3'-Dideoxy-S-fluoro-3'-thiacytidine (FTC) has been shown to be a potent and selective compound against human immunodeficiency virus type 1 in acutely infected primary human lymphocytes. FTC is also active against human immunodeficiency virus type 2, simian immunodeficiency virus, and feline immunodeficiency virus in various cell culture systems, including human monocytes. The antiviral activity can be prevented by 2'-deoxycytidine, but not by other natural nucleosides, suggesting that FTC must be phosphorylated to be active and 2'-deoxycytidine kinase is responsible for the phosphorylation. By using chiral columns or enzymatic techniques, the two enantiomers ofFTC were separated. The (-)-(o-enantiomer of FTC was about 20-fold more potent than the (+)-,i-enantiomer against human immunodeficiency virus type 1 in peripheral blood mononuclear cells and was also effective in thymidine kinase-deficient CEM cells. Racemic FEC and its enantiomers were nontoxic to human lymphocytes and other cell lines at concentrations of up to 100 ,uM. Studies with human bone marrow cells indicated that racemic FTC and its (-)-enantiomer had a median inhibitory concentration of >30 FiM. The (+)-enantiomer was significantly more toxic than the (-)-enantiomer to myeloid progenitor cells. The susceptibilities to FTC of pretherapy isolates in comparison with those of posttherapy 3'-azido-3'-deoxythymidine-resistant viruses in human lymphocytes were not substantially different. Similar results were obtained with well-defined 2',3'-dideoxyinosine-and nevirapine-resistant viruses.(-)-FTC-5'-triphosphate competitively inhibited human immunodeficiency virus type 1 reverse transcriptase, with an inhibition constant of 2.9 ,IM, when a poly(I) . oligo(dC)1-24 template primer was used. A two-to threefold decreased affinity was noted for the (+)-enantiomer. By using sequencing analysis, racemic FTC-5'-triphosphate was shown to be a potent DNA chain terminator when human immunodeficiency virus reverse transcriptase was used. These results suggest that further development of the (-)-1-enantiomer of FTC is warranted as an antiviral agent for infections caused by human immunodeficiency viruses.

A novel anti-hepatitis B virus (anti-HBV) agent, 2-fluoro-5-methyl-␤-L-arabinofuranosyluracil (L-FMAU), was synthesized and found to be a potent anti-HBV and anti-Epstein-Barr virus agent. Its in vitro potency was evaluated in 2.2.15 and H1 cells for anti-HBV and anti-Epstein-Barr virus activities, respectively. In vitro cytotoxicity in MT2, CEM, 2.2.15, and H1 cells was also assessed, and the results indicated high antiviral selectivities of L-FMAU in these cells.A number of nucleosides have been reported to be antihepatitis B virus (anti-HBV) agents, although none of them have yet been proven to be clinically useful. As a part of our antiviral drug discovery program for HBV, we recently have reported the syntheses and anti-HBV activities of dioxolane (14, 15) and oxathiolane (1,8,13) nucleosides. (Ϫ)-␤-L-Dioxolane-cytosine has been found to be the most potent anti-HBV agent (50% effective concentration ϭ 0.0005 M in 2.2.15 cells), although the compound was the most toxic (50% inhibitory concentration ϭ 0.26 M in CEM cells) among those tested (15). Among the oxathiolane cytosine nucleosides we evaluated, (Ϫ)-␤-L-oxathiolane-cytosine has the most potent anti-HBV activity (50% effective concentration ϭ 0.01 M) and a favorable cytotoxicity (Ͼ50 M in CEM cells) (1). Interestingly, the (Ϫ)-␤-L isomer of this compound resisted deoxycytidine deaminase, while the (ϩ)-␤-D isomer was deaminated under similar conditions (3).Recently, an increasing number of L nucleosides have been reported to be antiherpesvirus (23), anti-human immunodeficiency virus (anti-HIV) (2,6,13,14,16,19), and anti-HBV (1,7,8,10,11,15,16) agents. Furthermore, some of the L nucleosides have been found to be more potent than the corresponding D nucleosides (6,13,14). Previously, 2Ј-fluoro-5-methyl-␤-D-arabinofuranosyluracil (FMAU) and 2Ј-fluoro-5-ethyl-␤-Darabinofuranosyluracil (FEAU) were reported (5, 24, 27) to be extremely potent antiviral agents against herpesvirus and HBV, respectively. However, the myelosuppression and neurotoxicity of FMAU limit its usefulness as a clinically effective antiviral agent. In view of the discovery that several nucleosides with the unnatural L configuration are selective antiviral agents, it was of interest to synthesize several 2Ј-fluoro-substituted (arabino configuration) nucleosides with the L configuration as potential antiviral agents, anticipating that these nucleosides will give lower toxicities than the corresponding D isomers. Thus, we report here the preliminary syntheses and antiviral activities of several pyrimidine nucleosides for which the corresponding D isomers have been known to exhibit potent antiviral activities.Synthesis. 1-O-Acetyl-2,3,5-tri-O-benzoyl-␤-L-ribofuranose (compound 1) was prepared from L-ribose (Fig. 1). The fully protected L-ribose compound 1 was selectively debenzoylated at the C-2 position and then was converted to the 2-fluorinated sugar (compound 3) according to the method described for the corresponding D isomer (25). The 2-fluorosugar (arabino configuration) compound 3 was c...

The impact of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the causative agent of COVID-19, is global and unprecedented. Although remdesivir has recently been approved by the FDA to treat SARS-CoV-2 infection, no oral antiviral is available for outpatient treatment. AT-527, an orally administered double prodrug of a guanosine nucleotide analog, was previously shown to be highly efficacious and well tolerated in HCV-infected subjects. Here, we report the potent in vitro activity of AT-511, the free base of AT-527, against several coronaviruses, including SARS-CoV-2. In normal human airway epithelial cells, the concentration of AT-511 required to inhibit replication of SARS-CoV-2 by 90% (EC90) was 0.47 μM, very similar to its EC90 against HCoV-229E, HCoV-OC43 and SARS-CoV in Huh-7 cells. Little to no cytotoxicity was observed for AT-511 at concentrations up to 100 μM. Substantial levels of the active triphosphate metabolite AT-9010 were formed in normal human bronchial and nasal epithelial cells incubated with 10 μM AT-511 (698 ± 15 and 236 ± 14 μM, respectively), with a half-life of at least 38 h. Results from steady-state pharmacokinetic and tissue distribution studies of non-human primates administered oral doses of AT-527, as well as pharmacokinetic data from subjects given daily oral doses of AT-527, predict that twice daily oral doses of 550 mg AT-527 will produce AT-9010 trough concentrations in human lung that exceed the EC90 observed for the prodrug against SARS-CoV-2 replication. This suggests that AT-527 may be an effective treatment option for COVID-19.

The relative antiviral potencies of five nucleotide heterodimers of 3'-azido-3'-deoxythymidine (AZT), 3'-azido-3'-deoxythymidilyl-(5',5')-2'-3'-dideoxy-5'-adenylic acid (AZT-P-ddA), 3'-azido-3'-deoxythymidilyl-(5',5')-2',3'-dideoxy-5'-inosinic acid (AZT-P-ddI), and the corresponding 2-cyanoethyl congeners AZT-P(CyE)-ddA and AZT-P(CyE)-ddI, were determined in primary human peripheral blood mononuclear cells infected with human immunodeficiency virus type 1. The homodimer 3'-azido-3'-deoxythymidilyl-(5',5')-3'-azido-3'-deoxythymidilic acid (AZT-P-AZT) was also included for comparison. The potencies of the compounds were AZT-P-ddA greater than or equal to AZT-P-ddI greater than AZT-P(CyE)-ddA greater than or equal to AZT-P(CyE)-ddI greater than or equal to AZT greater than AZT-P-AZT. Whereas AZT-P-ddA and AZT-P-ddI had in vitro therapeutic indices greater than that of AZT, the homodimer of AZT had a low therapeutic index. AZT-P-ddI exhibited the lowest toxicity in peripheral blood mononuclear, Vero, or CEM cells. Combination studies between AZT and 2',3'-dideoxyinosine (ddI) at nontoxic concentrations indicated a synergistic interaction at a drug ratio of 1:100. At higher ratios (1:500 and 1:1,000), the interactions were synergistic only at concentrations that produced up to 75% virus inhibition. At higher levels of antiviral effects, this combination was antagonistic, as determined by the multiple drug effect analysis method. AZT-P-ddI was about 10-fold less toxic than AZT to human granulocyte-macrophage progenitor cells. However, no significant difference was apparent when the compounds were evaluated against cells of the erythroid lineage. The greater antiviral activity and lower toxicity of this compound could not be attributed to the extracellular decomposition of the dimer in media at physiological temperature and pH. However, in acidic solutions, AZT-P-ddI decomposed in a pH-dependent manner. Advanced preclinical studies with this heterodimer of two clinically effective antiretroviral agents should be considered.

The effects of 3'-azido-3'-deoxythymidine (AZT) and 9-(1,3-dihydroxy-2-propoxymethyl)guanine on myeloid and erythroid colony-forming cells were studied by clonogenic assays. Both consistently inhibited granulocytemacrophage CFU (CFU-GM) and erythroid burst-forming units in a dose-dependent fashion. Concentrations of AZT and 9-(1,3-dihydroxy-2-propoxymethyl)guanine required for 50% inhibition of CFU-GM were, respectively, 0.9 0.1 and 2.7 0.5 ,uM; those required for 90% inhibition were, respectively, 34.0 2.8 and 35.7 + 3.6 ,uM. Erythroid burst-forming units were less sensitive to high concentrations of AZT than were CFU-GM.3'-Azido-3'-deoxythymidine (AZT), an inhibitor of the human immunodeficiency virus reverse transcriptase, is, at the present time, one of the most promising drugs available for the treatment of acquired immunodeficiency syndrome (AIDS). 9-(1,3-Dihydroxy-2-propoxymethyl)guanine (DHPG), an inhibitor of the cytomegalovirus DNA polymerase (2), has shown encouraging results in the treatment of life-threatening cytomegalovirus infections, a frequent cause of severe morbidity in AIDS patients. The side effects of these agents and especially their myelosuppressive activities have been of some concern in preliminary clinical trials (3,13). Surprisingly, little attention has been given thus far to the effects of these antiviral drugs on such "host" cells as those of bone marrow. This has led to our study of the effects of AZT and DHPG on the growth of normal human hematopoietic progenitor cells (specifically granulocytes-macrophages and erythrocytes) in vitro, with the results set forth in the present report.Bone marrow cells obtained by aspiration from the posterior iliac crest of healthy volunteers who had given informed consent were anticoagulated with heparin and layered on a single-step Ficoll-Hypaque discontinuous gradient to remove mature erythroid and myeloid cells. The mononuclear cells, obtained from the interface of the gradient, were collected and washed twice in Hanks balanced salt solution.The culture assay for granulocyte-macrophage CFU (CFU-GM) was performed by using the modified bilayer soft-agar system described by Metcalf et al. (9). Briefly, the bottom or feeder layer was composed of 0.5% agar (BactoAgar; Difco Laboratories, Detroit, Mich.) containing McCoy 5A nutrient medium supplemented with 15% dialyzed fetal bovine serum (heat inactivated at 56°C for 30 min) (GIBCO Laboratories, Grand Island, N.Y.) and 10% (final concentration) giant cell tumor-conditioned medium (GIBCO) as a source of colony-stimulating factor. The desired drug concentrations (or drug diluent for the control) were added to * Corresponding author.this layer, and 1-ml volumes of this mixture were poured into 35-mm plastic petri dishes. After solidification, a top layer (1 ml) was prepared which contained 0.3% agar, McCoy SA nutrient medium supplemented with 15% fetal bovine serum, and 105 mononuclear cells per ml. The plates were incubated for 14 days in a humidified atmosphere of 5% C02-95% air at 37°C. Colonies c...