We also intended to obtain some information about qualitative differences between the effects of prednisone and deflazacort which was reported to produce less reduction of calcium absorption and bone degradation than other GC. Both GC were administered to probands in doses regarded as equivalent with respect to general GC effects by the manufactures of deflazacort (5 to 6 mg or multiples). One single dose of 10 mg prednisone increased the number of RM 3/1 M,I, from a basic level within 12 h reaching a maximum of about 60% at 24 h, declining thereafter. 12 mg deflazacort raised the number of RM 3/1 MO much slower reaching a maximum of 30% (average) after 48 h. The interindividuel variation was found to be mainly the time lag between dosage and maximum effect. Interindividuel differences of prednisone effects concerned mainly the maximal increase of RM 3/1 M4, after 24 h. The results show that in this test system deflazacort was found to be less effective than was expected. The difference might be due to pharmacokinetic characteristics of both compounds.

Institute of Pharmacology and Toxicology, RWTH Aachen, Wendlingweg 2, 5100 Aachen, F.R.G.

O.Rickinger, R. Hirschelmann, J. GieBler I t is generally accepted that the basis of action of nonsteroidal antiinf|ammatory drugs (NSAIDs) is related to their inhibition of prostaglandin biosynthesis.This hypothesis however may not account for all of the effects of NSAIDs at all dosages.Thus, i t may be argued that other processes are contributory to their antiinflammatory action. Reactive oxygen species (ROS) are involved in the pathogenesis of inflammation. Therefore one other mode of action is presumed to be an inhibition of the "respiratory burst" of inflammation cells. In our investigation we determined the influence of some substances on hydrogen peroxide production of rat pleural macrophages in v i t r o and correlated the results to the antiinflammatory potency in rat carrageenin paw edema, We found indeed an inhibition of HgO 9 production by NSAIDs. However, this inhibition is not s t r i c t l y correlated to the antiinflammatory a c t i v i t y and the therapeutic concentration of these compounds. Taken together, this mechanism seems to be of minor relevance for the action of NSAIDs. In experiments searching for immunomodulatory substances the poorly soluble compound T44 (1-benzoyl-2(naphth-l-yl-amino)-ethene) and its soluble derivative PSI2 (p-methyl-l(perhydroacepin-l-ylmethyl)-propiophenone) showed stimulating and depressing effects in several models of cellular immunity, e. g. contact hypersensitivity reaction to picrylchloride and experimental allergic encephalomyelitis in mice. Therefore we studied the humoral immune response of inbred mice to sheep red blood cells (SRBC) under the influence of p. o. treatment with T44 and PSI2. The compounds were given once a ~ay from day -2 to +6 of antigen booster injection. The antibody response was checked by Takatsy's microhemagglutination assay and the differentiation of the antibodies (IgG/IgM) was done by 2-mercaptoethanol treatment of the sera. The main effect of T44 as well as of its derivative PSI2 was a suppression of the antibody formation. Also we found differences of the effects between male and female animals. The results were depending also on time, dosis and strain of mice. Bradykinin is involved as a regulator peptide in various physiological and pathophysiological processes. Its influence on different target cells can be due, for instance, to the capability of inducing the expression of cytokines (Vandekerckhove et al. : Lymphokine & Cytokine Res. 10 : 285, 1991; Paegelow, Werner : Meth. & Find. Exptl. Clin. Pharmacol. 8: 91, 1986) . In this study we present the evidence that during a bradykinin (BK)-and histamine -induced contraction of isolated lung strips (human, mice, guinea pigs) different cytokines are released into the bath solution. Carbachol is not able to express such a cytokine secretion . The cytokines are released in a time-dependent manner during the first 3 hours with a maximum of 60 to 90 rain and could be identified as IL-I, IL-2, IL-3 and IL-6 by means of monoclonal antibodies and enzyme immunoassays. The quantification of the cytokines released by bradykinin into the bath solution shows a relation between IL-2 and IL-6 of 1:30 to 1:50. Obviously there is no clear correlation between the druginduced contraction of the lung strips and the cytokine release: carbachol elicits a dose-dependent contraction, but can not induce a release of cytokines investigated in this study. IL-6 released after the contact with BK and histamine seems to influence the contractility of the lung strips. It can be speculated that cytokines released by bradykinin from the lung strips could be integrated into the pathophysiology of the late phase of asthma.

Institute of Human neutrophils respond to phagocytic stimulation with the NADPH oxidase-mediated generation of great amounts of microbicidal reactive oxygen species, the so-called oxidative burst. During our experiments concerning the impact of redox processes on signal transduction, we observed that phenylarsine oxide (PAO), a compound which is reported to specifically react with vicinal SH-groups is a highly potent inhibitor of this oxidative burst. The IC50 after stimulation with the chemoattractant fMLP is about 100nM. The oxidative burst after stimulation with the Ca 2+ionophore ionomycin or the G-protein activating aluminium fluoride is inhibited with the same potency, whereas significantly higher concentrations of PAO are required upon stimulation with the phorbol ester PMA. Using a cell-flee system, we could demonstrate that PAO has no direct effect on the NADPH oxidase, pointing to the activation mechanism of this enzyme complex as the functional target of PAO. The effect of PAO on the oxidative burst could be completely reversed by addition of the vicinal dithiol 2,3-dimercaptopropanol, whereas the monothiol 2-mercaptoethanol is completely ineffective in this respect.

Other cell responses of human neutrophils like degranulation and shape change are only inhibited at much higher concentrations. PAO does not influence the fMLP-triggered Ca2+-signal which excludes the phospholipase C pathway as a possible target. Phospholipase D (PLD) is inhibited with an IC50-value of 500nM. Together with other results this indicates that PLD is also not the PAO-sensitive target. Platelet activating factor (PAF) is a ~hos~mlipid derivative that is produced by leukocytes, endothelial oells and platelets in response to immune and non~ stimuli. PAF se~m~ to play an i~portant role as mediator of allergy and inflammation. Intravenous injection of PAF into mice causes shock-like sy~, although mouse platelets are insensitive to direct activatLng effects of this substance. Therefore, PAF-induced mortality in mice has been suggested as a model for in vivo testing of drugs active against plateletindependent anal~hylactic reactions. Depending on the strain, source and age of ~3imals as well as sc~e other e~perimental conditions great v~riations in the susceptibility of mice to the lethal effects of intravenous PAF have been reported. We now obsel~;ed that overnight fasting also profoundly influences PAF-induced death in mice. While intravenous injection of i00 ~g/k~ PAF in fed animals always caused a mortality rate of more than 60%, after food deprivation never more than 10% of the mice died at the same PAF dosage. Reduction of blo~d gluoose levels by the oral application of tolhutamide failed tm reduoe sensitivity to PAF in fed animals. Inhibition of synthesis of llB-hydroxylated corticosteroids by the administration of metyrapone w~s also without any effect on hunger-induced protection against PAF toxicity. However, intraperitoneal injection of the ~-adrenoreoeptor antagonist propranolol (i mg/kg) cc~@letely abolished resistence to PAF mortality in fasted animals. In acoordance with previous studies these results indicate that food deprivetion like other manipulations which cause release of catecholamines fr~ the adrenal medulla have pronouned effects on PAF toxicity. 

bronchoconstriction and vascular leakage in the isolated perfused rat lung: role of thromboxane,

Platelet activating factor (PAF) is a proinflammatory mediator that induces pulmonary hypertension, bronchoconstriction and permeability edema in vh, o. The mechanism(s) of these actions of PAF are still not completely understood. Due to the complexities in v/re, PAF was studied in the isolated perfused and ventilated rat lung (IPL). In the IPL, a bolus of PAF provoked an immediate (<2rain) rise in vascular (ECs0=0.TnMol) and bronchial resistance (ECso=O.TnMol) and an increase of the capillary filtration coefficient (EC50=2.3nMol). Thus, the actions of PAF m v/re are reproduced in the blood-free perfused rat lung. The mediation of these effects by eicosanoids was studied. Pretreatment of the IPL with the lipoxygermse inhibitors diethylcarbamazine and VZ65 (4-(11-hydroxy-l,9-undecadiin)-brenzcatechin) afforded no protection ~ain~t a bolus injection of PAF. However, pretreatment of the IPL with either the cyclooxygnnase inhibitor acetyl salicylic acid or with the thromboxane receptor antagonist BM13177 (([4-[2-(phenylsnifonyl) amino]ethyl]phenoxy acetic acid) abofished the PAF-induced vase-and bronchoconstriction, yet was ineffective against the PAF-induced increase in v a s s a r leakage. Direct assessment of thromboxane showed that PAF dose-dependently released massive amounts of thromboxane into the perfume. The maximum thromboxane concentration in the lung efuate occuned 2 rain after injection of PAF, i.e. at the nadir of vase-and bronchoconstriction. We conclude that in the IPL PAF can stimulate the synthesis ofthromboxane within 2 min. This PAF-induced thromboxane is helie~,ed to he responsible for the PAF-induced vasoconstriction and bronchoconstriction. The mechanism of the increased vascular permeability after PAF remains unknown. 

The mitogenic plant leetin Concanavalin A {Con A) is known to activate T lymphocytes in vftro. In rico, Con A caused hepatic injury in several mouse strains, i.e. NMRI, BALB/c-, or C3H/HeJ-mice, when given intravenously in a dose of 2 0 mg/kg. The extent of liver injury was assessed by determination of s e r u m t r a n s a m i n a s e s a n d by histopathologlcat evaluation. The m a x i m u m of transaminase release was detected 24 h after Con A challenge. T h d p e r lymphoeytes as well as maerophages were identified in this model a s effector ceils of organ injury [ I] . IL-2 release into the circulation demonstrated t h a t T cell activation occured. Spleen ceils isolated l h after Con A challenge released substantial a m o u n t s of IL-I, IL-2 a n d TNF into the culture supernatant. Since this s u p e r n a t a n t was cytotoxic towards Isolated mouse hepatocytes, it is concluded t h a t the Con A-induced release of various cytokines is responsible for organ injury.

[1] O Tiegs, J Hentsehel, a n d A Wendel. 1992. J C/in Invest 90: 196-203 .

Biochemical Pharmacology, Faculty of Biology, University of Konstanz, W-7750 Konstanz.

Rodents challenged with endotoxin (LPS) develop hepatic failure as part of a septic shock-like multiorgan failure. Hepatic failure can be aggravated, e.g by sensitisation with galactosamine (GAIN), to a fulminant hepatitis. Rats or mice treated i.p. with a combination of galactosamine and LPS from Salmonella abortus equi developed fulminant hepatic failure within 8 hours. In order to assess the participation of cyclooxygenase products, animals were pretreated with a single dose of aspirin, ibuprofen or the thromboxane receptor antagonist BM 13.177 ([4-[2-(phenylsulfonyl) amino]ethyl]phenoxy acetic acid). These pretreatments significantly protected rats but not mice. We have recently described a cellular model based on the cocultivation of paranchymal and nonparenchymal rat liver cells, which shares basic mechanisms with LPSinduced hepatic failure in rive. In these cells, LPS-induced cytotoxicity was significantly reduced by four different cyclooxygenase inhibitors, three inhibitors of thromboxane synthase and the thromboxane receptor antagonist. On the other hand, U46.619 (9,11-Dideoxy-lla, gaepoxymethanoprostaglandin F2ce), a stable analogue of thromboxane A2~ significantly increased LPS-induced cytotoxicity. The cells released thromboxane as measured with an EIA technique in response to an LPS challenge. TNF, a central mediator in the in vivo as well as in the in vitro system, was not suppressed by the inhibitors mentioned above in LPSstimulated Kupffer cells. It is concluded that thromboxane is a mediator of LPS-induced liver injury independent of mediation by TNF in rats unlike in mice in vivo as well as in vitro.

GSH-depletlon h a s pleiotropic effects on lymphoeytes in tgtro as well as in v/vo, Preineubation of m o u s e spleen lymphoeytes {SLC) with GSH-depletors was shown to make these cells unresponsive to Con A-Induced proliferation. The aim of this study was to examine whethe_r GSH-depletion in v/vo would affect the sensitivity oI mice towards T-cell mediated liver injury induced by Con & Freshly prepared {SLC) were incubated with the GSH-Stransferase s u b , i r a t e l-ehlor-2,4-dinitrobertzene (CDNB, ~(~M), the synthesis inhibitor buthionine sulfoximine , 200pM) or a combination of both agents to deplete GSH. After polyclonal stimulation with Con A {5 pg/ml) the cells were incubated at 37°C a n d proliferation was m e a s u r e d after 48 h. GSH-depletors were given in v/vo intraperitoneally 0.5 h (CDNB, 100 mg/kg; Diethyl maleate, 400 mg/kg), 1.5 h {Phoronc, 125 mg/kg) or 9 h (BSO, 888 mg/kg] prior to intravenious administration of Con A (20 mK/kgL Liver injury was assessed by determination of serum enzymes {ALT, SDH}. In vioo pretreatment of mice with GSH-depleting compound~ fully protected against Con A-induced hepatitis. IL-2-produetion was changed by GSH-depletion ~n vitro b u t not in v/vo. The results suV4~est t h a t the lymphocyte mediated killing m e c h a n i s m ~ hepatoeytes need~ a n intact GSH status. We believe that GSHdepletion of the effetor cell (activated T-lymphocyte) is more likely to a c c o u n t for the protection in our model t h a n GSH-depletion in the target cell (i. e. the hepatoeyte).

In this study the effects of Con A upon hepatocytes were investigated in a mouse liver cell culture system. The lectin caused a dose-dependent cytotoxicity, as assessed by increased LDH-release and less tetrazolium dye reduction. Early events in the course of the leetin-induced cell death were stril0ng morphological changes characterized by irregular protrusions of the cells. The high affanity of Con A ~or marmose was identified as a necessary condition for the induction of cell death. Thus, inhibition of binding of the lectin to the cell surface abolished cytotoxicity. Neither dimeric Con A (Suecinyl-Con A) nor other tetrameric marmose-binding lectins caused hepatocyte death in our culture system, indicating a high specificity for tetrameric Con/k In primary liver cells from immunodeficient mice (e.g. nude mice; T-cell depleted mice) as well as in liver cell lines similar Con A-inducible cell death was observed. This suggests a direct interaction of the lectin with the hepatocytes without participation of accessory cells (such as liver lymphoeytes). Glomerular de0osition of immune complexes, followed by activation of complement and generation of chemoattrac.tants, may cause the influx of T-lymphocytes and monocytes seen in many forms of human glomerulonephritis. With primary cultures of HMC we determined the capacity of these mesenchyme-derived cells to express receptors for IgG. We employed polymerase chain reaction from reverse transcribed RNA (RT-PCR) of HMC, which have been growth arrested and stimulated for 16 hrs by rh interleukin-li~, platelet-derived growth factor, interferon-3, (IFN-3,) , lipopolysaccharideE.coli (LPS) and combinations thereof. Primer pairs specific for the three types of human Fc3' receptors (CD16, CD32 and CD64) were used and the identity of the resulting RT-PCR products controled by Southern blot hybridization with independent [e32p]-Iabeled Fc3,R cDNA-pr0bes. With HMC prepared from three different donors only the combination of IFN-3' + LPS (100 U/ml + 10 pg/ml) elicited a significant expression of hFc3,RIII (CD16) RNA, while RNA of Fc-~RI and Fc3,RII was not detectable in resting HMC. The remaining stimuli failed to induce any of the three Fc3, receptor types in growth arrested HMC. Unlike untreated HMC proliferating HMC treated with IFN-7 + LPS and additionally stimulated with cryoglobulin or IgG complexes exhibited an increased level of interleukin-6 RNA. Thus, the functional expression of low affinity Fc3,RIII was confirmed, Ongoing sequence analysis of the PCR-derived cDNA will show, whether it represents the transmembrane form, Fc-/RIIIA. Since the discovery of the in vitro glutathione peroxidase (GPO)-like activity of ebselen there has been continuing interest in small molecules with similar properties. Known examples include derivatives of ebselen, 2-methylselenenylbenzoic acids and phenylselenenylacetophenones. Ebselen reacts spontaneously with cellular thiol compounds, preferably reduced glutathione (GSH). The GPO-like activity of ebselen has been shown to start with the formation of a mixed selenodisulfide from GSH. An intermediate selenol formed by reduction of the selenodisulfide with a second GSH molecule then reacts with another selenodisulfide to form a diselenide which undergoes a reversible oxidation/reduction cycle in which peroxides are reduced (Cotgreave et al., Chem.-Biol. Interactions 84 (1992) 69-76) . On the other hand, ebselen reportedly reacts directly with peroxides by reversible formation of a Se(=O) compound (Dereu et al., Bull. Soc. Chim. Belg. 96 (1987) 757-768) While this reaction is unlikely in the presence of thiol compounds, biotransformation studies have shown that no unchanged ebselen could be recovered from animals indicating that some other intermediate of the ebselen catalytic cycle might be the point of reversal. We have prepared the diselenide of ebselen along with some other aromatic diselenides which are not capable of forming an selenium-containing heterocyclus. All diselenides have been tested for GPO-like activity using the standard GSSG / glutathione reductase assay (Mtiller et al., Biochem. Pharmacol. 33 (1984) 3235-3239) . When compared to ebselen all diselenides exhibited similar GPO-like activity when incubated with tort-butyl-, cumene-and hydrogenperoxide. For example, at room temperature the 2,2'-bis(carboxyphenyl)diselenide reacted at a similar rate as ebselen itself. Our data suggest that ebselen may be a starting compound for the GPO-like catalytic cycle, but is not an essential intermediate in the catalytic reduction of peroxides. The new antiasthmatic/antlallergic compound flezelastine (I(2H)phthalazinone,4-[(4-fluorophenyl) methyl]2-Ihexahydro-i -(2-phenylethyl) -1H-azepin-4-yl]-, monohydrochlorlde) exhibits Hi-receptor antagonism, 5-1ipoxygenase inhibition, and calcium antagonistic properties. In the present study we investigated the effect of the racemic compound flezelastine {FL) and its enantiomers (+-FL and --FL) on mediator release from different immunocompetent cells, which are thought to play important roles in the development and maintenance of allergic inflammation. Mast cells isolated from the pleural and peritoneal cavities of rats sensitized against whole egg white released histamine after allergen challenge. All of the three compounds exerted comparable inhibitory effects with IC50-values (lamol/l) of 6.4, 6.3, and 6.2 for FL, +-FL, and --FL, resp. After stimulation with LPS the human monocytic cell line THP-1 produced IL-135. The production of IL-lj5 was dose-dependently decreased by the three compounds. At the highest concentration of 12,5 ]amol/l the inhibition was 38%, 400/5, and 49°/o for FL, +-FL, and --FL, resp., We conclude that flezelastine is a potent inhibitor of mediator release of mast cells, and monocytes and may therefore be effective In the management of early-and late-phase symptoms of asthmatic/allergic diseases. Furthermore, both enantiomers contribute similarly to the pharmacological effects of the racemic flezelastine.

ASTA Medica AG, Dept, of Pharmacology, Weismfiller,,str. 45, D-6000 Frankfurt/Main, FRG

The distressing symptoms of rhinitis are mainly caused by extravasation of plasma from leaky capillaries in the nasal mucosa. This extravasation causes both, rhinorrhea and the edema of the mucosa with obstruction of the airways. The primary cause of this pathological chain reaction is inflammation, either induced by allergy or by infection. There is an increasing number of mediators involved in this inflammatory process and it is of major therapeutical value to find out, which inhibitors may releave or moderate the symptoms. In our model in anaesthetized rats we used constant perfusion with phosphate buffered saline and Evans Blue as plasma marker to evaluate extravasation (as an indicator of rhinorrhsa). Challenge was induced either by histamine, by capsaicin or, in sensitized rats, by allergen. Aim of the study was to investigate under these reproducible pathological conditions the effect of azelastine, a novel, multifaeetted drug with various antiallergic properties in comparison to standard compounds. Histamine-induced hypersacration was inhibited by azelastine with an IC50 of 0.25 pmol/I (mepyramine 7.4, antazotine 1.26).

Capsaicin, the active agent of hot pepper, causes release of substance P, via stimulation of sensory C-fibres. The so provoked neurogenic inflammation caused marked rhinorrhea in rats. Azelastine was able to reduce the amount of extravasation in a dose-dependant manner. IC50 was 20.5 pmol/l. With morphine, a compound which inhibits substance P release, an IC.~ 0 of 4.1 tJmol/1 was found. Though azelastine is less effective inhibiting the capsaicin induced rhinorrhea its dose dependant effect may contribute to the therapeutic action. Dicationic amphiphilic compounds are accumulated in lysosomes. They interfere with lysosomal degradation of sulfated glycosaminoglycans (sGAG) thereby inducing cellular lesions reminiscent of inherited MPS of man. The two hypotheses proposed to explain the underlying pathomechanisms are (a) formation of non-degradable drug-sGAG complexes (Grave et al., Toxicol. Appl. Pharmacol. 114: 215-224, 1992) , and (b) secretion of lysosomal enzymes with the consequence that the cells become deprived of enzymes (Gupta et al., Hoppe-Seyler's Z. Physiol. Chem. 365: 859-866, 1984) . In the present study, which was performed on cultured bovine and human fibroblasts exposed to the immunomodulatory drug tilorone, we have found that the two dr~g effects - Thalidomide (Thd) has been reported to be capable of therapeutically inhibiting inflammatory processes, such as leprosy reaction, and to suppress the graft vs host disease subsequent to bone marrow transplantations.

We have attempted to reveal possible mechanisms of these actions of Thd on components of the immune system. Such actions of Thd might also provide a clue for similar types of changes occurring on cells in the embryo. Initially, we performed the studies in non-human primates, and pronounced effects of Thd on defined receptors on lymphocytes, especially g2 integrins, were found in marmosets (Neubert et at. Arch Tox 1993 in press ). We have now tested whether similar effects may be induced by Thd in man, 5 mg Thd/kg body wt were given orally (divided in 3 doses per day) to a volunteer and a variety of surface receptors on white blood cells were analyzed with monoclonal antibodies and flow cytometry. Although pronounced effects on g2-integrins on various white blood cells have been found, here we only report on changes in gl-integrins, known as adhesion factors in cell-extracellular matrix interactions, on lymphocytes. Pronounced changes were seen in and CD3+CDw49d[ bri]) on T-cells. Also VLA~4 receptors on B-cells were affected (CD20+CDw49df°ril).

From the results it can be concluded that Thd is the first substance known to specifically interfere with the expression of some, but certainly not all, g~-and lh-integrin receptors in man. Treatment of cultured bovine and human fibroblasts by the dieationie amphiphilic compound tilorone (2,7-bis[2-(Diethylamino)ethoxy]-9-H-fluorene-9-one) and related substances causes intraiysosomal accumulation of sulfated glycesaminoglycens (GAG) due to impaired degradation. As previously observed in bovine corneal fibroblasts, tilorone-indnced GAG-storage is, at low filurone concentrations (5~tM), characterized mainly by aceumulation of dermatan sulfate (DS) and by specific cytologieai alterations, whereas at high filorone concentrations (20ttM) the contribution of heparnn sulfate towards total GAG-storage is increased. Aim of the present study" was to examine wether the same ceneentrationdependent shift of GAG-pattern is found also (a) in human desmodontal fibreblasts exposed to tilorone, and (b) in bovine eorneal fibroblasts exposed to two other compounds differing from tilorone with respect to the central ring system. Haman fibrobtasts were presently found to react towards tilorone in a similiar manner, but less pronounced then do bovine fibroblasts. Furthermore, bovine fibroblasts were treated with the acridine derivative 3,6-bis[2-(diethyl-amino)ethoxy]acridine and the anthraquinone derivative 2,6-bis[2-(diethyl-amino)ethoxy]anthraquinonc. Marked storage of DS was found after exposure to 3/tM of either compound, whereas above 5ttM (acridine derivative) and 10ttM (anthraquinone derivative) the concentration-depandent increase of GAG-storage is mainly due to the increase of HS-aceumulafion. The concentration-dependent shift of GAG-pattern seems to be characteristic of the disturbance of lysosomal GAG-degradatiun induced by these dicationie amphiphilic compounds. As the cytological alterations and biochemical findings at high drug concentrations resemble the effects of unspecific inhibitors of lysosomal function such as NH~C1 (10raM), the mechanism of interference might be elevation of intralysosomal pH resulting in a loss of enzymes. In contrast, at low concentrations an alternative, more specific mechanism may be responsible for the GAG-accumulation.

Dept. of Pharmacology, University Kiel, Hospitalstr. 4, D-2300 Kiel Immunocytocbemical methods were used to examine intracellutar compartments of chondrocytes involved in extracellular matrix biosynthesis. Bovine articular chondrocytes were cultured within an agarose gel under conditions in which the newly secreted matrix molecules are effectively trapped and accumulated according to Aydelotte and Kuettner 1988 (Connect. Tissue Res. 18, 205-222) . The aim of the present study was to investigate the effect of human recombinant IL-lo~ on articular chondrocytes using CLSM and several fluorescent lectins for the characterization of subcellular compartments. Wheat germ agglutinin (WGA) stained the perinuclear cytoplasm and therefore recognizes carbohydrates added to the core protein of proteoglycans in the Golgi, while concanavalin A (Con A) stained vesicles throughout the rest of the cytoplasm demonstrating mannose-rich oligosaccharides added co-translationally corresponding to endoplasmic subcompartments. Our results clearly demonstrate that exposure of IL-lc~ (10 U/ml) to articular chondrocytes can lead to a morphological change to stellate-shaped cells which displayed an intense staining reaction with Con A within the cell processes suggesting an accumulation of mannose-containing oligosaccharides. Our findings are in agreement with previous studies which have shown that IL-1-treated cells may not be able to incorporate proteoglycans into the matrix partly because of a defect in the proteogtycan molecules and partly due to other mechanisms regulating proteoglycan assembly. In our experiments IL-la did not alter binding of Con A to endoplasmic subcompartments, which gives evidence that the biosynthetic steps of core protein regarding the attachment of oligosaccharides may not be affected.

Institut ftir Pharmakologie und Toxikologie, Universi~t Bonn, Reuterstr. 2, and *Botanisches Institut der Universit~t Bonn, Venusbergweg 22, D-5300 Bonn 1

By simultaneous monitoring of lung function and mediator production the contribution of TXA 2 to bradykinin-induced bronchospasm was assessed in GP isolated lung. Challenge with 0.2 nmol bradykinin (bolus) increased TXB 2 release (median 0.9 ng/min in the 2 min period before; peak 3.4 ng/min in the period 2-4 min after challenge), and reduced air flow, but the two were only poorly correlated (Spearman coefficient 0.49, N=57).

(i) Inhibition of TX synthesis with a synthetase inhibitor doserelatedly reduced bronchospasm (Jonckheere's test p<0.05).

(2) ixl04/ml washed GP thrombocytes added to the lung perfusate prevented bradykinininduced bronchospasm without influence on TXB2, PGE 2 or PGFI~.

(3) When bradykinin challenge was repeated TXB 2 production and bronchospasm were similar after both challenges. If however, between bradykinin challenges the lungs (N=6) were also challenged twice with 0.i nmole NECA (adenosine agonist) the effects on TX production and bronchospasm were opposite. Bradykinin-induced TXB 2 release increased (medians for 12 min. post challenge: ist challenge 16.2 ng; 2nd challenge 23.0 ng, Wilcoxon p<0.05), whilst bradykinininduced bronchospasm decreased (median values ~ R L (max): ist challenge 0.36, 2nd challenge 0.15 cm H20/ml/sec , p<0.05). It is therefore concluded that thromboxane release alone does not account for bronchospasm induced by bradykinin, although thromboxane maybe able to act synergistically with other mediators/mechanisms to cause bronchospasm. The following study was aimed at assessing mechanisms and mediators involved in the acute phase of antigen (Ag)-induced increase of airway permeability in passively sensitized guinea pigs (GPs). Methods: (a) Male GPs were passively sensitized with a rabbit anti-ovalbumin (OA) serum (i.v.) 24 h before Ag; (b) anaesthetized GPs were pretreated with i.v. mepyramine (-10 rain at 20 ug/kg) for prevention of immediate death after OA-challenge; (c) 5 min after i.v. loading of Evans blue dye (EB; 25 mg/kg) GPs were challenged with CA at 1 mg/kg i.v.; (d) another 5 min after CA whole body perfusion with saline at 37~C and 100 mm Hg was performed in situ for 3 min; (e) GP airways were separated into trachea, main bronchi, intrapulmonary airways and lung parenchyma and the tissues analysed individually for EB in a modification of methods described in J. Neuro. Sci. 6,41-49, 1983; if) if not otherwise indicated test compounds were administered i,v. -15 min before CA. Results: (1) In sensitized GPs CA challenge induced at 5 min a significantly increased extravasation of EB into the trachea, main bronchi and intrapulmonary airways (45.04-2.9, 56.1+5.4 and 27.t4-3.4 ng EB/mg wet weight, respectively; n=32,) vs. sham sensitized, sham challenged GPs (10.4+1.1, 13.7+1.7 and 11.6+1.2 ng EB/mg wet weight, respectively; n =27, mean4-S.E.M., p<0.001).

(2) The % inhibition (% inh.) of EB extravasation into the trachea, main bronchi and intrapulmonar~ airways, respectively, for test compounds was determined as follows: (a) additional mepyramine at 1 mg/kg -10 min, 58.1, 61.1, 38.9 % inh., n=9; 52.8, 73.9, 77.1% inh., n=10; 63.7, 72.8, 77.4 % inh., n = 11 ; 61.8, 88.1, 107.0 % inh., n=9; (d) MP + fenoterol as above, 62.2, 98.5, 82.6 % inh., n=10; (e) Pal-antagonist bepafant (WEB 2170) at 1 mg/kg, 24.1,31.1, 40.7 % inh., n = 6; if) peptido-LT-antagonist IC1198.615 at 3 mg/kg, 11.9, 36.9, 55.6 % inh., n=10; (g) NK-antagonist FK 888 at 1 mg/kg, -10.7, 15.1,31.8 % inh., n = 10. In conclusion results suggest a relevant reduction of vascular permeability to EB by antihistamines, 8-adrenoceptor-agonists and glucocorticoids during the acute phase of GP-anaphylaxis. In contrast Paf, peptido-LTs and neurokinins do not seem to play an important role at this time point. The putative involvement of Paf, peptido-LTs and NK in a late-phase associated edema remains to be investigated. In order to investigate the mechanism and source of calcium (Ca) entry in histamine-(H) and KCMnduced tracheal smooth muscle contraction, we studied the effects of extracellular Ca depletion, of SK&F 96365 (a novel inhibitor of receptor-mediated calcium influx) and of nifedipine iN) on Hand KCl-induced tracheal smooth muscle contraction. The diameter change of epithelium-denuded isolated perfused guinea pig tracheal (GPT) tubes in response to H and KC1 were studied under the following conditions: 1. in Ca-containing buffer (1.7 raM; control), 2. Ca-free buffer intra-and extraluminally, 3. Ca-free buffer with EGTA 0.1 mM intra-and extraluminally, 4. Ca-containing buffer extraluminally and Ca-free buffer with EGTA 0.1 mM intraluminally, 5. Ca-containing buffer with SK&F 96365 (1-10 p.M) or N (10 gM) intraluminally or intra-and extralurninally. The contractile responses of GPT were expressed as percentage of methacholine (0.1 mM) induced reference contractions. Both H-and KC1induced contractions were inhibited by 50%-60% in Ca-free buffer without EGTA and 70%-90% in Ca-free buffer with EGTA. When Ca was present extraluminally, but not intraluminal[¥, the H-, but not the KCl-induced contraction was inhibited in=6, p<0.05). SK&F 96365 (10p.M) intraluminallv caused a greater rightward shift of the dose-response curve to H in=7, p<0.01) than N (10gM) in=9, p<0.05), but no inhibition of the maximum contraction. When SK&F 96365 (10t.tM) was present intra-and extraluminally, the maximum contraction to H was inhibited by 55%. In contrast, when N (10p.M) was administered in the same way, the maximal response to H was not affected. The spasmogenic activity of KC1 was almost completely abolished by both SK&F 96365 and N. We conclude that 1. extracellular Ca is crucial in H-and KCl-induced contractions of guinea pig airway smooth muscle, 2. H-induced Ca influx is at least partially mediated by receptor-operated calcium channels, 3. SK&F 96365 acts at least partially inhibitory on this receptor-mediated Ca channels. Ha 679 BR, a~ anticholinergic, with the chemical na~e: [7(S)-(i~, 28, 48, 5e, 78) ] -7-[ (hydroxydi-(2-thienyl) acetyl) oxy] -9,9dimethyl-3-oxa-9-azoniatricyclo [3.3.1.0~2, 4 ] nonane brcmLide, was synthetized by R. Banholzer, Boehringer Ingelheim. The pharmacological profile was evaluated in in vitro and in vivo experiments. In isolated tracheal smooth ~m/scle preparations of guinea pigs a concentration-dependent inhibition of methacholine induced contractions could be demonstrated. Schild analysis proved that Ba 679 BR is a highly potent anticholinergic with a slope deviating significantly frc~ -I. In vivo the bronchospasmolytic potency and duration of action was tested against the spasm~gen acetylcholine. In anaesthetized rabbits a dose dependent bronchospasmolytic effect was obtained. Ba 679 BR i.v. was 3 times more potent and markedly longer acting than ipratropium bromide (IPRA). The bronchospasmolytic and protective effects of Ha 679 BR and IPRA following inhalative administration of neh~/ized aqueous solutions was demonstrated in anaesthetized (plethysmographic method) and conscious guinea pigs (Kallos-Pagel). The bro~chodilating EC50-value was 0.13 % (Ha 679 BR) and 0.09 % (IPRA). The bronchoprotective effect in conscious animals lasted > 24 ho~rs (Ha 679 BR) and 4 -6 hours (IPRA). In the anaesthetized dog both oc~ had a concentrationdependent bronc~lytic effect afte~ inhalation with ECs0-values of 0.01% and 0.04 %, respectively. These pharmacological results characterize Ha 679 BR as an effective agent which completely inhibits acetylcholine induced bronchospasm with a very long duration of action. Inhaled Ha 679 BR proved to be free of cardiovascular side effects even when administered in dosages exceeding 100-fold the therapeutic ones. (650) 690 (900) The i.t. instilled bronchodilators act on both central and peripheral airways, but mainly centrally. Administered as powder, they act mainly in peripheral airways and at lower doses compared to i.t.solution. Administered as an aerosol with mass median diameter of ~3 um (ultrasonic) the bronchodilators act on both regions of the airways at low doses. But the lowest doses were needed when administering aerosols with mass median diameters of ~I pm (air-jet), which act mainly in the peripheral airways. Aerosols produced by air-jet nebulizers seem, therefore, to be the most effective method to administer drugs topically into the lung. Anticholinergics like ipratropiumbromide are used to treat chronic obstructive pulmonary disease successfully. Ipratropiumbromide shows a good clinical profile but has to be applied 3-6 times/ day. We studied the dissociation kinetics of the new antimuscarinic compound Ba 679 BR, [7(S)-(l~,-28,4B,5~,7~)]-7-[(hydroxydi-(2)-thienyl)acetyl)oxy]-9,9-dimethyl-3-oxa-9-azoniatricyclo-[3.3.1.02,4]nonane bromide, which showed in different pharmacological experiments a very long duration of action. Tritium Ba 679 BR was compared with tritiated ipratropiumbromide. Membranes from Chinese hamster ovary cells expressing the human muscarinic receptors Hm2 and Hm3 were used for time course experiments at 230C. After an association time period of two hours, dissociation was initiated by the addition of I0 BM atropine. From the half logarithmic plot the dissociation half lives were calculated.

Dissociation half lives (mean~_SD) 110:1189,1982) in rat Sertoli cells and an up to lOfold rise of activity of this isoenzyme was induced by the cAMP enhancing gonadotropin FSH. In order to study whether this regulation holds also for human tissue we chose a cell line derived from human keratinocytes. PDE isoenzyme content of HaCaT cells (provided by Dr.N.E.Fusenig, DKFZ,Heidelberg) was investigated by use of (1)selective inhibitors and by (2)FPLC chromatography. HaCaT cells contain predominantly PDE IV (17.4±0.8 pmol/mg x min) and small amounts of PDE III and V (3.2±1.2; 1.3±0.7 pmol/mg x min). Neither PDE I nor PDE II activity could be detected. Due to this isoenzyme pattern HaCaT cells offer a useful model to study the regulation of PDE IV activity. Incubation of HaCaT cells for 6 hours with either I~M salbutamol, IO0~M forskolin or 500~M dbcAMP resulted in an 1.5 -2 fold enhancement of PDE IV activity. The augmentation of PDE IV activity by salbutamol was synergistically amplified when the cells were coincubated with PDE IV activity inhibiting agents (IO~M rolipram, IO~M zardaverine; ImM theophylline). No change of PDE IV activity could be observed, when the cells were incubated with rolipram, zardaverine, theophylline alone. The increase in PDE IV activity could be mimicked by the proteinkinase A activator Sp-5,6-DCI-cBIMPS** (500~M), whereas preincubation with the proteinkinase A inhibitor Rp-8-Br-cAMPS** (500pM) reversed the increase of PDE IV activity elicited by salbutamol and rolipram by about 30~. IOpM cycloheximide diminished the increase of PDE IV activity after incubation with rolipram and salbutamol by about 50%. We conclude, that PDE IV activity in HaCaT cells can be enhanced by cAMP elevating agents. Proteinkinase A activation and de novo protein synthesis are pivotal for this induction. Concentrations of the intracellular messengers cAMP and cGMP are regulated by rate of synthesis and breakdown. Whereas control of the respective synthesizing cyclases has been intensively studied, only recently the phosphodiesterases (PDE) which catalyse hydrolysis of cAMP or cGMP have been focussed on. Five families of PDE isoenzymes with different properties regarding (a)substrate specificity, (b)allosteric regulation, (c)subcellular distribution and (d)sensitivity to selective inhibitors have been identified and denominated PDE I-V. The experience that selective PDE inhibitors may be targeted towards specific organs (e.g. positive inotropic PDE I I I inhibitors) demonstrated the tissue specific distribution of PDE isoenzymes. Several approaches to quantitatively analyse tissue specific isoenzyme patterns were hampered by It has been shown in clinical studies that the anticholinergic agent pirenzepine (P) is able to inhibit growth hormone (GH) secretion after various stimuli. Although P is an Ml-selective compound, the doses applied in these studies were rather high and therefore no conclusion can be drawn concerning the subtype of muscarinic recep'tor involved in the inhibition of GH secretion. It was the aim of the present study to characterize the receptor subtype involved, by investigation of the inhibition of GH secretion in conscious dogs by using the following muscarinic antagonists with different selectivity profiles: P (high M1 In conclusion, no e v i d e n c e was found for a l i n k a g e of M a u t o r e c e p t o r s to a d e n y l a t e c y c l a s e or g u a n y l a t e cyclase. The data suggest, however, a c o u p l i n g of M a u t o r e c e p t o r s t h r o u g h P T X / N E M -s e n s it i r e G T P -b i n d i n g p r o t e i n s to K + c h a n n e l s s e n s i t i v e to TEA, 3 , 4 -D A P and a-DTX.

Inst. Pharmacol. Toxicol., U n i v e r s i t y of Freiburg, H e r m a n n -H e r d e r -S t r . 5, W -7 8 0 0 Freiburg, G e r m a n y 

In the present experiments the isolated trachea of rats with intact extrinsic vagal nerves was used. After labelling of the 3 ~ r trachea with [ H]choline a subsequent pregangl'.onic ne ve stimulation (four 20 s periods at 15 Hz) causes the release of newly-synthesized [3H]ACh, that can be estimated from the increase in tritium efflux. Isoprenaline (IS0; 10-100 nmol/l) inhibited evoked [3H]ACh release in epithelium containing tracheae maximally by about 50%. A high concentration of I 9mol/l IS0 was ineffective, thus a bell-shaped concentration response curve was obtained. Propranolol and the selective B Iantagonist CGP 20712A (100 nmol/l) prevented the inhibitory effect of IS0. S2-Receptor selective agonists (fenoterol, formoterol and salbutamol) did not affect evoked [3H]ACh release. The inhibitory effect of IS0 could be prevented also by indomethacin (3 ~mol/l). Moreover, in epithelium-denuded tracheae ISO did not modify evoked [3H]ACh release. In separate experiments isolated rat tracheae were opened longitudinally and placed horizontally into an organ bath to record the tonic contraction of the tracheal muscle exposed to 36 mmol/l potassium chloride. IS0 reduced the smooth muscle tone concentration-dependently, the maximal effect was a 50% relaxation. It is proposed that in the rat airways B1-receptors localized most likely on the airway e~.ithelium inhibit the evoked release of newly-synthesized [ H]acetylcholine by the generation of prostanoids. [Wessler Iet al (1990) this journal 342:387]. In the present experiments it was tested, whether the release of endogenous, stored ACh is modified correspondingly; evoked ACh release was measured also in human bronchi. ACh was separated from the medium or the homogenized tissue (supernatant) by ion-pair extraction and measured by HPLC with fluorometric detection. In guinea-pig tracheae mechanical removal of the epithelium affected neither tissue content (930-1700 pmol/g) nor evoked release of ACh elicited by nerve (30 pmol/g) or transmural (280 pmol/g) stimulation. In rat tracheae neither tissue content (3300-4100 pmol/g) nor evoked release of ACh (180 pmol/g) elicited by nerve stimulation differed in the presence or absence of the epithelium. However, with transmural stimulation tissue content (11300 vs 4300 pmol/g) and evoked ACh release (3500 vs 1600 pmol/g) was reduced in epithelium-denuded tracheae. Human bronchi contained 2600 pmol/g ACh; after incubation with [3Hlcholine a specific radioactivity of I pmol/~00dpm was found for ACh. Electrical field stimulation (four 20 s periods at 15 Hz) of human bronchi caused an enhanced tritium efflux that was blocked by tetrodotoxin or by the removal of calcium. 0xotremorine inhibited evoked tritium efflux in a concentration-related manner; pirenzepine, methoctramine and p-fluoro-hexahydrosiladiphenidol (each I ~mol/l) antagonized the inhibitory effect of oxotremorine. It is concluded that [3H]ACh represents a very small and highly active transmitter pool. Release of newly-synthesized ACh in human airways is under the control of inhibitory presynaptic muscarine receptors. 805, 1990 ) and immunoprecipitation assays (Dfrje et al., Mol. Pharmacol. 40: 459, 1991) . However, their function and cellular location remain to be determined. The results of functional experiments, studying the contractile response to methacholine in rabNt peripheral lung strips (PLS) clearly indicated the presence of postjunctional M3 receptors (Vockert et al., Life Sci., in press) . The aim of the present study was to characterize the prejunctional mAChRs on noradrenergic nerves by investigating the mnscarinic modulation of noradrenaline (NA) overflow evoked b~Yoelectrical field stimulation (EFS). Rabbit PLS were incubated at 33 C in modified Krebs buffer (1.8 mM Ca 2÷) containing ~M): (-)-cocaine (10), corticosterone (10), rauwolscine (1), L-tyrosine (1), indomethacin (3) and ascorbic acid (57). The NA overflow evoked by EFS was determined by HPLC-ED-analysis. EFS (3 Hz, i ms, for 3 rain; S1 -$6) was applied every 30 rain. Sl (reference stimulation) increased NA overflow 23.39 (_+4.02)-fold above basal level (n=5). The selective muscarinic agonist arecaldine propargyl ester (APE, 0.01 -10,uM) suppressed the evoked overflow of NA in a dose-dependent and reversible manner with a maximum inhibition of 85.35 (e 4.39)% and an ECs0 of 0.23 (-+ 0.07)/~M (n =6). The antagonists atropine (0.03/~M, n = 3, pA~= 8.60 -+ 0.03), pirenzepine (0.3/~M, n = 3, p.A z = 7.15 _+ ~.36) and himbacine (0.3/zM, n = 3, pA,= 7.97 -+ 0.12), when present a/one, failed to affect the evoked overflow of NA but shifted the APE concentration-inhibition curve to the right whithout suppressing the maximum. These results demonstrate that prejunctional muscarinic heteroreceptors inhibit the release of NA from noradrenergic nerves in rabbit PLS. The pA~ values obtained for atropine, pirenzepine and himbacine favour these prejunctional mAChRs to be of the M2 or M4 subtype. There is evidence that indirect, indomethacin-sensitive mechanisms are involved in the muscarinic inhibition of noradrenaline release from the rat trachea (Brunn et al. this journal, 344:R74, 1991) . Moreover, acetylcholine, via muscarine receptors, can induce the release of 3H-arachidonic acid (metabolites) from isolated rat tracheae (Brunn et al. this journal, abstract in press). In the present study, effects of acetylcholine on the outflow of several prostanoids from the isolated rat trachea was studied. Isolated rat tracheae were incubated in Krebs-HEPES medium which was changed every 10 min. The outflow of 6-keto-PGF1, x, TXBa and PGE~ was determined by specific RIAs. In some experiments the'epitheliumwas removed mechanically prior to the start of the incubation. The spontaneous outflow of 6-keto-PGFl,,, TXB 2 and PGE 9 (determined 80 min after the onset of incubation) an~ffunted to 3400+3~0 (20), 259+ 27 (17) and 4574-60 (6) pg/10 min, respectively (means 4-SEM of (n) experiments. In control experiments, the outflow of TXB~ and PGE2 remained constant that of 6-keto-PGFl~ slightly declined with time. Acetylcholine (300/~M, a maximally effe6five concentration) increased the outflow of 6-keto-PGF-, TXB~ and PGE.. by about 500, 550 and 250 %, respectively. This inc~se wa~ antagonized completely by 10/~M p-fluorohexa-hydrusiladifenidol, but only slightly reduced by 10 /~M methoctramine. After mechanical removal of the epithelium the spontaneous outflow ofTXB 2 was reduced by 50 % and that ofPGE 9 by 30 %, whereas that of 6-keto-PGFx~ was not significantly affected. Moreover, removal of the epithelium almost completely abolished the effect of acetylcholine on the outflow of TXB.~ and partially reduced that on the outflow of PGE2, but did not affect th~ acetylcholine induced stimulation of the outflow of 6-keto-PGF. OT concentrationdependently inhibited the evoked release (-log EC50 8.0). Pirenzepine (0.1, I ~M), HHSiD, dicyclomine (DIC) (both 0.1 -3 NM) and AF-DX 116 (10 ~M) caused parallel shifts of the concentration response curves for OT without affecting the maxima. The slopes in the Schild plots for pirenzepine (0.98±0.12), HHSID (0.91±0.08) and DIC (1.09±0.13) did not differ from unity. Pirenzepine (pA 2 7.3), HHSiD (pA 2 7.3) and DIC (pA 2 7.0) had higher affinities to presynaptic autoreceptors in the trachea than to M2 receptors of guinea-pig atria (pA 2 for both DIC and HHSiD 6.3, for pirenzepine 6.8). This suggests that the presynaptic autoreceptor does not belong to the M2 subtype but is similar to the M4 sites found in rat striatum (McKinney et al., J Pharmacol Exp Ther 250: 565, 1989 ) and in rabbit lung (Lazareno et al., Mol Pharmacol 38: 805, 1990 In order to assess the structural requirements (including stereochemlcal aspects) for antimuscarinic potency and subtype selectivity of (R)-and (S)-phenglutarimide (R t = phenyl; R z = H; NR~ = diethylamino) a series of chiral analogues were investigated for their functional antagonistic properties (pA 2 values) at muscarinic M1 (rabbit vas deferens; !RVD), M2 (guinea-plg atria; GPA) and M3 receptors (guinea-pig ileum; GPI). In addition, affinities (pK. values) were determined in compe-3 t tition studies using [ Hi-N-methyl scopolamine binding to homogenates of NB-OK 1 ceils (M1) and rat heart (RH; M2), pancreas (RP; M3) and striatum (RS; M4). All compounds acted as pure competitive antagonists at the four receptor subtypes. The affinity profiles were controlled by the following structural parameters: the structure of 1 2 the ring systems "R ", the nature of "R " (H or CH ) and the structure of the amino group. Extremely high eudismic ratios (up to 16000-fo~) were observed for (R)-and (S)phenglutarimide. However, this stereoselectivity was not the same for all subtypes. (~2AR) is desensitized by a variety of mechanisms following activation by its ligand. Among these mechnisms there are some which desensitize receptor function within seconds to minutes. A second set of mechanisms causes receptor downregulation, a slow reduction of the total number of receptors. This downregulation correlates with reduction of the 132AR mRNA, which appears to be due to enhanced degradation. The receptor mRNA degradation increases, if the receptor is activated or if the cAMP level in the cell is increased by other mechanisms. In order to define the part of the I~2AR mRNA which is responsible for the cAMP sensitive degradation chinese hamster ovary cells were transfected with two [$2AR cDNA constructs: one expressing the full length receptor mRNA, including the 500 bp 3'untranslated region (3'UTR) and another one lacking the 3'UTR. Receptor mRNA half-life was determined after transcription blockade with acfinomycin D in the presence or absence of forskolin to elevate cAMP. Only the full length construct showed cAMP-dependent increased ~A R mRNA degradation. This shows that the 3'UTR is involved in cAMP-induced mRNA degradation.To further characterise the cAMP regulatory stretch, a set of receptor cDNA constructs were created by progressive truncation of the mRNA. We established an in vitro system, where the mRNAs could be compared in their degradation pattern. Cytosolic preparations of control and forskolin treated CHO cells were used to degrade the various mRNAs. These experiments define a stretch in the mRNA of the receptor that confers cAMP-sensitivity. Our findings suggest that down-regulation of ~2-adrenergic receptors occurs via specific sequences mediating cAMPdependent receptor mRNA degradation. Duo3 is a bis-dichlorobenzylether of a bispyridinium oxime. Duo3 stabilizes antagonist binding to M2-raceptors with rather high potency. In order to find structural elements essential for this allosteric action, derivatives of Duo3 were synthesized in which one end of the molecule was systematically shortened. To determine the stabilizing potency of the compounds, their effect on the rate of [3H]N-methylscopolamine ([3H]NMS)-dissociation was measured in pig cardiac membranes (3mM MgHPO4, 50mM Tds, pH 7.4, 37oc). As a measure of the allosteric potency served the concentration at which the rate of [3H]NMS-dissociation (tl/2 control ~ 2 min) was reduced by 50% (EC50, indicated together with the structural formulae). The B-adrenergic receptor kinase (BARK) is involved in mediating rapid homologous (agonist-specific) desensitization of the B2-adrenergic receptor (82AR). The kinase specifically phoshorylates only the agonistoccupied form of the B2AR. In analogy to the visual system, where lightbleached rhodopsin is phosphorylated by rhodopsin ldnase, BARK is also capable of phosphorylating rhodopsin. BARK has originally been purified in small quantities from bovine cerebral cortex. For expression of the kinase in insect cells the cDNA of bovine BARK was cloned into the baculovirus expression vector pVL1393. Cotransfection of that vector together with linear AcMNPV virus (baculovirus) DNA into Spodoptera frugiperda calls (sfg-cells) yielded recombinant viruses containing the cDNA of the kinase. Maximal expression of the kinase, which was up to 5% of the total cytosolic protein, was obtained three days after infection with recombinant viruses. The cytosol of lysed insect cells was applied onto a DEAE-Sephacel column in presence of 30 mM NaC1. The passthrough fraction was then applied onto a CM Fractogel cation exchanger column and eluted in steps from 20 mM to 120 mM NaC1. The kinase eluting at 80 mM NaC1 as a single band was estimated by Coomassie-blue staining to be 95% pure. Several mg of pure kinase could be produced from 200 ml Sf9-cell culture. Activity was evaluated by phosphorylation of light-activated rhodopsin. Agonist-occupied reconstituted B2AR was phosphorylated by purified BARK in the absence of G-proteins, but this phosphorylation was markedly stimulated by addition of I~Y G-protein subunits. The values Vmax and K m of BARK were determined for B2AR or rhodopsin as substrates, and for the cofactor ATP. The obtained values were comparable to those described for BARK purified from bovine cerebral cortex, indicating the functionality of the reconstituted kinase. The corticosterone-sensitive extraneuronal transport mechanism for noradrenaline (uptake2) removes the transmitter from the extracellular space by active transport into adjacent non-neuronal cells. Recently, we introduced an experimental model for uptake 2 (Caki-1 cells) which is based on tissue culture techniques (SchOmig and Sch6nfeld, 1990, Naunyn-Schmiedeberg's Arch Pharmacol 341:40,1-410). Clonal Caki-1 cells stem from a human renal cell carcinoma. Experiments on these cells disclosed some disadvantages of tritiated noradrenaline for the investigation of uptake 2. Although uptake~ is an important mechanism for the inactivation of catecholamines in intact organs, the transport rate of noradrenaline per single cell is rather low and intracellular noradrenaline is subject to rapid metabolism. Thus, we looked for an alternative substrate that is more suitable than noradrenaline itself.

Interestingly enough, the neurotoxin MPP* was found to be a good substrate of uptak%. Initial rates of specific ~H-MPP* transport into Cald-1 cells were saturable, the K~, being 24 (95% confidence interval: 11, 50) ~tmol/l and the V~, x being 420~50 pmol/(mg protein rain) (n=4). The rate constant of specific inward transport was 34 times higher than that of 3H-noradrenaline. The ratio specific over non-specific transport was considerably higher for ~H-MPP + (12.6) than for 3H-noradrenaline (3.0). 3H-MPP÷ transport into Caki-1 cells was inhibited by various inhibitors as well as substrates of uptake z. The highly significant positive correlation (p<0.001, r=0.986, n=7) between the IC~0's for the inhibition of the transport of ~H-noradrenaline and 3H-MPP*, respectively, proves the hypothesis that MPP ÷ enters Caki-1 cells via nptak% Trifiated MPP ~ is a new and convenient tool for the investigation of uptakez. MPP ÷ can be expected to greatly facilitate the characterization of uptake~ in tissues and tissue cultures that possess fewer carriers than Caki-1 cells. The question whether uptake~ is involved in the mechanism of MPP ÷ neurotoxicity is subject to further investigation. Various substances irthibit the carrier-mediated uptake of noradrenaline (NA) into isolated bovine adrenal ehromaffin granules (CG) with high affinity as shown by their IC 5o values between 10-~ and 10 -s tool/1. The inhibiturs belong to different groups of drugs: alpha-and beta-adrenoeeptor antagonists, calcium eharmel blockers, monoamine oxidase inhibitor~, 5hydroxytryptamine antagonists, opioids and dipyridamole. From these inhibitors in vitro, antihypertenstve acting drugs were administered i.v. to rabbits to examine whether in plasma the eoneentratinn of 3,4-dihydroxyphenylglycol (DOPEG) increases transiently due to inhibition of uptake into nerve granules. Reuptake of amines into neurones was blocked by desipramine (2 mg/kg t.v. 30 min prior to and 0.5 mg/kg 30, 90 and 150 rain after drug administration) to prevent the irdluanee of sympathetic activity on the formation of DOPEG. In anaesthetized rabbits (fluanisone/fentanyl 250 I~g/5 pg per kg i.v. 45 rain prior to drug administration) the central ear artery was carmulated to record heart rate and blood pressure and to allow blood sampling (2 ml at 1 min prior to and 14, 30, 60, 120 and 180 rain after drug administration) for the determination of DOPEG, 3,4-dflaydroxyphenylaeetie acid (DOPAC), NA and 3,4-dLhydroxyphenylalanine (DOPA) by HPLC with electrochemical detection. The i.v. administration of 150 l~g/kg reserpine (NA uptake into CG: IC 5o = 3 nmolZl) caused an increase of DOPEG from = 0.5 (controls) to 6 ng/ml plasma within 14 rain, followed by a decrease with t½ = 45 -60 rain. Doxazosin, prazosin, labetalol (1 mg/kg, each), ketanserin and verapamil (3 mg/kg, both) caused a similar time course of increase of DOPEG in plasma with peak concentrations (14 mha) of 1 -1.5 ng/ml. Surprisingly, verapamll (1 and 3 mg/kg) and dlpytidamule (1 mg/kg) evoked rather an increase of DOPAC (controls: 2 -3 ng/ml) than of DOPEG with a similar time course. Under the applied experimental conditions only the~e two drugs produced a short-time decrease of blood pressure. Conclusion: Various antlhypertensives eanse an increase of MAO products in plasma due to trdatbitton of amine uptake into subeellular stores. These effects are evoked by drugs which inflaibit the uptake of NA into isolated CG with IC ~o values _< 10 -6 tool/1. The extraneuronal transporter for noradrenaline (uptake2) has been demonstrated to exist in various sympathetically innervated peripheral tissues. However, the question is still open whether uptake2 exists in the CNS. The recent finding that uptake2 avidly transports 1-methyl-4-phenylpyridiniurn (MPP +) opened a new possibility in the search for uptake2 in the CNS.

Rat cerebral cortex slices were incubated for 15 rnin at 37°C with 100 nmol/1 3H-MPP+ in the presence of cocaine (10 gmol/1). The known inhibitors of uptake2, O-methylisoprenaline ( 100 gmol/1) and l-methyl-1 "isopropyl-2,4"-cyanine (1 gmogl), inhibited 3H-MPP+ uptake downto 68 _+ 11% and 32 __-8%, respectively (n=4). Furthermore, various primary cultured human glioma cells were incubated for 10 min at 37°C with 100 nmogl 3H-MPP+. All tested cell lines possessed 3H-MPP+ uptake which was sensitive to 1-methyl-1 "-isopropyl-2,4"-cyanine. One of these cell lines, namely the HTZ146 cell line, was investigated in more detail. Saturation of the 3H-MPP+ uptake in HTZ146 cells was measured. The Km amounted to 23 (95% confidence limits: 19, 28; n=`1) gmol/1 which perfectly fits the known Krn of 3H-MPP + transport by uptake2. The Vmax amounted to 116 _+ 8 pmol/(mg protein, rain) (n=4). Moreover, the effect of various substrates and inhibitors of uptake2 on 3H-MPP+ transport into HTZ146 cells was determined. There was a highly significant positive correlation between the IC50"s for the inhibition of 3H-MPP + transport into HTZ146 cells and the IC50"s for the inhibition of uptake2 (r=0.990, n=7, p<0.001).

These findings suggest that uptake2 indeed exists in glia cells of the human CNS. Uptake2 may well play a role in the inactivation of centrally released catecholamines. Mouse striatal sli~es preincubated with ~l~-do~nmi~ 25 r~ol/l were superfused with medium containi~ nc~ifensine i0 )~m~.!/1 and the effect of histamine on the electrically or 2~,~l,~a -' Ca -~v~ xritium overflow was studied. The electrically (3 Ilz) evoked tritit~ overflow was inhibited by histamine i0 )~oi/i by 18 %. A similar degree of inhibition (i.e. by 15-20 %) was obtained when (i) the concentration of 3H-dopamine used for preincubation was increased to I00 rmol/l or (2) the stimulation frequency was reduced to 0.3 Hz or (3) atropine was added to the superfusion medium, whereas addition of haloperidol ~ t/n~ effect of histalltine to 38 %. The effect of histamine, which was mimicR~d by the H 3 agonist R-(-)-~-methylhistamine, was abolished by the H 3 antagonist thioperamide 1 oI/i but not affected by the H 1 antagonist dimetindene the H 2 antagonist ranitidine, The Ca2+-induced tritium overflow (evoked ~ introducillg Ca2+ ions (1.3 l~mol/l) into Ca2+-free, Kr-rich (12 m~ol/l) medix~a containing tetrodotoxin) was not affected by histamine I0 jumol/l in the absence, but inhibited by 30 % in the presence, of haloperidol. The inhibitory effect of histamine was abolished by thioperamide 1 ~mlol/l. In another set of experiments, mouse brain oortex slices Ioere preincubated with ~ll-noradren~line 25 rmol/l and superfused with medium containir~ desipramine and rauwolscine. ~ne concentrationresponse cLxrve of histamine for its ir2%iq0itory effect on tb~ electrically (0.3 Hz) evoked tritium overflow (which was shifted to the right by thioperamide 0.1~umol/l) was not affected by nomifensine 32 )umol/l . The present resa/Its suggest that the dopamine~gic nerve terminals in the mouse striatum are endowed with inhibitory presynaptic H 3 receptors. The relatively small extent of inhibition (a) can be ~ by simLlltaneotls blockade of the dopamine autoreceptors and (b) Recently we found that, under pentobarbital anaesthesia, increases in blood pressure by hypervolaemia or noradrenaline (NA) infused intravenously diminish the release rate of endogenous NA in the locus coeruleus (LC); a fall of blood pressure elicited by hypovolaemia exerts the opposite effect (Singewald et al., Naunyn-Schmiedeberg's Arch Pharmacol, 1992;  in press). The effects of carotid occlusion, phenylephrine and chlorisondamine on catecholamine release in the LC, under pentobarbital or chloralose anaesthesia, were now investigated. In cats anaesthetized with sodium pentobarbital a push-pull cannula was stereotaxically inserted into the LC which was supeffused with artificial cerebrospinal fluid. The release of NA and dopamine iDA) was determined radioenzymatically in the superfusate. A bilateral carotid occlusion led to an immediate increase in the release rate of NA in the LC. The elevated NA release seemed to be due to the fall of blood pressure in the carotid sinus during occlusion, because hypovolaemia also enhances the NA release in the LC (see above). Intravenous infusion of phenylephrine led to a rise in blood pressure and bradycardia. These cardiovascular effects decreased the NA release in the LC. Under chloralose anaesthesia, the reflex bradycardia after infusion of phenylephdne was more pronounced, whereas the changes in blood pressure and release of NA in the LC were similar to anaesthesia with pentobarbital. The release rate of DA was not influenced substantially. Under chloralose anaesthesia, intravenous injection of chlodsondamine led to a hypotension but the release of NA was not influenced. The chlodsondamineinduced hypotension decreased the release rate of DA in the LC. It is concluded that noradrenergic neurons, and possibly dopaminergic nerve terminals, of the LC contribute to barorecaptor reflex. The responses of LC noradrenergic neurons to impulses from peripheral baroreceptors are similar under pentobarbital and chloralose anaesthesia. In rat iris adenosine (ADO) and ATP inhibited the evoked [3H]noradrenaline overflow (NSAP 345:417-423, 1992) . ADO, but not ATP, acted via prejunctional A~ purinoceptors. To find out which purinoceptor mediated the ATP effects, we superfused rat isolated iris (preloaded with [3H]-noradrenaline, ~H-NA) with Tyrode solution (containing cocaine, corticosteron, phentolamine, and as an A 1 antagonist, DPCPX, 1,3-dipropyl-8-cyclopentylxanthine), and determined the overflow of 3H-NA evoked by field stimulation (3 Hz, 2 min). The effect of P2 selective antagonists on the ATP induced inhibition of evoked 3H-NA overflow was considered to reflect the type of prejunctional P2 purinoceptor involved. ATP inhibited the evoked 3H-NA overflow in a concentration dependent manner by up to 80 _+ 2 % (100 gM, n=4) compared to controls. The ATP concentration response curve was unaffected in the presence of the relatively P2x selective antagonist suramin 100 ~tM and the relatively P2z selective 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid 50 gM (DIDS). The relatively P2v selective antagonist cibacron blue 3GA 30 and 100 FM shifted the ATP concentration response curve to the right by a factor of about 3 and 6, respectively, leading to an approximate -log K s of 4.7, In separate experiments (no DPCPX present) cibacron blue 3GA 30 gM failed to affect the inhibition of evoked 3H-NA overflow induced by ADO, indicating an antagonism selective for the ATP effects. Cibacron blue 3GA, but not suramin or DIDS, enhanced by about 40% the evoked overflow when present alone thus demonstrating a modest tonic activation of inhibitory P2y purinoceptors by an endogenous agonist. The results are compatible with the idea of a prejunctional inhibition of sympathetic neurotransmission by P~y purinoceptors which are tonically activated by endogenous purines under our conditions. Tramadol is a centrally acting analgesic with 10w opioid receptor affinity which is routinely used in clinical practice. Our previous work on spontaneously active central noradrenergic neurones of the nucleus locus coeruleus (LC) indicates that the (+)-and (-)-enantiomers of tramadol (TRAM), as well as the (+)-and (-)-enantiomers of its principal metabolite O-desmethyltramadol (DMTRAM) concentrationdependently inhibit the firing rate (J. Sevcik et al., Naunyn-Schmiedeberg's Arch. Pharmacol. 346:R22, 1992) . In this extracellular study, the effect of (-)-TRAM was abolished by the ~2-adrenoceptor antagonist rauwolscine while the effect of (+)-DMTRAM disappeared in the presence of the opioid/z-antagonist naloxone. The other two enantiomers became inactive only in the combined presence of naloxone and rauwolscine. The aim of the present study was to investigate the effects of (-)-TRAM and (+)-DMTRAM on LC neutones by intracellular recording techniques. (-)-TRAM (100gM) and (+)-DMTRAM (10#M) produced a hyperpolarization of approximately 10 mV and in consequence inhibited the firing rate. Both compounds decreased the input resistance. The effect of (-)-TRAM was abolished in the presence of rauwolscine (lgM) while the effect of (+)-DMTRAM disappeared in the presence of naloxone (0. lgM). Furthermore, (-)-TRAM (100#M) but not (+)-DMTRAM (10,aM) increased the hyperpolarizing effect of exogenously applied noradrenaline (30~M). The results confirm that (-)-TRAM and (+)-DMTRAM modulate the function of LC neurones by different mechanisms. (-)-TRAM inhibits the uptake of endogenous noradrenaline which then in turn, activates c~2-adrenoceptors. On the other hand, (+)-DMTRAM directly activates opioid ~-receptors. The common post-receptor mechanism involves the opening of K + channels followed by hyperpolarization and inhibition of spontaneous firing. Since LC neurones appear to modulate pain transmission, a change in their activity may play a role in the analgesic effect of tramadol. To determine whether mediators of the immune system may influence the activity of the adjacent "innervating" sympathetic fibers, we investigated the effects of various human recombinant interleukins ills) on exocytotic noradrenaline (NA) release in the rat isolated, vascularly perfused (Tyrode solution) spleen. The overflow of endogenous NA into the venous effluent was measured by HPLC-EC. The overflow evoked by the second and third period of perivascular electrical stimulation ($2-$3, 30 min apart; 4 Hz, 2 rain, 2 msec, 16-28 mA) after perfusion for 90 rain with any IL is expressed relative to the overflow evoked by the first period ($1) applied 90 rain before $2 in the absence of any compound. All ILs tested caused a significant inhibition of the evoked NA-overflow ($2) persisting washout of the IL ($3). Indomethacin 3 gM inhibited evoked NAoverflow to a similar extent iS2 = 66.7 ± 5.7 % of $1, n = 5) indicating a facilitatory effect of cyclooxygenase products on evoked overflow, in the continuous presence of indomethacin the inhibition by IL-1~3 remained unaffected iS2 = 58~7 ± 6.0 % of $1, n = 4).

In conclusion, mediators of the immune system such as lEe modulate the evoked NA release in the spleen. The mechanism for the slowly developing (90 min perfusion) inhibition is unclear, but does probably not involve an interaction with the cyclooxygenase pathway. Having in mind the differences in the purinergic contribution to the neurogenic contractile response, between the epididymal and the prostatic portions of the rat vas deferens, the present study was undertaken to compare, in these two tissues, the modulator role of endogenous adenosine on noradranaline (NA) release evoked by electrical stimulation (5 Hz, 2 ms, I00V, 9 rain). During perifusion with Krebs containing 1,uM yohimbine and 400aM desipramine, two or four periods of electrical stimulation were applied. The influence on NA overflow of the adenosine uptake inhibitor S-(p-nitrobenzyl)-6-thioinosine (NB1;50nM) in the absence and in the presence of the A 1 receptor antagonist 1,3-dipropyl-8-cyclopentylxanthine (DPCPX; 20nM), as well as that of the adenosine receptor agonists R-N6phenylisopropyladenosine (R-PIA; 0.1-3,uM) and 2-[p-(2-carboxyethyl) phenethylaminol-5"-N-ethylcarboxam~deadenosine (CGS 21680;3-100aM) was investigated. In the epididymal portion, NBI increased NA overflow, the $2/S1, $3/S1 and $4/S1 ratios being 1.12 -+ 0.1M, 1.26 + 0.06, and 1.35 + 0.05, respectively, (n=6), whereas, in the prostatic portion, it decreased NA overflow, the S2JS1, $3/S1 and $4/S1 ratios being 0.67 _+ 0.06, 0.62 + 0.06, and 0.60 _+ 0.05, respectively, (n=6). In the presence of DPCPX, NBI increased NA overflow in both portions ($2/S1_ ratio of 1,30 -+ 0.09, and 1.41 + 0.07 in the prostatic and in the epididymal portion, respectively, n=4). R-PIA reduced and CGS21680 increased NA overflow both in the epididymal and in the prostatic portions. These results show that,both A1 and A2 adenosine receptors are present in both portions. However, inhibition of nucleoside uptake favours much more the inhibitory effect of endogenous adenosine in the prostatic than in the epididymal portion, what may indicate that the the purinergic modulation of NA release is different at the two levels. Experiments on rat vas deferens led to the view that there are two subtypes of t~ 1adrenoceptor which differ both pharmacologically and in that one is coupled to Ca 2+ entry (CtlA), the other to intracellular Ca 2+ mobilization (alB). Adrenergic neumgenic contractions of rat vas deferens consist of a Ca 2+ entry and a Ca 2+ mobilization component. Are these two components in fact mediated by ~IA and ~XlB-adrenoceptors, respectively? -Neurogenic contractions were elicited by a single electric pulse (p) followed 10 s later by 3 p/100 Hz and 5 rain later by 50 p/5 Hz; the three l~ulse patterns were used in order to obtain well measurable Ca 2+ enlzy-and Ca z+ mobilization-mediated contractions. The adrenergic part was isolated by addition of suramin 300 I.tM; Ca 2+ entry and Ca 2+ mobilization components were isolated by ryanodine 20 ~tM and nifedipine 10 I.tM, respectively. The o~-adrenoceptor antagonists prazosin, WB 4101, 5-methylurapidil and HV 723 reduced all contractions in a concentration-dependent manner. WB 4104 was 5 -10 times more potent than prazosin against both Ca 2+ entry-and Ca 2+ mobilizationdependent contractions (as it is at CtlA-adrenoceptors; at ~tlB-adrenoceptors it is at least 6 times less potent than prazosin). 5-Methylurapidil was eqoipotent with prazosin against Ca 2+ mobilization-dependent contractions and 5 -7 dines less potent than prazosin against Ca 2+ entry-dependent contractions (again similar to O~lA-adrenoceptors; at C~lB-adrenoceptors 5-methylurapidil is about 100 times less potent than prazosin). HV 723 was approximately equipotent with prazosin against both Ca 2+ entry-and Ca 2+ mobilization-induced contractions (HV 723 has been used for another kind of al-adrenoceptor subelassificadon; Ohmura et al., Br J Pharmacol 107: 697-704, 1992) . In vasa deferentia preincubated with 3Hnoradrenaline, the antagonists at the highest concentration tested did not alter the overflow of tritium evoked by 6 p/100 Hz but increased by 12 -66% the overflow evoked by 50 p/5 Hz. -Tile results ii]dicate that in rat vas deferens neurally released noradrenaline acts on a single Ct.l-adrenoceptor, apparently CtlA, to activate both Ca 2+ entry and intracellular Ca z+ mobilization. Sympathetic transmission in the vas deferens is due to at least two co-transmitters, noradrenaline and ATP. The purinergic component of the tetanic response of the vas deferens to trains of high f~equency pulses is characterized by a twitch which rapidly fades to a much lower plateau. Experiments were carried out to examine two possible reasons for the fade, desensitization of postjanctional P2-purinoceptors and decline of ATP release. -Electrica~ stimulation (180 pulses; 10 Hz) of vasa deferentla superfused with medium containing prazosin (0.3 ~tM; to isolate purinergic component) elicited the typical biphasic tetanus. Application of either 3.0 IxM ~,~-methylene-ATP or 1.0 mM ATP during the plateau phase induced a contraction comparable to that observed when either agonist was administered 30 rain before or after electrical stiranlafion./n order to study transmitter release, vasa deferenfia were preincubated with 3H-noradrenaline and then superfused and sdmalated electrically with 30, 60 and 90 pulses at 10 Hz. ATP was measured with the luciferin-luciferase technique. In some experimeuts contractions were suppressed by prazosin and suramin to exclude extra-neural ATP release. Results are shown in the table. Irrespective of whether contractions were blocked or not, the overflow of tritium per pulse decreased only slightly with increasing pulse number, whereas the overflow of ATP per pulse decreased markedly. The three train lengths were applied in all six possible different sequences. a Twitch height was independent of pulse number and averaged 68.6 + 2.6 mN (n = 36). b Prazosin 0.3 IJM and suramin 0.3 mM in medium.

The results indicate that the rapid fade of the neurogenic purinergic tetanus is not due to desensitization of postjunctional P2-purinoceptors. It may be due at least partly to a decline in ATP release in the course of a pulse train, a decline that is observed both when total and when neural ATP release is measured. . It has been suggested that adenosine and adenine nucleotides inhibit the release of noradrenaline (NA) in rat vas deferens via a common presynaptic pudnoceptor (P3; Forsyth et al., J Pharmacol Exp Ther 256:821-826, 1991) . In contrast, in mouse vas deferens separate presynaptic purinoceptors for adenosine (P1) and adenine nucleotides (P2y-~ike) seem to exist (yon K~gelgen et al., Naunyn-Schmiedeberg's Arch Pharmacol 340:522-532, 1989) . We, therefore, compared the effects of chloroadenosine, I~,.,/-methylene-ATP (~'~nATP), 2-methylth(o-ATP, the adenosine A 1-receptor antagonist 8-cyclopentyl-l,3-dipropylx~nthine (DPCPX) and the Pp_-purinoceptor-antagonist suramin on the release of [°H]-NA i~ slices of mouse and rat vas deferens. The slices were preincubated with [ H]-NA and then superfused with medium containing desipramine 1 pmol/I, corticosterone 10 p.moVI and yohimbine 1 ~tmoVI. In most experiments they were electrically stimulated by trains of 60 pulses/1 Hz. Chloroadenosine (0.1 -10 p.mol/I), I~h'rnATP (3 -300 lamol/I), and 2-methylthio-ATP (30 .u.moVI) reduced the evoked overflow of tritium. In either species the concentration-response-curve {crc) of chloroadenosine was not changed by suramin 300 ~mol/I but markedly shifted to the right by DPCPX 10 nmol/l. Moreover, in either species the crc of I~nATP was also shifted to the right by DPCPX, and to an extent similar to the crc of chloroadenosine. The two species differed, however, in the interaction of I~/rnATP with suramin: suramin only slightly attenuated the effect of I~TrnATP in the rat but shifted the crc of ~3,rnATP about tenfold to the right in the mouse. The difference was also observed when crcs for I~'?mATP were determined in the presence of both suramin and DPCPX: suramin caused little antagonism beyond that obtained with DPCPX in the rat but caused a marked additional rightward shift of the I~'¢mATP crc in the mouse. Given alone, DPCPX failed to change the evoked overflow of tritium in either species. Suramin, however, when given alone, increased the evoked tritium overflow by up to 45 % in mouse vas deferens and by up to 20 % in rat vas deferens. An even greater increase, by up to 65 %, was obtained in the mouse but not the rat, when the stimulation frequency was increased to 8 Hz (60 pulses). The basal outflow of tdtium was not changed by suramin. The results confirm that at least in the mouse vas deferens, in addition to the known adenosine Al-receptor, a separate presynaptic purinoceptor for nucleotides (P2) exists which, in contrast to the A 1-receptor, is blocked by suramin. There is no need to postulate the existence of the putative P$-purinoceptor. The experiments also suggest that the presynaptic, suramin-sensit~ve P2-purinoceptor is activated by an endogenous ligand, possibly the NA cotransmitter ATP, during trains of pulses. This may represent a novel presynaptic autoinhibitory feedback loop for the control of sympathetic cotransmitter release. Pharmakologisches Institut, Hermann-Herder-Strasse 5, D-7800 Freiburg, FRG

We demonstrated last year that DHEC decreased dose-dependently blood pressure (BP) of conscious and anaesthetized SHR (W. H/iuser, Naunyn-Schmiedeberg's Arch. Pharmaenl. Suppl. 345: R82, 1992) . In contrast, in pithed SHR BP was increased after administration of DHEC. We suggested that this dual effect is due to a clortidine like action of DHEC. The aim of the present study was to characterize the types of adrenoeeptors involved in the modulation of BP in anaesthetized and pithed SHR and to demonstrate the agonistie and/or antagonistic action of DHEC.

Male SHR (200 -250 g, strain: SHR/NCrl BR, Savo, KiBlegg) were pithed by a standard method. Polyethylene-catheters were inserted into femoral veins (PE-10) for drug administration and into a carotid artery (PE-50) for measurement of BP and heart rate (HR). Dose-response-curves were installed during DHEC-infusioas (70 lag/kg/5min) by using noradrenaline (0.01 -1000 ~tg/kg), the C~l-agonists norfenefrine (1 -1000 pg/kg) and methoxamine (10 -1000 !ag/kg) and the ¢t~-agonist clonithne (0.3 -1000 ~tg/kg). In separate experiments, SHR received reserpine (5 mg/kg) i.p. on two consecutive days to investigate the partial agonistie effect of DHEC. On the 3rd day, rats were pithed as described above or anaesthetized with Thiobutabarbital (100 mg/kg i.p.). Dose-reponses to DHEC (1 -1000 ~tg/kg) of BP and HR were determined during i.v. bolus injections.

DHEC decreased maximal BP responses (Er~x) to noradrenaline, to the cq-agonists norfenefrine and methoxamine and to the c~agonist cloindine. The ECs0-values of these dose-response-curves were slightly increased. In pithed SHR, pretreatment with reserpine enhanced Em~, and shifted the DHEC-dose-reponse-eurve to the left. In anaesthetized SHR pretreatment with reserpine led to a reversed effect of DHEC: i.v. administration caused a dosedependent increase in BP. -Our results demonstrate an antagonistic action of DHEC on both, cq-and w~-adrenoceptors. However, in reserpinized SHR an agonistie effect on mainly c~-adrenoeeptors could also be observed. Thus DHEC may be characterized as a partial agonist at ct-adrenoeeptors. (eg Schw~rer et al. (1992 ) Gastroenterology I02:1906 . In the present experiments we studied the effects of purines on the spontaneous release of 5-HT from the porcine intestinal mucosa. Strips of the jejunum were incubated in Krebs solution and the outflow of 5-HT was measured by HPLC with electrochemical detection. The spontaneous outflow of 5-HT amounted to about 300 pmol/g*5min. Adenosine and adenosine 5'-triphosphate (ATP, each 0.i-I mmol/l) reduced the 5-HT outflow by maximally about 50%, respectively. Both effects were not affected by tetrodotoxin. The adenosine A2 receptor selective antagonist 9-chloro-2-(2-furanyl)-5,6-dihydro-[!,2,4)-triazolo(l,5-c)quinazolin-5-imine (CGS 15943A; 0.3 ~mol/l) blocked the inhibitory effect of adenosine, but not that of ATP.~,~ -methylene-ATP (0.01-0.i mmol/l) did not affect 5-HT release. The effect of I mmol/l ATP, but not that of adenosine, was abolished by apamin (30 nmol/l]. In conclusion the release of 5-HT from porcine ECC is modulated by inhibitory P1 purinoceptors of the A2 subtype and P2 purinoceptors, probably of the P2y subtype. The latter appears to be linked to apamin-sensitive potassium channels. In acute myocardial ischemia (MI) 'large quantities of endogenous catecholamines are presynaptically released. Despite these desensitizing concentrations of catecholamines present at the receptor site, acute MI leads to an increase of functionally coupled ~-adrenergic receptors ~AR) in part due to an abolition of receptor internalization. To test if an impairment of EAR phosphorylation may be responsible for this sensitization at the recepter site, phosphorylation of ~]AR in isolated rat hearts perfused according to Langendorff after acute ischemia (5-15 min) was compared to (Iso)-induced (Iso 10-M, 10 min) phospho~lation. Isosproteranol 5

Using peffusion with carrier free [32PIP i the intracellular ATP-Pool was labelled to a specific activity of 30 mCi/mmol ATP. Membrane bound ,BAR were solubilized with 1.2 % Digitonin followed by immuno-precipitation using a rabbit antibody directed against a peptide (16 amino acids) specific for the human/~IAR. The specificity of the antibody was demonstrated by Weatemblot analysis. Perfusion with Iso for 10 minutes resulted in significant increase of receptor phosphorylation by 350 %. Acute MI for 5 minutes did not alter phosporylation of /~AR compared to the normoxic control. However, in contrast to expectation, acute MI for 10 and 15 minutes significantly increased/~AR phosphorylation in a time-dependent fashion.

This is the first demonstration of a /~-agonist-induced phosphorylation of mammalian ~IAR receptors in intact heart. Acute MI leads to a comparable phosphorylation of ,8~AR despite the complete impairment of receptor internalization resulting in the clinically unwanted sensitization of ,BAR in the infarcted heart. Further studies have to investigate which kinase may be involved in this ischemia-induced phosphorylation process. The biogenic amines, histamine and serotonin (5-HT), have been regarded as pure mediators of inflammation for a long time. However, it is now supposed that 5-HT plays a modulator role. The reports are controversial in this respect.

In the present paper we deal with 5-HT in rat acute and chronic inflammation models. We determined the participation of 5-HT in carrageenin paw edema and adjuvant arthritis of Wistar and Lewis rats by administration of the potent 5-HT 2 receptor antagonist ritanserin (2,5 mg/kg p. o.) and of leflunomide (HWA 486) (20,0 mg/kg p. o.), a novel immunomodulating compound. We measured a significant increase of free plasma 5-HT level during both diseases and inhibitory effects after application of the drugs. The changes will be discussed. It was of special interest if 5-HT might be responsible, at least in part, for strain differences regarding the incidence of secondary adjuvant arthritis lesions. 5-hydroxtryptamine (5-HT) release was studied in the awake rat using the microdialysis technique. Animals were placed in a stereotaxic frame, and a guide cannula was implanted. Two days later a microdialysis probe was inserted into the guide shaft. The tip of the probe extended 2.7 mm beyond the guide shaft to reach the B1/B3 cell region (N. raphe magnus and pallidus; coordinates of tip with Lambda as reference: 0.0 lateral, 10.4 ventral. 1.0 caudal). A microinjection pump, with a gas-tight sydnge, perfused Ringer solution (containing the 5-HT uptake inhibitor citalopram 1 pmol.I "1) through the microdialysis probe at a rate of 2 pl.min "1. The microdialysis experiments started 2 hours after insertion of the probe during which time the basal extracellular 5-HT level was stabilized. Then the perfusate was collected in 20 rain fractions. 5-HT and its main metabolite, 5-hydroxyindole acetic acid (5-HIAA) were measured using hplc with electrochemical detection. The basal efflux of 5-HT and 5-HIAA during 20 rain was 5-20 fmol and 250-800 fmol, respectively. Elevation of K + to 100 mmol.1-1 for 40 rain in the dialysis buffer caused a sharp, about 10fold increase in the fractional release of 5-HT. The excitatory glutamate analog kainic acid, when added to the perfusion medium, also increased the release of 5-HT at 100 pmol.I -t, but not at 10 pmol.I -t. To test a possible modulation of 5-HT release by somatodendritic 5-HTIAreceptors, the 5-HT1A-receptor agonist 8-hydroxy-dipropylamino-tatralin (8-OH-DPAT) was introduced into the buffer 40 rain before kainic acid at a concentration of 10 pmol.1-1. The kainic acid-induced facilitation of 5-HT release was completely prevented by the agonist. By contrast, the 5-HTIA -receptor antagonist spiroxatrine (10 pmol.1-1) increased 5-HT release induced by kainic acid. There was no effect of 8-OH-DPAT on basal efflux of 5-HT. However, spiroxatdne exhibited a faciltitatory action on spontaneous release of 5-HT in some experiments. Radioligand binding studies have shown the presence of a 5-HT binding site different from 5-HT~, 5-HTm, 5-HTlc and 5-HT m in pig caudate membranes (Waeber et al., this journal, 337, 595-601, 1988) . This site had a low affinity for 5-carboxamidotryptamine (5-CT). We now looked for a transduction mechanism of this putative receptor, Pig caudate membranes were prepared by conventional methods and frozen at -30" C. Adenylate cyclase activity was measured using a-[~zP]ATP as previously described (Schoeffter et al., this journal, 337, 602-608, 1988) , with the following modifications: 5'-guanylylimidodiphosphate 10 ttmol/1 instead of GTP and absence of NaCI. Under these conditions, 5-HT produced a concentration-dependent stimulation of adenylate cyclase activity in pig caudate membranes. Maximal stimulation was 20-25 % over basal activity. 5-Methoxytryptamine (5-MeOT) and 5-CT mimicked the effect of 5-HT, reaching about the same maximal effect. The rank order of potency was (mean pEC~o values): 5-HT (7.1) _> 5-MeOT (6,9) > 5-CT (5.4). Sumatriptan was a weak partiaJ agonist (pECso -< 5). The 5-HT4 receptor agonist, renzapride, the 5-HTt~ receptor agonist, 8-hydroxy-2-(di-n-propylamino) tetraJha (8-OH-DPAT) and the 5-HTz receptor agonist, 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane (DOI) had no significant activity. Methiothepin (0.1 #mol/1) and ketanserin (10/~rnol/1) shifted the concentration-response curve to 5-HT to the right, with mean pKn values of 8.0 and 5.0, respectively. Tropisetron Currently, 5-HT receptors are classified into 3 groups, 5-HT 1, 5-HT: and 5-HT3, according to Bradley et al. (Neuropharmacol, 1986, 25:563) . However, emerging evidence from biochemical pharmacology (radioligand binding and second messenger studies) and molecular biology (receptor cloning work) indicates that there are more groups than anticipated and that within each group, more receptor subtypes exist. The criteria adopted for the new classification include operational (based on drug action), transductional and structural information. On this basis, we can distingnistl 5zHT~ to 5-HTs receptors which are members of the G-protein coupled receptor superfamily, with the exception of 5-HT3 receptors which belong to the ligand-gated ion channel receptor superfamily. The 5-HT~ group comprises 5-HTIA, 5-HT1B (rat equivalent to human 5-HTI~e), 5-HT~ and 5-HT~t~ ~ (both human clones express 5-HT~D pharmacology), 5-HT m and 5-HTmz. Their genes are introuiess and the receptors are negatively linked to adenylyl cyclase. High affinity for 5-CT is not an absolute criterion. The 5-HT: group includes 5-H%A (the classic 5-HT: receptor), 5-HT:~ (the recently cloned fundus receptor) and 5-HT:c (the former 5-HT~c receptor). They activate phospholipase C and their genes have introns. High affinity for ketanserin is no longer considered an abolute criterion for inclusion in the 5-HT~ group. 5-HT3 receptor subtypes remain to be identified, although clear (pharmacological) species differences exist. Murine and rat homologue cDNA's of one subunit have been cloned, functionally expressed and co~lfirm species differences. The 5-HT~ receptor is positively linked to adenylyl cyclase; no structural information is available, but it is expected from operational data that this class is different from the 5-HT~ group. Subtypes remain to be described. 5-HT5 receptors display a '5-H%-like' pharmacology, their genes in contrast to those for 5-H% receptors have introns, but their operational and transductional features are not known. Some orphan receptors (e.g. various vascular 5-HT~-like receptors) remain unassigned until further transductional and structural information is available. Drugs acting at 5-HTIA receptors are of use in the treatment of anxiety and have been implicated in the central control of blood pressure. In the hippoeampus and the dorsolateral septal nucleus (DLSN), activation of 5-HT1A receptors elicits a neuronal hyperpolarisation accompanied by a reduction in input resistance [Van den Hooff & Galvan (1991), Eur. J. Pharmacol., 196, 291-298; Boeijinga et al., (1992) Pflfigers Archly. Suppl. 420:R4]. The present experiments were performed to further characterise these receptors in rat and guinea-pig hippocampal CA1 neurones using compounds described as full 5-HT1A agonists in DLSN neurones. Conventional intraceliular recordings were made in transverse hippocampal slices maintained in vitro at 32°C; compounds were bath applied in fixed concentrations. In rat CA1 neurones, 5-HT (1-100 p.M) and N,N-dipropyl-5-carboxamJdotryptarnine (DP-5-CT; 3 paVl) behaved as full agonists, whereas 8-hydroxy-2-(di-n-propylamino) tetralin (8-OH-DPAT; 3-30 btM), flesinoxan (3-30 gM) and 5-methyl-urapidil (3-30 ILM) evoked maximal hyperpolarisations which were sigrdficantly smaller than those elicited by 5-HT. These compounds also attenuated the responses to subsequent applications of 5-HT, indicative of partial agonism. The estimated relative intrinsic activities were: 5-HT=I, DP-5-CT=I.I, 8-OH-DPAT=0.7, flesinoxan=0.5 and 5-methyl-urapidil=0.3. In guinea-pig CA1 neurones, 5-HT (3-100 gM) and DP-5-CT (3 I.tM) produced maximum responses which were not significantly different. Surprisingly, 5-methyl-urapidil (3 pM) failed to elicit a hyperpoladsation or a reduction in resistance but still reduced responses to 5-HT suggesting that in these neurones this compound is an antagonist. Thus, the difference between the observed maximum responses to DP-5-CT and 5-methyl-urapidil in the guinea-pig hippocampus appears to be more marked than in rat hippocampus. Differences in apparent efficacy of 5-HT1A ligands have already been described for the receptors at pre-and postsynaptic sites [Van den Hooff & Galvan (1991) ]. The present results suggest variance in efficaeies for 5-HT1A receptors localised at two postsynaptic sites, namely the hippocampus and the DLSN. These data support the hypothesis that heterogeneity of 5-HT1A receptors or their coupling mechanisms to second messengers exists. The drug discrimination procedure provides a useful method to test for interoceptive properties of drugs. The discriminative stimulus properties of a drug appear to be specific with tittle or no generalization between drugs with different mechanisms of action. One group of female Wistar rats (n = 12) were trained to discriminate the full 5HT1A-agonist 8-OH-DPAT (8-hydroxy-2-(di-n-propylamino)tetraline; training cue 0.4 mg/kg ip) from saline administered 15 min prior to daily (Monday to Friday) two-lever food-reinforced (FR10) operant sessions. Generalization of a range of doses of 8-OH-DPAT to the training dose occurred at an EDso-dose of 0.3 mg/kg ip. The prototypic serotoninergic anxioiytics buspirone, a partial 5HT1A-agonist with D2-antagonistic activity, and ipsapirone, a selective partial 5HT1A-agonist, generalized to the 8-OH-DPAT cue with ED50-values of 6.3 mg/kg ip and 3 mg/kg ip, respectively. EMD 56551 (5methoxy-3-[4-(4-(4-methoxyphenyl)-l-piperazinyl)butyl] indole), a full 5HT1A-agonist, generalized to the 8-OH-DPAT cue with an ED50dose of 0.6 mg/kg ip. The atypical neuroleptic/antidepressant roxindole, which has 5HT1A-agonistic, 5HT-re-uptake inhibiting and D 2autoreceptor agonistic activity, generalized to the 8-OH-DPAT cue with an ED50 of 1 mg/kg ip. To verify whether EMD 56551 has stimulus properties of its own, another group of rats (n = 6) was trained to discriminate EMD 56551 (1 mg/kg ip) from saline; EMD 56551 generalized to the training dose with an ED50 of about 0.35 mg/kg ip. 8-OH-DPAT generalized to the EMD 56551 cue with an ED50-dose of 0.3 mg/kg ip. In comparison to the full 5HT1A-agonists 8-OH-DPAT and EMD 56551, which both produced an interoceptive stimulus and generalized almost equipotent to each other, the partial agonists buspirone and ipsapirone required higher doses for generalization; roxindole displayed the profile of a full 5HT1A-agonist, Ligand-gated ion channels are sites of action of anaesthetics IBarann et el. (1993) Naunyn-Schmiedebergg Arch Pharmacol , in press] and ethanol. We investigated, whether cation influx through the 5-HT3 receptor channel is modified by ethanol and whether the cation influx through the fast, voltagedependent sodium channel is affected in the same way. The effects of ethanol on both channels were investigated in N1E-115 neuroblastoma cells and the organic cation ~C-guanidinium was used as a tool to measure cation influx. Cells were pro-incubated in Ca* + -and Na+-free buffer (sodium replaced by choline) for 15 min and then exposed for 2 min to 5/~mol/I t4C-guanidinium plus either 30-100 pmol/I 5-HT or 0.1-1mmol/l veratridioe. After washing the cells, the ~4C content was determined by liquid scintillation counting. The t4C influx elicited by 30/~mol/I 5-HT was increased by ethanol 3-100 mmol/I (by 100% at 30 mmol/I and by 300% at 100 mmol/I), whereas the veratridine-induced influx was either not or considerably less increased. The concentration-effect curve for the 5-HT-induced t4C-guanidinium influx was bell-shaped with the maximum influx at 100-300/anol/I 5-HT. Ethanol (100 retool/I} increased the 5-HT-induced ~4C-guanidinium influx at all 5-HT concentrations studied. Pro-incubation of the cells with 30/tmol/I 5-HT for 2-16 rain resulted in a decrease of the 5-HT-effect I 30 ~mol/I, 2 min) which was the more pronounced the longer the period of pro-incubation was, indicating desensitization of the receptor. This decrease could not be altered by addition of 100 mmol/I ethanol to the pro-incubation buffer indicating that the amplifying effect of ethanol may not be due to a deceleration of 5-HT 3 receptor desensitization.

Institut fQr Pharmakologie und Toxikologie der Universit~t Bonn, Reuterstr.2b, 5300 Bonn, BRD 498 HUMAN ARTERY CONTRACTILE 5-HT1-LIKE RECEPTORS RESEMBLE MORE 5-HT1D~3 THAN 5-HT1Do~ RECEPTORS A.J. Kaumann, A.A. Parsons* and A.M. Brown* 5-Hydroxytryptamine (5-HT) constricts human large arteries (cerelgral, coronary) through 5-HTl-like receptors or a mixture of 5-HTl-like and 5-HT 2 receptors (coronary artery). It has been suggested that 5-HTl-like receptors are similar to the 5-HT1D receptor of which two subtypes of similar characteristics, 5-HT1Do~ and 5-HT1DI3, have been cloned. To establish whether 5-HT1Dc~ or 5-HT1DI3 receptors could be arterial contractile 5-HTl-like receptors we compared our data from human isolated cerebral and coronary arteries with our data from cloned and transfected 5-HT1D ~ and 5-HT1D ~ receptors. Sumatriptan was a partial agonist with an intrinsic activity of 0.7 with respect to 5-HT in cerebral arteries; its contractile potency (pEC50 = 6.3) matched its blocking potency (pKp = 6.2) against 5-HT-induced contractions. On coronary arteries sumatriptan was a full agonist with respect to 5-HT for 5-HTl-like receptors characterised in the presence of ketanserin (to block 5-HT 2 receptors). Ketanserin 1 gM failed to block 5-HT-induced contractions in cerebral and on occasion in coronary arteries but methiothepin caused blockade with a pK B of 9.0. Unlike 5-HT1Dc~ receptors (pK D = 7.2-7.5) 5-HT1D[5 receptors were found to have only negligible affinity for ketanserin (pK D < 5.0). Because cerebral and on occasion coronary arteries are resistant to blockade b~ a ketanserin concentration (1 gM) that should saturate 5-HT1Dc~ receptors, these can be excluded as candidate for 5-HTl-like receptors, plCs0 values of partial agonists as inhibitors of forskolin-stimulated adenylyl cyclase coupled to 5-HT1D ~ receptors matched the corresponding pEC50 values for cerebral arteries. The results suggest that contractile arterial 5-HT 1 -like receptors are 5-HT 1DI3 receptors. Pharmacol. 328:301-3; 1985) we intended to investigate the effects of higher concentrations of methiothepin and ketanserin (> 0,1 pmol/l) upon serotonin induced contractions of human temporal arteries 0-ITA). The vessels were obtained during neurosurgical operation. After dissection of connective tissues arteries of 600-800 gm diameter were cut into tings of 4 mm length. The vessels equilibrated for 1-1½ hours in oxygenated Krebs-Hanseleit buffer (20% 0 2, 5% CO 2 in N2) and were mounted isometrically using 1 g tension. The ECs0 for serotonin in vessel preparations with intact respectively removed endothelium amounted to 72.38 ± 6.77 nmol/l or 23.13 4-2.59 nmol/l. Pretreatment of the HTA with methiothepin (0.001 -30 nmol/l) and ketanserin (0.003-30 nmol/l) resulted in an suppression of the serotunin -dose response curves. According to the results obtained after precontraction of the vessels with serotonin prior addition of ketanserin and methiothepin we consider both substances as non-competitive antagonists. The IC50 of ketanserin and methiothepin was 2.34 4-0.55 nmol/l and 11.43 ± 0.53 nmol/l in presence of endothelium and 2.97 4-0.23 nmol/l respectively 1.35 ± 0.18 nmol/l after removal of the endothelium. However, higher concentrations of ketanserin as well as methiothepin were surmounted by serotonin concentrations above 30 grrtol/l. In conclusion our experiments revealed that both ketanserin and methiothepin act as non-competitive and surmountable antagonists upon human temporal arteries. The measurement of locomotor activity and dopamine (DA) release in the nucleus accumbens (NAC) by in viva microdialysis was used to examine the effects of a highly selective k-antagonist, nor-binaltorphimnine (nor-BNI), on the development of morphine-induced sensitization to both the behavioral and neurochemicai effects of the prototypic ~-opioid receptor agonist morphine. Sprague-Dawley rats were pretreated with a single dose of nor-BNI (30/xg, i.c.v.) followed by saline or ascending doses of morphine (10-120 mg/kg s.c.) for 10 days. Three days after the last morphine injection, a challenge dose of morphine (10 mg/kg s.c.) was administered. Morphine-induced changes in motor activity and dopamine release were then assessed. Morphine produced a biphasic effect on motor activity in drug naive animals. Depression of activty daring the first 2 h post-morphlne injection was followed by a significant increase in activity (3-4 h post-injection). Chronic morphine resulted in a significant enhancement of morphine-induced stimulation of activity indicating the occurence of behavioral sensitization. In animals which received the nor-BNl pretreatment the stimulant effect of morphine was significantly potentiated [F(1,48) = 10,4; P<0.011. On the neurochemical level, similar effects were observed: Morphine (10 mg/kg s.c.), given 3 days after the chronic exposure to morphine, produced a greater increase in DA release in the NAC in the morphine experienced rats compared to naive rats. In nor-BNI pretreated rats, DA release was enhanced even further compared to animals which received only the chronic morphine treatment [F(1,72) = 6,4; P<0.05]. These results suggest that a blockade of endogenous k-opioid systems may potentiate the development of morphine-indnced sensitization to motoractivating as well as mesolimbic DA-releasing effects. The latter process is thought to be critically involved in the initiation and reinstatemant of drug-seeking behaviour. Exposure of organisms to stressful stimuli is associated with activation of the hypothala~c pituitary adrenal (HPA) axis. In the adult, this also activates the mesolimbie dopamine (DA) system. However little is known about neonatal stress effects on the mesolimbie systtala and DA depemtent behaviors in adulthood. Such information is however of particular importance in view of recent evidence suggesting that stress or the responsiveness of an individual to such stimuli may be a causal factor in compulsive drug-seeking behavior. The present study was thus conducted to examine these questions, Rat pup litters were separated from their mothers for lh sessions daily bstwean postnatal days 16 and 20. After weaning (d23) male pups were kept for behavioral testing starting at day 40. ha the first ex~rin~nt rats were exposed to one of three durations of restraint stress: 0, 20 or 60rain. Immediately after termination of stress, cortieosterone levels in serum, and DA and DOPAC levels in pundaes from the nucleus aecurnbens and the prefrontal cortex were determined. In a further experiment locomotor acaJvity in an open field was assessed; with and without prier saline injections. These rats were then conditioned with 3.0 or 5.0 mg&g morphine in a plaee preference conditioning paradigm. Dopamine utilization (ratio of DOPAC to DA) was significantly reduced after restraint stress in stressed but not control animals, whereas corticostorone levels woe significantly increased at~ stress in both groups. Habituation to the open field was impaired in the post-natully str~sed animals and the raotivatiorml eff~eta of low, but not of high, doses of morphine were increased, as assessed by place conditioning.

These results indicate that postnatal stress has long-lasting effects on DA neurotransmission and DA-related behaviors. Furthermore they suggest that if such experience occurs early in development, drug-seeking behavior may be significantly altered. respectively. After equilibration for at least 30 min the agonist potency for morphine and M6G was redetarmined. All tested antagonists resulted in a rightward displacement in the dcee-response curve with no decreasing maximum effect, a pattern typical of cempetitive, reversible antagonism. The IC50-, pA2-, and KD-values were computed employing the pseudo-Hill plot, the Schild plot, and the Lineweaver-Burk plot, respectively. The ICs0-values of morphine and M6G were found to be 0.07.10 -6 and 0.065-10 -6 M, respectively. In presence of nalexonazine the IC50-values increased 1.5-6 fold and 1.3-7 fold for morphine end M6G, respectively. Norbinaitorphimine resulted in a rightward-shift of the IC50values between t-4 fold and 1-3 fold and Naltdndois betwesn12-30 fold and 10-25 fold for morphine and M6G, respectively. Our data demonstrate that M6G is generally not more potent than morphine in displacing opioid antagonists in the guinea-pig ileum. At least three cDNA's, named SSTR1-3, coding for SRIF receptors have been cloned (Kluxen et at, Proc Natl Acad Sci USA, 1992, 89, 4618; Yamada et at, ibidem, 251; Yasuda et at, J Biol Chem, 1992, 267, 20422) . The mRNA distribution of these SRIF receptors was studied in rat brain by in situ hybridization. Oligonucleotide (40-43 mars) probes complementary to SSTR-1 (mouse), SSTR-2 (rat) and SSTR-3 (mouse) cDNA's were used.

Highest levels of SSTR-1 mRNA were found in lateral septal nucleus, entorhinal and frontal cortex, and cortical layers V-V1. Moderate signals were detected in taenia tecta, cortical layer IV, dentate gyrus and medial amygdaloid nucleus. Most other brain regions including superficial cortical layers, hippocampus, basal ganglia, locus coeruleus and cerebellum sl~owed comparatively low hybridisation signals. High densities of SSTR-2 hybridisation signals were found in inferior/superior colficulus, cerebral cortical layer IV, dentate gyrus, medial habenular nucleus, taenJa tecta and subiculum. Intermediate levels were detected in entorkinal and frontal cortex, superficial grey layer of the superior colliculus, all regions of Amman's horn 0~ppocampus), and in median mammilary nucleus. Moderate signals were found in primary olfactory cortex, basal ganglia and deep cortical layers. Other regions such as locus coeruleus, nucleus accumbens, cortical layers I-Hi and cerebellum showed very low signals. High levels of SSTR-3 hybridisation signals were detected in the olfactory bulb and granular layer of the cerebellum. Intermediate densities were found in primary olfactory cortex, media/ habenular nucleus, locus coeruleus, CA3, CA2 and CA4 fields of Amman's horn, lateral habenular nucleus, islands of Calleja and mammilothalamic tract. Moderate signals were localised in frontal and entortfinai cortex, nucleus accumbens, CA1 field, inferior colliculus, taenia tecta, dentate gyrus, accumbens nucleus, basal ganglia and subiculum. The wide and differential brain distribution of the three SRIF mRNA's suggests that brain SRIF receptors are not restricted to neuroendocrine functions.

Preclinical Research, SANDOZ Pharma Ltd, 4002 Basel, Switzerland 5O5 BINDING PROPERTIES OF THREE SOMATOSTATIN RECEPTOR SUBTYPES C. Bruns, L. Rohrer, D. Hoyer, H. Ltibbert Somatostatin (SRIF) is an important modulator of endocrine and exocrine secretion. In addition, SRIF was shown to act as a neurotransmitter. SRIF induces its biological actions by interacting with specific membrane receptors. Previous biochemical and pharmacological studies provided evidence that at least two subtypes of SRIF receptors exist. Recently, the genes encoding three different SRIF receptor subtypes have been isolated from various cDNA and genomic libraries. To investigate the binding properties of the three cloned SR/F receptors (SSTR 1-3) the genes encoding the three cloned receptors were transiently expressed in COS-1 monkey kidney cells. Each receptor subtype was labelled with [~zSI]Tyr~-SRIF-14, a specific ligand for SRIF receptors.

A single class of high affinity binding sites was identified for each receptor subtype.

[125I]Tyr~LSRIF-14 binding to SSTR 1-3 receptors reached equilibrium by 30 minutes. The natural forms of SRIF, SRIF-14 and SRIF-28 exhibited high affinity binding to SSTR i-3. Strikingly short synthetic SRIF analogs like Octreotide, MK 678 (Seglitide), RC-160 (Vapreotide) or BIM 23014 (Angiopeptin) displayed high affinity binding to only one subtype, SSTR2 (IC5o=3-6x10 -~° mol/L). In contrast these SRIF analogs showed low affinity binding to SSTR 1 (IC~o>=10 "~ mol/L) and a moderate affinity to SSTR 3 (IC~0 = 3-6x10 -8 mol/L). The binding data obtained so far with SSTR 1-3 indicate significant differences in binding properties. No differences in binding affinities were observed comparing the various synthetic SRIF analogs which have been reported earlier as having different selectivities for SRIF receptor subtypes. We have generated stable transfected cell lines expressing the cloned SRIF receptors SSTR 1-3. These cell lines are now used as tools to screen receptor ligands and to determine the receptor second messenger coupling. Within the mesolimbic dopaminergic system of the rat multiple and very complex interactions exist between the transmission of cholecystokinin (CCK) and dopamine (DA). The interpretation and comparison of the plenty of described CCK effects necessarily requires a further classification into the brain area, the pathway and the receptor subtype. Based upon former results, indicating a CCK-B receptor-mediated increase in the anterior part of the nucleus accumbens (N.acc.) after microinjection of CCK into the ventral tegmental area (VTA) (1), and the finding, that CCK injected into the VTA led to behavioral alterations via the CCK-B receptor when combined with apomorphine (APO) (2, 3), we investigated the influence of APO and CCK administered simultaneously into the VTA on the DA release in the anterior part of N.acc. of anaestethised rats using differential pulse voltammetry. Microinjection of 1 ng APO into the VTA caused an increase of peak 2 (which corresponds to DA in pargyline pretreated animals) whereas 10 and 100 ng were ineffective. Simultaneous microinjection of 1 ng APO and I ng CCK augmented this increase of peak 2 whereas 100 ng CCK suppressed the APO (1 ug) induced increase of peak 2. Neither the low dose (lug) nor the high dose (100 ng) of CCK were able to change the DA signal produced by APO (10 or 100 rig). The data suggest both synergistic and antagonistic effects on the DA release in the N.acc. after microiujection of various doses of APO and CCK into the VTA, A mouse somatostatin (SS) receptor cDNA was cloned from neuroblastoma x glioma (NG108-15) cells. The sequence is almost identical to that of the mouse SSTR2 receptor (Yamada et al. Proc.Natl.Acad.Sci. USA 89, 251 (1992) ) but lacks about 300 nucleotides between transmembrane domain VII and the C-terminus. This spliced variant of SSTR2 (designated SSTR2B) encodes a protein which is 23 residues shorter than that predicted from the SSTR2 sequence (designated now as SSTR2A), and differs in 15 amino acids at the C-terminus. mRNA corresponding to SSTR2B occurs in various mouse tissues and cell lines as revealed by Northern blot analysis and polymerase chain reaction. The SSTR2B receptor binds SS-14, SS-28 and the stable SS-analogue SMS 201-995 with a similar high affinity (KD between 0.5 to 1 nM) when transiently expressed in COS-7 monkey kidney cells or stably in chinese hamster (CHO-K1) cells. Activation of adenylate cyclase by forskolin in stably transfected CHO cells but not in mock-transfected cells is markedly inhibited by SS-peptides. Moreover, pre treatment of the transfected CHO cells by pertussis toxin completely prevents the inhibition by SS-peptides of the forskolin-induced activation of the adenylate cyclase. These findings indicate that the SSTR2B receptor is negatively coupled to adenylate cyclase by a pertussis-eensitive G-protein. In contrast, the SSTR2A receptor is not linked to adenylate cyclase (Rens-Damiano et al. 1992 J.PharmacoI.Exp.Ther. 42,28 (1992 . It, thus, appears that one physiological consequence of the alternate splicing of SSTR2 gene is to regulate the coupling of the receptor to differeot effector systems.

Physiologisches Institut, University of Munich, Pettenkoferstral$e 12, D-8000 MQnchen, Germany Cholecystokinin (CCK) receptors can be divided into CCK-A and CCK-B types. CCK-A receptors show an enhanced sensitivity to CCK-8s, while those of CCK-B type display a high affinity for CCK-4 and CCK-8s. Recently, a role of CCK in learning and memory processes was suggested. The aim of the present study was to investigate the influence of post-trial peripherally administered CCK-8s and CCK-4 on habituation of rats to the novelty of environmental stimuli in two different tests. First, habituation was measured in 5-min periods in an open field on three consecutive days and rearings were recorded. Second, rats were exposed to a hole-board for two 10-min periods, spaced 24h apart and head-dips and locomotor activity were measured. CCK-8s and CCK-4 (both 0.2-25 ~g/kg, i.p.) administered immediately post-trial resulted in a reduction of rearings in the open field, indicating enhanced habituation and thus, facilitation of memory. This treatment failed to influence habituation using the hole-board method. The administration of CCK-8s or CCK-4 with a delay of at least 2.5h did not alter habituation in both tests. Thus, both CCK fragments can facilitate learning, but the effectiveness in promoting habituation learning depends on the task used. Cholecystokinin (CCK) is one of the so-called brain-gut peptides and belongs to the CCK/gastrin peptide family. Tiae cholecystokinin octapeptide (CCK-8) seems to be the main molecular subform found in several regions of the vertebrate brain so far. It could be shown that CCK-like IR mainly was found in neurons of the mesencephalon as well as in basal ganglia of the forebrain. In rats, it could be proved, that some of the mesolimbic neurons contain both CCK and DA. It was demonstrated, that these neurons terminated in the Nc. accumbens (Nc. acc.) posterior, whereas no colocalization was found in Nc. acc.anterior. That as well as other findings substantiated the hypothesis of cotransmission of a neuropeptide and a classical transmitter. In the last years many attempts were made to confirm that concept but the results obtained so far are not sufficient to construct a conclusive theory. Findings from literature concerning anatomical as well as functional studies give a rather confusing picture. Since cireadiane rhythms are known to influence pharmacological effects (chronopharmacology) to a remarkable degree, the circadiane rhythms of both, CCK and DA were investigated in Nc.acc. and striatum of rat brain. Following results were obtained: 1. There are significant time-dependent fluctuations of CCK and DA content in all the brain regions investigated. 2. The CCK rhythm is in opposite to the DA rhythm. 3. There is a significant negative correlation between the time course of CCK and DA content. Based on these results we have affected the dopaminergic system of rats by gamma butyrolactone (GBL, 750 mg/kg) at 9 a.m. and 1 p.m.,respectively. Both, the CCK and DA content were measured after 0,5h and 2h in Nc.acc..Following results were obtained: 1. There are significant (remarkable) differences in CCK and DA content after drug administration at 9 a.rn. in comparison to the same treatment at I p.m.. 2. The changes in CCK content are not in coincidence with DA changes (at least partly). This mode of CCK,rDA correlation not known so far is discussed on the base of the cotransmission hypothesis. [Michel (1991) TiPS 12:389] . Autoradiographic studies revealed that rat cortex and hippocampus contain predominantly Y1 and Y2 binding sites, respectively [Dumont et al. (1991) Eur. J. Pharmacol. 191:501] . It was the aim of the present study to confirm the presence of the two binding sites in these tissues by means of radioligand competition binding experiments and to study the effect of different ion compositions of the buffer on the binding behaviour of selective NPY agonists. Displacement experiments were performed using 35 pM [125I]NPY in a Ca2+(7.5 raM) or Mn2+(5 raM) containing Tris/NaCl buffer. 1 we can conclude that displacement curves with slopes close to unity could only be obtained when Ca 2+ or Mn 2+ was added to the cortex or hippocampus assay, respectively. Under these conditions the Ca2+/cortex is a Y1 assay since the Y1 agonist [Leu31,Pro34]NpY displays a high affinity. On the contrary the Y2 selective C-terminal fragments of NPY exhibit high affinity in the Mn2+/hippocampus assay indicating the presence of Y2 binding sites. Accordingly, the identification of NPY receptors in the rat brain depends not only on the brain region under investigation but strongly on the experimental conditions. The principal brain cholecystoldnin receptor (CCK-B/gastrin) has been implicated in mediating anxiety, panic attacks, satiety, the perception of pain, and memory. Specific antagonists provide a potential means to alleviate CCK-induced neuropsychiatric disorders. We have recently cloned the canine and the human CCK-B/gastrin receptors. Although these receptors are 90% identical at the amino acid level and have a similar agonist binding profile, comparison of receptor ligand affinities has revealed a 25-fold higher affinity for the nonpeptide antagonist L365,260 (3R(+)-N-(2,3-dihydro-1-methyl-2-oxo-5-phenyl-1H-1,4-benzodiazepin-3-yl)-N'-(3-methylphenyl)urea) in the human as compared to the canine receptor. Conversely, the structurally related antagonist L364,718 (devazepide) was preferentially recognized by the canine receptor. Using 125I CCK-8 competition experiments with chimeric constructs bewcen these two receptor subtype homologues, we have identified a domain which determines high affinity antagonist binding. The critical amino acids within this domain were identified by sequentially mutating candidate residues. Our results demonstrate that amino acid sequence divergence underlies species differences in receptor affinity for both L365,260 and L364,718. Of note, agonist-derived dipeptoid antagonists including PD135158 [Hughes J et al. (1990) We reported earlier [Doods and Krause (1991) Eur.J.Pharmacol 204:101] that prejunctional NPY receptors in the rat vas deferens belong to the Y2 subtype whereas those present in rabbit vas deferens are of the Y1 type, It was the aim of the present study to characterize the prejunctional NPY receptor in the guinea-pig trachea and we used nonselective, Y1 selective as well as Y2 selective NPY/PYY analogs. Tracheal strips from guineapigs (DHP, 500-600 g) were suspended under a tension of 10 mN in a Krebs-Ringer solution. Contractile responses were elicited by electrical field stimulation (supramaximal voltage 0.5 ms, 0.5 Hz) of the tissue and cumulative concentration-response curves to the agonists were constructed. NPY and PYY inhibited the contractile response to electrical field stimulation in a concentration dependent manner with -log EC50 values of 7.89 and 8.73, respectively. The YI selective agonists [pro34]NpY, [Pro34]pyY and [Leu31,Pro34]NpY and the Y2 selective agonlsts NPY(13-36) and NPY(18-36) were at least 300-fold less active in comparison to NPY and PYY. However, PYY(3-36) possesses a -log EC50 value of 8.00. In conclusion: the low potency of the Y1 selective agonists exclude the involvement of YI receptors. The presence of Y3 receptors is unlikely because PYY is even more potent than NPY in the guinea-pig trachea whereas PYY possesses a low affinity for the putative Y3 receptor [Michel (1991) TIPS. 12:389] . Although the C-terminal fragments of NPY exhibit a low potency the high potency of the Y2 selective agonist PYY(3-36) [Feth et al. (1992) Three binding sites for neuropeptide 5' (NPY) have been identified so far. These binding sites can be distinguished by their differential affinity for certain NPY analogues. [Leu31,Pro34]NpY for instance possessess high affinity for YI sites whereas C-terminal fragments like NPY(13-36) and NPY(18-36) display selectivity for Y2 sites [Michel (1991) TiPS 12:389]. The present study was performed in order to investigate receptor selectivity and possible species selectivity of these ¥1 and Y2 selective agonists. For Y1 receptor analysis receptor binding studies were performed using cell lines expressing the cloned human (CHO-cells) or rat (293-cells) receptor as well as naturally expressed Y1 receptors in human SK-N-MC cells and rat cortex. With respect to the 5"2 receptor subtype we used rabbit kidney, rat hippocampus and the human SMS-KAN neuroblastoma cell line. In addition, we tested the affinity of the putative 5"1 antagonist PYX-2 [Tatemoto et al. (1992) Proc.Natl.Acad.Sci. USA 89:1174] for the Y1 and Y2 binding sites. In agreement with the literature NPY displays tenfold higher affinity for the Y2 receptor, [Leu31,Pro34]NPY is roughly 10-50 fold Y1 selective. The C-terminal fragments are more than 100 fold Y2 selective. The tested compounds do not show species selectivity for the Y2 receptor subtype. With regard to the Y1 receptor subtype, only the human Y1 receptor expressed in CHO-cells displays 10-fold lower affinity for NPY (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) compared to the other membranes tested (SK-N-MC, rat Y1-293, rat cortex). A different glycosylation pattern in the CHO cells might influence this receptor/ligand interaction. The putative antagonist PYX-2 did neither bind to the Y1 receptor subtype nor to the Y2 receptor and it did not show antagonistic properties in a Ca 2+ mobilization assay, In conclusion: the NPY analogues tested are subtype selective and, with one exception, species nonselective. Furthermore, the compound PYX-2 did not display NPY receptor affinity or antagonistic effects. Substance P (SP) (NK1 receptor agonist) stimulates the hydrolysis of inositol phospholipides in peripheral tissue and in the brain. In mammalian tissues, three tachykimn receptor subclassesfNK1, NK2, NK3) have been identified; CP 96,345-1 [(2 S,3 S)-cis-2 -(diphenylmethyl)-N-((2-methoxyphenyl)methyl)-1 -azabicyelo[2.2.2] octan-3-amme] has been proven to be a NK1 specifin receptor antagonist. The purpose of our study was to investigate the effect of CP 96,345-1 on phosphoinositide hydrolysis induced by activation of central NK 1 receptors.

Male Wistar rats were saerified, their brains were rapidly removed and dissected on ice. Hypothalamic slices were cross chopped using a Me Ilwain tissue chopper and incubated with myo-2[3H]-inositol for 2.5 hours, then LiCI (10 raM) was added. 20 minutes later, one of the following agonists, SP [10"6M], the NK1 specific receptor agonist [Sar 9, Met(O2)ll]-sP [10"6M] and carbachol [10"4M] were added and the incubation was continued for 1 hour. The antagonist CP 96,345-1 [2x10"4M] was added 10 minutes before the agonists. Incubation was terminated by addition of perchloric acid (HC104) (1%). Lipides and the inositol phosphates were extracted. SP, [Sar 9, Met (02 )11].Sp and earbachol stimulated significantly phosphoinositide hydrolysis in the hypothalamus. CP 96,345-1 abolished the [Sar 9, Met(O2)ll]-sP induced phosphoinositide hydrolysis and tended to inhibit that of SP. Furthermore, the carbachol induced phosphoinositide hydrolysis was inhibited by the NK1 specific receptor antagonist. When used alone, C]? 96,345-1 did not alter phosphoinositide metabolism.

Our results demonstrate that the compound CP 96,345-1 inhibits the NK1 specific receptor agonist induced phosphoinositide hydrolysis and diminishes that of SP. Furthermore, these findings may indicate that carbachol induced stimulation of phosphoinositide hydrolysis is partially dependent on an activation of NK1 rec~tors. Calcitonin gene-related peptide (CGRP) potently enhances mucosal blood flow (MBF) in the rat stomach. The aim of this study was to examine whether CGRP also dilates extramural arteries supplying the stomach and whether the vasodilator effect of CGRP involves nitric oxide (NO). Gastric MBF in urethane-anaesthetized rats was measured by the hydrogen gas clearance technique. Blockade of NO synthesis by N 6nitro-L-arginine methyl ester (L-NAME, 20 ~mol kg 1, injected close arterially) significantly enhanced mean arterial blood pressure and abolished the increase in MBF caused by close arterial infusion of rat CGRP-a (30 and 60 pmol min -1) to the stomach, The inactive enantiomer D-NAME (20/~mol kg -1) was without effect. Rat CGRP-u (30-1,000 pmol kg -l, i.v.) led to a dose-dependent increase in blood flow through the left gastric artery as determined by an ultrasonic transit time technique. L-NAME (20/~mol kg -1, i.v.) reduced the activity of CGRP to augment blood flow in the left gastric artery by a factor of 5, whereas D-NAME was without effect. The ability of vasoactive intestinal polypeptide to increase left gastric arterial blood flow remained unaltered by L-NAME. Experiments on the isolated perfused bed of the rat left gastric artery precontracted with methoxamine showed that rat CGRP-a was 3 times more potent as a dilator than rat CGRP-g. The dilator action of rat CGRP-a in this isolated preparation was not affected by L-NAME (40 gM). These data show that CGRP is a potent dilator of mucosal and extramural resistance vessels in the rat stomach. The mechanism of vasodilatation is regionally different inasmuch as NO is essential for the CGRP-evoked hyperaemia in the gastric mucosa, whilst the CGRPinduced dilatation of extramural arteries involves both NO-dependent and NO-independent mechanisms. In the periphery, angiotensin II ( ANG II ) promotes trophic changes in vascular and heart smooth muscle cells by inducing the expression of transcription factors like Fos and Jun. However, little is known about trophic effects of ANG I~ in the central nervous system ( CNS ). Therefore, the aim of our study was to investigate the influence of ANG II on the expression of the immediate-early gene products Fos, Jun and Krox 24 in the CNS using immunohistochemical techniques. Male conscious Wistar rats were injected intracerebroventricularly ( iev ) with ANG II ( 1 rig, 10 ng, 100 ng) or 0.9% NaC1 as control. Slices of fixed brains were incubated with polyclonal antibodies against Fos, Jun and Krox 24. ANG II induced a dose-dependent expression of the proteins Fos and Jun in the ependyma of the lateral ventricle ( LV ), median preoptic area ( MnPO ), subfornical organ ( SFO ), paraventricular nucleus ( PVN ) and supraoptic nucleus ( SON ) with a threshold dose of 1 ng; controls showed no expression in those regions. A distinct expression of KROX 24 was detected in Mr&O, SFO, PVN and SON, while the ependyma of the LV showed no clear staining. The ANG II-AT1 receptor antagonist losartan ( 5 ~tg/~tl ), given icv five minutes before iev application orANG II ( 100 ng ), blocked the expression of Fos and Jun in the MnPO, SFO, PVN and SON, while the expression in the ependyma of the LV remained unaffected. Our data show for the first time that ANG II can induce expression of transcription factors in distinct brain nuclei via an ANG II-AT1 receptormediated pathway. This expression could promote a neuronal plasticity, which may consist of generation of new proteins, receptors or molecular neuroanatomical changes. Department of Pharmacology, Experimental Hypertension Research and *Department of Physiology II, University of Heidelberg, 6900 Heidelberg, Germany INVOLVEMENT OF SEROTONINERGIC PATHWAYS IN TIqE CENTRAL CARDIOVASCULAR AND BEHAVIORAL EFFECTS OF SUBSTANCE P H. Decker, J. Culman, S. Klee and Th. Unger Substance P (SP) applied into the lateral ventricle was shown to induce a distinct cardiovascular response associated with increased locomotor activity and excessive grooming. Interactions between serotonni (5-HT) and SP systems in certain forebrain regions have been postulated, however, the functional implications of such interactions have not yet been established. We studied the effects of perturbation of the rat brain serotoninergic system on the cardiovascular and behavioral responses to SP (50 pmol) injected intracerebroveatricolarly (icy). Blood pressure (BP) and heaxt rate (/dR) were measured through an intraattedal catheter, the behavioral response (face washing (FW), hind limb grooming and biting (HLGB) and wet dog shakes (WDS)) were recorded over a 30 rain period after icy SP injection. A depletiou of brain 5-HT was achieved by icv injections of 5,7-dihydroxytryptamine (150 ug injected twice ), which decreased the 5-HT content in the hypothalamns, hippocampos, striatum, lower brain stem and spinal cord by 80.3 %, 93.8%, 86.2 %, 47,1% and 78.2 %, respectively.This procedure markedly potentiated the cardiovascular but not ~e behavioral response to icv SP (BP by 95 % and HR by 100 %, respectively). Increases in 5-HT neurotransimssioo by intraperitoncal (i.p) administration of 5-HT reaptake blooker fluoxetine (10 rng/kg, 1 h before icv SP) did not affect the cardiovasealar and behavioral responses to icy SP. In order to selectively characterize the functional interaction between SP and 5-HT in forebrain struetores, the cerebral aqueduct was obstructed by a cream plug in one group of rats. The cardiovascular and behavioral (FW and HLGB) responses to iev SP were enhanced in plug-animals compared to sham operated controls (BP, p<0.05; HR, p<0.01), fluoxetine prareatment markedly reduc~l both, the cardiovascular and behavioral responses to icv SP. The brain serotoninergic system exerts a tonic inhibitory effect on cardiovascular effects of iev SP, since the depletion of the brain 5-HT led to a poten "tmtion of the cardiovascular response to icv SP. However, increased 5-HT neurotransmission in the brain inhibits the eeotral cardiovascular effects of SP only when SP cannot interact with its receptors localized in the lower brainstem or spinal cord.

Department of Pharmacology, University of Heidelberg, Im Neueraheimer Feld 366, 6900 Heidelberg, Germany.

T. Barth, A. Domeney*, M.E. Kelly* and H. Fink

The two-compartment black and white test box was validated for mice and used to study anxiety-related behavior and anti-anxiety properties of drugs (Costall et el., Pharmacol. Biochem. Behav. 32: 777, 1989) . In a twocompartment box divided in a dark and a brightly illuminated area mice show aversion to the light and preference for the black area. The aim of this study was to validate this method for rats. There exist only sporadic findings in literature concerning the use of this method in rats (Onaivi et el., Prog. Neuropsychopharm. Biol. Psych. 13: 963, 1989) . The validity of the black and white test box to measure changes in exploratory behavior relevant to assessment of anxiety was investigated by variation of illumination within the white and the black area, starting point, test duration and different times of day, the use of different strains of rats and drug treatments (benzodiazepines and 5-HT1A-agonists). Furthermore, the influence of handling, pre-training and habituation was assessed. Our data reveale that two of eight measured parameters (time spent in and latency of first reentry into the white area) and two calculated ratios (number of line crossings as well as rearings in the white area/ both areas) are suitable to assess anxiety-related behavior and anxiolytic drug effects. Our results suggest that this test may be a useful addition to the battery of rat models of anxiety. C.Tsch6pe 1, J. Cuhnan 1, P.Picard 2, A.Prat 2, D.Rcgoli 3, R.Coutore 2 and Th.Unger 1 Substance P (SP) and neurokinin A (NKA) are non-selective naturally occaring agonists for NK-1 and NK-2 tachykinin receptors, respectively. In eonscions rats, intracerebroventrienlarly (icy) injected NKA is as potent as SP in elicifing cardiovascular and behavioral responses. The mechanisms of central NKA action is a matter of controversy since NK-2 receptors have not been convincingly demonstrated in the adult rat brain. We studied the cardiovasenlar and behavioral response to icy SP and NKA after iev pretrealment with specific NK-1 (CP 96,345 [eis-3-(2-eis-methoxybenzylamino)-2-benzhydryl-quinudidine] ), NK-2a (MEN 10,376 [H-Asp-Try-D-Trp-Val-D-Trp-D-Trp-Lys-NH2] ) and NK-2b (R 396 [Ac-Leu-Asp-Gln-Trp-Phe-Gly-NH2] ) receptor antagonists. Rats bearing an intmarterial catheter were pretreated iev with specilic antagonists (CP 96,345 -5000 pmol; MEN 10,376 and R 396 -650 pmol) 10 rain before iev application of SP and NKA (25 pmol), respectively. Blood pressure (BP) and heart rate (HR) were measured and behavioral activity, face washing (FW), hind limb grooming and biting (HGB) as well as wet dog shakes (WDS) were recorded over 30 toni period atter peptide injection. The cardiovaseulur responses and FW to icv SP but not to NKA were attenuated by pretreatment with the NK-1 receptor antagonist, CP 96,345. The cardiovascular and behavioral effects of NKA (except of WDS) were inhibited by the icv pretreatment with the NK-2b receptor antagonist, R 396. The NK-2a antagonist, MEN 10,376, failed to affect the central cardiovascular and behavioral effeets of SP and NKA, We conclude: The cardiovascular and behavioral responses to iev SP are mediated by NK-1 receptors. In contrast, NKA in the brain interacts with another type of taehykinin receptor not identical with NK-1 receptors, which may belong to the NK-2b receptor subclass. Our findings provide pharmacological evidence for the presence of functionally aaive NK-2 receptors in the rat brain.

2Dept. Physiol., Univ. Montreal, Montreal, Canada and 3Dept. Pharmacol., Univ. Sherbrooke, Sherbrooke, Canada. 1Department of Pharmacology, University of Heidelberg, Im Neuenheimer Feld 366, 6900 Heidelberg, Germany 520

A.Rex, H,Flnk and C.A.Marsden"

The elevated plus maze is a well established anlmal model of anxiety. It is considered a valid and reliable behavloural model for detecting "anxlogenlo"/"anxlolytic" effects of suPstances acting at the benzodlazeplne/GABAreceptor. Testing serotonerglc drugs does not dlsplay such homogenous effects.

Substances affecting the dopamlnerglc system, are rarely tested.

The effects of a variety of acutely admlnlstered drugs, acting at the dopamlne receptors, were examined in the elevated plus maze in comparison to dlazepam as an "anxlolytic" and mCPP as an "anxlogenic" reference-compound. Dlazepam Kindling, a phenomenon involving the progressive development of seizures following the repeated administration of an initially subconvulsive stimulus provides an appropriate animal model for studying epilepsy and neuronal plasticity.

Recently, we could demonstrate a deficit in pentylenetetrazol (PTZ)-kindied rats to acquire an active avoidance reaction in a two-way shuttle-box. We tested a competitive (CGP 43 487) and a non-competitive NMDA-receptor antagonist in this model. The substances were injected during the course of kindling development before each P-I-Z (45 mg/kg i.p.) injection. Both antagonists potently inhibit the development of kindled seizures. They also inhibit fully kindled seizure activity. However, the learning deficit was prevented by treatment with MK 801 only, but no effect was found using the competitive antagonist (CGP 43 483 One candidate for the neuronal mechanism of kindling is the facilitation of excitatory synaptic transmission and the obviously participation of N-methyI-D-aspartate (NMDA) receptors in synaptic transmission following chemical kindling, Rats were kindled by repeated application of pentylenetetrazol (PTZ; 45 mg/kg i.p., 10 times for 20 days) or in combination with either MK 801 a noncompetetive antagonist of NMDA receptor or the competitive antagonist CGP 43 487 given before each P'f-Z injection. PTZ kindling is associated with the development of potentiation effects in the field potentials evoked in the dentate area of the hippocampus by single test stimuli in the perforant pathway (pp). In fully kindled rats test dosis of PTZ induced a long-lasting potentiation of the population spike, which lasted for at least 24 hours. Control animals showed an increase of the population spike amplitude only immediately after P r z injection. Long-term potentiation (LTP) induced by tetanic stimulation oft.he pp is blocked by MK 801. In PTZ kindled animals blockade of NMDA receptor functions during the kindling procedure prevents also the development of long-lasting potentiation effects in the dentate area but had different effects on the deficit in the shuttle box learning. The specific binding of (3H)-L-glutamate to its receptors was investigated on crude synaptic membrane preparations from hippocampus of pemylenetetrazol (P'TZ)-kindled rats as well as from kindled rats pratreated with diazepam or CGP 43 487 using binding assay technique on glass-filters. PTZ-kindling induced by 10 intraperitoneal (i.p.) applications of 45 m~/kg over a period of 20 days resulted in a significant increase of both tile convulsive susceptibility of animals to the convulsant and the specific (3H)-L-glutamate binding in hippocampus. Kinetic studies indicated that the enhanced (3H)-Lglutamate binding to hippocampai membrer~s from kindled rats reflects changes in the density of the glutamate binding sites since we observed an increased in Bma x by about + 50% (p<0.05) whilst the KD-value remained unchanged. Acute generalized convulsions induced by application of 60 mg/kg P'1-Z did not change the specific amino acid binding. It is concluded that the increase in glutamate receptor density of P'TZ-kindled rats may by the expression of a specific enhancement of susceptibility of glutarnatergic systems to this excitatory amino acid developing in the course of formation of PTZ-induced kindling. The protreatment of kindled rats with the anticonvulsant drugs was followed by different effects on (3H)-L-glutamate binding. Whereas diazepam was not able to alter the P'IZ kindling induced increase of the amino acid binding, the competetive NMDA receptor antagonist CGP 43 487 blocked this increase by more than 60%. Aggregate cell cultures derived from fetal rat brain and cultivated in well defined serum free medium has been characterized using various biochemical, morphological and electrophysiological techniques. Validity of the use of such cultures for in vitro studies of various physiological, pharmacological and toxicological ph~n has also been demonstrated. Some experimental observations suggesting the usefulness of such cultures for the pharmacological studies directed towards understa~ling excitatory amino acid induced pathological processes and for in vitro screening of potential neuroprotective agents will be presented. Fetal rat brain aggregated cell culture of different ages (in vitro) were incubated with 1 mM glutamic acid or with the excitatory amino acid (EAA) ligand N-methyl-D-aspartic acid (h~MDA) for varying periods and the effects of such treatment on lactate dehydrogenase (LDH) release and on various cellular marker enzymes were studied. Both amino acids tested showed dose and time dependent neurotoxic effects as judged by LDH-release and also decreased the glutamate decarboxylase contents of the aggregates. ~3qe specific excitotoxic effects of the amino acids were not only found to be dependent on the age of aggregates in the culture but also on the magnesium and glycin content of the medium. Further, after coincubation of the cultures with 1 pM dizocilpine (MK-801) and i00 pM h~3A no excitotoxic effects of the amino acids were observable. Such findings clearly indicate that useful pharmacological models for the study of excitatory amino acid induced r~urotoxicity can be developed using aggregate cell cultures. Efforts ere, therefore, now being made in our laboratories to define more precisely the various excitatory reoeptor systems present in such cultures. Therapy of cerebral ischemia usually starts hours after the ischemic event.

Therefore, we examined at what time after glutamate intoxication the administration of nimodipine, dizocilpine or NBQX (2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo(F) quinoxaline) still causes a neuroprotective effect. We used primary cultures of hipposampal neurons from newborn rats.

After 14 days in culture the neurons were exposed to imM glutamate for 30 min. Viability was assessed by using the trypan blue excluding method 18 hours after intoxication. Exposure of hippocampal cultures to 1 mM glutamate for 30 min resulted in a decreased number of viable, trypanblue-unstained neurons. Compared to control cultures treatment with nimodipine, dizocilpine or NBQX 30 min before or immediately after intoxication showed a significantly (p<0.001) increased number of viable neurons, but only the dizocilpine-treated cultures reached the control level of viability.

When the drugs were added 30 or 60 min after intoxication, they revealed a slightly decreased but a still statistically significant effect (p<0.001).

Administered 4 h after glutamate incubation only NBQX (p<0.01) and nimodipine (p<0.05) were capable of protecting the neurons against damage. No effect was demonstrable when the drugs were added 6 hours after glutamate intoxication, Summarizing, it was demonstrated that the drugs used were capable of protecting cultured hippocampal neurons against damage when they were added at the latest 4 h after glutamate intoxication. By means of the push-pull technique, the release of GABA, taurine and glutamate was studied in the posterior hypothalamus (PH) of the conscious rat. The outflow of the three amino acids was investigated dudng basal conditions, as well as in response to centrally applied drugs or to peripherally induced blood pressure changes. After an initial exponential decline, constant release rates of GA6A, taudne and glutamate were reached after 40-50 rain of superfusion and lasted at least 6h (maximum duration investigated). Fluctuations in the release rates point to the existence of ultradian rhythms. Superfusion of the PH with a potassiumrich (50 or 90 mmol/I) artificial cerebrospinal fluid enhanced the release of GABA (30 or 100 fold), glutamate (6 or 24 fold) and taurine (2.5 or 4 fold) in a concentration dependent way. Supeffusion with veratridine (1 or 10 pmol/I) had a similar effect. Superfusion with tetrodotoxin ( I/Jmol/I) decreased the release of GABA and taudne (approximately -40%) and, less pronounced but still significantly, the glutamate outflow. A dee in blood pressure (45 mm Hg) elicited by intravenous infusion of noradrenaline was associated with an increased release rate of GABA and taurine in the PH. The outflow of glutamate was not altered. Conversely, a fall of blood pressure (30 mm Hg) caused by intravenous infusion of nitroprusside diminished GABA and taurine outflow in the PH. On the other hand, the release rate of glutamate was enhanced by the hypotension. The results suggest that, in vivo, at least part of the GABA, taurine and glutamate output originates from hypothalamic neurons. The alterations in the release rates of GABA and taurine in response to experimentally induced blood pressure changes indicate that these amino acids play an important role as neurotransmitters with a hypotensive function in the PH of the rat. On the other hand, glutamate in the PH seems to act hypertensive in response to a lowered blood pressure. Male rats, treated with apomorphine (APO; 0,6-5 mg/kg s.c.), showed individually different motility patterns, each expressed by oral stereotyped behaviour and enhanced locomotor activation and reproduciable in a second test 4 days later (Havemann et al., Psychopharm. 90, 1986) . The rats were injected with 2 mg/kg APO s.c. and tested in an Animex-Motility-Meter. Part of the rats predominantly showed stereotyped sniffing with an increased locomotor activation, S(L,G)-rats. The other part could be classified as mainly licking and gnawing rats, L(S,G)-rats and G(L,S)-rats, with less increase of locomotor activation. In literature, clozapine antagonizes the locomotor activity but not the stereotyped behaviour in rats. Therefore, in this study the different types of rats were treated with the atypical neuroleptic clozapine (CLO; 10 and 15 mg/kg i.p.) in presence of APO. The locomotor activation of the S(L,G)-and the G{L,S)-rats was completely antagonized by 10 and 15 mg/kg CLO, whereas the stereotyped behaviour of these rats was not antagonized but showed a characteristic shift: the predominant stereotypy in both types of rats, quantified by a special scoring system, changed from sniffing (S(L,G)-rats) and gnawing (G(L,S)-rats) to licking. In the L(S,G)-rats the predominant licking stereotypy was not changed by 10 and 15 mg/kg CLO and even the locomotor activity of these rats was affected by 16 mg/kg CLO only. In contrast classical neuroleptics, e.g. haloperidol (HAL; 0,2 and 0,4 mg/kg i.p.), had a clear dose dependent antagonizing effect on Apomorphine-induced stereotypies: a pronounced effect on L(S,G)-and G(L,S)rats and less on S(L,G)-rats. Pretesting of rats with APO probably points to a preexisting different sensitivity of the rats to dopam]nergic stimulation and inhibition. The S(L,G)-rats seemed to be predominantly "mesolimbic active" rats, the L(S,G)-rats more "nigrostdatal active" rats and the G(L,S)-rats seemed to be most "nigrostriatal active". CLO, in contrast to HAL, might have a modulatory effect on the individual different degree of stimulated nigrostriatal and/or mesolimbic dopaminergic systern.

Psychiatric Hospital of the University of G6ttingen, von-Siebold-Str, 5, D-3400 GSttingen The hypothalamus possesses the highest histaminergic innervation of the whole brain. Cholinergic fibers are also present in this brain region. Both neurotransmitters seem to be involved in hypothalamic regulation of autonomic functions. By means of the push-pull technique, we investigated the influence of cholinergic neurons on the in rive release of histamine in the hypothalamus. Since brain histamine is present in neuronal and non-neuronal sites, the odgin of histamine released in the supedusate was also studied. Superfusete was collected continuously in time pedods of 10 min. The anterior hypothalamus was superfused with drugs for 10 rain. Histamine was determined by HPLC with fluorimetric detection. Superfusion with compound 48/80 did not increase histamine release rate, indicating that mast cells do not contribute to the release of histamine in the antedor hypothalamus. Superfusion with the H 3 receptor antagonist thioperamide (10 pmol/I) increased, whereas the H3 receptor agonist R-c~-methylhistamine (10 pmol/min) decreased the release rate of the amine. These effects show that, under the experimental conditions used, basal release of histamine is mainly of neuronal origin and that histamine outflow is modulated by H3 autoreceptors. The acetylcholine receptor antagonist atropine (5-10 .umol/I) caused a concentration-dependent increase in the release rate. Thus, in the anterior hypothalamus, histamine release is modulated by muscarinic acetylcholine receptors, probably located on histaminergic neurons. At least in the hypothalamus of the conscious rat, acetylcholine released from its neurons reaches these receptors and inhibits tonically histamine release. To further characterize the muscarinic receptors involved in the release of histamine, experiments with specific agonists and antagonists are in progress. The kappa opioid receptor agonist U50,488H (trans-(+)-3,4-dichloro-N-methyl-[2-(1-pyrrolindinyl)-cyclohexyl]-benzeneacetamide) has recently been reported to induce dystonia after s.c. administration of 5-10 mg/kg in guinea pigs. The dystonic movements observed in response to U50,488H resembled those previously reported to occur spontaneously or in response to mild environmental stimuli in a mutant hamster model of paroxysmal generalized dystonia. This prompted us to study the effects of opioid receptor antagonists and of U50,488H in the mutant hamster model. Naloxone and naltrexone, 1 and 10 mg/kg i.p., were either ineffective or tended to induce prodystonic effects in the mutant hamsters. In contrast, U50,488H markedly reduced the severity of dystonic movements at doses of 1-10 mg/kg s.c. In non-dystonic hamsters, U50,488H produced reduction of locomotor activity but no dystonic-like symptoms. The data in mutant hamsters demonstrate that stimulation of kappa opioid receptor is a powerful means of attenuating dystonic movements in a genetic animal model of idiopathic dystonia. The antidystonic effects of U50,488H might relate to interactions between kappa receptors and dopaminergic and glutamatergic neurotransmission.

Department of Pharmacology, Toxicology and Pharmacy, School of Veterinary Medicine, Bfinteweg 17, W-3000 Hannover 71, Germany Recently, nitdc oxide has been found to be released from neurons upon activation of glutamate receptors and to act as a neuronal messenger in the brain. A substantial role of the cofactor tetrahydrobiopterin in nitdc oxide synthesis in the brain has also been proposed. Furthermore, tetrahydmbiopterin was shown to increase the release of various neumtransmittere by a still unidentified mechanism. Now the influence of nitric oxide and of the tetrahydmbiopterin precursor sepiaptedn on acetylcholine release in the ventral striatum was studied. The ventral striatum of conscious, freely moving rats was superfused through a push-pull cannula with artificial CSF containing 1 ,umol/~ neostigmine . Acetylcholine in the superfusate was determined by HPLC combined with electrochemical detection. The experimental strategies included inhibition of nitric oxide synthase by superfusion with the irreversible inhibitor NG-nitro-Larginine, the nitric oxide donor SIN-l, the inhibitor of guanylate cyclase methylene blue and the tetrahydmbiopterin precursor sepiaptedn. Superfusion with NG-nitro-L-arginine (100 pmol/I) caused a substantial decrease in activity of nitric oxide synthase of the striatum and decreased the release rate of asetylcholine by about 40% of basel release. Supedusion with SIN-1 (200 ,umol/I) increased the acetylcholine release. Presuperfusion with methylene blue abolished the effect of SIN-1 indicating that nitric oxide, produced from decomposition of SIN-l, acts via cGMP. Hence, it seems that in the ventral striatum nitric oxide is synthesized continuously and stimulates the release of acetylcboline from its neurons. Superfusion with the tetrahydrobiopterin precursor sepiapterin also increased the acetylcholine release. This effect was not prevented by presuperfusion with NG-nitro-L-arginine. Thus, sepiaptedn increases acetylcholine release via a mechanism independent of nitdc oxide synthesis. Dextromethorphan is a widely used nonopioid antitussive which has been reported to show anticonvulsant activity against maximal electroshock and _ran_ ygdala-kindled seizure~. D~xtro-metho_rp_h_an and his metabolite dextrorphan are potent inhibitors of NMDA-induced convulsions and there is some evidence that interaction with the NMDA receptor complex may contribute in part to their pharmacological action. In fully-kindled rats, a chroni~ model for focal seizures with secondary generalization, only relatively high doses of dextromethorphan which lead to side-effects were reported to reduce the seizure stage significantly. In the presen~ study the ability of dextromethorphan and his metabollte dextrorphan to increase the threshold-for induction of afterdisehar_ges (ADT) was tested in fully-kindled rats: Dextrometho~lian in doses from 7.5 to 30 mg/kg i.p. increasea the ADT sig __" ~¢antly, whereas dextrorphan was only active at 30 mg/kg. After 7.5 to 15 mg/kg dextromethorphan the rats showed slight or no side-effects at all, 30 mg/kg lead to spontaneous seizdre activity in some rats. The adverse effects (fiminly motor impairment) occ_urrin~g after dextrorphan were more pronounced. Thirty mg/,k K ~.p. induced spontaneous seizure activity in some animals sin~lar to dextrome_tliorphan.

. We studied the behavioural effects of both drugs in kindled ano non-kindled rats because in recent experimenfs we observed a greater sensitivity of kindled rats to side-effects of NMDA antagonists. There were only slight differences at.30 .mg/kg dextromethorphan and dextrorphan between kindleo ano nonkindled rats. The data suggest that particularly dextromethorphan mig~. t be a potential anticonvulsant for the treatment of. drug-res'_~'tent partial epilepsy with secon ~dary generalization o.eeause of the significant e~f~cts against kindled seizures in ooses without prononnced side-effeCts.

Dept. of Pharmacol., Toxicol., and Pharmacy, School of Veterinary Medicine, Biinteweg 17, D-3000 Hannover 71

The novel anticonvulsant drug ucb L059 ([(S)-(~-ethyl-2-oxo-l-pyrrolidlneacetamide], an ethyl analogue of the nootropic drug piracetam, was evaluated in several rodent models of partial and generalized seizures. Ucb L059 (27-108 mg/kg i.p.) increased the thresholds for tonic electroconvulsions and myoclonic and clonic seizures induced by timed intravenous infusion of pentylenetetrazol (PTZ), but was ineffective in the traditional maximal electroshock seizure and s.c. PTZ seizure tests in mice and rats up to doses of 500 mg/kg, Anticonvulsant potency of ucb L059 in seizure threshold tests was similar to standard drugs, such as vaiproate. In anaygdaiakindled rats, ucb L059 exerted potent anticonvulsant activity against both focal and secondarily generalized seizures at doses of 13-108 mg/kg. Adverse effects of ucb L059 were quantitated in the open field and by standard tests for motor impairment, such as the rotarod and chimney tests. Ucb L059 exerted only minimal effects on behaviour, e.g. slight hyperactivity, and was devoid of muscle impairing activity in the rotarod test at doses up to 1700 mg/kg i.p. The data indicate that ucb L059 is an interesting new anticonvulsant agent with a broad spectrttm of activity and high therapeutic index. vitro. However, primary neurons are very rarely, if at all infected by HIV-I. The finding that exposure of rat cultured neurons to gpl20 leads to cell death was unexpected. It has been suggested that this is the result of a gpl20-induced increase in intracellular free calcium. After incubation of rat cortical cell cultures with the human immunodeficiency virus type 1 (HIV-I) coat protein gpl20 for 12 h, cells showed fragmentation of DNA at internucleosomal linkers, the characteristic feature of apoptosis. Simultaneously, the percentage of viable cells decreased from 94% to 33%. Memantine (l-amino-3,5-dimethyladamantane), a drug currently used in the therapy of spasticity and Parkinson's disease as well as the NMDA antagonist MK-801 prevented the effects of gpl20 on cortical cell cultures at micromolar concentrations. In human cultured astrocytes, gpl20 was ineffective with respect to DNA fragmentation and cell toxicity. The cytoprotective effect of memantine in cortical cell cultures may qualify the drug for the treatment of AIDS-related dementia. § Institut for Physiologische Chemie, Abteilung "Angewandte~ Molekularbiologie", Universit~t, ~uesbergweg 6, D-6500 Mainz, Germany;

Max-Planck-lnstitut for Hirnforschung, Deutsch-ordenstraBe 46, D-6000 Frankfurt (M) 71, Germany; * Division of AIDS Virus, AIDS Research Center, National

Institute of Health, Musashimurayama, Tokyo 208, Japan. The substantia nigra is in current discussion to play a critical role in the propagation of epileptic seizures. Several autors reported that the enhancement of GABAergic transmission in the substantia nigra pars reticulata (SNR) can suppress the motor manifestation of experimental induced seizures in animals (c.f. Bonhaus et al., Brain Res. 545, 41-48, 1991) . We were interested in the effects of the anticonvulsant agent valproic acid (VPA) and its metabolite trans-2-en-VPA on the activity of GABAergic SNR neurons. The experiments were performed in female wistar rats under chloral hydrate (CH, 360 mg/kg i.p.) anesthesia. The rats were ventilated and carbon dioxide in expiratory air, heart rate, blood pressure, and rectal temperature were continuoulsy monitored to assure stability of the anesthesia. To maintain the anesthetic state, CH was continuously infused at a rate of 140-160 mg/kgtk The spontaneous firing rate of one identified GABAergic neuron in the SNR was monitored over 10 minutes after stabilization and for further 40-50 rain after i.v. application of 50, 100 or 200 mg/kg VPA. Each dose was tested in a group of 7-10 animals, one group treated with saline served as control. The mean basic firing rate of SNR neurons was 18.8+6.2 Hz. VPA suppressed the activity in a dose dependent manner, yielding a maximal depression of 8% with 50 mg/kg, 14% with 100 mg/kg, and 32% with 200 mg/kg. The neurons recovered slowly over the time but the depression lasted longer than 50 rain and was significant at 100 and 200 mg/kg. In addition, valproic acid deepened the anesthesia in a dose dependent manner, To maintain a stabil anesthetic state the infusion rate of CH had to be reduced to 86 mg/kg/h after application of 50 mg/kg VPA and to 70 mg/kg/h after 100 and 200 mg/kg VPA. During injection of VPA over the period of 2 rain, but not during the application of saline, a depression of blood pressure and heart rate was evident but disappeared within 3 rain after termination of the application. To compare the effect of VPA with its major The intention of the antiepileptic therapy is to decrease the incidence and severity of spontaneous seizures. But epilepsy as well as long-term therapy with antiepileptics can cause disturbances of cognitive function. Therefore substances with an association of anticonvulsant and cognition-enhancing potency are of interest for new ways in the treatment of epilepsy. Awake and free-moving adult femal Wistar rats with chronic implantated electrodes (sensomotoric cortex) were pretreated with a nootropic drug (piracetam [PIR, 300 mg/kg body weight] or meclofenoxate [MEC, 300 mg/kg]), with an antiepileptic drug (carbamazepine [CARB, 10 mg/kg], clonazepam [CLON, 0.06 mg/kg] or phenobarbital [PB, 20 mg/kg]) or with a combination of one neotropic and one antiepileptic drug (PIR+CARB; PIR+CLON, PIR+PB, MEC+CARB, MEC+CLON, MEC+PB) 30 minutes before intraperitoneal PTZ administration. Electroencephalogram (EEG) and behaviour were investigated. The pretreatment with PIR, PIR+CLON or MEC+CLON showed a prolongation of latency up to the first occurance of petit real or absence like EEG-patterns induced by 20 mg/kg body weight PTZ. But the PTZ-effect was increased br unchanged after the other pretreatments. PB, CLON and CARB were very effective in the suppression of tonic-clonic seizures induced by 40 mg/kg body weight PTZ. PIR and MEC increased the antiepileptic effect of PB or CARB by prolongation of latency up to the first occurance of spike waves and sharp waves. There is growing evidence that carbamazepine has antimanic, antidepressant and prophylactic efficacy in manic-depressiv illness. However, the mechanism of action of carbamazepine underlying its therapeutic efficacy is still unclear. Many recent investigations studied changes of second messenger systems in man under carbamazepine treatment. To investigate these systems in man it is necessary to use peripheral tissues as model systems for negronal signal transduction. In a previous report we could demonstrate that carbamazepine inhibits cyclic guanosinemonophosphate (cGMP) accumulation in human lymphocytes (Schubert et al., Psychopharmacology 104, 45-50, 1991 ) . The intracellular calcium concentration ([Ca2+]) is also a central event in sgnal transduction, t s known that a varety of classcal Ca 2+ channel antagonists like verapamil or diltiazem seem to inhibit Ca 2+ mobilization after mitogen stimulation in human lymphocytes. Furthermore, using patch clamp techniques DeCoursey et al. (J. of Neuroimmunol, 10, 71-95, 1985) showed that these substances block K + channels and that the inhibition of mitogenesis parallels K + channel block. Findings of Cai et al. (Eur. J. of Pharmacol. 21_~9, 53-57, 1992) suggest that carbamazepine blocks the elevation of [Ca z +]~ induced by membrane depolarisation in a similar manner as classical Ca z + channel blockers involving voltage-dependent Ca 2 + channels in neuronal culture, To extend these findings to human lymphocytes and to another neuronal cell system we investigate intracellular Ca 2+ mobilization on two levels: 1. in peripheral human lymphocytes as a possible model system in man and 2. in dissociated mouse brain cells as neuronal system. Intracellular Ca 2+ levels are measured using the fura2 method and are calculated from the ratio of two excitation wavelengths (340 and 380 nm) according to the method of Grynkiewicz et al. (J. of Biol. Chem. 260, 3440-3450, 1985) .

Our preliminary findings confirm that carbamazepine at therapeutical concentrations (10 rag/I) has an inhibitory effect on phytohemagglutinine- . As phospholipase (PL) C-mediated PI-hydrolysis is known to supply the protein kinase C activator DAG, we investigated the possibility, whether preceding stimulation of one of the receptors coupled to PI-metabolism affects inositolphosphate (IP) formation induced by another receptor. Except of a moderate increase of M 1receptor mediated IP-formation as a consequence of preceding a 1receptor stimulation, we could not find evidence for interactions between the different signal transduction pathways. These results suggest that P~-, adrenergic a 1-and muscarimc Ml-receptors initiating the hydrolysis-of PI are acting on separate lipid'pools and generate second messengers wich do not influence the operation of other PImobilizing pathways in astrocytes. Cholera toxin (CTX; 200 ng/ml; 6 hours) stimulates PI-hydrolysis mediated by P?-, al-or Ml-receptors and increases iafracellular cAMP levels. Stimulation bf the astroghal g-receptors by 10" ~ M isoproterenol led to increased formation of cAMP without influencing ATP-, NA-or CCh-evoked PI-metabolism. CTX-induced increases of P2", al" and Ml-receptor mediated IP-breakdown must therefore be explained in terms of G-protein activation and cannot be due to elevated cAMP levels. Prestimulation of the astrocytes by ATP or CCh did not affect greceptor mediated increases of intracellular cAMP. This part of results shows the failure of interactions between PLC-mediated and adenylate cyclase-mediated signalling processes in astrocytes. Free radicals seem to take part in many aspects of brain aging. We have investigated a-lipoic acid (thioctic acid), a lipophilic radical scavenger passing the blood brain barrier to test the hypothesis that protection against free radicals might ameliorate age-related cognitive deficits. Young (3 months) and aged (23 months) female NMRI mice were treated orally with i00 mg ~-lipoic acid/Mg body weight once daily for 15 days. On the last day of treatment animals were placed in an open field for 3 minutes to measure horizontal activities. Open field exposure was repeated after 15 minutes and 24 hours. Habituation as a form of simple non-associative learning is seen as an effect of repeated measures (RM) in ~hNOVA. Any effect on learning is expressed as an interaction between RM and the intervention. Untreated aged mice showed long term memory deficits expressed by an interaction between RM and age combined with 24 hour activities close to intial ones. Treatment with thioctic acid improved long term memory in aged animals. There was a highly significant interaction (p<0.O001) between RM and treatment with 24 hour activities close to 15 minute ones. No such effect was evident in young animals. Deficits in NMDA receptor density in aged female NMRI mice were ameliorated without any effect on muscarine, benzodiazepine, and alpha2-adrenergic receptors. Our previous studies demonstrated that interleukin-1 (IL-1) is a potent activator of NGF secretion from rat neonatal cortical astrocytes maintained in primary culture (~arman-Kr~an M. et.al. (1991) J. Neurochem. 56:636) . In the present study we have used IL-1 receptor antagonlst,+(IL-lra), in addition to activators and inhibitors of Ca = -dependent intracellular signalling systems to elucidate the mechanisms whereby IL-1 stimulates NGF secretion. The results show 2+ that IL-1 stimulation of NGF secretion is Ca dependent. Treatment of cells with phospholipase A 2 inhibitor mepacrine (30 t~i) inhibits basal and IL-1 (10 U/ml) stimulated NGF secretion. Indomethacin (10 1~1), a cyclooxygenase inhibitor, produced a slight increase in basal NGF secretion, but failed to inhibit NGF secretion stimulated by IL-1. In contrast, treatment of cells with nordihydroguaiaretic acid (NDGA), a lipoxygenase inhibitor, blocked in a concentration-dependent manner IL-1 stimulation of NGF secretion. Recombinant human IL-lra inhibited IL-1 stimulated NGF secretion.

The results suggest that o~e mechanism through which IL-1 stimulates NGF secretion from astrocytes involves IL-1 receptor mediated activation of the phospholipase A 2lipoxygenase pathway.

Department of Pharmacology, Medical Faculty, Korytkova 2, 61000 Ljubljana, Slovenia Our previous studies have shown that intracerebroventricular (i.c.v.) administration of endothelin-1 (ET-1) produces antinociception as measured by hot plate test in mice (Nikolov et al., 1992, Meth Find Exptl Clin Pharmacol 14: 229) . Now we report for an antinociceptive effect of ET-1 and ET-3 in two other antinociceptive tests: tail flick latency and acetic acid-induced writhing in mice. After determination of the baseline response latencies, mice were injected i.c.v, with 1.25, 2.5 and 5pmol/mouse of ET-1 and with 2.5, 5 and 10 pmol/mouse of ET-3 in a volume of 0.01 ml. Tail flick latencies were reassessed 15, 30, 60, 90 and 120 min later. In the writhing test mice were injected i.c.v. with 0.5, 1, 2.5 and 5 pmol/mouse of ET-1 or 1.25, 2.5, 5, and 10 pmol/mouse of ET-3, and 15 rain later they received 1% v/v acetic acid (0.1 ml/10 g b.w.i.p.). The number of writhings was counted for a period of 20 rain starting after the acetic acid administration. I,c.v. administration of ET-1 and ET-3 led to a statistically significant increase of the tail flick latency, Maximum effect was observed at the 30th min by 5 pmol/mouse of ET-1 (54.5%, p<0.05), and b~¢ 10 pmol/mouse of ET-3 (48.2%, p< 0.05). The duration or the effect was about 120 rain.The EDna for ET-1 in the writhing test was 2.2 (1.5-2.9) pm~l'/mouse and for ET-3 a maximum effect of 45.2% was achieved at 5 pmol/mouse. Naloxone (1 mg/kg i.p.) did not antagonize significantly the ET-1 and ET-3-induced antinocicepfion in the tail flick test. The present results support our previous data on an antinocieeptive effect of ETs after central administration, not mediated by naloxone-sensitive receptors. Thiamine (vitamin El) is well known as a coenzyme in glucose metabolism. However, evidence from a number of electrophysiological studies suggests that the di-and triphosphates (TDP, TTP) of thiamine might play an additional role in regulating nervous excitation (for review see Cooper and Pincus, 1979) . Using whole-cell voltage clamp recordings we studied the effect of intracellularly applied TDP (5 raM) on the run-down of Na* and K + currents (In,, I~ in response to repetitive electrical stimulation applied to cultured primary central neurons. Cells were stimulated every 2 minutes to up to one hour by polarising the holding potential (-60 mV) briefly (50 ms) to values more positive and more negative than the holding potential to elicit Na + and K ÷ currents. The amplitudes of the peak Na + inward and the steady-state delayed rectifier outward current in response to each individual stimulus sequence were measured for n=8 control and n=8 TDP-treated cells. In control cells, both, INs and IK showed a timedependent decline (run-down) to 30% (Ir~) and 40% (It3 of the starting value after 60 rain recording time. In the presence of TDP the run-down for Ir~, was attenuated to 60% of the starting value after 40 rain, and 80% after 60 min recording time. For IK the run-down appeared only transient, and currents showed a tendency to recover to starting values. The effect became apparent after 5 rain recording time which might be explained by the conversion of TDP to the more potent TTP by a membrane-bound phosphorylase. Our results could be interpreted as art attenuation of exhaustion of cellular excitability in the presence of intracellular TDP. Endothelin-1 (ET-1) is a vasoactive peptide first isolated from porcine endothelial cells, with a strong and long-lasting vasoconstrictive action. Numerous studies have shown that ET-1 could evoke cerebral vasospasm and significant CBF reduction. However, no experiments have been performed to evaluate how ET-1 modifies the cerebral infarct area, which was the aim of the present study. We studied also the direct effect of ET-I on the cultured neurons under normoxic and hypoxic conditions in vitro. Focal cerebral ischemia was produced by permanent ligation of the middle cerebral artery (MCA) in mice. The experiments in vitro were performed on primary neuronal cultures of chick embryo cerebral hemispheres. Intracisternal injection of ET-1 (5 and 10 pmol/mouse, 10 min pretreatment time) augmented significantly the infarcted surface area (by 15.5%, p<0.05, and by 23.5%, p<0.001, respectively) in focal cerebral ischemia. ET-1 (0.01, 1, and 100 nM) addedto the primar), cultures of neurons for 1 h and 24 h did not influence the viability of the cells and the protein content of the cultures kept in normoxic conditions. When applied simultaneously with 1 mM sodium cyanide for 30 min, ET-1 (0.01, 1, and 100 nM) did not modify the hypoxiainduced changes. The present results show that exogenously applied ET-1 could exacerbate cerebral ischemia, due probably to its vasoconstrictive properties and not to a direct neurotoxic effect. Atracurinur, a peripheral muscle relaxant, is degraded in vivo by non-enzymatic Hofmann elimination and ester hydrolysis. One metabolite, laudanosine (N-methyltetrahydropapaverine), that is also an alcaloid from papaver somulferttm, elicits convulsant activity in several animal models. Clinical examinations demonstrated detectable amounts of laudanosine in the curebrospinal fluid after administration of atracurium. Therefore it was of interest, whether laudanosine could potentially disturb synaptic transmission. Here we report that landanosine induced paroxysmal depolarizing events (PDs) in cultured spinal cord neurons while its prodrug atracurium and papaverine were ineffective. -Electrophysiological experiments were performed in the whole-cell mode. Spontaneous synaptic activity recorded in current clamp consisted of randomly evoked inhibitory and excitatory potentials. The neurons developed PDs when laudanosine was applied. Depending on its concentratior~ these events became more and more organized, while their duration declined at higher euncentrations. -To elucidate the underlying mechanism the interaction of landanosine with chemoseusitive and voltage-dependent membrane currents was investigated using voltage clamp reeorcling. Glycine and GABA, resp., were applied to single neurons by pressure ejection which caused a strychnine-or bicacaffine-sensitive current. Ouly the glychae-evoked current was reduced by laudanosine (ECs0:35/~M). The GABA-evoked current was, if at all, only slightly affected. Of the voltage-dependent currents ouly the sodium current was decreased by landanosiae concantration-depandently, while the potassium current was unaffected. The PDs and the suppression of the chloride a~d sodium currents were fully reversible after withdrawal of laudanosine. We postulate that an interaction of laudanosine with the glycine-operated chloride channel is responsible for its convuisant effect and that the block of the sodium current that occurred at higher concentrations of landanosine eventually suppresses the PDs. -This examination also cotffirms the usefulness of neuronal cells in culture to investigate convuisant and anficonvulsant activity.

Medicaf School of Hannover, Institute of Toxlcology, 3000 Haunover 61, Germany. tUuiversity of Giessen, Rudolf-Buchheim-Institute, 6300 Giessen, Germany.

54O CHARACTERIZATION OF THE SELECTIVE AND REVERSIBLE M e N . -AMINE OXIDASE B INHIBrrOR LU 53439 IN VITRO AND IN V1VO

Under identical experimental conditions inhibitory potency and selectivity were more pronounced in compurison to selegilhle (IC~o for MAO B 7.8 nM, for MAO A 2000 aM) and lazabemide (22 nM, and > I0000 raM), respectively. No activity of LU 53439 on the reuptake of neurotrausmitters has been detected up to 1-0 "s molfl. No substantial affinity

Ex rive MAO inhibition was tested after oral admlni~tration of LU 53439 in mice

6 mg/kg) the concentrations of the hOur,transmitters dopamine iDA), serotonln and noradreanline and the MAOdependent metabolites 3,4-dihydroxyphenylaeetic add (DOPAC) and 5-hydroxyindolaeetle add (5-t-IIAA) were unaffected. However, after complete inhibition of MAO A by the selective and reverf~ble inhibitor esuprone (10 mg/kg p.o.), ~drnlnl. stration of LU 53439 (3.16 mg/kg p.o.) induced a further decrease in the content of DOPAC and 5-HIAA in the corpus striatum of rats. In rive LU 53439 (10 mg/kg i.p.) did not change the extraceaular levels of DA, DOPAC, homovanlllle add (HVA), 3-methoxytyramine (3-MT) and 5-HIAA in the striatum of anaesthetized rats measured by ruler,dialysis and differential normal pulse voltammetry (DNPV). However, if MAO A is inhibited by elorgyline (10 mg/kg i.p.) or esuprone (10 mg/kg i.p.), administration of LU 53439 (10 mg/kg i.p.) led -as in ex rive experiments -to a further decrease of 5-HIAA and HVA and a further increase of DA and 3-MT. The small size of carbon fiber alectrodes made the dorsal rxphe nucleus (DRN) accessible to in rive metabolic measurements. This region, whleh is responsible for the serotoncrgic imaervatiun of the corpus striatum, had been shown to contain a relatively high portion of MAO B

Acknowledgment: This research was started in the laboratory of Dr. Nell M. Nathanson who is gratefully acknowledged for his support.The nutritional intake of polyunsaturated fatty acids (FAs) was shown to have beneficial influences on many aspects of cardiovascular diseases. In this study we compared the potency of various oil supplements admininstered for 10 weeks to rats for preventing ischemia-induced ventricular tachycardia and fibrillation in the isolated perfused heart. Moreover, the contribution of the sympathetic system which might be involved in the observed antiarrhythrnic effects should be investigated. Wistar rats were fed a low fat standard diet, which was enriched with either coconut oil (saturated FAs), corn oil (0-6 FAs), linseed oil or fish oil (0-3 FAs) (10 % (w/w) each) for 10 weeks. Isolated perfused hearts were prepared from these rats and the arrhythmias occurring within 20 rain after occlusion of the left anterior descending coronary artery were recorded. As parameters for the sympathetic tone in hearts of identically treated rats the 13-adrenoceptors were characterized by binding studies with the [3t-specific antagonist [3H]-CGP 12177. Furthermore, cardiac homogenates were analyzed for catecholamine content and tyrosine hydroxylase activity by HPLC and electrochemical detection, and for cAMP by radioimmuno assay. The incidence of ventricular fibrillations was significantly reduced by the diet containing fish oil (10 %) compared to all other treatments (40 -75 %). Also the diets enriched with polyunsaturated FAs (corn oil, linseed oil) caused lower rates of arrhythmic events in comparison to the low fat or coconut oil diets. Taken together all groups, the values ranged for Bmax: 37.9-42.3 frnol/mgprot., for KD: 0.23 -0.28 nM, for noradrenaline: 1.1 -1.4 ng/mgProt. and for cAMP: 2.24 -2.47 pmol/mgprot. None of these parameters revealed a statistically significant difference between the groups. In conclusion, fatty acid enriched nutrition does not influence the sympathetic tone in cardiac muscle and therefore the observed antiarrhythmic activity induced by fish oil diet may be caused by a different mechanism.Institut fOr Pharmakologie, Medizinische Universitat zu LObeck, Ratzeburger Allee 160, D-2400 Lt~beck, *Institut ~ar Physiologie II, Universit~.t Tgbingen, Gmelinstr. 5, D-7400 TObingen

The muscarinic agonist WAL 2014 FU (l-Azabieyclo[2.2.2] octane,3-(2-propynyloxy)-,(R)-,(E)-2-butenedioate) has been selected as a candidate for cholinergie replacement therapy in Alzheimer°s disease because of high oral bioavailability, lack of metabolism and ability to penetrate the blood brain barrier readily.According to experiments in cell linns expressing human muscarinie receptor subtypes, in isolated organs, and in whole animals, WAL 2014 FU acts preferentially at Ml-reeeptors, and is a partial agonist at M2-and M3-reeeptors (Enslnger et al., 1993, Life Sci., in press) . We confirmed its lack of bronchospastic activity even at i.v. doses depressing spontaneous respiration (up to 128 mg/kg) in a modified Konzett-R6ssler preparation of the guinea pig which responded to arecoline with full bronehospasm.As these results did not reconcile with the constrictor effect of WAL 2014 FU in isolated guinea pig tracheal muscle (EC50 ~ 6.7 #M, intrinsic activity 0.63, carbachol ~ 1.00), we wondered whether the bronchospastie potency of WAL 2014 FU in whole animals might be masked by stimulation of the sympathetic nervous system via neuronal receptors. Indeed, after blockade of fi-adrenoceptors in the guinea pig with the nonselective B-blocker toliprolol (5 mg/kg i.v.), intravenous injection of WAL 2014 FU induced full bronchospasm with an EDS0 of 1.8 mg/kg. Toliprolol pretreatment shifted the EDS0 value of intravenous arecoline from 0.048 to 0.012 mg/kg.These results suggest that a 5-fold higher affinity for MI-over M3-receptors of a muscarinic agonist as assessed in isolated organs, combined with limited intrinsic M3-activity, suffices to prevent M3-mediated bronchospasm in guinea pigs probably by activation of the sympathetic system via Ml-receptors. A contribution to this sympathetic activation by the weak nicotinic action of WAL 2014 FU seems possible. The inlfibitiun by antagonists with selectivity for muscarinic receptor subtypes (M1, M2, M3) of field-stimulated (FS, postganglionic) and vagus nerve-stimulated (VS, pregangliunic) contractions of the rabbit main bronchus preparation (20 V, 0.3 ms, 7-20 Hz) was compared with their affinities (PA2) at MI-, M2-receptors in vas deferens (RVD) and at M3-receptors in traachea (RT) of tlie rabbif. At Fewer concentrations, M2-selective antagonists (AQ-RA 741 > himbacine = naethoctramine > > gallamine) first potentiated, but at concentrations 10-fold higher than their affinity at M3-receptors attenuated contraction toFS. M1-/M3-unselective antagonists, atropine, thiaziilamium, HHSiD exhibited nearly the same mhihitory potency (plC ) against FS and VS. However also M -/M -discrimi-.... 50 . . . ' c 1 .3 natmg antagonists, (+)-blpenden, telenzep~ne, plrenzepme, UH-AH 37 , consistently showed no preferential inhibition of the VS over the FS responses (IC~: 0 ratio FS:VS < 4):pIC50 Multiple muscarine receptor subtypes have been described in the rabbit airway tissue by binding studies and immunobiological methods (e.g. Drrje et al. Mol Pharmacol 40:459, 1991) . In several species, it has been shown that the sympathetic nerve endings in the airways are endowed with inhibitory muscarine receptors (e.g. Rack6 et al. Br J Pharmacol 107:3, 1992) . The aim of the present experiments was to characterize pharmacologically the muscarine receptor subtype involved in the inhibition of noradrenaline release in the rabbit trachea. Isolated rabbit tracheae were incubated in Krebs-HEPES medium (conraining 1 /imol/1 yohimbine, 10 #mol/1 tyrosine and 1 ~mol/1 desipramine) and the outflow of endogenous noradrenaline was determined by HPLC with electrochemical detection. Two periods of electrical field stimulation (S1, $2, 540 pulses at 3 Itz) were carried out. Electrical stimulation induced the release of about 30 pmol/g noradrenaline, corresponding to about 5 % of the tissue noradrenaline. Oxotremorine inhibited completely the evoked release of noradrenaline (IC5o: 82 nmol/1). In the presence of several subtype selective muscarine receptor antagonists, scopolamine ( 0.1 #reel/l); 5,11-dihydro-11-(((2-(2-dipropylamino-ethyl-1 -piperidinyl)ethyl)amino)-carbonyl) -6H-pyrido (2,3-b)(1,4) benzo-diazepin-6-on methansulfonate (AF-DX 384, 1 /~mol/1); p-fluoro-hexahydrosiladifenidol (p-FHI-ISiD, 1 and 10 #mol/l); methoctramine (1 /lmol/l) and pirenzepine (1 ~mol/1) the concentration response curve of oxotremorine was shifted to the right. The rank-order of potency (-log KB) was: sc.o. polamine (8.4) > AF-DX 384 (8,1) > > methoctramine (5.7) = pirenzeplne (5.7) > p-FHHSiD (5.3). Indomethacin (3 /~mol/l) did not affect the inhibitory effect of oxotremorine (ICs0:75 nmol/1) nor the antagonistic potency of methoctramine or p-FHHS1-D.In conclusion, the muscarinic inhibition of the release of noradrenaline from the rabbit isolated trachea is mediated by an atypical, M2-1ike muscarine receptor. The enterochromaffin cells (ECs) of several species are endowed with stimulatory muscarine receptors and a particularly marked effect of muscafine receptor agonists on the release of 5-HT is observed in the rabbit intestine (Reimann & Hering, this journal 345:Rl11, 1992) . The aim of the present experiments was to characterize the muscarine receptors on ECs in this species. Rabbits were killed by a blow to the back of the neck and exsanguinated. The intestine was arterially perfused (10 ml/min) in situ for 20 rain with Krebs-HEPES solution. Thereafter, ileal segments were removed, opened along the mesentedc border and incubated in Ussing-chambers. The medium (1 ml at the luminal and serosal side, respectively) was changed every 10 rain. 5-HT was determined by I-IPLC with electrochemical detection. Oxotremorine was added from the 50th min, antagonists and/or tetrodotoxin CI'FX) were present from the onset of incubation.In the absence of drugs, the spontaneous serosal outflow of 5-HT, determined between 40 -50 min of incubation, amounted to 1.5 ___ 0.1 pmol/10 min/cm 2 (mean + S.E.M,, n=67). Oxotremorine increased the outflow of 5-HT in a concentration dependent manner up to 23 __. 3 pmol/10 min/ 2 cm at 10/~M (ECs.0:0.5 ~M). In the presence of 1/~M TTX, the sponta-2 neous outflow was ~.7 + 0.1 pmol/10 min/cm (n=99) and oxotremorine increased it maximally to 23 _+ 4 pmol/10 min/cm 2 (ECs0:0.9 /~M). In the presence of TTX, several subtype selective muscarine receptor antagonists, para-fluoro-hexahydrosiladifenidol (p-FHHSiD, 0.1 and 1 ~M), pirenzepine (0.3 and 3 ~M), AQ-RA 741 (11-((4-(4-diethylamino)butyl)-1-piperidinyl)acetyl)-5,11-dihydro-6H-pyrido(2,3-b)(1,4) benzo-diazepin-6-on hydrochloride, 0.3 and 3/iM) and ethyl)amino)-carbonyl)-6H-pyrido (2,3-b)(1,4) benzo-diazepin-6-on methansulfonate, 3 ~M) shifted the concentration response curve of oxotremorine to the right. The rank-order of potency (-log Ka) was: p-FHHSiD (7.3) > pirenzepine (6.7) > AQ-RA 741 (6.5) = AFTDX 384 (6.4).In conclusion, 5-HT release from the rabbit ECs is stimulated via muscafine M3 receptors. The conclusions about a large receptor reserve at presynaptic ot2-adrenoceptors in rat neocortex according to the method of Furchgott (1966, Adv Drug Res 3:21-55) of a recent publication (Agneter et al. 1993, Br J Pharmacol, in press) were examined. In the case of exogenous noradrenaline (NA) as agoalst to induce inhibition of transmitter release activation af only 17% of the autoreceptors was sufficient for a maximal inhibition of NA release. The activation of 17% of all autoreceptors induces a maximum inhibition, when an all-or-nothing response to the activation of at least 1 of 25 assumed autoreceptors per terminal is given (binomial distribution). When, however, a graded inhibition of release per terminal exists, reflecting a high multiplicity of 6 autoreceptors per terminal, then the activation of 17% (~ •00/6) of the autoreceptors at each terminal corresponds to an occupation of 17% of all the autoreceptors in the tissue, Then, the mayamum inhibition per terminal reflects the overall inlaibition of release from all terminals. In the first case a slope of 0.787 in the point of half-maximum effect of the concentration-response curve can be computed. In the case of a graded response per terminal an even higher slope of 1.055 in the point of half-maximum effect of a highly non-symetrieal sigmoid is calculated. The slopes of the concentration-response curves of exogenous NA obtained before or after destruction of spare receptors were compared by means of nonlinear regression analysis to that of a bimoleeular binding sigmoid with a slope of 0.567 in its point of half-maximum effect, The confidence intervals of the so-called slope factor c, expected to increase from 1 (spare receptors abolished) to about 1.366 (=0.787/0.576) or 1.832 in case of a fully reestablished receptor reserve, never allowed to assume a c > 1. The congruence of the experimental concentration response curves with a binding sigmoid is a clear contradiction to the assumed receptor reserve and suggests a direct proportionality between receptor occupation and response. The aim of this study was to investigate the 13-adrenoceptors (13AR) in the myocardium from patients suffering from hypertrophic obstructive cardiomyopathy (HOCM). 13AR were determined in surgically excised septal myocardium. 13AR densi~ (Bmax) and 13AR affimty (Kd) were • 1251 measured in radioactive hgand binding assays w~th Icy P ( ).-Iodo ano indolol ICYP To discriminate M-and I]2-AR subtype~s ICYP was displaced with CGP 20712 A (1-(2-(3-¢arbarnoyl-4-hydroxy) phenoxyethylamino)-3-(4-(methyl-4-trifluoromethyl-2-imidazolyl) phenoxy)-2-propanol-methansulfonate).Epinephrine (E) and norepinephrine (NE) were determined electrochemically in the same tissue (tE, tNE) and in the plasma of the same patients (pE, pNE) after separation with high performance liquid chroraatograjohy. For comparison papillary muscles from patients undergoing matral valve replacement because of mitral regurgitation NYHA III (MR) were studied• In MR a down-regulation of 13AR density, probably due to elevated pNE, is well known. Results: Bmax=maximal ICYP-binding in fmol/mg protein; Kd=affinity in pmol/1; M=M-AR-subtype in % ot Bmax; pE and pNE = nmol/1 (reference: 0.16-0.46 and 1. 43-+3 16+-3 61-+8 0.9.+0.4 3.0.+0.5 2.5.+0.5 47_+6 gAR density was comparable reduced in HOCM and MR. There was also no significant difference in 131:132-ratio and 13AR affinity to ICYP in these tissues. Tissue catecholamines and pE did not show any difference in both groups, whereas pNE was elevated onl~v in MR, but significant lower and slightly below the normal range in HOCM. Conclusion: 13AR density is reduced in HOCM, which is not secondary due to elevated plasma norepinephrine. • The aim of this study was to investigate whether the effects of halothane on the sensitivity of myocardium to catecholamines are due to changes in the l~adrenoceptors (BAR) of the cardiomyocytes. Therefore, we isolated cardiomyocytes from adult rats by perfusing their hearts with collagenase (lmg/ml). An aliquot of cells was immediately taken for reference (Ref), the remaining cells of each heart were divided into a control (Ctr) and a halothane (Hal) group. Both groups were incubated at 37°C in Medium 199, containing additional 0.375 mmol/1 (0.64 vol%) halothane in Hal group. Aliquots were taken every hour up to 4 h and quickly frozen in liquid nitrogen. B A R were determined by a radioactive ligand binding 125~r assay with I-Iodocyanopindolol (ICYP). 131-and 132-AR-subtypes were discriminated by ICYP-displacement with CGP 20712 A (1-(2-(3carbamoyl-4-hydroxy)phenoxyethylamino)-3-(4-(methyl-4-trifluoror , methyl-2-imidazolyl)phenoxy)-2-propanol-methansulfonate). Na~"/K ~-ATPase activity (NaK) was used as a membrane marker (Jones et al. J Biol Chem (1979) The neuronal noradrenaline transporter which is responsible for the rapid inactivation of released noradrenaline (NA) is a primary target of tricyclic antidepressants.The gene for the human noradrenaline transporter (hNAT) is a potential candidate for depression, thus it was of interest to determine its chromosomal localization. Human genomic DNA was isolated from SKN-SH SY5Y cells, digested with the restriction enzymes Ncol (N), PstI (P), BamHI (B) and EcoRI(E), separated on agarose gel and transferred to nylon membranes. Hybridization of restriction enzyme-digested DNA with Digll-dUTP-labelled cDNA of the hNAT revealed a single signal at 5 kb with E-digested DNA, whereas 2, 3 and 4 signals were obtained with N-, P-and B-digested DNA, respectively. Since the cDNA of the hNAT has no cleavage sites for P and B, the multiple band pattern indicates intron-exon structure of the hNAT gene. Hybridization of a human chromosome mapping panel (from E-digested DNA of rodent/human somatic cell hybrids) with labelled hNAT cDNA revealed an assignement of the hNAT gene to chromosome 16. This localization was confirmed by fluorescent in situ hybridization of hNAT cDNA to human metaphase chromosomes; in addition, this analysis showed that the gene is localized to band q12. [519] [520] [521] [522] [523] [524] [525] [526] [527] [528] [529] [530] [531] 1991) . The aim of the present study was to investigate the effect of the selective uptake 1 inhibitor (+)-oxaprotiline (OXA) on the plasma kinetics o1 noradrenaline (NA) and AD in conscious as well as in anaesthetized rabbits.[SH]NA and [3H]AD were infused i.v. and plasma concentrations of endogenous and radiolabelled NA and AID were measured.Results obtained in conscious and anaesthetized rabbits were similar. OXA 0.2, 0.6 and 1.8 mg kg "1 i.v. dose-dependently reduced the clearance of [SHJNA from the plasma. The clearance of [3H]AD was only slightly reduced. The spillover of endogenous NA was decreased by up to 35 %. In contrast, the spillover of AD was, if anything, slightly enhanced, Prazosin 0.1 and 1 mg kg "t was injected i.v. in a second part of each experiment in order to examine the reason for the difference. It lowered the blood pressure to the same extent in animals that had received OXA and in those that had not. The prazosin-induced hypotension led to a marked increase in NA spillover, by up to 300 % in OXA-treated animals, but not to any increase in AD spillover, neither in unpretreated nor in OXA-pretreated animals.The results are compatible with the following hypothesis on central adrenergic regulation of sympathetic tone. Sympathetic outflow to most tissues is mainly subject to central cte-adrenergic depression with little influence from cqadrenoceptors, and uptake1 inhibitors depress sympathetic outflow to such tissues by enhancing the ~2-adrenergic depression (Szabo et al. 1991) . Sympathetic outflow to the adrenal medulla, in contrast, is subject to both central tz2-adrenergic depression and a marked central cq-adrenergic excitation. Uptake t inhibitors cause little change in AD release because they simultaneously reinforce the e~adrenergic depression and the ~h-adrenergic excitation. Prazosin produces a reflex increase in NA release but not the expected reflex increase in AD release because it blocks central al-adrenoceptors and, hence, removes the central tx 1adrenergic drive of sympathetic outflow to the adrenal medulla.Pharmakologisches Institut der Universit&t Hermann-Herder-Strasse 5, D-7800 Fre[burg i. Br., FRG [Ekas et al. (1982) Clin Sci 63:309], using the u2adrenoceptor agonist tramazoline, has suggested that prejunctional u2-adrenoceptors are supersensitive in SHR. The a~m of the present study was to reinvestigate prejunctional u2-adrenoceptor modulation of noradrenaline release in SHR (11-12 weeks) and age-matched WKY kidneys by using the selective u2-adrenoceptor agonist bromoxidine and the selective u2-adrenoceptor antagonist rauwolscine. The renal nerves were stimulated at 1 Hz for 30 s. The stimulation-induced (S-I) outflow of endogenous noradrenaline was measured by high pressure liquid chromatography with electrochemical detection.The S-I outflow of noradrenaline was significantly greater in SHR (200 ± 13 pg/g/min, n=39) than in WKY (138 ± 9 pg/g/min, n=47). Rauwolscine (0.I and 1.0 ~M) enhanced the S-I outflow of noradrenaline to a greater extent in SHR than in WKY. However, bromoxidine (0.01 -1.0 ~M) caused a similar maximal inhibition of the S-I outflow of noradrenaline of 90 % with a pEC50 of 7.49 (7.32 -7.68) and of 91% with a pEC50 of 7.46 (7.30 -7.74) in SHR and WKY, respectively. The results suggest that prejunctional u2-adrenoceptors of the isolated kidney are not supersensitive in SHR as compared to WKY. The more pronounced facilitatory effect of rauwolscine in SHR may simply result from an increased autoinhibition due to an enhanced release of noradrenaline. . In order to characterize presynaptic et 2autoreceptor of human kidney cortex, affinity estimates of 11 ct-adrenoceptor antagonists and of the ~2-adrenoceptor agonist oxymetazoline were determined. Slices of human kidney cortex were preincubated with 3H-noradrenaline and then superfused in presence of desipramine 1 ~M. Four periods of eleclrical stimulation were applied (60 pulses/1 Hz) and cumulative concentration-response curves for the a-adrenoceptor figands were obtained. All antagonists except (-)-miansedn and BRL41992 enhanced electrically evoked tritium overflow. Negative logarithms of concentrations that increased evoked overflow by 30 % were calculated as affinity estimates (pEC 30; Table) . Oxymetazoline inhibited the evoked overflow (EC50 = 20 nM). Partial receptor inactivation by o~-Antagonist pEC30 pretreatment with phenoxybenzamine 3 I.tM shifted the concentration.response curve to the right and diminished the maximal Rauwolscine 9.0 effect of oxymetazoline. From equieffective concentrations WB4101 7.9 without and with phenoxybenzamine pretreatmem a K A value Phentolamine 7.3 of 630 nM was calculated for oxymetazoline. Comparison of BRL44408 7,0 the observed affinity estimates with binding affinities from the (+)-Mianserin 6,9 literature indicates that the tx2-autoreceptor in human kidney ARC239 In contrast, there is only a poor correlation with ~2A, tX2B and ~X2D binding sites as well as with rat kidney ~t2-autoreceptors (r=0.78; 6 antagonists). Moreover, the low affinity of oxymetazoline also argues against U2A and ~2D properties. It is concluded that the ~z 2antoreceptor in human kidney is different from that in rat kidney and belongs to the C~2c subtype. The exocytotic noradrenaline (NA) release can be inhibited via o~ 2autoreceptors and putative imidazoline receptors in the rabbit perfused head. Pretreatment of the hearts with the c~ 2-and imidazoline receptor agonist aganodine 2 #tool/I, a guanidine derivative, desensitized both, the prejunctional imidazoline and the ~2-adrenergic receptors (Schwarz and Fuder NSAP 346:R31, 1992) . Attempting to desensitize the imidazoline receptors without affecting the sensitivity of the ~2-adrenoceptors, we now used as a tool exposure to the c~2-adrenoceptor antagonist and imidazoline receptor agonist BDF 6143 (4-chloro-2-[2-imidazoline-2-ylamino]isoindoline). Rabbit heads were perfused with Tyrode solution containing 3 pmol/I cocaine and corticosterone and 0.1 ~mol/I rauwolscine. Endogenous NA was released by stimulation of the right sympathetic nerves (0.66 Hz, 80 pulses). Pretreatment of the heads with BDF6143 (2 and 10 ~mol/I, 30 min, followed by washout of 80 min) decreased significantly the inhibitory effect on evoked NA overflow of aganodine 10 nmol/I mediated by imidazoline receptors (67.0+1.9% and 45.4_+2.'1% inhibition, n=4, compared to 75.4+1.0% in 4 control hearts). In the pretreated heads the inhibitory effect of the ~t2-agonist oxymetazoline 30 nmol/I mediated by incompletely blocked a2-adrenoceptors was also decreased significantly (51.4+1.6% and 36.5+5.8% inhibition, n=3-5, compared to 72.7+3.3% in 3 control heads). Upon BDF6143 pretreatment however, a decrease in oxymetazoline potency at the a2-adrenoceptors could be explained by a persisting o~ 2adrenoceptor antagonism of the drug which may not be easily washed out of the tissue. Supporting this view the concentration-inhibition curve of oxymetazoline was shifted to the dght bya factor of 12 after BDF 6'143 10#mol/I pretreatment (without BDF6143:IC50 11 nmol/I, maximum 90.7+2.3%, n=4; after BDF6143 exposure: IC50 134 nmol/I, maximum 67.2_+2.4%, n=4). At the present, no unequivocal condition can be presented under which a selective desensitization of prejunctional imidazoline receptors occurred without effects on the sensitivity of the ~2-autoreceptors. Among several hypothesis, Starke(Ann Rev Pharm Tox 21:7-30,198&) admitted the possibility that alpha--autoreceptors may represent vestiges of evolution that continue to exist because they do us no harm. On the other hand, vascular adrenergic nerves do not fully mature until well after birth and neuronal transmitter release is the last step in the developmental sequence (Suet al. Blood Vessels 14:12-24,1977) .This led us to look for the relevance of both alpha-receptor-mediated inhibition and beta-receptor-mediated facilitation of noradrenaline release in the canine saphenous vein of neonates, a tissue where both mechanisms are well represented in adult animals. The selective alpha2-agonist 5-bromo-6-(2-imidazolin-2-ylamino)quinoxaline (UK-14304) (I0-i000 nM) reduced and the selective alpha2-antagonist yohimbine (30-300 nM) enhanced the overflow of tritium evoked by electrical stimulation, showing that alpha~-adrenoceptor-mediated negative modulatory mechaSism operates in neonates. However, in strips preloaded with ~H-noradrenaline, the beta-agonist isoprenaline (50 nM) increased the overflow of tritium in strips of adults but had no effect in ~trips of neonates;and in strips preloaded with H-adrenaline, the beta-antagonist propranolol (I ~M) reduced the overflow of tritium in adults, but had no effect in neonates.Additionally it was observed that isoprenaline (i0-i000 nM) failed to cause relaxation on strips previously contracted by phenylephrine while forskolin (0. Neurotransmitters, hormones and growth factors cause profound changes in phospholipid metabolism.The role of phospholipase C mediating hydrolysis of inositol phospholipids is well known, but extensive evidence is accumulating to indicate that phospholipase D (PLD) is a key-enzyme of a novel signal transduction pathway. PLD utilizes phosphatidylcholine (PC) as substrata and leads to the formation of choline, phosphatidic acid (PtdOH) and, in the presence of a primary alcohol, such as propanol, to phosphatidylpropanol (PP) (transphosphatidylation reaction).In the present study, PLD activity was determined in superfused hippocampal slices of newborn and adult rats by measuring the formation of labeled PtdOH and PP. PC was labeled with [3H]-glycerol. It WaS found that 0.1 and 1.0 mM L-glutamate~ in contrast to NMDA, increased the formation of both [~H]-PP (up to 270% of control; N=7) and [3H]-PtdOH. The partial antagonist DL-2-amino-3-phosphonopropionic acid (AP3;I mM) which selectively binds to the "classical" metabotropic glutamate receptor failed to antagonize the glutamate-induced increase of[3H]-PP and [3H]-PtdOH. The effect of glutamate was present in newborn, but not in adult rats. Similar to glutamate, the phorbol ester 4B-phorbol-12S,13e-dibutyrate (PDB) enhanced PLD activity. There is some evidence that protein kinase C and PLD form a positive feedback loop (L6ffelholz, Biochem. Pharmacol. 38:1543 , 1989 . Glutamate which is the dominant excitatory transmitter in the hippocampus,, has been associated with long-term aspects of cellular control and development.We speculate that PLD-mediated signal transduction could be involved in these glutamatergic functions. Pharmakologisches Institut der Universit~t, 0bere Zahlbacher Str. 67, W-6500 Mainz, Germany

6. Weinhetmer, 0. Lang, W. Sauermann, T. 6inap and T.J, Feuerstein Ascorbic acid (AscA) is widely distributed in tissues of most animals. Relatively high concentrations were found in mammaIlian brain. It has been reported that neuronally released glutamate induced release of AscA (~henk et el. 1982, Brain Research 253,353) .The intraceIlular concentration of AscA seems even to reflect the release of excitatory amino acids in the brain (O'Neill et el. 198, 1) . Under pathological conditions both AscA concentrations and NMDA receptor activity is increased. The aim of the present stuoy was to evaluate the possible interaction between AscA and the NMDA receptor. For this purpose we investigated the effects of AscA on NMDA-stimulated Acetylchollne (Ach) release in slices of rabbit brain, Slices of caudete nucleus were incubated with [3H]choline, then superfused with modified Krebs-Henseleit buffer and stlmulated either with NMDA or elactricelly. When NMDA was applied a Mg++-free buffer was used, The superfusate was collected in 5 rain samples, Each slice was stimulated twice, after 15 rain ($I) and 55 rain ($2) superfusion.[3H]-Efflux and [3HI-content of the tissue was determined by liquid scintillation counting. In slices superfused with AscA free buffer and stimulated with NMDA ( I OIaM) for 4 rain ($i) tritium overflow was about 2.24 ± 0.06%. of tissue [3HI. The release by the second stimulation was about 1.66 ± 0.07%, AscA added 35 min before 52 produced a concentration dependent inhibition of Ach release, The inhibitory effect of the NMDA antagonist AP-5 [(+)2-amino-5-phosphopentanoic acid] was larger when experiments were performed in AscA free buffer than in the presence of AscA (PA2:4.77 versus 4.38). Ach release evoked by electrical stimulation was not inhibited by AscA, This stuff# revealed that AscA interacts with the NMDA receptor by attenuating the NMDA receptor-mediated Ach release. AscA decreased the inhitory effect of the NMDA antagonist AP-5. In future studies with NMDA this interaction should be taken in consideration.Neuropharmakologischas Labor der Neurologiachen Universit~itsklinik, clo OiJdeske AO, 7800 Freiburg, F,R.O.

Enkephalin-induced electroencephalographic seizures have been partly attributed to an increase in the excitability of hippocampal pyramidal cells, which is brought about by decreasing the gammaaminobutyric acid mediated early and late postsynaptic potentials. In the present study it was investigated how this inhibition of interneurons by opiods affect excitatory synaptic transmission. On this purpose both the non N-methyl-D-aspartate (NMDA) antagonist CNQX (6-cyano-7-nitroquinoxaline-2,3 diane) and the noncompetitive NMDA antagonist MK 801 (5-methyl-10,11-dihydro-5Hdibenzo[a,d]cyciohepten-5,10-imine) were employed in order to antagonize the effects produce~ by the selective mu-opioid agonist DAMGO (D-Ala~-NMePhe ~, Gly-ol-enkephalin) in the hippocampal formation. Local cerebral glucose utilization, which can be closely correlated to neuronal activity, was most pronouncedly increased in the strata radiata and strata pyramidale of field CA 1 and CA 3 in the ventral hippocampus after intracerebroventricular injection of 0.5#g DAMGO. MK 801 (0.1mg/kg, i.v.) and CNQX (12#g, i.c.v.) antagonized these changes differently. CNQX proved most effectively in field CA 3, that is the effects of DAMGO were blocked completely. In field CA 1, CNQX produced only nonsignificant reductions; MK 801, in contrast, reversed the changes in this area by 60-80%. Thus, activation of mu opioid receptors entails activation of both non-NMDA and NMDA receptors. Non-NMDA receptors, however, appear to be primarily involved in CA 3, NMDA receptors in CA 1. In vitro studies on brain slices showed that not only anticholinergic but also indirect dopaminomimetic properties contribute to the antiparkinsonian effects of budipine (BUD), whereas biperidene (BIP) exhibited mainly anticholinergic effects (Jackisch et al., J Pharmacol Exptl Ther, in press, 1993 In previous studies it was shown that effects of apomorphine on behavior and EEG can be conditioned: when the drug effect is repeatedly associated with well-defined conditional stimuli (CS), the presentation of the CS alone produces effects which resemble drug effects. Although acute administration of apomorphine activates postsynaptic dopamine receptors and concomitantly decreases DA release, the conditioned apomorphine effects might be due to an increase in DA release. This hypothesis was tested in the striatum by using microdialysis in freely moving rats for estimation of extracellular DA, its metabolites DOPAC and HYA, and the 5HT metabolite 5HIAA. Conditioned rats received apomorphine (I mg/kg s.c.) in the test cages in association with the CS (an auditory and an olfactory stimulus) for 7 times, in alternating sessions saline in association with the home cages. Pseudoconditioned animals were treated in the reverse manner. The control group received saline in the test and in the home cages. ~Vhereas acute administration of apomorphine decreased extracellular DA and HVA, the CS in conditioned rats decreased DA only. The results suggest that the occurance of conditioned apomorphine responses cannot be explained by an enhancement of striatal DA release, but might be due to processes downstream of the dopaminergic terminals. However the effectiveness of this long-term treatment is progressively reduced and, in addition to that, the benefit of this treatment is diminished by the appearance of well-known side effects. Peripheral catechoI-O-methyltransferase (COMT) inhibitors have been developed under the idea that inhibition of peripheral L-DOPA metabolism provides the opportunity to reduce the daily dose of L-DOPA/Carbidopa without reducing the effectiveness of the treatment. We used the 6-hydroxydopamine (6-OHDA) model to study the effect of the centrally and peripherally acting COMT inhibitor OR-486 (3,5-dinitrocatechol) on rotational behavior produced by daily IP doses of L-DOPNCarbidopa (18/2 mg/kg ) in comparison to L-DOPNCarbidopa (27/3 mg/kg), given alone. Drug treatments started three weeks after the lesions. During the ten days of teeing L-DOPNCarbidopa produced progressive increase in contralateral rotations which were higher in the group with the higher dose of L-DOPNCarbidopa. Combined treatment of OR-486 plus the lower dose (18/2 mg/kg IP) of L-DOPNCarbidopa produced about the same amount of contralateral rotations as the higher dose (27/3 mg/kg IP) of L-DOPNCarbidopa, given alone. This demonstrates that the peripherally and centrally acting COMT inhibitor OR-486 may increase the central effect of L-DOPNCarbidopa. However, it is unclear whether this effect is solely due to increased central availability of L-DOPA or in addition to that, by decreased production of 3-O-methyldopa.

In rats with electrodes chronically implanted into the skull above the cerebral cortex, the effects of two psychomotor stimulants, damphetamine and cocaine, were studied on behaviour and electrocorticographic (ECoG) activity, which was telemetrically recorded and on-line quantified in its power spectrum. In low doses, d-amphetamine (AMPH 0.4 

(3.0 mg/kg s.c.), a selective agonist for D-1 receptors, elicited similar ECoG effects, which were characterized by a general decrease in power in all frequency bands in their power spectra. In contrast, high doses of AMPH (4.0 mg/kg i.p.) and COC (30 mg/kg i.p.), which predominantly produced locomotor activity and stereotypy activity, significantly increased power in the alpha-1 band (7.00 -9. 50 Hz) to a/eve/of 140 % of baseline activity. The power in all of the other frequency bands decreased, except in alpha-2 band (9,75 -12.50 Hz), which was not significantly changed. In previous studies with apomorphine (0.5 mg/kg s.c.J, a mixed dopamine receptor agonist preferentially acting on D-2 receptors, a comparable alpha-1 band activity was observed. Transplants containing fetal dopaminergic cells are capable of inducing functional recovery in rats with unilateral lesions of the nigrostriatal pathway. The grafts were placed to the lesioned side, but there are some indications for a functional recovery after implanting to the non-lesioned side. The behavioral and neurochemical effects of intraventricularly transplanted solid fetal mesencephali¢ grafts in rats with unilateral 6-hydroxydopamine (6-OHDA) lesions of the nigrostriatal pathway were investigated. The rats received the graft either ipsilateral or contralaterai to the lesioned side and were tested for their rotational behavior induced by apomorphine (0.25 mg/kg i.p.) or amphetamine ($ mg/kg i.p.) 8 times up to 29 weeks after transplantation. Afterwards the animals were tested by in viva voltammetry for the amphetamine (5 mg/kg Lp.) induced dopamine (DA) release in the an~eromedial striatum. Only rats with ipsilateral grafts showed a stable long term effect of compensation of the rotational behavior after apomorphine challenge, but a transient compensation in rotations was seen in rats with contralateral grafts too. During the tests with amphetamine most of the ipsilaternlly transplanted animals showed compensation in the number of turns over a period of 29 weeks whereas the contralaterally placed grafts produced a marked increase of rotations. Using differential pulse voltammetry the preliminary data indicate a small increase of peak 2 (which corresponds to DA in pargyline pretreated rats) in the lesioned and grafted striatum after systemic injection of amphetamine. Independent on the side of transplantation the amphetamine-induced DA signal in the non-lesioned striatum was smaller in grafted rats than in nongrafted controls.

Our data indicate that changes in the release of DA in the unlesioned strlatum may, at least in part, contribute to a functional recovery of lesioned and grafted rats. The dopamlne (DA)-precursor L-DOPA has been shown to be beneficial in the treatment of Parkinson's desease via reducing the DA-deficit in the basal ganglia. However, the high level of L-DOPA itself and its metabolites are suspected to be involved in the production of effects other than substituting the DA-deficit. We used the methods of K +-and amphetamine (AMPH)-stimutated [3H]DA-release from synaptosomes isolated from rat striatum to study the effect of L-DOPA and its metabolites upon the dopaminergic autoreceptor as well as upon the uptake carrier. None of the substances affected the K+-evoked [3H]DA-release under conditions where the DA-uptake was inhibited by nomifensine indicating that dopamine autoreceptors are not involved in their action. However, in the absence of nomifensine L-DOPA was found to increase both the K +and AMPH-stimulated [3H]DA-release dose-dependently. This may be indicative for at least two mechanisms of L-DOPA on in vitro DA-release: 1. Ca++-dependent (K+-stimulus) and 2. Ca++-independent (AMPHstimulus) DA-release. Recently, Snyder and Zigmond (1990) have demonstrated increased DArelease following electrical stimulation of superfused brain slices after the incubation with L-DOPA. Thus, it can be assumed that L-DOPA does increase DA-release evoked by three different stimuli affecting different intracellular DA-pools (Fairbrother et aL, 1990) . However, it remains an open question whether this increased DA-release is due to only increased DA-synthesls or due to both increased synthesis plus fascilitated release. Snyder, G. and M.J. Zigmond, Brain Res. 508, 181-187 (1990 ) Fairbrother, I.S. et al., J. Neurochem. 54, 1834 -1851 (1990 Dept Earlier studies indicate that taurine after intracerebrovenl:ricular administration similarly to GABA reduces the release of striatal dopamine iDA) (Panula-Lehto et al. Naunyn-Schmiedeberg's Arch Pharmacol 346:57-62, 1992) . In this experiment effects of intrastriatal and intranigral taurine infusions on striatal dopamine metabolism were compared in microdialysis study in anesthetized rat. A microdialysis probe was implanted into the rat striatum (A + 1,0, L +2,7, D -6,5) under halothane anestesia. For intranigral infusion another probe was implanted into the ipsilateral substantia nigra (A -5,5, L + 2.0, D -8.5). Extracellular concentrations of striatal DA, 3,4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA) were assayed by HPLC with electrochemical detection. 50raM, 150 raM,and 450 mM taurine solutions were infused via microdialysis probe into the striatum or into the substantia nigra for four hours (2 pl/min), The release of striatal DA was clearly increased by intrastriatal taurine; 450 mM the increase (by 10 fold) lasted only for 30 rain, whereas at 50 or 150 mM the increased release (by 2-4 fold) lasted up to four hours. Extracellular DOPAC was increased to 160 % of the control values, whereas HVA was decreased to 90 %, When taurine was infused into substantia nigra, it decreased the release of DA in the ipsilateral striatum down to 40% of the control values. Extracellular DOPAC and HVA were both increased to 150 % and 170 % of the control, respectively. All effects seemed to be dose-dependent, In microdialysis studies intranigral GABA changes the striatal DA metabolism similarly to taurine in this study (Reid et al., Eur J Pharm. 147:411-420, 1988) . In conclusion, these results show that taurine seems to enhance DA release when given into the striatum. But, when taurine is given into the cell body region of nigrostriatal DAergic pathway it reduces DA release, Repeated systemic application of the neurotoxin MPTP causes a profound disruption of nigrostriatal dopamine iDA) neurons in primates and in certain rodent species. It is assumed that the conversion of MPTP to MPP + by MAO B and the subsequently uptake by the DA carrier system is responsible for the specific neurotoxicity of MPTP. But it is not well known which mechanisms are involved in the acute local effects of MPTP. We investigated the local effects of MPTP on the dopaminergic synaptic transmission in the striatum of rats by means of microdialysis. Chloral hydrate-anesthetized male Wistar rats were imp/anted with 2 mm microdialysis probes into the striatum and were perfused with arteficial CSF (Ibtl/min). The dialysates were collected at 25 rain intervals and analyzed for DA, DOPAC, HVA, 3-MT and 5-HIAA by HPLC with electrochemical detection. The intrastriatal infusion of 50 p.M MPTP (25 rain) induced by marked increase of DA release (approx. 2000 % of basal level), an elevation of the extracellular 3-MT with a delayed maximum seen after 50 rain and a continuously decrease of DOPAC and HVA to 40 % of basal values, The level of 5-HIAA was not altered. The MPTP induced excessive DA efflux could be partially inhibited by nomifensine (20 mg/kg b. wt. i. p.) and by using a calcium-free CSF.These results indicate that also the local effects of MPTP on the dopaminergic transmission are mediated, at least partially, by the metabolite MPP ÷.

Flunarizine has been shown to accelerate removal of extracellular lactate accumulated during Spreading Depression (SD, Scheller et al. (1992) , Naunyn-Schmiedebergs Arch. Pharmacoh 345:R122) which might be attributed to an enhancement of cerebral blood flow (CBF). However, its effects on CBF have been discussed controversial (Jansen et al. (1991), Eur. J. Clio. Pharmacol. 40:7-15 ). The following experiments, therefore, were performed to reinvestigate possible vascular consequences of flunarizine treatment. The skull of male Wistar rats (220 to 250 g, urethane 1.75 to 1.9 g/kg) was trepanized and a LD probe (1 mm diameter, TSI) was positioned above the dura. Basal local cerebral blood flow (ICBF) was recorded for 90 min. In group I, either flunarizine (2.5 mg/kg in 10 % cyclodextrin, n=5) or vehicle (0.5 ml 10% cyclodextrin, n=4) was infused i.v. (within 10 rain), in group iI a control SD was induced by a needle prick several mm remote to the LD probe. 10 rain after SD had occurred, either flunarizine (n=5) or vehicle (n=4, as above) was infused. 90 rain after the 1st SD a 2nd SD was induced. Flow values (arbit. units) are expressed as % of ICBF in control period. Statistical analysis was performed using the Wilcoxon-Mann-Wittney-U-test. In group I, ICBF increased to 131.8 + 21.97 % during the first 60 rain after flunarizine infusion. In group II, ICBF rose to maximally 282.9 + 23.78 % during SD and then decreased to 69.3 ± 11.46 %. Hypoperfusion lasted for 90 to 120 min. Flunarizine normalized SD-mediated hypoperfusion within 10 to 15 rain and caused ICBF to rise to 161.9 + 56.54 % of pre-SD level. During the following SD, ICBF maximum was not different from controls. However, a SD-mediated hypoperfusion was no longer observed. The data show, that flunarizine given i.v. enhances cortical blood flow acutely. Moreover, it antagonizes the SD-mediated, long lasting hypoperfusion. This could be of relevance for stroke and migraine therapy since the occurrence of SDs during infarction and during an aura phase of migraine was shown recently ( Male mice were exposed to normobar hypoxia (7% 02) for 1 hour. One, 2, 4, 7, 14, and 21 days later the seizure susceptibility to pentylentetrazol (PTZ) (58 mg/kg i.p.) was investigated. Shamexposed control animals, which were placed in the hypoxia chamber at normal air conditions, and naive controls were used for comparison. Seven days after hypoxia a protection against PTZ seizure was significantly developed in hypoxia-exposed animals. In contrast, a more strong normobar hypoxia (5,5% 02, 1 h) did not alter the seizure susceptibility after PTZ. Ten min after hypoxia an administration of MK 801 (0.1, 0.2 or 0.3 mg/kg i.p.) was not able to modify the hypoxia-induced protection, whereas the seizure susceptibility in hypoxia-exposed animals was enhanced to the level of both control groups following the pretreatment with the adenosine antagonist DPX (1,3-Diethyl-8-phenylxanthin) (0.1 or 0.5 mg/kg i,p.). This studies indicates that functional changes of adenosin system are involved in hypoxia-induced decrease of seizure susceptibility after PTZ. Models of global cerebral ischemia in the rat and gerbil have been widely used for the evaluation of potentially protective drugs. Predominantly, the percentage of damaged hippocampal neurons was taken as an indicator for a putatively protective compound. However, chronic ischemic brain damage to other than hippocampai areas is difficult to quantify by cell counts, but could be a most relevant aspect for the efficacy of pharmacoprotection after stroke. For this reason, we have measured the regional fresh volumes in postischemic brains. Forebrain ischemia was induced by reversible occlusion of both common carotid artades and simultaneous lowering of blood pressure to 40 mmHg for 10 minutes. Blood pressure was restored to normal levels by reinfusion of the shed blood. After survival of 3 months (n=10) rats were perfusionfixed. The entire brains were sedally sectioned, the brain regions were delineated on the Nissl-stained sections and the areas ware planimetersd using an image analysis system. Volumes of the brain regions were calculated from the planimetered areas and the distance between the sections. Each brain was individually corrected for shrinkage due to fixation, Compared to unlesioned rat brains, reduction of volume (30-40 %) was biggest where neuronal damage was most severe, i.e. in the hippocampus. However, also in regions where ischemic damage cannot be easily quantified by ceil count, like in cortical and extrapyramidal regions, significant reductions occurred as well ranging from 10 to 30 %.The results show that determination of fresh volumes is a sensitive tool for pharmacological studies seeking to evaluate the pharmacological treatment of experimental stroke.Anatomisches Institut, Gertrudenstral~e 9, 0-2500 Rostock Recent investigations have demonstrated that a mild normobar. hypoxia protects rats and mice against pentylentetrazol-induced convulsions 7 days after hypoxia. In order to clarify the mechanism of this protection we recorded monosynaptic evoked ~eld potentials in the CA1 region of transversal hippocampai slices from rats pretreated in 3 different ways: 1: Rats were exposed 7 d prior to the experiment with mild hypoxia (1 hour, 9%, 02). 2: active control rats were exposed for 1 hour normal airconditions in the hypoxia-chamber 3. passive controls sheltred in there home cage. The peffuaion with Mg++-free medium induced in all slices a potentiation of the population spike of the evoked field potential which out lasted the perrusion time of 30 rain.In slices of hypoxiaexposed rats this potentiation was significant greater than in slices of control animals. It may be concluded that normobar hypoxia induces modifications in the hippocampal NMDA-system which seems not to be correlated with the processes of protection in vivo. This difference should be clarilied by further experiments. 

Electreshock-induced amnesia has been proposed as a technique for producing retrograde anmesia. The aim of our study was to investigate whether calcit~n channel blockers, nitrendipine and felodipine, can prevent memory disruption caused by ele-ctl~oshock (ECS) application.The study was carried out on Hannover Wistar rats weighing 150-200 g. Various doses of nitrendipine or felodipine (0.03; 0.i; 0.~; 1.0 mg/kg) were given i.p. Thirty minutes later the passive avoidance task was performed according to the stepthrough procedu~e. A two-compartment apparatus with a grid floor which could be electrified was used. During the learning trial the rat was Dlased in the light compartment. Aftee SO seconds the guillotine door was opened and the latency between the door opening and the entrance into the dark com~ pa~tment was measured. When the rat walked into the dark part of the apparatus it received a foot shock, l~nediately after the training trial the animal receivedanEC$ through the ears, The retention trial was carried out 24 hours later. It has been found that both tested drugs were effective in reversing the memory deficits in ECS-treated rats. Statistically significant enhamcement of the passive avoidance behavior was preducedbyO.l; 0.S and 1.0mg/kg of nitrendipine and 1.0 mg/kg of felodipine.Department of Ph~u~nacelogy, School of Medicine, University of Rijeka, 51000 Rijeka, Croatia. The ventromedial globus pallidus and adjacent magnocellular gest that cholinergic neurotransmission activity of central nervous system is strongly involved in memory processes and particullary vulnerabile to hypoxia. Therefore, the present study was designed to examine if hypoxia-induced annesia could be reversed by cholinergic agonist arecoline. The study was carried o~t on Hsnnover~Wistar rats weighing 250 g. Arecoline (0.1;0.3;1.0 mg/kg ) was given i,p. The contrel greup received saline i.p. Ten minutes later passive avoidance behavior was studied according to the step-through procedure. A two compartment apparatus with a grid floor which could be electrified was used. Du~ing the learning trial the rat was placed in the illu~ninated compartment. Ten seconds later the guillotine door was raised and the latency between the door opening and the entrance into the da~k compartment wa~ meastu~'ed. After entering into the dark part of the apparatus, the animals received an unavoidable foot shock. I~mediatelly after the training, the rats were subjected to a period of oxygen deprivation hypoxia. The retention trial was carried out 24 hours later. , Our results demonstrate that hypoxia produced significant im~ pairment of paSSive avoidance behavior. A~ecoline was effective in reversing the memory deficits in hypoxia exposed rats. Namely, all tested doses of mentioned drug produced significa]t improvement of the passive avoidance task in hypoxic animals. These results support the hypothesis that impairment of cholinergic neurotransmission activity is involved in memory deficit in hypoxi c rats.Department of Pharmacology, School of MedSe%ne, University of Rijeka, 01ge Ban 22, 51000 Rijeka, Croatia. neurones of basal forebrain (NB) provide the ma~or cholinergic innervation to the neocortex in the rat. The pL~rpose of this study was to examine the passive avoidance behavior twenty days and six months after the NB lesion. Male Wistar rats weighing 200-250 g were used. Unilateral or bilateral electrolytic NB lesions were made stereotaxically. The site of the lesions was histologically verified. Twenty days later the passive avoidance task according to the procedure of Ashford and Jones (1976) was performed. A chamber with a grid floor continously electrified %ras used. The central area of the platform contained a wooden platform. Each rat wee placed on the platform and left in the apparatus for t~ree minutes. The total time spent on the platform was recorded. A total of four consecutive daily sessions were given. Six months later the passive avoidance task was repeated in bilateral lesioned rats. Results of our experiments show that both unilateral and bilateral lesions of the NB inlmaired the passive avoidance responses in the rat. The effect of bilateral lesions was statistically significant. Six months later significant imprevement of passive avoidance behavior was noted in bilateral lesioned rats. It is concluded that destruction of the NB neurons can be compensated by the plasticity of the residual NB neurons and other central cholinergic and noncholinergic structures which 8re involved in learning and memory processes. The mechanisms underlying hyperosmolar diabetic coma are not yet understood. Since changes in NT release may be involved, the aim of the present study was to investigate whether high D-glucose (GLUC) modifies the release of two representative NTs, noradrenaline (NA) and ~'-aminobutyric acid (GABA), in the cerebral cortex. Temporoparietal cortex specimens either from rats or from leftovers of pieces which had to be removed from patients undergoing neurosurgery were used to prepare slices or (from rat tissue only) synaptosomes. The preparations were preincubated with ZH-NA or 3H-GABA and subsequently superfused. ZH overflow was stimulated by elevation of the K + concentration by 15-25 mmol/I. In human cortical slices an increase in GLUC concentration in the superfuslon fluid by 10-100 mmol~l concentrationdependently enhanced ZH-GABA release (apparent maximum: by 73 %; apparent ECso: 17 retool/I), whereas SH-NA release was inhibited (by maximally 98 %) in the presence of the same GLUC concentrations (ICso : 85 mmol/1). GLUC induced the same effects on the release of these NTs in rat cortical slices, but in synaptosomes an increase in GLUC concentration by up to 100 mmol/l only increased 3H-GABA release, leaving that of 3H-NA unaffected. The possibility that the inhibition of SH-NA release observed in slices may be related to the increase in GABA release, was supported by experiments with the GABA~ receptor antagonist CGP 35348 (100 pmol; p-(3-aminopropyl)-p-diethoxymeth¥/-phosphinic acid) which abolished the GLUC-induced inhibition of 3H-NA release in slices from rats and humans, in rat cortical slices, the effect of high GLUC was mimicked by equiosmolar NaC/or D-fructose, whereas the freely membrane-permeable dimethylsulfoxide was ineffective.In conclusion, (1) the increase in GABA release in the cerebral cortex by high GLUC is probably due to hyperosmolarity and the osmotic pressure gradient at the neuronal membrane, (2) the inhibition of NA release seems to be the consequence of the increased GABA release which appears to induce activation of presynaptic inhibitory GABA B receptors, and (3) Some of the A-ring reduced steroids, including the natural occuring metabolite 5a-Pregnane-3o~, 21-diol-20-one (5a-THDOC), are potent modulators of C1-fluxes mediated by GABAA receptors. 50~-THDOC is synthesized in the hypothalamus and related structures of the limbic system, following stress induced secretion of deoxycorticosterone from the adrenal cortex. The effects of 5~-THDOC on GABA-activated C1-currents were investigated in rat hypothalamic neurons in primary cell culture. Neurons were recorded with the whole cell voltage clamp (giga-seal technique). GABA (1 -1011.M) and/or 5~-THDOC (10 nM-1 ~M) were locally applied from a multibarrel/single tip superfusion system for 1 to 5 s. The amplitude of the GABAactivated C1-currents was measttred at a holding potential of -70 mV. To increase the driving force, the patch pipettes were filled with a high CI-solution (150 raM). Both, the amplitude and the kinetic of the GABA-evoked inward currents were dose-dependent. Responses to low GABA concentrations (up to 1 ~tM) increased slowly (1 -1.5 s) and attained up to 200 pA a constant level during the time of application. The currents induced by GABA higher than 1 ~tM (exceeding 200pA) reached a maximum (up to 3 nA, 100p.M) within 500 ms and decayed to a plateau of 10 to 30 % of the peak amplitude. In 54 % of the neurons (20/37) the currents evoked by 10 ~tM GABA (1.1 + 0.4 hA) were increased by 5~-THDOC (10 nM-1 ~tM) to 140+ 37 %. The currents induced by 1 paM GABA (60 -600 pA, 230_+ 156 pA) were increased by 1 p.M 5c~-TI-IDOC to 185_+49 % (n = 17). In contrast, 10nM 5c~-THDOC decreased the current induced by 1 I~M GABA to 80 % (9/17; p < 0.05). Our data show that 5~-THDOC modulates GABA-activated C1-currents in a complex dose-dependent manner: the direction of the modulatory effect depends on both the GABA and the steroid concentration. Responses to high GABA concentrations are augmented at all concentrations of 5~-THDOC. However, responses to low concentrations of GABA are attenuated at low (10 nM) and augmented at higher (100 nM-1 I.tM) concentrations of 5c~-THDOC. Thus, the neurosteroid 5~-THDOC determines the signal to noise ratio of GABAergic transmission, a mechanism which may have physiological relevance for adaptive changes of the CNS. Various actions of steroids on the brain link the endocrine to the central nervous system. 5~-pregnane-3c~,21-diol-20-one (5~-tetrahydrodeoxycorticosterone; 5c~-THDOC) is metabolized within the brain from adrenal deoxycorticosterone passing the blood-brain-barrier. The increase in 5~-TI-IDOC levels in cerebral cortex and hypothalamus of the rat following stress stimulation suggests a role in adaptive responses to stress. In addition to regulation of gene expression as the classical cellular mechanism of steroid effects, certain steroids modulate membrane channels and receptors. 5~-THDOC binds to the GABA A receptor complex and subsequently enhances 121-flux through the associated ionopbore. We investigated the effects of 5c~-THDOC on electrophysiological properties of neocortical pyramidat neurons in a slice preparation of the rat frontal neocortex. Cells were recorded in layers IUIII by conventional intracellular impalement or by whole cell patch cl~anp. Postsynaptic potentials were elicited by orthodromic stimulation with a bipolar stainless steel electrode located in layers V/VI. GABA receptor agonists were admirdstered by microiontophoresis using multibarrelled pipettes placed in the vicinity of the recorded cell. In convenlionally recorded neurons 5~-THDOC (10 .5 M) potentiated and prolonged the inhibitory postsynaptic potential 0PSP). The amplitude of the early IPSP (measured at 15 ms after stimulus) was increased to about 200 %. Estimations of the synaptic conductance revealed an average enhancement to 700 % after 15 minutes of steroid application. The effect of 5~-THDOC on the IPSP was observed in spite of the impairment of intmcellular cascades by perfnsion of the cytoplasm by the patch pipette. The amplitude of the postsynaptic responses to the iontophoretically applied GABA a receptor agonist muscimol increased to approximately 170 % within 15 to 30 minutes after steroid administration. Excitatory postsynaptic potentials and passive membrane properties such as resting membrane potential, input resistance and action potential amplitude remained unchanged. These results suggest that the eniaancement of synaptic inhibition by 5~-THDOC observed in the in vitro slice preparation is due to a modulatory action mediated via postsynaptic GABAa receptors.Ma~-Planck-/nstitute of Psychiatry, Clinical Institute, Clinical Neuropharmacology, Kraepelinstr. 2, 8000 Mtinchen 40, F.R.G.

GLUCOCORTICOID REGULATION OF SPECIFIC NEUROTROPHIN EXPRESSION IN HIPPOCAMPAL NEURONS J.L. Scully, and U. Otten Glucocorticoid holInones (GCs) are secreted by the adrenal medulla in response to stress. The hippocampus is a particular target for GCs. Prolonged exposure to the hormones exacerbates damage to hippocampal pyramidal neurons due to ageing or neurological insults, while on the other hand adrenalectomy (which removes the source of GCs) destroys hippocampal granule ceils in vivo. The mechanisms of the GC effects are unknown, but the synthetic GC dexamethasone is known to affect levels of neurotrophins (NTs) in brain. Members of the neurotrophin family are strongly implicated in the development and maintenance of the hippocampus, and it is possible that GC effects are mediated via NTs. We were interested to see whether GCs could regulate NT expression in immortalized hippocampal cell lines which show many characteristic features of neurons and which have been shown to carry type II corticosteroid receptors. Using reverse transcription and amplification of the resulting cDNA by the polymelase chain reaction (PCR), we have found that, in a cell line derived from embryonic hippocampal neurons (HN10), NGF and NT-3 mRNA levels were significantly increased after administration of dexamethasone. The increase is dose-dependent and is visible with 10-8M dexamethasone; maximum effect is seen with 5xl0-6M dexamethasone. Only at the highest concentration tested (10-4M) is there evidence of neurotoxicity. The time course of increase differs between the two NTs, NT-3 peaking earlier (2h after GC treatment) and failing off more rapidly than NGF. A cell line derived from posmatal hippocampus, HN25.2, shows no change in NT-3 mRNA. NGF mRNA does rise, with a longer time course compared to the NGF rise in HNI0. Hence the change in NT expression in response to GCs may be developmentally regulated, and may protect neurons against the harmful effects of exposure to GCs. The preprocholecystokinin gene is expressed in a subpopulation of GABAergic interneurons in rat cortex. Its expression, measured as immunoreacfivity or as mRNA coding for preprochulecystokinin, is enhanced in a model of temporal lobe epilepsy, i.e. after a single i.p. injection of 10 mg/kg kainie acid (Meyer et al., Neurosci, Lett. 69:208 (1986); Olenik et al., Synapse 4:223 (1990) ). The mechanism of this effect is unclear. Kainic acid may act directly on the cholecystokinin interneurons or indirectly via stimulation of their neuronal afferents. The further analysis in vivo does not seem feasible due to the complex neuronal interactions which may occur. Therefore, in search for an in vilxo model which is organotypic and allows the analysis of neuronal interactions, slices of rat sematosensory cortex were explanted at postnatal day 1 and cultivated for 12 + 2 days using the method of Stoppini et al. (L Neurosci. Meth. 37:173 (1991) ). In these slice cultures, in situ hybridization showed that the neurons containing cholecystokinin-mRNA were grouped in two layen. Tl3.is arrangement is typical for the localization in situ. Light-and eleetmn-uncroscol~C examination combined with immunoeytoehemical techniques showed immtmopositive edls with small round somata and dense networks of immonoreactive fibers. Only symme~c synapses were formed by immunoreactive terminals. All these morphological features are typical for cholecystoldnin neurons in situ. In a first serie~ of experiments, the effect of bicuculline (70 I~M) was tested. The agent, which can hardly be used in vivo due to its epileptogenie effects, was applied for 90 rain. Twelfe hours thereafter no change in the levels of eholecystokinin-mRNA was observed indicating the absence of GABAergic inhibition via GABAA receptors.Taken together, we have good evidence that cholecystokinin neurons in slices of rat somatosensery cortex retain their biochemical and morphological propexfies, whm cultured in vitro. These organotypical slice cultures should make possible a thorough analysis of the effects of excitatory aminoacids on the regulation of neuropeptide synthesis in interneurons. In recent years computer-aided instruction in various disciplines in the sciences has multiplied in response to progress in the development of hardware and adequate software. "Hightech" software allows almost all interested people, whether they are computer specialists or not, to develop programs dealing with multimedia. It has been our aim for many years to incorporate computer-assisted programs into our curriculum for medical students. The first target was to reduce the number of animal experiments in the pharmacology course. In this respect, we were generously assisted by Dr.R.D. Purves who developed computer animations for the teaching of introductory pharmacology, published under the title "Computer Club" in TIPS, Dec. 1984 (496-498) .

Our further efforts concentrated on computer animations of complicated enzyme reactions, such as hydrolysis of acetylcholine by acetylcholinesterases, the catalytic action of carboanhydrase, and on those animations which allow the demonstration of basic organ functions which are difficult to explain verbally. In this respect we were able to produce animations of heart muscle contraction, the pharmacology of EKG, and action of diuretics, especially those influencing the counter current mechanism at the level of the loop of Henie.