key: cord-0808455-15yhf955
authors: Abdou, S.; Collomb, J.; Sallas, F.; Marsura, A.; Finance, C.
title: Beta-cyclodextrin derivatives as carriers to enhance the antiviral activity of an antisense oligonucleotide directed toward a coronavirus intergenic consensus sequence
date: 2014-04-05
journal: Arch Virol
DOI: 10.1007/s007050050182
sha: 8b0dc9983d3905e515b2feb7ded520dcc7a4057d
doc_id: 808455
cord_uid: 15yhf955

The ability of cyclodextrins to enhance the antiviral activity of a phosphodiester oligodeoxynucleotide has been investigated. Aff18-mer oligodeoxynucleotide complementary to the initiation region of the mRNA coding for the spike protein and containing the intergenic consensus sequence of an enteric coronavirus has been tested for antiviral action against virus growth in human adenocarcinoma cells. The phosphodiester oligodeoxynucleotide only showed a limited effect on virus growth rate (from 12 to 34% viral inhibition in cells treated with 7.5 to 25 μM oligodeoxynucleotide, respectively, at a multiplicity of infection of 0.1 infectious particle per cell). In the same conditions, the phosphorothioate analogue exhibited stronger antiviral activity, the inhibition increased from 56 to 90%. The inhibitory effect of this analogue was antisense and sequence-specific. Northern blot analysis showed that the sequence-dependent mechanism of action appears to be the inhibition of mRNA transcription. We conclude that the coronavirus intergenic consensus sequence is a good target for an antisense oligonucleotide antiviral action. The properties of the phosphodiester oligonucleotide was improved after its complexation with cyclodextrins. The most important increase of the antiviral activity (90% inhibition) was obtained with only 7.5 μM oligonucleotide complexed to a cyclodextrin derivative, 6-deoxy-6-S-β-D-galactopyranosyl-6-thio-cyclomaltoheptaose in a molar ratio of 1:100. These studies suggest that the use of cyclodextrin derivatives as carrier for phosphodiester oligonucleotides delivery may be an effective method for increasing the therapeutic potential of these compounds in viral infections.

The use of antisense oligonucleotides is an interesting approach to study cellular and viral gene function and to block gene expression [2, 3, 60] . Oligonucleotides have been considered as ideal agents for inhibiting viral replication due in part to their high speci®city for viral DNA and RNA sequences. These agents have been successfully employed to block the replication and cellular expression of several viruses, including herpes simplex virus [20] , cytomegalovirus [43] and especially human immunode®ciency virus [33] . However, the vulnerability of oligonuleotides to nucleases' attack, combined with their intracellular distribution and uptake properties, have limited their therapeutic potential. To avoid the problem of nucleases' digestion and to improve the cellular uptake of oligonucleotides, either chemical modi®cations of the internucleoside phosphodiester linkages [35] or of the ribose sugar [25] and various oligonucleotide conjugates have been tested. One chemical analog that is widely employed for enhanced stability is the phosphorothioate oligonucleotide. But, although this analog has increased resistance to degradation in vivo and in vitro [1, 46, 59] it has exhibited sequence-speci®c and non sequence-speci®c activities [11, 54, 55] . The oligonucleotide conjugates include oligonucleotide adsorbed onto poly (isobutylcyanoacrylate) nanoparticles [40] , encapsulated in liposomes [23] or conjugated to the cholesterol [18, 26] . These strategies have resulted in an increase of the cellular uptake of oligonucleotide and / or stability against nucleases. However, the inherent properties of the cholesterol, liposomes or nanoparticles carriers led to increase the immunogenicity of the oligonucleotide complex and the tendency of the complex to be phagocytosed by the reticuloendothelial system [44] .

Investigations in cyclodextrin chemistry ®eld have been increased since several decades. Cyclodextrins are molecular hosts acting as carriers. It frameworks describes a group of cyclic polysaccharides containing 6 to 8 glucose units in a- (1, 4) linkage (a-CD, b-CD and g-CD have respectively 6, 7 or 8 glucose units). They possess an hydrophobic internal cavity which can include various small molecules and a peripheral structure containing a large number of hydroxyl groups making the molecule water soluble. The cyclodextrins form complexes with numerous compounds and b-cyclodextrin is the most employed for numerous applications. A wide variety of cyclodextrins and derivatives has been studied to enhance the solubility and the availability of various water-soluble and water-insoluble pharmacological agents [6, 9, 21, 38, 64] . This approach using cyclodextrin derivatives has been recently studied to increase the cellular uptake of phosphorothioate oligonucleotides [65] . In this work, we have investigated the ability of native b-and gcyclodextrins and b-cyclodextrin derivatives as carriers to enhance the antiviral activity of a phosphodiester oligonucleotide directed toward the initiation region of the mRNA coding for the spike protein and containing the intergenic consensus sequence of an enteric coronavirus.

The native cyclodextrins were purchased from Roquette-fre Áres (Dijon, France) for bcyclodextrin (b-CD) and from Wacker-chemie GmbH, (Lyon, France) for g-cyclodextrin (g-CD). The chemically modi®ed cyclodextrins were synthetised in our laboratory according to [17] for 6-deoxy-6-S-b-D-galactopyranosyl-6-thio-cyclomaltoheptaose (S-Gal-b-CD) and heptakis (6-deoxy-6-S-b-D-galactopyranosyl)-6-thio-cyclomaltoheptaose (7(S-Gal)-b-CD) or to [39] for heptakis [6-deoxy-6-(1-2-amino) ethylamino]-cyclomaltoheptaose (7(ethane-diamino)-b-CD). All the derivatives were soluble in distilled water and sterilized by ®ltration through a 0.22 mm membrane (Millipore, France) before use. Their chemical structures are presented in Fig. 1 .

Oligonucleotides were purchased from Appligene (Strasbourg, France). An antisense sequence (3 H TGGTATTAGATTTGTACA 5 H ), synthetized as natural phosphodiester (O À oligo) and phosphorothioate analog (S À oligo), a sense sequence (5 H ACCATAATC-TAAACATGT 3 H ), synthetized as phosphorothioate analog (S oligo) and a random sequence (5 H GAATTGTGATATTTCATG 3 H ), synthetized as phosphorothioate analog (S r oligo) have been tested. All these oligonucleotides were puri®ed by HPLC, lyophilized and the powder was dissolved in distilled water.

The F15 strain of Bovine Enteric Coronavirus (BECV) was obtained from Laporte (INRA, Jouy-en-Josas, France) and propagated in the human rectal adenocarcinoma tumor cell line (HRT-18) grown in RPMI 1640 medium (ICN Biomedicals Inc., Costa Mesa, CA, USA) supplemented with 2% foetal calf serum (FCS) (Life Technologies, Eragny, France), 10 mg/ ml o¯oxacin as antibiotic (Roussel Uclaf, Romainville, France) and incubated at 37 C for 3 or 4 days in a humidi®ed atmosphere containing 5 % CO 2 [30, 31] .

The complexes were prepared taking in account the b-and -cyclodextrins solubility data according to Higuchi and Connors [19] . Cyclodextrins are regular cyclic oligosaccharidyl molecules exhibiting a tronconic ring which are for simplicity represented by a truncated cone with the wider side formed by the secondary 2-and 3-hydroxyl groups and the narrower side by the primary 6-hydroxyl A volume of 7.5 mM phosphodiester oligonucleotide (O À oligo) was added to 75 mM b-CD, 75 mM g-CD, 7.5 to 750 mM S-Gal-b-CD, 75 mM 7(S-Gal)-b-CD or 75 mM 7(ethanediamino)-b-CD in a 0.5 ml total volume of RPMI 1640 medium. The solutions were mixed overnight at 25 C by vortexing at 200 rpm with a Rotatest 74 401 apparatus (Bioblock-Scienti®c, Illkirch, France) and the resulting complexes were immediately tested [51] . Complex formation has been controlled by circular dichroõ Èsm measurement and observation of oligonucleotide absorbance intensity at 260 nm (data not shown).

The cellular viability was evaluated by the modi®ed MTT assay as previously described [4] . HRT-18 cells were grown at con¯uence in 96-well¯at bottom culture plates (Costar, Cambridge, MA, USA) at 4Â10 4 cells per well and treated with varying concentrations of oligonucleotides (1 to 100 mM) or native cyclodextrins (1 to 8 mM) or CD derivatives (0.1 to 1 mM) for 16 h at 37 C. At the end of the incubation period, the cytotoxic effect of the drug was evaluated by determination of the mitochondrial dehydrogenase activity using the MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium Bromide (Sigma Chemical Co., St Louis, MO, USA). The absorbance of the solubilized material was determined using a Titertek Multiscan MCC/340 MK II apparatus (Labsystems, Helsinki, Finland). A cytotoxic effect corresponds to a decrease in purple formazan production. Results shown are the average from three experiments.

HRT-18 cells were exposed for 1 h at 37 C to various concentrations of free oligonucleotides or oligonucleotides complexed with cyclodextrins and, at the same time, were inoculated with a BECV suspension at multiplicities of infection (MOI) of 0.1, 1 or 10 PFU / cell (Plaque Forming Units / cell). At 16 h post-infection, the virus titration was performed by the plaque forming assay using crystal violet for visualization, as previously described [30, 31, 61] . In interpreting viral titers, con®dence intervals were determined and were taken into consideration [7] .

Intracellular RNAs were extracted from infected HRT-18 cells by the acid guanidinium thiocyanate-phenol-chloroform method [12] using a commercial``Message systems''kit (Bioprobe Systems, Montreuil-sous-Bois, France). The RNA pellets were dissolved in nuclease free water and adjusted to 2.2 M formaldehyde ± 50% formamide in MOPS (morpholinepropane sulfonic acid) buffer (20 mM MOPS pH 7.0, 5 mM sodium acetate, 1 mM EDTA). Samples were denatured at 65 C for 10 min and analyzed by electrophoresis in 1 % agarose gel and 2.2 M formaldehyde ± MOPS buffer as previously described [37] . After electrophoresis, the gel was treated with 50 mM NaOH for 20 min, soaked in DEPC (diethylpyrocarbonate) treated water, then in 20 Â SSPE (1 Â SSPE is 3.6 M NaCl, 0.2 M sodium phosphate pH 7.7, 0.02 M EDTA) for 45 min. RNAs were transfered overnight on a hybond-N nylon membrane (Amersham, Buckinghamshire, England) in 20 Â SSPE, then baked at 80 C for 2 h. Nylon membrane was prehybridized for 2 h at 42 C in plastic bags containing a solution of 50% formamide ± 5 Â Denhardt's solution (1 Â is 0.02% Ficoll, 0.02% polyvinylpyrrolidine, 0.02% bovine serum albumine) ± 0.5% SDS ± 5 Â SSPE and 200 mg/ml of sonicated calf thymus DNA. Hybridization was done overnight in the same solution at 42 C using a [ 32 P] DNA probe prepared by random-primed labelling of gel puri®ed BECV 3 H end speci®c cDNA fragment (1.8 Kb) as previously described [13, 47] . The membrane was washed twice at room temperature for 10 min in 2 Â SSPE, 0.1% SDS then once at 65 C for 15 min in 1 Â SSPE, 0.1% SDS and then once at 65 C for 10 min in 0.1 SSPE, 0.1% SDS. The membrane was air-dried and autoradiographed on X-ray ®lm (Kodak X Omat AR-5, N.Y, USA).

In¯uence of oligonucleotides on the cell viability These effects were evaluated by the MTT assay after 16 h of treatment of the HRT-18 cells with increasing concentrations (1 to 100 mM) of each oligonucleotide. Figure 2 shows that S r , S À and O À oligonucleotides exhibited no inhibitory effect on the viability of the cells at any concentration tested.

The antiviral activities of the antisense phosphodiester (O À oligo) and its phosphorothioate analog (S À oligo) were evaluated as described in Materials and methods. HRT-18 cells were exposed to various oligonucleotide concentrations (2.5 mM to 25 mM) and, at the same time, were infected by a BECV suspension at a MOI of 0.1 PFU/cell for 1 h at 37 C. At 16 h postinfection the virus titers of the treated infected cells were determined comparatively to those obtained with control infected cells. We observed In¯uence of oligonucleotides on cell viability. Cells grown to con¯uence (4.10 4 cells/well) were treated for 16 h with S r (1), S À (2) and O À (3) oligonucleotides at concentrations ranging from 1 to 100 mM. Cell viability was evaluated by their dehydrogenase activity (MTT assay). Each point is the mean of three independent experiments performed in triplicate and error bars represent standard error of mean ( Fig. 3 ) about 88 to 66% of viral production when cells were treated with 7.5 to 25 mM of O À oligo. In contrast the S À oligo signi®cantly reduced the ability of BECV to form plaques on HRT-18 cells. When cells were treated with 7.5 to 25 mM of the S À oligo, from 44 to 10% of viral growth were observed comparatively to control infected cells. These results con®rm that the inhibition is dose-dependent and that the phosphorothioate analogue exhibited a more pronounced antiviral activity than the phosphodiester oligodeoxynucleotide.

In¯uence of the viral multiplicity of infection The antiviral activity of 25 mM of S À oligo in presence of various viral MOI (0.1, 1 or 10 PFU/cell) was also evaluated. The results are presented on Fig. 4 . Respectively 10% or 34% of virus production were observed in cultures infected with 0.1 or 1 PFU/cell. In contrast in presence of an MOI of 10 PFU/ cell, the use of 25 mM of S À oligo inhibit weakly the virus infectivity, we obtained 66% of viral growth.

Sequence speci®c effect of the phosphorothioate oligonucleotide on the viral growth In order to study the speci®city of the antiviral activity of the phosphorothioate analogue, sense (S oligo), antisense (S À oligo), and random (S r oligo) oligonucleotides were comparatively tested. Using oligonucleotide concentrations of 15 or 25 mM and a MOI of 0.1 PFU/cell we showed (Fig. 5 ) that in presence of S r oligo and S oligo, no signi®cant antiviral effect was observed, whereas 25 mM S À oligo allowed only 10% of viral production. These results con®rm the antisense sequence speci®city of the S À oligo antiviral activity. (MOI of 0.1 PFU/cell) or mock-infected (medium containing 2% FCS) were exposed to a 25 mM either S À oligo, S r oligo or S oligo. Each oligonucleotide and the viral suspension were added together at the same time of infection. After 7 h post-infection intracellular RNA extracts were separated by electrophoresis, transfered to a membrane and hybridized to a cDNA probe representing the 3 H end of the BECV genome (including almost all the N and a part of the M gene) [13] Autoradiograms (Fig. 6) showed that in infected cells, 7 RNA species could be detected at 7 h post-infection. These mRNAs were transcribed in nonequimolar amounts. They were named mRNA 2 to 7 (2, 2a, 3, 4, 5, 6, 7) according to decreasing order of size coding for viral proteins ns2, HE, S, ns3, ns4, M and N respectively, as already described [24] . Although the oligonucleotide had a sequence complementary to mRNA3, the synthesis of all mRNAs was signi®cantly reduced when cells were treated with S À oligo. In the same conditions, mRNA pro®les or production were not affected when cells were treated with sense oligo (S oligo) and random oligo (S r oligo) comparatively to untreated infected cells.

In¯uence of cyclodextrins on the cell viability The effect of native CD and cyclodextrin derivatives on cell viability have been tested. Native b-CD and g-CD (Fig. 7) did not inhibit the cellular viability at concentrations varying from 1 to 8 mM. On the contrary, the b-CD stimulated the cell dehydrogenase activity, we obtained an increase of cell viability with concentrations ranging from 1 to 4 mM. Concerning the CD derivatives, it can be observed from Fig. 8 that S-Gal-b-CD and 7(S-Gal)-b-CD exhibited no Fig. 6 . Northern blot analysis of subgenomic RNAs. Intracellular RNAs were extracted at 7 h post-infection, separated by formaldehyde-1% agarose gel electrophoresis and transfered to a nylon membrane. The probe was prepared by random primed [ 32 P] labelling of enteric coronavirus-speci®c cDNA fragment corresponding to the 3 H end of genomic RNA. Autoradiograms were exposed overnight at À80 C with an intensifying screen. Infected cells (MOI 0.1 PFU/cell) (1), infected cells treated with S r oligo (2) or S oligo (3) or S À oligo (4). Each oligonucleotide was used at a concentration of 25 mM Table 1 ) that with O À oligo the viral production was 75% which was similar to previously obtained (88%). After complexation with cyclodextrins, the inhibition was improved, 50% of viral production was obtained with [O À oligo-

In¯uence of the molar ratio between oligonucleotide and S-Gal-b-CD on the improvement of the antiviral activity We have evaluated the in¯uence of the molar ratio between the O À oligo and the S-Gal-b-CD on the antiviral activity of the phosphodiester oligonucleotide. O À oligo was used at a concentration of 7.5 mM and S-Gal-b-CD at concentrations ranging from 7.5 mM to 750 mM. We showed (Fig. 9 ) that at molar ratio 

The effect of the S-Gal-b-CD on the viral multiplication was evaluated. The results ( Table 2) showed that no inhibitory effect in viral production was HRT-18 cells were inoculated with viral suspension (MOI 0.1 PFU/cell) in presence of S-Gal-b-cyclodextrin at various concentrations. After 16 h post-infection the virus titers were determined by the plaque forming assay. L 1 Lower 95% limit of con®dence interval, L u upper 95% limit of con®dence interval b-cyclodextrin derivatives observed. Nevertheless with 750 mM, there was a signi®cant stimulation of about 180% in viral growth.

Coronaviruses which have an original replication procedure can be candidates for antisense strategy. These viruses cause acute and / or persistent infections and might be etiological agents for autoimmune diseases [16, 41, 45] . They are enveloped viruses which contain a single-stranded, positive-sensed polyadenylated genome. They possess three or four major structural proteins. All coronaviruses have spike (S), membrane (M) and nucleocapsid (N) proteins and some have a haemagglutinin-esterase (HE) protein. In addition, there are at least four genes that encode nonstructural proteins (ns1 to 4) [29, 58] . During the replication in infected cells, genomic RNA is transcribed into genomesized, negative-stranded RNA which serves as template for the synthesis of a 3 Hcoterminal nested set of 7 to 8 subgenomic mRNAs [52, 56] and only the 5 Hproximal region which is not present in the next smaller RNA of the set is used for translation [48] . Another unique feature of coronavirus mRNAs is that the 5 H ends of genomic RNA and of all mRNAs contain a 72±77 nucleotide-long leader sequence [27, 28] . Studies made on the murine coronavirus mouse hepatitis [10, 50] have shown that the 3 H end of the leader sequence contains a repeated UCUAA pentanucleotide sequence included in the unique UCUAAAC consensus sequence (or similar sequence) observed in all intergenic regions of genomic RNA. Thus the mRNA transcriptional start is the result of the interaction between the leader sequence of the positive-stranded RNA and the complementary antileader sequence of the negative-stranded RNA [49, 50] .

This intergenic consensus sequence, common to the transcription initiation sites of mRNAs could be an interesting target for an antisense oligonucleotide. We have targeted it by a 18-mer antisense oligodeoxynucleotide complementary to the initiation region of mRNA coding for S protein and which contains the unique UCUAAAC intergenic consensus sequence which is the promoter of genomic RNA and subgenomic mRNAs transcription [36] .

We have shown that 25 mM phosphorothioate antisense (S À oligo) oligonucleotide is able to inhibit 90% of enteric coronavirus growth in HRT-18 cells at a MOI of 0.1. The intensity of the inhibitory effect of S À oligo is a function of its concentration and of the multiplicity of viral infection. In previous studies, ef®cient inhibition of gene expression by antisense sequence in cells requiring oligonucleotide concentrations in the range of 1 to 100 mM have been reported. The great varieties in the biological models investigated render comparisons dif®cult [3, 8, 20, 32, 33, 43, 57] .

In the same experimental conditions, only 34% of the viral production is inhibited by the unmodi®ed phosphodiester antisense (O À oligo) oligonucleotide. The absence of a more signi®cant inhibitory effect of this oligonucleotide may be related to its rapid degradation by nucleases and to its reduced cellular uptake. Our observations about the cellular uptake of¯uorescein-conjugated oligonucleotides have revealed (data not shown) that only 1 mM of S À oligo is quickly (`1 h) internalized by HRT-18 cells and that¯uorescence is present both in the cytoplasm and in the nucleus of cells. In contrast, intracellular localization of even 20 mM of O À oligo can not be noticed. Studies have shown [8, 22, 62] that the uptake of oligonucleotides is a complex process which depends on many factors including cell type, structure and chemical modi®cations of oligonucleotides, temperature and time of incubation. This uptake is a saturable process that probably occurs through an 80 kD protein but the true role of this protein is still unknown [5, 34] . The stronger inhibitory effect observed with S À oligo could be related to an increase in stability and in lipophilicity resulting from the replacement of the oxygen atom by sulfur [15] . It has thus been reported that phosphorothioate oligonucleotides are more strongly associated with cell membranes than the corresponding phosphodiester [8] .

As previously observed [3, 20, 32, 33, 43] the antiviral activity of S À oligo is not due to an oligonucleotide induced cytotoxicity. Assays to measure host-cell viability reveal that the oligonucleotides tested do not affect the level of cellular metabolism even at high concentration (100 mM) which is 4-fold in excess of inhibitory concentration (25 mM) .

It is noteworthy that non-sequence-speci®c oligonucleotides can inhibit HIV replication especially at high concentrations [53±55]. So, a random unrelated phosphorothioate oligonucleotide (S r oligo) with the same base composition as S À oligo but without homology with any known sequence in genomic RNA shows no activity. Therefore, the inhibition of virus growth does not seem to be due to a non-sequence-speci®c effect. In addition a sense oligonucleotide (S oligo) at any concentration employed does not inhibit virus growth showing that the antiviral mechanism of action of S À oligo is antisense. These results are con®rmed by northern blotting analysis in which all transcripts are signi®cantly reduced in infected cells when 25 mM of S À oligo is used. In contrast RNAs levels are unaffected by non-speci®c control (S r oligo) or by sense (S oligo) phosphorothioate oligonucleotides. In addition, northern blotting shows that the antisense phosphorothioate S À oligo acts through a transcription inhibition mechanism in which the duplex made by hybridization of positive-stranded genomic RNA and S À oligo stops the synthesis of the full length negative stranded-RNA preventing so a subgenomic and genomic RNA synthesis. The non-inhibitory effect of the sense oligonucleotide could be explained by considering the viral replication mechanism. The transcription of the mRNA is initiated after the hybridization between the leader and anti-leader sequences. The anti-leader sequence contains the anti-consensus sequence which is the target of the sense oligonucleotide. Consequently, we assume that a competition occurs between the leader sequence and the sense oligonucleotide for hybridization to their target. This competition might be in favour of the leader sequence mainly due to its total complementarity to its target and due also to a highest af®nity. Our results provide evidence that the coronavirus intergenic consensus sequence of the positive-stranded genomic RNA is a good target for an antisense oligonucleotide antiviral action. The inhibiting effect on virus growth is mediated by the ability of the oligonucleotide to interfere with mRNA transcription.

However, some potential disadvantages may be inherent to the use of phosphorothioate analogs. There is no assurance that substitution of a phosphorothioate for a phosphodiester in an oligonucleotide would preserve its ability to bind with similar high af®nity to its target. In addition studies have shown [14] that the sulfur on phosphorothioate may provide a novel reactive center on a DNA molecule for drug interactions and this interaction may affect the ef®cacy of the antisense compound. It may also act by viral polymerase inhibition or by interference with viral adsorption or penetration or by inhibition of cellular enzymes or growth factors [55] . Moreover commercially phosphorothioates tend to be more expensive than the phosphodiester oligonucleotides. To avoid these problems, we attempted to improve the properties of the phosphodiester oligonucleotide by complexation with cyclodextrin supermolecules. We compared the biological activity of the phosphodiester antisense oligonucleotide (O À oligo) either free or complexed with native b-CD and -CD or b-cyclodextrin derivatives. When the O À oligo is complexed either with native b-, g-CD, or with a 7(ethane-diamino)-b-CD derivative, 50% decrease in viral multiplication is observed. The inhibition is higher (75%) after its complexation with S-Gal-b-CD or 7(S-Gal)-b-CD. Moreover, the antiviral activity of O À oligo complexed with S-Gal-b-CD is improved to 90% if the complexation is performed in the [O À oligo-CD] molar ratio of [1:100; 7.5 mM: 750 mM]. The same level of inhibition is obtained in presence of 25 mM of S À oligo, a concentration three fold up to that used in the complex.

It has been demonstrated in the last year that substituted cyclodextrins were no longer biological amorphous compounds, but could exhibit biological activities such as antiviral activity [42, 63] . We checked that the chemical modi®cation of b-CD does not induce any cyclodextrin antiviral nor anticellular effect. S-Gal-b-CD induces no inhibitory effect on the viral multiplication at the concentrations used for the preparation of the complexes. Moreover, assays to measure cell viability show no cytotoxic action of the cyclodextrins. These observations exclude the possibility that the antiviral activity obtained with the [O À oligo-S-Gal-b-CD] complexes is due to an effect of the S-Gal-b-CD itself.

From these results it might be concluded that the complexation of the oligonucleotide with the cyclodextrins protect the O À oligo from its degradation by the nucleases perhaps by protecting the phosphodiester group. Cellular uptake studies recently reported that the¯uorescent intensity of (FITC)conjugated phosphorothioate oligonucleotides was greatly increased when the oligonucleotide was complexed with cyclodextrin derivatives, indicating that complexation increases cellular uptake [65] . In our case it might be expected that the chemical modi®cation of the native b-CD by substitution of the hydroxyl of the primary alcohol of one glucose unit on the ring with a thiogalactose residue improves the cellular uptake of the O À oligo by a ligand-carbohydrate cell receptors interaction. As the surface of numerous normal and malignant eukaryotic cells are covered with galactose-speci®c lectins, galactose-substituted b-CD might be used as drug carriers for targeting oligonucleotides in vivo. Another possibility which may also contribute to the enhancement of the antiviral activity of O À oligo complexed to cyclodextrin is that the complexation could modify the intracellular distribution of the antisense oligonucleotide. Probably oligonucleotide-cyclodextrin complexes enter the cells by a process different from endocytosis pathway in which the oligonucleotides may be trapped within endosomal or lysosomal vesicles. Further studies have to be achieved to study the intracellular distribution of oligonucleotide after treatment with [oligonucleotide-cyclodextrin] complexes and to elucidate the mechanism of the interaction between the oligodeoxynucleotide and cyclodextrin.

Modi®ed oligonucleotides as therapeutic and diagnostic agents

Antisense oligonucleotides: towards clinical trials

Phosphorothioate oligonucleotides inhibit the replication of lentiviruses and type D retroviruses, but not that of type C retroviruses

Recognition ability and cytotoxicity of some oligosaccharidyl substituted b-cyclodextrins

Cationic lipids enhance cellular uptake and activity of phosphorothioate antisense oligonucleotides

Piroxicam b-cyclodextrine: ame Âlioration de la cine Âtique des me Âdicaments par les cyclodextrines

Viruses in water systems: detection and identi®ca-tion

Modi®ed oligonucleotides in rabbit reticulocytes: uptake, stability and antisense properties

The potential use of cyclodextrins in parenteral formulations

Three intergenic regions of nucleotide sequence that is homologous to the 3 H end of the viral mRNA leader sequence

Potent inhibition of Epstein-Barr virus by phosphorothioate oligodeoxynucleotides without sequence speci®cation

Single-step method for RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction

Radioactive and enzymatic cloned cDNA probes for bovine enteric coronavirus detection by molecular hybridation

Reaction between metabolically activated acetaminophen and phosphorothioate oligonucleotides

In vitro toxicology and pharmacokinetics of antisense oligonucleotides

Characterization of monoclonal and polyclonal antibodies of bovine enteric coronaviruses: establishment of an ef®cient Elisa for antigen detection in feces

Synthesis of new oligosacchridyl-thio-b-cyclodextrins (CDS): A novel family of potent drug-targetting vectors

Effect of structural variations in cholesteryl conjugated oligonucleotides on inhibitory activity toward HIV-1. Bioconjugate Chem

Phase-solubility techniques. In: Reilley CN (ed) Advances in analytical chemistry and instrumentation

Inhibition of viral growth by an -oligonucleotide directed to the splice junction of herpes simplex virus type 1 immediate early pre-mRNA species 22 and 34

Beta-cyclodextrins as vehicles in eye-drop formulations: an evaluation of their effects on rabbit corneal epithelium

Cellular uptake of antisense oligodeoxynucleotides

Liposomes as a drug delivery system for antisense oligonucleotides

Temporal regulation of bovine coronavirus RNA synthesis

Synthesis and hybridization properties of oligonucleotides containing 2 H -O-modi®ed ribonucleotides

Modi®cation of antisense phosphodiester oligodeoxynucleotides by a 5 H -cholesteryl moiety, increases cellular association and improves ef®cacy

Characterization of leader RNA sequences on the virion and mRNAs of mouse hepatitis virus, a cytoplasmic RNA virus

Presence of leader sequences in the mRNA of mouse hepatitis virus

Mouse hepatitis virus A59: mRNA structure and genetic localization of the sequence divergence from hepatotropic strain MHV-3

Une ligne Âe cellulaire particulie Árement sensible a Á la re Âplication du Coronavirus ente Ârique bovin : les cellules HRT-18

vitro culture of bovine enteritic coronavirus (BEC). Les colloques de l'INSERM. Viral Enteritis

Inhibition of in¯uenza virus replication by phophorothioate oligodeoxynucleotides

Speci®c inhibition of human immunode®ciency virus type 1 replication by antisense oligonucleotides: an in vitro model for treatment

Characterization of oligonucleotide transport into living cells

Inhibition of HIV-1 in vitro by C-5 propyne phosphorothioate antisense to rev

Effect of intergenic consensus sequence¯anking sequences on coronavirus transcription

A system for study of coronavirus mRNA results from intergenic site insertion

Piroxicam b-cyclodextrin in the symptomatic treatment of primary headache. Meeting of the Interdisciplinare di Cardiologia e Neurologia Selecta Neurologia 449±453

Catalytic Properties of Cu (II) complex with a modi®ed cyclodextrin

Pharmacokinetics and biodistri± bution of oligonucleotide adsorbed onto poly (isobutylcyanoacrylate) nano particles after intravenous administration in mice

Bovine viral diarrhoea virus-speci®c bovine monoclonal antibody

Modi®ed cyclodextrin sulphates (mCDS11) have potent inhibitory activity against HIV and high oral bioavailability

Potent antiviral activity of an antisense oligo± nucleotide complementary to the intron-exon boundary of human cytomegalovirus genes UL 36 and UL 37

Immunogenicity of immunoliposomes

Experimental production of bovine respiratory tract disease with bovine diarrhea virus

Biodistribution and metabolism of internally 3 H-labelled oligonucleotides

Sequence and analysis of BECV F15 matrix protein

Coronavirus transcription: subgenomic mouse hepatitis virus replicative intermediates function in RNA synthesis

Minus-strand copies of replicating coronavirus mRNA contain anti-leaders

The 5 H -end sequence of the murine coronavirus genome implications for multiple fusion sites in leader-primed transcription

Solubilization of ovine growth hormone with 2-hydroxypropyl-bcyclodextrin

Coronavirus: structure and genome expression

Phosphorothioate oligodeoxynucleotides are potent sequence non-speci®c inhibitors of de novo infection by HIV

Does antisense exist

Phosphorothioate antisense oligodeoxynucleotides: questions of speci®city

Coronavirus multiplication strategy. II. Mapping the avian infectious bronchitis virus intracellular RNA species to the genome

Reduction of expression of the multidrug resistance protein (MRP) in human tumor cells by antisense phosphorothioate oligonucleotides

The molecular biology of coronaviruses

Pharmacokinetics, biodistribution and stability of capped oligodeoxynucleotide phosphorothioates in mice

Antisense nucleic-acids. Prospects for antiviral intervention

Plaque assay for titration of bovine enteric coronavirus

Transport of oligonucleotides across natural and model membranes

Synthetic cyclodextrin derivatives inhibit HIV infection in vitro

Cyclodextrin as building blocks for supramolecular structures and functional units

Use of cyclodextrins and its derivatives as carriers for oligonucleotide delivery

This research was supported by the French Ministe Áre de la Recherche et des Enseignements Supe Ârieurs, the Re Âgion Lorraine and the French Association pour la Recherche Me Âdicale.

Authors' address: Dr. C. Finance, GEVSM, EA 1123, Univ. H. Poincare Â, Faculte  de Pharmacie, B.P. 403, F-54001 Nancy Cedex, France.Received December 13, 1996