key: cord-0874571-8hwexeia
authors: Manta, Stella; Tsoukala, Evangelia; Tzioumaki, Niki; Goropevšek, Aleš; Pamulapati, Ravi Teja; Cencič, Avrelija; Balzarini, Jan; Komiotis, Dimitri
title: Dideoxy fluoro-ketopyranosyl nucleosides as potent antiviral agents: Synthesis and biological evaluation of 2,3- and 3,4-dideoxy-3-fluoro-4- and -2-keto-β-d-glucopyranosyl derivatives of N(4)-benzoyl cytosine
date: 2009-01-29
journal: Eur J Med Chem
DOI: 10.1016/j.ejmech.2009.01.020
sha: 1d94975c65d36c72fb7f6fe29e1ee26c4a432ab6
doc_id: 874571
cord_uid: 8hwexeia

The synthesis of the dideoxy fluoro ketopyranonucleoside analogues, 1-(2,3-dideoxy-3-fluoro-6-O-trityl-β-d-glycero-hexopyranosyl-4-ulose)-N(4)-benzoyl cytosine (7a), 1-(3,4-dideoxy-3-fluoro-6-O-trityl-β-d-glycero-hexopyranosyl-2-ulose)-N(4)-benzoyl cytosine (13a) and their detritylated analogues 8a and 14a, respectively, is described. Condensation of peracetylated 3-deoxy-3-fluoro-d-glucopyranose (1) with silylated N(4)-benzoyl cytosine, followed by selective deprotection and isopropylidenation afforded compound 2. Routine deoxygenation at position 2′, followed by a deprotection-selective reprotection sequence afforded the partially tritylated dideoxy nucleoside of cytosine 6, which upon oxidation of the free hydroxyl group at the 4′-position, furnished the desired tritylated 2,3-dideoxy-3-fluoro ketonucleoside 7a in equilibrium with its hydrated form 7b. Compound 2 was the starting material for the synthesis of the dideoxy fluoro ketopyranonucleoside 13a. Similarly, several subsequent protection and deprotection steps as well as routine deoxygenation at position 4′, followed by oxidation of the free hydroxyl group at the 2′-position of the partially tritylated dideoxy nucleoside 12, yielded the desired carbonyl compound 13a in equilibrium with its hydrated form 13b. Finally, trityl removal from 7a/b and 13a/b provided the unprotected 2,3-dideoxy-3-fluoro-4-keto and 3,4-dideoxy-3-fluoro-2-ketopyranonucleoside analogues 8a and 14a, in equilibrium with their gem-diol forms 8b and 14b. None of the compounds showed inhibitory activity against a wide variety of DNA and RNA viruses at subtoxic concentrations, except 7a/b that was highly efficient against rotavirus infection. Nucleoside 7a/b also exhibited cytostatic activity against cells of various cancers. BrdU-cell cycle analysis revealed that the mechanism of cytostatic activity may be related to a delay in G1/S phase and initiation of programmed cell death.

The synthesis of the dideoxy fluoro ketopyranonucleoside analogues, 1-(2,3-dideoxy-3-fluoro-6-O-tritylb-D-glycero-hexopyranosyl-4-ulose)-N 4 -benzoyl cytosine (7a), 1-(3,4-dideoxy-3-fluoro-6-O-trityl-b-Dglycero-hexopyranosyl-2-ulose)-N 4 -benzoyl cytosine (13a) and their detritylated analogues 8a and 14a, respectively, is described. Condensation of peracetylated 3-deoxy-3-fluoro-D-glucopyranose (1) with silylated N 4 -benzoyl cytosine, followed by selective deprotection and isopropylidenation afforded compound 2. Routine deoxygenation at position 2 0 , followed by a deprotection-selective reprotection sequence afforded the partially tritylated dideoxy nucleoside of cytosine 6, which upon oxidation of the free hydroxyl group at the 4 0 -position, furnished the desired tritylated 2,3-dideoxy-3-fluoro ketonucleoside 7a in equilibrium with its hydrated form 7b. Compound 2 was the starting material for the synthesis of the dideoxy fluoro ketopyranonucleoside 13a. Similarly, several subsequent protection and deprotection steps as well as routine deoxygenation at position 4 0 , followed by oxidation of the free hydroxyl group at the 2 0 -position of the partially tritylated dideoxy nucleoside 12, yielded the desired carbonyl compound 13a in equilibrium with its hydrated form 13b. Finally, trityl removal from 7a/b and 13a/b provided the unprotected 2,3-dideoxy-3-fluoro-4-keto and 3,4-dideoxy-3-fluoro-2-ketopyranonucleoside analogues 8a and 14a, in equilibrium with their gem-diol forms 8b and 14b. None of the compounds showed inhibitory activity against a wide variety of DNA and RNA viruses at subtoxic concentrations, except 7a/b that was highly efficient against rotavirus infection. Nucleoside 7a/b also exhibited cytostatic activity against cells of various cancers. BrdU-cell cycle analysis revealed that the mechanism of cytostatic activity may be related to a delay in G1/S phase and initiation of programmed cell death.

Ó 2009 Elsevier Masson SAS. All rights reserved.

Nucleosides are structural modules of nucleic acids and therefore of fundamental importance in all living systems [1, 2] . They have been playing a major role in treating tumor and virus either as selective inhibitors of certain obligatory enzymes for cancer or viral replication [3] , or as nucleic acid chain terminators, which interrupt the replication of cancer cells or a virus [4] [5] [6] . Pyrimidine nucleosides have attracted increased attention as a result of their important physiological and biogenetic properties. Among them, nucleosides bearing cytosine as a base have been developed as potent antitumor agents, which are effective not only on leukemias and lymphomas, but also on a wide variety of solid tumors in vitro as well as in vivo [7] [8] [9] [10] [11] [12] [13] .

The last decades, nucleosides with a six-membered carbohydrate moiety have been evaluated for their potential antiviral [14] [15] [16] [17] [18] , antibiotic [19] and antioxidant [20] properties and as building blocks in nucleic acid synthesis [21, 22] . A series of sixmembered nucleoside analogues, the ketonucleosides, proved to be key intermediates in synthetic and biosynthetic processes and exhibited interesting antitumor and antiviral properties [23] [24] [25] [26] [27] . These nucleosides are known to inhibit DNA, RNA, and protein synthesis [28] , to interact with sulfhydryl groups of cellular proteins and enzymes [29] , while they were found to be highly cytotoxic in vitro [24, 30, 31] and to exert powerful inhibitory action against L1210 leukemia in vivo [32, 33] .

In view of these facts and taking into account that introduction of fluorine into the sugar moiety of nucleosides enhances activity [34] [35] [36] [37] [38] [39] , we have previously synthesized various unsaturated fluoro ketopyranonucleoside analogues, which proved to be efficient as tumor growth inhibitors and showed to have a promising potential in combating the rotaviral infections [40, 41] .

Thus, in continuation of our studies on the synthesis of sugarmodified pyranonucleosides and based on the interesting biological properties of deoxyhexopyranose nucleosides [14] [15] [16] [42] [43] [44] [45] , we decided to design and synthesize a new class of fluoro ketopyranonucleoside analogues, the fluorinated 2,3-dideoxy-4-keto and 3,4-dideoxy-2-keto nucleosides of N 4 -benzoyl cytosine, in order to assess their biological activity. Chemical synthesis and biological activity of these compounds are presented herein.

Two different synthetic routes were investigated for the synthesis of the title compounds, 1-(2,3-dideoxy-3-fluoro-b-D-glycero-hexopyranosyl-4-ulose)-N 4 -benzoyl cytosine (8a) and 1-(3,4-dideoxy-3-fluoro-b-D-glycero-hexopyranosyl-2-ulose)-N 4 -benzoyl cytosine (14a). In the first reaction scheme, condensation of 3-deoxy-3-fluoro-1,2,4,6-tetra-O-acetyl-glucopyranose (1) [46, 47] (Scheme 1) with silylated N 4 -benzoyl cytosine followed by selective deprotection and specific acetalation using 2,2-dimethoxypropane in N,N-dimethylformamide (DMF), gave the 4 0 ,6 0 -isopropylidene derivative 2 [40, 41] .

Phenoxythiocarbonylation of 2 under a commonly used condition, phenyl chlorothionoformate (PhOC(S)Cl) and an excess amount of 4-dimethylaminopyridine (DMAP) in acetonitrile [48] [49] [50] , afforded the 2 0 -O-phenoxythiocarbonyl derivative 3, in 70% yield. The deoxygenation of compound 3 was carried out with tri-nbutyltin hydride (Bu 3 SnH) to afford the 2 0 -deoxynucleoside 4, in 60% yield [50, 51] . The 4 0 ,6 0 -O-isopropylidene group of 4 was removed upon treatment with aqueous trifluoroacetic acid (TFA) and selective protection of the primary 6 0 -hydroxyl group with a trityl group yielded the partially protected derivative 6, in 65% yield. Oxidation of the fluoro tritylated dideoxy precursor 6 was performed with pyridinium dichromate (PDC)/acetic anhydride (Ac 2 O) and afforded 4 0 -keto form, 1-(2,3-dideoxy-3-fluoro-6-O-trityl-b-D-glycero-hexopyranosyl-4-ulose)-N 4 -benzoyl cytosine (7a), in equilibrium with its hydrated analogue 7b in a 3:7 ratio, as it was clearly determined by integration of the related 1 H NMR signals. It appears that the presence of the electron-withdrawing fluorine atom a to the carbonyl, causes the easy hydration and gem-diol formation [52] . Finally, detritylation of 7a/b in CH 2 Cl 2 /formic acid, 1:1, led to the corresponding dideoxy carbonyl compound, 1-(2,3-dideoxy-3-fluoro-b-D-glycero-hexopyranosyl-4-ulose)-N 4 -benzoyl cytosine (8a) in equilibrium with its gem-diol 8b.

Compound 2 [40, 41] was the starting material for the synthesis of 1-(3,4-dideoxy-3-fluoro-b-D-glycero-hexopyranosyl-2-ulose)-N 4 -benzoyl cytosine (14a) (Scheme 2). Acetylation of the nucleoside 2, followed by deisopropylidenation and finally by selective tritylation of the primary hydroxyl group gave compound 9 [41] , which was then converted to the 4 0 -deoxy nucleoside analogue 11 via 4 0 -O-phenoxythiocarbonyl derivative 10 by the similar methodology as described above. Deacetylation of 11 was performed using NaOH/ethanol (EtOH)/pyridine [53] to afford compound 12, in good yield (88%). Oxidation of the free hydroxyl group in 2 0 -position of the sugar moiety of nucleoside 12 with PDC/Ac 2 O afforded the desired 2 0 -keto form, 1-(3,4-dideoxy-3-fluoro-6-O-trityl-b-D-glycero-hexopyranosyl-2-ulose)-N 4 -benzoyl cytosine (13a) in equilibrium with its hydrated analogue 13b in a 6:4 ratio, clearly determined by integration of the related 1 NMR signals. Finally, deprotection of the primary hydroxyl group of 13a/b was performed using CH 2 Cl 2 /formic acid, 1:1, and furnished ketone 14a in equilibrium with its gem-diol 14b. It should be mentioned that all attempts to remove the benzoyl group of the target nucleoside analogues, were unsuccessful and only degradation products were obtained [23, 54, 55] .

All compounds were well characterized by NMR and UV spectroscopies, mass spectrometry and elemental analysis. The 1 H NMR spectrum of 7a/b showed a doublet at 6.35 ppm (J 1 0 ,2 0 ¼ 10.1 Hz) and a doublet of doublet of doublets at 5.22 ppm (J 3 0 ,F ¼ 47.0 Hz, J 2 0 a,3 0 ¼ J 2 0 b,3 0 ¼ 6.9 Hz), which are assigned to H-1 0 and H-3 0 protons, respectively, of keto compound 7a, as well as a doublet at 5.94 ppm (J 1 0 ,2 0 ¼ 10.6 Hz) and a doublet of doublet of doublets at 4.72 ppm (J 3 0 ,F ¼ 48.1 Hz, J 2 0 a,3 0 ¼ 5.0 Hz, J 2 0 b,3 0 ¼ 5.1 Hz), assigned to H-1 0 and H-3 0 protons, respectively, of gem-diol 7b. Similarly, in the 1 H NMR spectrum of 13a/b, H-1 0 signals of keto 13a and gem-diol form 13b appeared as two sharp singlets at 6.41 and 5.86 ppm, respectively and H-3 0 signals of 13a and 13b appeared as two doublet of doublet of doublets at 5.41 ppm (J 3 0 ,F ¼ 47.5 Hz, J 3 0 ,4 0 a ¼ J 3 0 ,4 0 b ¼ 7.4 Hz) and 4.82 ppm (J 3 0 ,F ¼ 48.3 Hz, J 3 0 ,4 0 a ¼ 5.5 Hz, J 3 0 ,4 0 b ¼ 5.3 Hz), respectively. Thus, in both cases, 1 H NMR spectra confirmed the equilibrium between keto and gemdiol forms. However, 1 H NMR spectra of 8a/b as well as 14a/b in methanol-d 4 (CD 3 OD) solution, showed a doublet at 5.94 ppm (J 1 0 ,2 0 ¼ 11.0 Hz) assigned to H-1 0 proton for 8b and a sharp singlet at 5.85 ppm, assigned to H-1 0 proton for 14b, while H-3 0 signals appeared as a doublet of doublet of doublets, for both compounds, at 4.72 ppm (J 3 0 ,F ¼ 48.4 Hz, J 2 0 a,3 0 ¼ 4.8 Hz, J 2 0 b,3 0 ¼ 4.9 Hz) and 4.94 ppm (J 3 0 ,F ¼ 48.5 Hz, J 3 0 ,4 0 a ¼ 5.5 Hz, J 3 0 ,4 0 b ¼ 5.6 Hz), respectively, indicating the presence of gem-diol forms exclusively. Finally, the target compounds 7a/b, 8a/b and 13a/b, 14a/b were characterized by the absence of the H-4 0 proton and H-2 0 proton respectively, in their 1 H NMR spectra. The results of the antiviral assays against rotavirus on the newly synthesized dideoxy fluoro-pyranonucleoside analogues 7a/b, 13a/ b and their detritylated analogues 8a/b and 14a/b, respectively, are summarized and compared with AZT in Table 1 .

In the neutralization assay, 8a/b was found to neutralize rotavirus before its attachment to Caco-2 cells at the concentration of 2.1 mM (vide IC 50 ) as compared to AZT (75 mM). Similar results were observed for 7a/b in inhibiting infectivity, following virus attachment (Table 1) ; analogue 7a/b was actually more potent than its congener 8a/b as it produced the same effect at a much lower concentration (0.63 mM) (treatment B). Interestingly, both 7a/b and its detritylated analogue 8a/b, although active in the same antiviral assay, seem to exert their action by a different mechanism. In contrast to compounds 7a/b and 8a/b, analogues 13a/b and 14a/b did not show any antiviral potency. We speculate that the position of the keto group in the ring skeletons of compounds 7a/ b and 8a/b is critical in inhibiting the rotavirus replication in the cell as well as for neutralizing rotavirus before its attachment. According to the antiviral assay used, it can be suggested that the tested compounds 7a/b and 8a/b showed specific activity against rotavirus, and were capable of inhibiting the rotavirus infection in cells at significantly lower concentrations than AZT.

Compared to the recently reported unsaturated fluoro ketopyranonucleoside analogues [40, 41] , the dideoxy fluoro ketopyranonucleosides presented herein, were found to be more potent and thus more suitable candidates in anti-rotavirus therapy. Thus, it can be suggested that the incorporation of the carbon-carbon double bond into the ring system of the dideoxy fluoro ketopyranonucleosides has a marked reducing effect on their antiviral activity. It seems that this structural feature of the carbohydrate moiety influences the ability of the nucleoside analogue to participate in the antiviral mechanism.

The cytotoxic potential of compounds 7a/b, 8a/b, 13a/b and 14a/ b was studied on H4 normal human intestinal cells and HEL human lung fibroblast cells, and on a series of various other human tumor cells, such as human colonic adenocarcinoma derived Caco-2 cells, gastric cancer derived AGS cells, and epithelial breast cancer derived MCF-7 cells, Vero African green monkey cells and HeLa human cervix carcinoma cells. The growth inhibition of Caco-2 cells, caused by the new compounds, was measured by determining the 50% growth inhibitory concentration (IC 50 ). The results are summarized in Table 2 and compared with the values obtained for 5-fluorouracil (5FU).

All of the tested compounds were cytotoxic on malignant cells of different origin and had a substantial tumor selectivity index ( Table 2) .

From the tested dideoxy fluoro-pyranonucleoside analogues, compounds 7a/b and 13a/b exhibited a noteworthy cytotoxicity, with good selectivity indices, particularly on AGS cells.

In comparison to the control compound, 5FU, compounds 7a/b, 13a/b and 14a/b were more effective on MCF-7, AGS and Caco-2 cells; compound 8a/b was more effective than 5FU only on AGS cells.

From the results obtained, it is clear that some of the new compounds show selectivity toward specific tumor cell lines. This is not unprecedented as there are examples in the literature, which demonstrate that the mode of inhibitory action especially on the target enzymes in carcinogenic cells is not always similar even among nucleoside antimetabolites which have the same nucleoside base [56] .

In the growth inhibition assay compounds 7a/b and 13a/b were capable of inhibiting the growth of malignant cells in a range comparable to 5FU (IC 50, Table 2 ). In spite of the fact that differences were observed among the tested compounds in the growth inhibition activity in colon carcinoma cells, all analogues were strong inhibitors of malignant cell growth of other cancer types.

Conversely to the previously reported unsaturated fluoro ketopyranonucleoside analogues [40, 41] , the new dideoxy fluoro-pyranonucleosides, especially compounds 7a/b and 13a/b, showed satisfactory, but similar, cytotoxicity against Caco-2 cells and MCF-7 cells. It appears that the existence of a carbon-carbon double bond in the sugar residue does not seem to be a structural prerequisite for cytotoxic and growth inhibition activity. On the contrary, the presence of the keto-fluoro system is indispensable.

Many mechanisms have been implicated in 5FU resistance and/ or sensitivity, such as pharmacokinetic resistance, decreased accumulation of activated metabolites, and altered effects on thymidylate synthase (TS) [57, 58] . Because fluoropyrimidine treatment leads to cell cycle arrest [59, 60] , other parameters, such as cell cycle distribution can also be used to predict ketonucleosides sensitivity. In order to find out, whether the mechanism of antitumor activity and/or resistance to dideoxy fluoro ketopyranonucleoside analogues was similar to the mechanism of action of 5FU, the influences on cell cycle distribution and induction of apoptosis were studied by BrdU-cell cycle analysis. Our results clearly showed that in Caco-2 cell line, 5FU equitoxic treatment resulted in an increase of the percentage of cells in G0/G1 and in % decrease of cells in the G2/M phase (Table 3 , compounds 7a/b and 13a/b, Fig. 1 ). However, BrdU analysis showed that a cell cycle delay took place specifically at the G1/S interface as previously reported for 5FU (Table 3 , Fig. 1 ) [59, 60] . Exposure of cancer cells to compounds 7a/b, 13a/b and 5FU led to a % BrdU-positive cells or labeling index (LI) increase. However, LI modifications were only related to the G1/S phase accumulation, and not to a real enhanced S-phase fraction. In fact, BrdU analysis showed that G1 cells, as determined by 7-AAD analysis, were composed of BrdU-positive and BrdU-negative cell subpopulations (Fig. 1) . These subpopulations cannot be distinguished using 7-AAD analysis alone. This is viewed as a consequence of the cell fraction that was able to incorporate BrdU and maintain its G1 DNA content. This fraction was considered to be in phase G1, by 7-AAD analysis, whereas it was included in the labeled subpopulation by BrdU analysis. Thus, cells performing repair synthesis of DNA, with no net DNA increase, would be included in the LI score, but could still have a DNA content that was indistinguishable from G1 cells.

Synchronization of cells, caused by a delay in G1/S, may be related to the programmed cell death initiation. Induced delay in G1/S was representative of apoptosis involvement and of 5FU treatment sensibility. Actually, sub-G1 peak, representative of apoptotic cells was detected using flow cytometry already in untreated cells but was enhanced in cells treated with compounds 7a/b and 5FU.

The apoptotic activity of the tested compounds was observed in the p53-mutated (Caco-2) cell line, suggesting that ketonucleosideinduced apoptosis is mediated by a p53-independent mechanism. This finding could have important implications in a clinical setting, since many tumors have p53 alterations and mutated p53 has been shown to be associated with a poor response to chemotherapy.

In summary, the newly synthesized fluorinated 2,3-dideoxy-4keto and 3,4-dideoxy-2-keto nucleosides of N 4 -benzoyl cytosine showed good antiviral activity toward rotavirus in very low concentrations as compared to AZT or/and previously synthesized corresponding unsaturated fluoro ketopyranonucleoside analogues [40, 41] . We speculate that modification in sugar component has a very important role in this case. Furthermore, all newly synthesized compounds showed antitumor activity against tested cell lines, especially gastric cancer derived AGS cells. Compounds 7a/b, 13a/b and 14a/b were even more cytotoxic to all cancer cell lines tested than control compound 5FU. Compounds 7a/b and 13a/ b also showed to be as effective in their antitumor activity and in growth suppression of colon carcinoma cells as previously synthesized corresponding unsaturated fluoro ketopyranonucleoside analogues [40, 41] .

We have also shown, using flow cytometric BrdU-cell cycle analysis, that the mechanism of antitumor activity of the newly synthesized compounds may be related to delay in G1/S and programmed cell death initiation. Moreover, our results indicated that compound 7a/b may have strong implication in clinical settings, especially for treatment of those tumors poorly responsive to the existing chemotherapy agents.

In view of these considerations, there has been increasing evidence suggesting a close correlation between the biological activity and the presence of the a-fluorocarbonyl moiety. However, the mechanism of action of the newly synthesized nucleosides is thus far unclear. It would therefore be interesting to further explore the structure-activity relationship by modifying the base part (keeping or removing the benzyl part) as well as replacing the fluorine group by different chemical entities.

Melting points were recorded in a Mel-Temp apparatus and are uncorrected. Thin-layer chromatography (TLC) was performed on Merck precoated 60F 254 plates. Reactions were monitored by TLC on silica gel, with detection by UV light (254 nm) or by charring with sulfuric acid. Flash column chromatography was performed using silica gel (240-400 mesh, Merck). 1 H, 19 F and 13 C NMR spectra were obtained at room temperature with a Bruker 400 spectrometer at 400, 376 and 100 MHz, respectively using CDCl 3 and CD 3 OD with internal tetramethylsilane (TMS) for 1 H and 13 C and internal trifluorotoluene (TFT) for 19 F.

The chemical shifts are expressed in parts per million (d) and following abbreviations were used: s ¼ singlet, br s ¼ broad singlet, d ¼ doublet, ddd ¼ doublet doublet doublet, dtr ¼ doublet triplet, dq ¼ doublet quartet and m ¼ multiplet. Mass spectra were obtained with a Micromass Platform LC (ESI-MS). Optical rotations were measured using Autopol I polarimeter. All reactions were carried out in dry solvents. CH 2 Cl 2 was distilled from phosphorous pentoxide and stored over 4 Å molecular sieves. CH 3 CN and toluene were distilled from calcium hydride and stored over 3 Å molecular sieves. Pyridine was stored over pellets of potassium hydroxide. All reactions sensitive to oxygen or moisture were carried out under nitrogen atmosphere. stirred for 90 min at room temperature. The reaction was quenched by adding water (1.1 mL). The solvent was removed under reduced pressure, the obtained residue was dissolved in EtOAc (500 mL), washed with water (1 Â 30 mL), dried over anhydrous magnesium sulfate (MgSO 4 ), evaporated under reduced pressure, and purified by flash column chromatography using ethyl acetate-n-hexane (4:6) as eluant to afford 3 as a yellow syrup. Yield: 2.72 g (70%); 

Compound 3 (2.72 g, 4.9 mmol) was coevaporated three times with anhydrous toluene, dissolved in toluene (106 mL), and degassed with nitrogen for 30 min. To this solution were added Bu 3 SnH (3.95 mL, 14.7 mmol) and 1,1-azobis(cyclohexane carbonitrile) (0.36 g, 1.47 mmol) and the mixture was heated to 80 C for 3 h. After cooling to room temperature, the mixture was evaporated, the residue was purified by flash column chromatography using ethyl acetate-n-hexane (6:4) 

To a solution of compound 5 (0.77 g, 2.12 mmol) in dry pyridine (10.6 mL) was added triphenylmethyl chloride (0.89 g, 3.18 mmol) and a catalytic amount of DMAP. The mixture was stirred at 100 C for 1 h and then concentrated. Purification by flash column chromatography using ethyl acetate-n-hexane 

A mixture of 6 (0.83 g, 1.38 mmol; dried by co-evaporation with toluene), PDC (0.78 g, 2.07 mmol) and Ac 2 O (0.65 mL, 6.9 mmol) was refluxed at 55 C in dry CH 2 Cl 2 (12 mL) for 90 min. After cooling, ethyl acetate (1.6 mL) was added and the resulting slurry was transferred on the top of a silica-gel column packed in ethyl acetate. The solution was filtered through the column and washed with ethyl acetate (30 mL) until the product was eluted completely. The solvent was evaporated and the residue was rendered free of Ac 2 O and pyridine by co-evaporation with dry toluene (3Â). Finally, purification by flash column chromatography using ethyl acetate-nhexane ( 

Compound 7a/b (0.42 g, 0.69 mmol) was dissolved in a mixture of CH 2 Cl 2 (2.45 mL) and formic acid (2.45 mL, 90%). The solution was stirred for 20 min at room temperature, diluted with toluene, and co-distilled several times with the same solvent to avoid ester formation [61] . The concentrated residue was purified by flash column chromatography using ethyl acetate-methanol (9.8:0.2) as eluant and compound 8a/b was obtained as a white powder. Yield: 0.14 g (52%); R f ¼ 0. 15 

Compound 11 (0.83 g,1.28 mmol) was dissolved in EtOH-pyridine (12.8 þ 3.85 mL), 2 M NaOH (0.86 mL) was added and the mixture stirred for 1 h at 0 C. Amberlite IR-120 (H þ ) was added to neutralize the base. The suspension was filtered, the resin was washed with EtOH and pyridine (10 þ 10 mL) and the filtrate was evaporated. The solid residue was purified by flash column chromatography using ethyl acetate as eluant and 12 was obtained as a yellowish foam. Yield Cytosine derivative 14a/b was synthesized from 13a/b by the same methodology as described for the synthesis of 8a/b. 

Dideoxy fluoro ketonucleosides were freshly prepared in sterile dimethyl sulfoxide (DMSO) at the concentration of 0.5 mg/mL. The final concentration of DMSO was below 0.1% of cell culture medium. All solutions were protected against light.

AZT (Retrovir Ò ) GlaxoSmithKline, USA, a drug used for antiretroviral therapy (ART) was used as a standard compound in biological experiments, prepared in the same way as dideoxy fluoro ketonucleosides.

5FU (fluorouracilum), Lederle Arzeinmittel GmbH, BRD, a drug used for antitumor therapy was used as standard compound (control) in growth inhibition and cytotoxicity assessment of compounds, prepared in the same way as dideoxy fluoro ketonucleosides.

The rotavirus RF strain was propagated in Caco-2 cells in the presence of trypsin (1 mg per mL of DMEM) as described previously [62] . Supernatant containing the virus was collected from the flasks when cytopathic effect (CPE) was observed (24-48 h at 37 C) by microscopy and clarified by centrifugation. Virus was stored at À70 C until used.

The potential antiviral activity of the newly synthesized compounds was tested against rotavirus by investigating:

a) The inhibition of infectivity following virus attachment: Washed monolayers of Caco-2 and H4 cells were first incubated with rotavirus for 1 h at 37 C in the atmosphere of 5% CO 2 (time for virus to attach to cell membrane receptors). After incubation the remaining virus was washed off with DMEM without supplements and the monolayer was treated immediately with the tested compounds or AZT added in 3-fold serial dilutions (initial concentration of 0.5 mg/mL). After 72 h of incubation for rotavirus, the plates were stained with Crystal Violet in EtOH, rinsed with water, and destained with 10% (v/v) acetic acid. The A 590 was measured, and the results were expressed, for each dilution, by the mean ratios (%, AESD) of absorbances in virus-infected wells (n ¼ 6) to those in control (only virus infected) wells (n ¼ 6). The 50% inhibitory concentration (IC 50 ) of the tested compounds was obtained from the concentration-effect curve. b) The neutralization of the virus in solution before attachment: 3fold dilutions of each of the tested compounds or AZT (initial concentration of 0.5 mg/mL) were first co-incubated with rotavirus in DMEM supplemented with trypsin for 12 h prior the infection of Caco-2 and H4 cell monolayer at 37 C and 5% CO 2 . Residual viral infectivity was measured after 72 h post infection. Rotavirus alone was treated in the same way as the control. After 72 h of incubation, the plates were stained with Crystal Violet in EtOH, rinsed with water, and destained with 10% (v/v) acetic acid. The A 590 was measured, and the results were expressed, for each dilution, by the mean ratios (%, AESD) of absorbencies in virus-infected wells (n ¼ 6) to those in control (only virus infected) wells (n ¼ 6). The 50% inhibitory concentration (IC 50 ) of the tested compounds was obtained from the concentration-effect curve.

Caco-2 cells, H4 cells, MCF-7 (breast carcinoma) and gastric cancer derived AGS cells (6X10 6 cells per plate) were seeded in P96 plates and treated with the compounds or AZT at 3-fold serial dilutions of each compound (initial concentration of 0.5 mg/mL). Then, the cells were incubated at 37 C in the humidified incubator for 72 h. The plates were stained with Crystal Violet in EtOH, rinsed with water, and destained with 10% (v/v) acetic acid. The A 590 was measured, and the results were expressed, for each dilution, by the mean ratios (%, AESD) of absorbances in treated wells (n ¼ 2) to those in control wells (n ¼ 24). The 50% cytotoxic concentration (CC 50 ) of the tested compounds was obtained from the concentration-effect curve.

It was performed on Caco-2 cell line by modified method described previously [63] . Briefly, in 96-well plates, six wells of 2-fold dilutions of each compound or 5FU (initial concentration of 0.5 mg/ mL) were applied to 10 cells/well in Dulbecco's modified Eagle's medium (DMEM)/10% fetal bovine serum. Incubation was performed at 37 C in the humidified incubator for 10 days. The colonies were counted in each well and the results were expressed, for each dilution, by the mean ratios (%, AESD) of colony number in treated wells (n ¼ 2) to those in control wells (n ¼ 24). The 50% growth inhibitory concentration (IC 50 ) of the tested compounds was obtained from the concentration-effect curve.

Caco-2 cells in Dulbecco's modified Eagle's medium (DMEM)/ 10% fetal bovine serum were seeded in P96 plates (1 Â10 6 cells per well) and treated with the compounds at concentrations yielding 50% growth inhibition (IC 50 ) or left untreated. Then, the cells were incubated at 37 C in the humidified incubator for 24 h. During the last hour of incubation cells were pulsed with 10 mM BrdU solution.

Samples for flow cytometry were processed using FITC-BrdU flow kit (BD Biosciences) according to the manufacturer protocol. Flow cytometry was performed on BD LSR II (BD Biosciences) instrument. The data were analyzed using FACSDiva software (BD Biosciences) on cells gated according to their (FSC versus SSC) scatter properties. Bivariate distributions of BrdU content (FITC) versus DNA content (7-AAD) were analyzed. Cell cycle distribution in G0/G1, S, G2/M phases and % of apoptotic (sub G1) cells was determined according to gating represented in Fig. 1 . Also the G1/S subpopulation, corresponding to BrdU-positive cells containing G1 DNA, was determined. Labeling index (LI) corresponded to S phase or percentage of BrdU-positive cells (Fig. 1 ).

The compounds were evaluated against the following viruses: herpes simplex virus type 1 (HSV-1) strain KOS, thymidine kinasedeficient (TK À ) HSV-1 KOS strain resistant to ACV (ACV r ), herpes simplex virus type 2 (HSV-2) strains Lyons and G, vaccinia virus Lederle strain, respiratory syncytial virus (RSV) strain Long, vesicular stomatitis virus (VSV), Coxsackie B4, Parainfluenza-3, Reovirus-1, Sindbis, Punta Toro, influenza virus type A (H1N1, H3N2) and type B and feline corona virus. The antiviral, other than anti-HIV, assays were based on inhibition of virus-induced cytopathicity in human embryonic lung (HEL) fibroblasts, African green monkey cells (Vero), human epithelial cervix carcinoma cells (HeLa), Crandel feline kidney cells (CFKC) and Madin-Darby canine kidney cells (MDCK). Briefly, confluent cell cultures in microtiter 96-well plates were inoculated with 100 CCID50 of virus (1 CCID 50 being the virus dose to infect 50% of the cell cultures). After a 1-2 h adsorption period, residual virus was removed, and the cell cultures were incubated in the presence of varying concentrations of the test compounds. Viral cytopathicity was recorded as soon as it reached completion in the control virus-infected cell cultures that were not treated with the test compounds. Antiviral activity was expressed as the EC 50 or compound concentration required to reduce virus-induced cytopathicity by 50%.

Human CEM cell cultures (w3 Â 10 5 cells mL À1 ) were infected with w100 CCID50 HIV-1(III B ) or HIV-2(ROD) per mL and seeded in 96-well (200 mL/well) microtiter plates, containing appropriate dilutions of the test compounds. After 4 days of incubation at 37 C, syncytia formation was examined microscopically in the CEM cell cultures.

The cytostatic concentration (for MDCK and CRFK cells) was calculated as the CC 50 , or the compound concentration required to reduce cell proliferation by 50% relative to the number of cells in the untreated controls. CC 50 values were estimated from graphic plots of the number of cells (percentage of control) as a function of the concentration of the test compounds. Alternatively, cytotoxicity of the test compounds (for HEL, Vero and HeLa cells) was expressed as the minimum cytotoxic concentration (MCC) or the compound concentration that caused a microscopically detectable alteration of cell morphology.

Antimicrob

This work was supported in part by the Greek General Secretariat for Research and Technology Program No. 3245C and bilateral scientific-technology project between Greece and Slovenia: ''Synthesis and biological evaluation as potential anticancerogenic and antiviral agents of amino-fluoronucleosides of adenine and cytosine''. The authors are grateful to Dr. K. Antonakis (Ecole Nat. Sup. de Chimie de Paris; CNRS-FRE 2463) for encouraging this work, to Dr. T. Halmos (Ec. Nat. Sup. de Chimie de Paris) for the unstinted and fruitful aid, to Lidija Gadišnik, Leentje Persoons, Frieda Demeyer, Vicky Broeckx, A. Camps, S. Carmans and L. Van den Heurck for the excellent technical performance as well as to ELPEN pharmaceuticals for financial support.