Synthesis of 4′-Substituted-2′-Deoxy-2′-α-Fluoro Nucleoside Analogs as Potential Antiviral Agents

Nucleoside analogs are widely used for the treatment of viral diseases (Hepatitis B/C, herpes and human immunodeficiency virus, HIV) and various malignancies. ALS-8176, a prodrug of the 4′-chloromethyl-2′-deoxy-2′-fluoro nucleoside ALS-8112, was evaluated in hospitalized infants for the treatment of respiratory syncytial virus (RSV), but was abandoned for unclear reasons. Based on the structure of ALS-8112, a series of novel 4′-modified-2′-deoxy-2′-fluoro nucleosides were synthesized. Newly prepared compounds were evaluated against RSV, but also against a panel of RNA viruses, including Dengue, West Nile, Chikungunya, and Zika viruses. Unfortunately, none of the compounds showed marked antiviral activity against these viruses.


Introduction
Modified nucleoside and nucleotide analogs are now the cornerstone of antiviral and anticancer chemotherapies [1,2] and among them, 4 -substituted nucleosides have attracted a great deal of attention ( Figure 1). Balapiravir (1), the prodrug of 4 -azidocytidine, was one of the early hits identified as a potent and selective inhibitor of hepatitis C virus (HCV) RNA polymerase [3]. Further, 4 -ethynyl-2-fluoro-2 -deoxyadenosine (2) (EFdA/MK-8591/islatravir), in its triphosphate form, is a highly potent nucleoside reverse transcriptase translocation inhibitor (NRTTI) which is right now evaluated for the treatment and pre-exposure prophylaxis of HIV-1 infection via subdermal implant [4]. In addition, 4 -C-cyano-2-amino-2 -deoxyadenosine (CAdA) (3) [5] was also reported as a highly potent inhibitor of both HBV and HIV-1 replication while E-CFCP (4), another 4 -C-cyano nucleoside analog, was reported to be a subnanomolar inhibitor of HBV replication [6]. ALS-8176/lumicitabine (6), a prodrug of ALS-8112, a 4 -chloromethyl-2 -deoxy-2 -fluorocytidine analog, was evaluated in a phase 2 clinical trial for the treatment of respiratory syncytial virus (RSV) infections which was terminated for unclear reasons [7]. We recently reported that ALS-8112 also displayed potent anti-Nipah virus activity in vitro while also displaying in vitro toxicity [8]. Based on the potential of ALS-8112, we wish to report herein, the synthesis and the antiviral evaluation of new 4 -substituted-2 -deoxy-2 -fluoro cytidine nucleoside analogs. Although numerous 4 -substitutions have already been introduced on the ALS-8112 scaffold, these modifications remained basic and included mostly simple groups such as N 3 , alkyls, vinyl, ethynyl, cycloalkyl, ethers and thioethers. Through this work we focused on small groups that had never been introduced on the 4 -position of a nucleoside analog. These modifications included
Targeted 4′-substituted-2′-deoxy-2′-α-fluoro nucleoside derivatives 11,14,17,20,25, and 26 were prepared from key intermediates 8 and/or 9 obtained from commercially available 2′-deoxy-2′-αfluorocytidine 7 following the chemistry described by Wang et al. [10] (Scheme 1). The synthesis of 4′-difluoromethoxy analog 11 was achieved by the reaction of 9 with a reactive Cu-difluorocarbene complex obtained by the reaction of CuI with FSO2CF2CO2H [11], followed by removal of the trityl groups in 80% aqueous acetic acid (Scheme 2). We first thought to prepare the desired azetidine analog 14 by reacting an activated 5′-methyltriflate intermediate with azetidine in presence of an organic base (Et3N or pyridine). However, under these conditions, we were unable to observe formation of the desired compound. We hypothesized that the relatively bulky azetidine ring could not reach the sterically hindered 5′-position due to the presence of the nearby large 5′-and 3′monomethoxytrityl groups. Therefore, we subsequently evaluated an intramolecular reductive

Chemistry
Targeted 4 -substituted-2 -deoxy-2 -α-fluoro nucleoside derivatives 11,14,17,20,25, and 26 were prepared from key intermediates 8 and/or 9 obtained from commercially available 2 -deoxy-2 -α-fluorocytidine 7 following the chemistry described by Wang et al. [10] (Scheme 1). The synthesis of 4 -difluoromethoxy analog 11 was achieved by the reaction of 9 with a reactive Cu-difluorocarbene complex obtained by the reaction of CuI with FSO 2 CF 2 CO 2 H [11], followed by removal of the trityl groups in 80% aqueous acetic acid (Scheme 2). We first thought to prepare the desired azetidine analog 14 by reacting an activated 5 -methyltriflate intermediate with azetidine in presence of an organic base (Et 3 N or pyridine). However, under these conditions, we were unable to observe formation of the desired compound. We hypothesized that the relatively bulky azetidine ring could not reach the sterically hindered 5 -position due to the presence of the nearby large 5 -and 3 -monomethoxytrityl groups. Therefore, we subsequently evaluated an intramolecular reductive cyclization via the use of a primary halogeno alkylamine. The oxidation of 9 to the corresponding aldehyde with Dess Martin periodinane followed by reaction with 3-bromopropylamine in the presence of MgSO 4 led to the formation of imine intermediate 12 which was subsequently reduced with NaBH 4 . Finally, the newly formed amine displaced the terminal bromine to form the desired azetidine derivative 13 [12,13]. Treatment of 13 under acidic conditions gave the targeted compound 14 (Scheme 3). The 4 -oxetane analog 17 was obtained from 9 by, first, oxidation to the corresponding aldehyde followed by a Johnson-Corey-Chaykovsky epoxidation and consecutive ring-expansion. Thus, compound 9 was oxidized by treatment with Dess-Martin periodinane to the corresponding aldehyde which was treated with 10 equivalents of trimethyloxosulfonium iodide in presence of t BuOK for 4 days to provide oxetane derivative 16 as a single isomer. Final deprotection under acidic conditions afforded the desired 4 -oxetane analog 17 in 48% yield over 3 steps (Scheme 3). Stereochemistry of the oxetane Molecules 2020, 25, 1258 3 of 13 ring in compound 17 could not be assessed with certitude by NMR analysis, therefore, crystals were grown from methanol by slow evaporation. Results from X-ray structure determination of 17 led us to ascertain the S-configuration of the 5 -carbon ( Figure 2). Synthesis of 4 -isoxazole analog 20 was achieved from intermediate 9 by first oxidation to the 5-aldehyde intermediate followed by a Van Leusen cyclization reaction using tosylmethyl isocyanide (TosMIC) in the presence of K 2 CO 3 [14] and final deprotection with acetic acid.   potency, of a methyl group on the 5′-methylene portion of ALS-8112, a modification known to be tolerated by other viral polymerases [9].
Targeted 4′-substituted-2′-deoxy-2′-α-fluoro nucleoside derivatives 11,14,17,20,25, and 26 were prepared from key intermediates 8 and/or 9 obtained from commercially available 2′-deoxy-2′-αfluorocytidine 7 following the chemistry described by Wang et al. [10] (Scheme 1). The synthesis of 4′-difluoromethoxy analog 11 was achieved by the reaction of 9 with a reactive Cu-difluorocarbene complex obtained by the reaction of CuI with FSO2CF2CO2H [11], followed by removal of the trityl groups in 80% aqueous acetic acid (Scheme 2). We first thought to prepare the desired azetidine analog 14 by reacting an activated 5′-methyltriflate intermediate with azetidine in presence of an organic base (Et3N or pyridine). However, under these conditions, we were unable to observe formation of the desired compound. We hypothesized that the relatively bulky azetidine ring could not reach the sterically hindered 5′-position due to the presence of the nearby large 5′-and 3′monomethoxytrityl groups. Therefore, we subsequently evaluated an intramolecular reductive Scheme 1. Synthesis of key intermediates 8 and 9 from commercially available 2 -deoxy-2 -αfluorocytidine 7.
Molecules 2020, 25, x FOR PEER REVIEW 3 of 13 cyclization via the use of a primary halogeno alkylamine. The oxidation of 9 to the corresponding aldehyde with Dess Martin periodinane followed by reaction with 3-bromopropylamine in the presence of MgSO4 led to the formation of imine intermediate 12 which was subsequently reduced with NaBH4. Finally, the newly formed amine displaced the terminal bromine to form the desired azetidine derivative 13 [12,13]. Treatment of 13 under acidic conditions gave the targeted compound 14 (Scheme 3). The 4′-oxetane analog 17 was obtained from 9 by, first, oxidation to the corresponding aldehyde followed by a Johnson-Corey-Chaykovsky epoxidation and consecutive ring-expansion. Thus, compound 9 was oxidized by treatment with Dess-Martin periodinane to the corresponding aldehyde which was treated with 10 equivalents of trimethyloxosulfonium iodide in presence of t BuOK for 4 days to provide oxetane derivative 16 as a single isomer. Final deprotection under acidic conditions afforded the desired 4′-oxetane analog 17 in 48% yield over 3 steps (Scheme 3). Stereochemistry of the oxetane ring in compound 17 could not be assessed with certitude by NMR analysis, therefore, crystals were grown from methanol by slow evaporation. Results from X-ray structure determination of 17 led us to ascertain the S-configuration of the 5′-carbon ( Figure 2). Synthesis of 4′-isoxazole analog 20 was achieved from intermediate 9 by first oxidation to the 5aldehyde intermediate followed by a Van Leusen cyclization reaction using tosylmethyl isocyanide (TosMIC) in the presence of K2CO3 [14] and final deprotection with acetic acid.  Targeted 5′-methyl derivatives 25 and 26 were prepared by following the chemistry described in Scheme 4. Protected intermediate 8 was oxidized under Pfitzner-Moffatt conditions and then reacted with MeMgCl to give the desired methylated intermediate as a 1/1 mixture. This compound was then oxidized to the corresponding ketone 21 under Pfitzner-Moffatt conditions and the tertbutyldimethylsilyl (TBS) group was removed using tetra-n-butylammonium fluoride (TBAF). 22 was then reacted with Tf2O in pyridine to form a triflate intermediate which was directly treated with LiCl or LiBr in DMF to give the corresponding halogeno derivatives 23 and 24, respectively. Finally, reduction with NaBH4 and removal of the monomethoxytrityl groups under acidic conditions afforded the desired compounds 25 and 26 as 1/1 mixtures of isomers at the 5′-position.

Antiviral and Toxicity Evaluation
Based on their structural similarity with ALS-8112, compounds 11,14,17,20, 25 and 26 were tested against RSV replicon-containing adenocarcinomic human alveolar basal epithelial A549 cells (kindly provided by Apath, L.L.C, New York, NY, USA), but unfortunately none of them display antiviral activity in this system when evaluated up to 10 µM. It is worth noting that they did not show

Antiviral and Toxicity Evaluation
Based on their structural similarity with ALS-8112, compounds 11, 14, 17, 20, 25 and 26 were tested against RSV replicon-containing adenocarcinomic human alveolar basal epithelial A549 cells (kindly provided by Apath, L.L.C, New York, NY, USA), but unfortunately none of them display antiviral activity in this system when evaluated up to 10 µM. It is worth noting that they did not show toxicity either, up to 100 µM, in a panel of cell lines, including primary human peripheral blood mononuclear (PBM) cells, human lymphoblastoid cells (CEM), African Green monkey (Vero) cells and human liver hepatocarcinoma (HepG2) cells. The excellent safety profile of these compounds led us to further evaluate them against a panel of RNA viruses (Dengue (DENV), West Nile (WNV), Chikungunya (CHIKV), and Zika viruses (ZIKV)) but, once again, none of them displayed antiviral activity when tested up to 20 µM for ZIKV or 30 µM for DENV, WNV, or CHIKV.

Synthesis
Anhydrous solvents were purchased from Aldrich Chemical Company, Inc. (Milwaukee, Wisconsin, USA). Reagents were purchased from commercial sources. Unless noted otherwise, the materials used in the examples were obtained from readily available commercial suppliers or synthesized by standard methods known to one skilled in the art of chemical synthesis. 1 H, 13