PII: S0014-5793(99)00473-1


Neuronal nAChR stereoselectivity to non-natural epibatidine derivatives

Sonia Bertranda, Jo« rg T. Pattb, Jo« rg E. Spangb, Gerrit Westerab, P. August Schubigerb,
Daniel Bertranda;*

aDëpartement de physiologie, Centre Mëdical Universitaire (Facultë de Mëdecine), 1 rue Michel Servet, CH-1211 Geneva 4, Switzerland
bCenter for Radiopharmaceutical Science, Swiss Federal Institute of Technology Zu« rich, Paul Scherrer Institute Villigen and Department of Radiology,

Clinic of Nuclear Medicine, University Hospital, Zu« rich, Switzerland

Received 17 March 1999

Abstract The frog toxin epibatidine is one of the most powerful
ligands of the neuronal nicotinic receptors and derivatives show
promising possibilities for labeling in positron emission tomo-
graphy studies. In an attempt to reduce epibatidine toxicity, new
methyl derivatives were synthesized, tested in positron emission
tomography imaging and in electrophysiology. labeling as well as
physiological experiments highlighted the differences in sensitiv-
ity of the neuronal nicotinic acetylcholine receptors between two
methyl enantiomers and the reduction in sensitivity caused by
introducing the methyl group. At present, epibatidine derivatives
seem the most promising compounds for in vivo labeling of
neuronal nicotinic acetylcholine receptors.
z 1999 Federation of European Biochemical Societies.

Key words : Nicotinic acetylcholine receptor ; Brain ;
Epibatidine ; Pharmacophore ; Imaging

1. Introduction

With the introduction of tobacco plants on the second voy-
age of Christopher Columbus its use and smoking addiction
quickly spread amongst a large fraction of the population.
Although it was recognized earlier that the alkaloid nicotine
is the active compound in tobacco leaves and that it is a
potent agonist of the cholinergic system at the neuromuscular
and ganglionic synapses, the mechanisms by which it causes
addiction remained obscure [1,2]. The identi¢cation of an en-
tire family of genes encoding the neuronal nicotinic acetylcho-
line receptors (nAChRs) that are speci¢cally expressed at the
neuromuscular junction or in nerve cells shone a new light on
the mechanisms by which nicotine can exert its action in the
central nervous system [3,4]. One of the critical ¢ndings was
the observation that neuronal nAChRs display a wide pattern
of physiological and pharmacological pro¢les and that a given
set of subunits is expressed in a well-de¢ned brain area [5^7].
For instance, it was shown that the major brain neuronal
nAChR results from the assembly of the K4 and L2 subunits.
Moreover, a high level of expression of this class of receptor
was observed on dopaminergic neurons that are known to
participate in reward and addictive behaviors [8,9].

With the introduction of in situ hybridization using speci¢c
mRNA probes it has become possible to map the pattern of
expression of the identi¢ed neuronal nAChR subunits in rat,

mouse or human brains [7,10]. Although allowing a high spa-
tial resolution the use of this technique cannot be employed
for in vivo measurements. The development of new tools is
therefore required to establish the relationship between the in
situ pattern of neuronal nAChR expression in a de¢ned brain
area and physiological behavior. The availability of positron
emission tomography (PET) ligands would constitute a ¢rst
alternative to postmortem hybridization or biochemical stud-
ies and make it possible to correlate nicotine labeling with
cognitive or behavioral status.

Initial studies performed with radioactive nicotine have re-
vealed the inadequacy of this compound in PET studies and
that more speci¢c ligands must be identi¢ed to obtain proper
spatial and temporal resolutions [11,12]. The identi¢cation of
a new frog toxin, epibatidine, which displays a very high af-
¢nity for the neuronal nAChRs, opened new possibilities to
design synthetic compounds of potential interest for PET
monitoring [13^17]. We produced N-methylated epibatidine
for this purpose [13]. Recently this and other methylated epi-
batidine analogs were synthesized, with very similar biological
properties (e.g. receptor a¤nity and toxicity) [16]. The aim of
this work is to examine the properties of the optical isomers of
N-methylepibatidine on reconstituted neuronal nAChRs and
to determine their potential application for PET investiga-
tions.

2. Materials and methods

2.1. Chemistry
All chemicals were purchased in analytical quality from Aldrich,

Fluka, Merck or Sigma unless otherwise mentioned. The 1H NMR
and 13C NMR spectra were recorded on Varian Gemini 200 and
Gemini 300 spectrometers. Mass spectra were recorded on a Trio
2000 spectrometer (VG Organic, UK) using the positive ion mode
with electrospray (ES+).

The separation of the enantiomers was performed by HPLC on a
semi-preparative ChirobioticT column (250U10 mm) from Astec and
Merck LaChrom equipment (D-7000, L-7100, L-7250, L-7450). Meth-
anol was used in preparative HPLC quality from Merck. Optical
rotations were measured with a Perkin Elmer 241 polarimeter at
23³C using a sodium lamp (D-line) and are reported as speci¢c rota-
tion [K] in degrees. The purity is reported in percent of enantiomeric
excess (%ee). The 11C-labeled compounds were produced as described
in detail elsewhere [18] by methylation of the desmethyl precursor
with [11C]MeI.

2.2. PET studies in the rat
Male Sprague-Dawley rats (300^400 g) were anesthetized with Do-

mitor and Hypnorm (0.25 ml/kg body weight), according to the pro-
cedure required by the veterinary department. Anesthesia was main-
tained during the study by subcutaneous injection of Hypnorm/
Domitor (0.12 ml/kg body weight). A catheter was introduced into
the tail vein. The rat was placed in the PET scanner (GE-Advance)
with the whole body of the rat in the ¢eld of view. Prior to the PET

0014-5793 / 99 / $20.00 ß 1999 Federation of European Biochemical Societies. All rights reserved.
PII : S 0 0 1 4 - 5 7 9 3 ( 9 9 ) 0 0 4 7 3 - 1

*Corresponding author. Fax : (41) (22) 702 54 02.
E-mail : bertrand@cmu.unige.ch

Abbreviations : ACh, acetylcholine ; nAChR, nicotinic receptor ; CNS
central nervous system ; PNS, peripheral nervous system

FEBS 21938 4-5-99 Cyaan Magenta Geel Zwart

FEBS 21938 FEBS Letters 450 (1999) 273^279



scans a transmission scan was performed. An amount of 1^2 MBq of
the 11C compound was injected. Data acquisition was started and 22
frames were collected 10U0.5 min, 5U1 min, 2U5 min, 4U10 min
and 1U20 min.

Regions of interest (ROI) were drawn directly on the `transaxial'
PET scans, which are sagittal as a result of the positioning of the rat
in the ¢eld of view. The activity in the ROIs was calculated from the
absolute count rate by multiplication with area and slice thickness
(0.425 cm). The result is displayed as percent injected dose. The
amount of activity injected into the rat was calculated from a whole
body ROI and the percentage of 11C tracer in the brain was deter-
mined as a function of midframe time. The result of the whole body
ROI was compared to the amount of activity injected into the rat
determined by measuring the syringe before and after injection in a
dose calibrator. The values determined by whole body ROI were in
agreement with the 11C dose determined with the dose calibrator.

2.3. Electrophysiology
Xenopus oocytes were prepared according to the standard proce-

dure [19] and nuclear-injected with 10 nl of bu¡er containing equal
concentrations (0.1 Wg/Wl) of K and L subunit cDNAs or 0.2 Wg/Wl of
K7. Recordings were made using the two-electrode voltage clamp
technique. During the experiments, oocytes were continuously super-
fused with control solution containing : 82.5 mM NaCl, 2.5 mM KCl,
5 mM HEPES, 2.5 mM CaCl2, 1 mM MgCl2, pH 7.4 adjusted with
NaOH. All chemicals were obtained from either Fluka or Sigma
(Buchs, Switzerland).

Dose-response relationships were adjusted to the empirical Hill
equation

y� 1
1� EC50

x

� �nH �1�
Fig. 1. Chemical structure of epibatidine and its methyl derivatives.
The upper panel presents the two enantiomers of epibatidine or its
methyl derivative. Chemical synthesis of the radiolabeled compound
is schematized in the lower panel.

Fig. 2. Currents evoked by epibatidine and its derivatives on the major brain K4L2 nAChR. A : Typical currents evoked in a responsive oocyte
by short application (2 s) of ACh, (+)-epibatidine and methyl-(+)-epibatidine at saturating concentration. B : Dose-response relationships for
(+)-epibatidine and methyl-(+)-epibatidine (mean of four cells). Lines through the data points correspond to the best ¢ts obtained with Eq. 1.
C : Currents evoked, in another cell, as in A for ACh, (3)-epibatidine and methyl-(3)-epibatidine. D : Dose-response relationships for (3)-epi-
batidine and methyl-(3)-epibatidine (mean of four cells, values in Table 1). Data in B and D were normalized with respect to currents evoked
by a saturating ACh concentration.

FEBS 21938 4-5-99 Cyaan Magenta Geel Zwart

S. Bertrand et al./FEBS Letters 450 (1999) 273^279274



where y = the fraction of activated current, EC50 = concentration of
agonist evoking a current of half maximal amplitude, nH = Hill coef-
¢cient, x = agonist concentration.

Dose-response curves for (+)- and (3)-epibatidine and their analogs
were normalized on an acetylcholine (ACh) saturating concentration.
All values indicated throughout the text are given with their respective
S.E.M. Cells were held throughout the experiments at 3100 mV.

2.4. Toxicology
The acute intravenous toxicity of methylepibatidine was estimated

in mice and rats by a specialized outside agency (RCC Research and
Consulting Co. Ltd., Itingen, Switzerland). A racemic mixture of
methylepibatidine was intravenously administered to groups of ¢ve
male and ¢ve female animals. The doses were 0.04, 0.4, 8 and 80
Wg/kg for the HanIbm :WIST (SPF) rats and 0.05, 0.5, 10 and 100
Wg/kg for the HanIbm :NMRI (SPF) mice. Surviving animals were
examined four to ¢ve times on day 1 and once daily on days 2^15
for clinical signs. Mortality/viability were recorded together with clin-
ical signs at the same time intervals. All animals were necropsied and
examined macroscopically.

3. Results and discussion

Neuronal nAChRs are integral membrane proteins that are
thought to result from the assembly of ¢ve subunits around
an axis of pseudosymmetry [20]. Earlier physiological and
pharmacological experiments have revealed that ganglionic
receptors from the peripheral nervous system (PNS) display
patterns of agonists and antagonists clearly distinguishable
from those of the muscle receptor [2,20]. Reinforced by
many recent studies, these di¡erences can be attributed to a

particular expression of neuronal subtypes (reviewed in [20]).
Furthermore, it was shown that ganglionic receptors contain
at least the K3 and L4 subunits while central nervous system
(CNS) receptors are thought to result from the assembly of K4
with L2 [7,21]. A less clear distinction can be made when
considering receptors containing the K7 subunit which have
been shown to be present in both the CNS and the PNS [6,22^
24]. To examine the e¡ects of epibatidine and its N-methyl
derivatives on the CNS and PNS nAChR types we determined
the action of these compounds on neuronal receptors recon-
stituted in Xenopus oocytes.

In a recent contribution Horti et al. [16] studied N-alkylated
epibatidine derivatives with various halogens, £uorine, bro-
mine and iodine, at the position of the chlorine on the pyr-
idine ring. They found a¤nities to rat brain homogenates,
mouse brain distributions and toxicity towards mice which
were all very similar. We concentrated on the di¡erences be-
tween the enantiomeric N-methyl compounds in their dynamic
behavior in the rat brain and in their activation of various
nAChR subtypes.

A schematic representation of the two enantiomers of epi-
batidine and N-methylepibatidine employed throughout this
work is given in Fig. 1.

3.1. Sensitivity of the rat K4L2 to epibatidine and
methylepibatidine

Oocytes were injected with rat cDNA combinations K4L2
and their physiological properties were examined using a dual-

Fig. 3. Currents evoked by epibatidine and its derivatives on the ganglionic K3L4 nAChR. A and C : Typical currents evoked in a responsive
oocyte by a short application (2 s) of ACh, epibatidine and its derivative at saturating concentration as in Fig. 2A,C. B : Dose-response rela-
tionships for (+)-epibatidine and methyl-(+)-epibatidine (mean of four cells). Lines through the data points correspond to the best ¢ts obtained
with Eq. 1. D : Dose-response relationships for (3)-epibatidine and methyl-(3)-epibatidine (mean of four cells, values are indicated in Table 1).
All procedures were as described in Fig. 2.

FEBS 21938 4-5-99 Cyaan Magenta Geel Zwart

S. Bertrand et al./FEBS Letters 450 (1999) 273^279 275



electrode voltage clamp. Typical currents evoked, in a respon-
sive oocyte, by acetylcholine (ACh), (+)-epibatidine and meth-
yl-(+)-epibatidine are illustrated in Fig. 2A. The equal ampli-
tude of currents evoked by these saturating test pulses
indicates that these three compounds act as full agonists on
the CNS receptor. Determination of dose-response curves
(Fig. 2B) over a broad range of agonist concentrations re-
vealed that the K4L2 receptor is about 15-fold more sensitive
to epibatidine than its equivalent methyl derivative. When the
same experiments were repeated with the (3) enantiomers a
di¡erent picture was observed. Although no di¡erences in
amplitude could be observed between the currents evoked
by ACh and (3)-epibatidine a small reduction was observed
with methyl-(3)-epibatidine (Fig. 2C). On average, however,
no signi¢cant di¡erences could be detected between these
three agonists. The small reduction observed on these currents
can be partly attributed to the fastest desensitization of the
receptor to methyl-(3)-epibatidine. Surprisingly, however,
dose-response relationships (Fig. 2D) yielded no di¡erences
in apparent a¤nity between (3)-epibatidine and methyl-(3)-
epibatidine indicating that introduction of the methyl side
chain does not alter the receptor sensitivity for the (3) enan-
tiomer (data summarized in Table 1).

3.2. Sensitivity of the ganglionic K3L4 epibatidine and its
derivatives

Analysis of the currents evoked by the two enantiomers of
epibatidine and methylepibatidine were carried out as de-
scribed above for the CNS nAChR subtype. As shown in

Fig. 3A, the (+) enantiomers of epibatidine and methylepiba-
tidine both evoked robust currents of comparable amplitude
to the natural agonist ACh. Moreover, as seen for the K4L2
nAChR, the K3L4 receptor displays a sensitivity roughly 30
times lower for the (+)-methyl compound than for the native
toxin (Fig. 3B). Assessing the properties of the (3) enantiom-
ers yielded results comparable to those observed with the
K4L2 nAChR. Namely, while both epibatidine and methyl-
epibatidine evoked currents of similar amplitude as those re-
corded in response to ACh (Fig. 3C), exposure to (3)-epiba-
tidine and methyl-(3)-epibatidine yielded approximately
similar half activation values (Fig. 3D). Thus, whereas this
class of receptor does not distinguish between the (+)- and
(3)-epibatidine enantiomers it is about 31-fold more sensitive
to the (3) form of the methyl derivative than to its (+) enan-
tiomer (see Table 1).

3.3. Sensitivity of the homomeric K7 nAChR
It is well documented that the homomeric K7 receptor is

strikingly less sensitive to epibatidine than the heteromeric
receptors [25]. For instance, whilst the major brain receptor
displays a more than two orders of magnitude higher a¤nity
for epibatidine than for ACh, the K7 sensitivity remains in the
micromolar range. In addition, it is also known that this re-
ceptor shows a higher desensitization rate when exposed to
epibatidine than when challenged with ACh. Current traces
recorded in response to a fast application pulse display the
typical pro¢le reported earlier for the homomeric K7 receptor
[23,26,27]. Unlike the heteromeric K4L2 or K3L4 receptors,

Fig. 4. Sensitivity of the K7 nAChR to epibatidine and its methyl derivatives. Traces recorded in response to (+) enantiomers are shown in A
while currents evoked by the (3) enantiomers are shown in C. Note the di¡erences in amplitude of the currents evoked by (+)- and (3)-epiba-
tidine. B and D : Ddose-response relationships of the homomeric K7 nAChR (mean of four cells, see Table 1) to epibatidine. Data were ob-
tained and processed as in Fig. 2.

FEBS 21938 4-5-99 Cyaan Magenta Geel Zwart

S. Bertrand et al./FEBS Letters 450 (1999) 273^279276



Fig. 5. PET images of [11C]N-methyl-(+)- and (3)-epibatidine in rats. The region of the liver, urinary bladder and brain showed a high uptake
of the tracer. The (+) isomer (A) showed a higher uptake in the region of liver, stomach and spleen compared to the (3) isomer (B). Because
of the limited resolution of the PET images, only the total the brain activity curve is shown as % injected dose versus time in Fig. 6.

FEBS 21938 4-5-99 Cyaan Magenta Geel Zwart

S. Bertrand et al./FEBS Letters 450 (1999) 273^279 277



however, both (+)-epibatidine and methyl-(+)-epibatidine
evoked only a fraction of the currents caused by ACh expo-
sure (Fig. 4A). Furthermore, as illustrated in Fig. 4B, adjunc-
tion of the methyl residue further decreased the apparent af-
¢nity of the K7 receptor to epibatidine by more than one
order of magnitude. Data obtained with the (3) enantiomers
are roughly comparable to the observation made on the het-
eromeric receptors (Fig. 4C). A marked di¡erence is, however,
found in the partial agonist mode of action of the (3)-methyl
compound (see Fig. 4D and Table 1), which is not observed
with the other receptor subtypes.

Taken together these data highlight the ¢nding that intro-
duction of a methyl side chain causes a decrease in the e¤cacy
of the (+) enantiomer and introduces a stereoselectivity that is
not observed with the native epibatidine. It is of interest to
recall that although the neuronal nAChRs display a poor
selectivity in the separation of enantiomers a di¡erence be-
tween (3)- and (+)-nicotine has been widely documented.
However, another chemical modi¢cation of epibatidine, the
removal of the chlorine atom, introduced no stereoselectivity
in the molecular recognition by the receptors [28]. Thus, the
stereoselectivity caused by the introduction of a methyl sub-
stitute onto the nitrogen bridge of the bicyclocyclohexyl ring
indicates that this segment of the molecule must be in closer
interaction with the ACh binding site than the chlorine atom.

3.4. PET studies with N-[11C]methyl(+)- and (3)-epibatidine
To characterize further the possibility of using these epiba-

tidine derivatives in PET studies, a ¢rst set of experiments was
conducted on rats. Pictures obtained after injection with N-
[11C]methyl(+)-epibatidine reveal that the concentration of
this compound transiently increased in the brain and then
progressively diminished (Fig. 5A). In contrast, when the
same experiment was repeated with the (3) enantiomer a
di¡erent pattern was revealed. While the signal observed in
the liver or kidneys progressively decreased the brain uptake
continued to increase (Fig. 5B). Comparison of the PET im-
ages and the brain uptake curves showed interesting di¡er-
ences in receptor binding properties of the methylated epi-
batidine stereoisomers. [11C]N-methyl-(3)-epibatidine binds
predominantly to one region in the rat brain. In contrast,
two binding sites were seen for [11C]N-methyl-(+)-epibatidine.
The additional binding site seen very clearly for the (+) isomer
was most likely the eyes of the rat, but because of limited
resolution of the PET scan some other brain region such as
olfactory bulb or frontal cortex may also show an enrichment
of the tracer. Considerable di¡erences were seen when com-
paring the time course of brain uptake for the enantiomers of
[11C]N-methyl-epibatidine. While [11C]N-methyl-(3)-epibati-
dine showed both a fast and slower uptake and an increasing
brain activity with a time constant of about 40 s, the (+)
enantiomer showed only a transient peak and a slow decay
(Fig. 6). Washout of the ligand could not be determined on
this fast time scale measurement. A typical washout time con-
stant of 3.8 h was measured with single photon emission
computed tomography for radioiodinated epibatidine cerebel-
lum clearance in baboon [15]. The slowest clearance measured
by these authors for this compound was in the thalamus with
a clearance time of 11 h.

In the toxicity measurements no deaths occurred in the
animal population with the two lower doses (0.04, 0.4 Wg/kg
for rats and 0.05, 0.5 Wg/kg for mice), while many animals
died at the two higher doses (8, 80 Wg/kg for rats and 10,
100 Wg/kg for mice). The LD50 in rats was 20 Wg/kg (95%
con¢dence limits 8^180 Wg/kg) and for mice 15 Wg/kg (95%
con¢dence limits 3^120 Wg/kg). To break the data down for
males and females makes no statistical sense. The clinical
signs of toxicity observed in the two high dose groups in-
cluded sedation, convulsions, ventral recumbency and dys-
pnea indicating an e¡ect on the nerve system. No clinical signs
were observed in the low dose groups. No organ abnormal-
ities were observed on necroscopy.

Thus, it seems that the toxicity of N-methylepibatidine is
possibly somewhat less than that of the analogous N-methyl-
norchloro£uoroepibatidine [29]. Although adjunction of a
methyl side chain to epibatidine reduces the toxic e¡ects
somewhat, this is still not su¤cient to ensure the complete

Fig. 6. Time course of [11C]N-methyl-(+)- and (3)-epibatidine bind-
ing. Quanti¢cation of PET measurements was done as described in
Section 2. Squares indicate the values measured with [11C]N-methyl-
(+)-epibatidine while data obtained with [11C]N-methyl-(3)-epibati-
dine are represented by circles. Lines through data points were
drawn to guide the eye.

Table 1
Sensitivity of epibatidine and derivatives to rat K4L2, K3L4 and K7 nAChRs

cDNA type (+)-Epibatidine Methyl-(+)-Epibatidine (3)-Epibatidine Methyl-(3)-Epibatidine

EC50 (WM) nH EC50 (WM) nH EC50 (WM) nH EC50 (WM) nH

Rat K7 2.5 þ 016 1.45 þ 0.1 30 þ 0.24 1.60 þ 0.2 2.03 þ 0.5 1.43 þ 0.2 2.5 þ 0.26 2.73 þ 0.2
Rat K4L2 0.021 þ 0.005 1.24 þ 0.1 0.36 þ 0.11 0.45 þ 0.1 0.023 þ 0.007 1.30 þ 0.1 0.024 þ 0.008 1.25 þ 0.1
Rat K3L4 0.036 þ 0.013 1.52 þ 0.4 1.1 þ 0.2 1.60 þ 0.1 0.019 þ 0.006 1.67 þ 0.2 0.035 þ 0.017 1.70 þ 0.1

Half activation values (EC50) and Hill coe¤cients for three receptor subtypes of the four chemical compounds tested are indicated. Values are
the means of 3^7 cells with their respective S.E.M.

FEBS 21938 4-5-99 Cyaan Magenta Geel Zwart

S. Bertrand et al./FEBS Letters 450 (1999) 273^279278



safety required for human administration, which is in agree-
ment with the conclusions by Horti et al. [16]. Thus, although
indicating the feasibility of the method, new epibatidine de-
rivatives await to be designed to accommodate the properties
required both for PET and for human safety.

These data illustrate, however, the importance of isolating
the enantiomers and of physiological testing for their charac-
terization.

Acknowledgements : We are very grateful to J. Patrick for kindly pro-
viding rat cDNAs. This work was supported by grants from the O¤ce
Fëdëral de l'Education et des Sciences and the Swiss National Science
Foundation, 3100-0371.91 to D.B. and 3200-046939.96 to G.W.

References

[1] Langley, J.N. (1905) J. Physiol. 33, 374^413.
[2] Ascher, P., Large, W.A. and Rang, H.P. (1979) J. Physiol. 295,

139^170.
[3] Benowitz, N.L. (1996) Annu. Rev. Pharmacol. Toxicol. 36, 597^

613.
[4] Corrigall, W.A., Coen, K.M. and Adamson, K.L. (1994) Brain

Res. 653, 278^284.
[5] Cimino, M., Marini, P., Colombo, S., Andena, M., Cattabeni, F.,

Fornasari, D. and Clementi, F. (1995) J. Neural Transm. Gen.
Sect. 100, 77^92.

[6] Deltoro, E.D., Juiz, J.M., Smillie, F.I., Lindstrom, J. and Criado,
M. (1997) Dev. Brain Res. 98, 125^133.

[7] Wada, E., Wada, K., Boulter, J., Deneris, E., Heinemann, S.,
Patrick, J. and Swanson, L.W. (1989) J. Comp. Neurol. 284,
314^335.

[8] Nitta, A., Katono, Y., Itoh, A., Hasegawa, T. and Nabeshima,
T. (1994) Pharmacol. Biochem. Behav. 49, 807^812.

[9] Balfour, D.J. (1994) Addiction 89, 1419^1423.
[10] Zoli, M., Le, N.N., Hill, J.J. and Changeux, J.P. (1995) J. Neuro-

sci. 15, 192^1939.

[11] Nyback, H., Halldin, C., Ahlin, A., Curvall, M. and Eriksson, L.
(1994) Psychopharmacology (Berl.) 115, 31^36.

[12] Muric, R.F., Berridge, M.S., Friedland, R.P., Zhu, N. and Nel-
son, A.D. (1998) J. Nucl. Med. 39, 2048^2054.

[13] Patt, J.T., Westera, G., Buch, A., Fletcher, S.R. and Schubiger,
P.A. (1995) J. Lab. Comp. Radiopharm. 17, 355^356.

[14] Villemagne, V.L. et al. (1997) J. Nucl. Med. 38, 1737^1741.
[15] Musachio, J.L., Villemagne, V.L., Sche¡el, U., Stathis, M., Fin-

ley, P., Horti, A., London, E.D. and Dannals, R.F. (1997) Syn-
apse 26, 392^399.

[16] Horti, A.G. et al. (1998) J. Med. Chem. 41, 4199^4206.
[17] Kassiou, M., Sche¡el, U.A s , Ravert, H.T., Mathews,, W.B.,

Musachio, J.L., London, E.D. and Dannals, R.F. (1998) Life
Sci. 63, L13^18.

[18] Patt, J.T., Spang, J.E., Westera, G., Buck, A. and Schubiger,
P.A. (1999) Nucl. Med. Biol. (in press).

[19] Bertrand, D., Cooper, E., Valera, S., Rungger, D. and Ballivet,
M. (1991) Methods Neurosci. 4, 174^193.

[20] Bertrand, D. and Changeux, J.P. (1995) Semin. Neurosci. 7, 75^
90.

[21] Boyd, R.T., Jacob, M.H., McEachern, A.E., Caron, S. and Berg,
D.K. (1991) J. Neurobiol. 22, 1^14.

[22] Conroy, W.G. and Berg, D.K. (1995) J. Biol. Chem. 270, 4424^
4431.

[23] Sëguëla, P., Wadiche, J., Dineley-Miller, K., Dani, J.A. and Pat-
rick, J.W. (1993) J. Neurosci. 13, 596^604.

[24] Alkondon, M. and Albuquerque, E.X. (1995) J. Pharmacol. Exp.
Ther. 274, 771^782.

[25] Gerzanich, V., Peng, X., Wang, F., Wells, G., Anand, R., Fletch-
er, S. and Lindstrom, J. (1995) Mol. Pharmacol. 48, 774^782.

[26] Couturier, S. et al. (1990) Neuron 5, 847^856.
[27] Gopalakrishnan, M. et al. (1995) Eur. J. Pharmacol. Mol. Phar-

macol. 290, 237^246.
[28] Sche¡el, U., Taylor, G.F., Kepler, J.A., Carroll, F.I. and Kuhar,

M.J. (1995) NeuroReport 6, 2483^2488.
[29] Ding, Y.S., Molina, P.E., Fowler, J.S., Logan, J., Volkow, N.D.,

Kuhar, M.J. and Caroll, F.I. (1999) Nucl. Med. Biol. 26, 139^
148.

FEBS 21938 4-5-99 Cyaan Magenta Geel Zwart

S. Bertrand et al./FEBS Letters 450 (1999) 273^279 279