Cytidine 5′-triphosphate

Chemical Synthesis of the Antiviral Nucleotide Analogue ddhCTP

James M. Wood, Gary B. Evans, Tyler L. Grove, Steven C. Almo, Scott A. Cameron, Richard H. Furneaux, and Lawrence D. Harris*

ABSTRACT:

3′-Deoxy-3′,4′-didehydro-cytidine triphosphate (ddhCTP) is a novel antiviral molecule produced by the enzyme viperin as part of the innate immune response. ddhCTP has been shown to act as an obligate chain terminator of flavivirus and SARS-CoV-2 RNA-dependent RNA polymerases; however, further biophysical studies have been precluded by limited access to this promising antiviral. Herein, we report a robust and scalable synthesis of ddhCTP as well as the mono- and diphosphates ddhCMP and ddhCDP, respectively. Identification of a 2′-silyl ether protection strategy enabled selective synthesis and facile purification of the 5′-triphosphate, culminating in the preparation of ddhCTP on a gram scale.

■ INTRODUCTION

The frequent occurrence of epidemics in the past two decades (SARS-CoV-1, MERS-CoV, EBOV) and the emergence of the COVID-19 pandemic has brought widespread attention to the impacts viruses can have on human life and healthcare systems. Viruses are implicated in the majority of epidemic and pandemic diseases, which reflects their high transmissibility and the paucity of effective antiviral therapies.1 In the absence of vaccines, treatment of emergent and re-emergent viruses relies on the repurposing of existing drugs, and therefore, the development of broad-spectrum antiviral drugs is of great value.2 Attainment of this goal is challenging as viruses are obligate parasites which employ host cell machinery and present fewer druggable targets compared to other pathogens. Nevertheless, a number of nucleotide analogues which target viral proteins such as RNA-dependent RNA polymerases (RdRp) and DNA polymerases do exhibit broad-spectrum activity and constitute an important class of antiviral drugs (Figure 1a).3
In 2018, Gizzi et al. reported that viperin (Virus Inhibitory Protein, Endoplasmic Reticulum-associated, Interferon-inducible), also known as RSAD2 (radical SAM domain-containing 2), catalyzes the formal dehydration of cytidine triphosphate (CTP, 1) to 3′-deoxy-3′,4′-didehydro-cytidine triphosphate (ddhCTP, 2) (Figure 1b).4 Viperin expression is induced by interferon pathways5 and has been shown to inhibit replication of both RNA and DNA viruses by diverse mechanisms, including protein−protein interactions and modulation of immune signaling, thereby contributing to the innate antiviral response.6 The role of ddhCTP (2) in the antiviral activity of viperin is still not completely understood; however, it has been shown to compete with CTP (1) for incorporation by RdRps of the Flaviviridae (dengue virus, West Nile virus, Zika virus, and hepatitis C virus), resulting in obligate chain termination and inhibition of viral genome replication.4 A recent study has also shown that ddhCTP (2) is efficiently utilized by the SARS-CoV-2 replicase.7 Importantly, ddhCTP (2) production does not affect the viability or growth rate of Vero and HEK293T cells and is specific for non-native RNA polymerases.4,8
In light of these findings, ddhCTP (2) represents a promising platform for the development of broad-spectrum antiviral agents; however, elucidation of its antiviral activity is ongoing. Previous studies have been conducted with small quantities of ddhCTP (2) prepared enzymatically from CTP (1) using isolated viperin; however, purification of ddhCTP (2) to remove unconverted CTP (1) poses a significant challenge.4 To circumvent complications arising from CTP (1) impurities, we sought to establish a chemical synthesis of ddhCTP (2) that would provide sufficiently pure material for biophysical studies and evaluation of antiviral activity. We envisioned ddhCTP (2) would be accessible by phosphorylation of an appropriately protected ddh-nucleoside 3, which in turn could be prepared by formal dehydration of 4-Nbenzoylcytidine (4). We anticipated chemical synthesis of ddhCTP (2) from a ddh-nucleoside 3 would be more scalable than the existing bioenzymatic synthesis and would also provide the opportunity to prepare differently phosphorylated forms of the parent nucleoside for elucidation of biochemical pathways. Herein, we report the development of a protecting group strategy to provide convenient access to the ddhC nucleotide derivatives ddhCMP, ddhCDP, and ddhCTP (2).

■ RESULTS AND DISCUSSION

Our development of a suitable synthetic strategy toward the ddhC-nucleotides commenced with preparation of ddhC (7) from commercially available 4-N-benzoylcytidine (4) following a method reported by Petrová et al.9 This approach involved oxidation of the 5′-alcohol to an aldehyde, which facilitates base-catalyzed elimination of a 2′,3′-orthoformate at the 3′position to establish the 3′-deoxy-3′,4′-didehydroribo-ring system (Scheme 1a). While following this work, we identified a number of procedural modifications which improved the practicality of the synthesis on a large scale, including the replacement of two chromatographic purification steps; orthoester 5 and alcohol 7 were both obtained cleanly by trituration. The reduction of aldehyde 6 was best performed with only 0.5 equiv of sodium borohydride to prevent reduction of the cytosine ring, while benzoyl cleavage with ammonia avoided a 4-N-methylcytidine side product that arose when the deprotection was carried out using the prescribed methylamine. More significantly, the Moffatt−Pfitzner oxidation to aldehyde 6 was poorly reproducible in our hands. Conversion to the aldehyde was efficient, yet it proved unstable during workup despite being stable once isolated; typical isolated yields ranged from 40−47%, compared with the literature reported yield of 69%.9a Exploration of alternative oxidation methodologies was largely precluded by the limited solubility profile of alcohol 5. However, through optimization we found the reaction went to completion with fewer equivalents of the EDC-DMSO reagent system. This in turn facilitated purification of aldehyde 6 but failed to improve isolated yields.
During our optimization of the Moffat−Pfitzer oxidation, we identified aldehyde 6 (Scheme 1) as a potentially convenient intermediate to attempt selective protection of the 2′-alcohol. It was anticipated the didehydroribo system would be incompatible with the deprotection conditions most commonly used in benzyl- or allyl-based protecting group strategies. Additionally, Petrová et al. have reported that 2′O-DMF acetals of didehydroribo systems readily undergo Ferrier-type allylic rearrangement as well as syn-elimination across C1′−C2′ to form furan products (Scheme 1b).9b An acyl or carbonate protecting group therefore presented the risk of undesirable reactivity. A tert-butyldimethylsilyl protecting group strategy was ultimately adopted, with the view that a lipophilic silyl group might aid purification of intermediates by both normal and reverse-phase chromatography. Standard protection conditions of silyl chloride and imidazole afforded the 2′-O-TBDMS ether 8 in high yield; however, careful control of reagent stoichiometry was required to avoid formation of N,O-acetal 9, and the reaction was capricious, particularly when attempting to scale up. Recovery of the aldehyde from N,O-acetal 9 required prohibitively harsh conditions,10 and so imidazole was substituted by triethylamine in the synthesis of silyl ether 8 to deliver more consistent results. Clean reduction of the aldehyde could be achieved with 0.5 equiv of borohydride, affording alcohol 3 in 92% yield.
Having identified 3 as a suitable intermediate to attempt phosphorylation reactions toward ddhC-nucleotides, an alternative, reliable, and scalable synthesis of this intermediate was sought. Accordingly, we envisioned a strategy involving elimination of a 3′-iodide in the xylo configuration to construct the 3′-deoxy-3′,4′-dehydroribose ring system.11 Regioselective silyl protection12 of 4-N-benzoylcytidine (4) and then iodination with methyltriphenoxyphosphonium iodide13 afforded 3′-iodide 11 (Scheme 2). A more obvious reagent system of PPh3, I2, and imidazole was also investigated for this iodination,14 but necessitated the use of high temperatures at which the 4-N-benzoyl group was labile. A crystal structure obtained for 11 confirmed the iodination proceeded with stereoinversion, setting the stage for base-mediated elimination of HI. Accordingly, 5′-O-silyl cleavage and then treatment with DABCO provided 4-N-benzoyl-2′-O-TBDMS ddhC (3). This intermediate formed the linchpin for the synthesis of all ddhCbased targets.
Synthesis of ddhCMP (15) was first attempted by applying the Yoshikawa method (POCl3 in trimethylphosphate) to intermediate 3;15 however, a complex mixture of products was obtained, implying the phosphorodichloridate intermediate was unstable (Scheme 3). Phosphoramidite chemistry was instead employed to access the monophosphate,16 delivering difluorenylmethyl phosphate 13 in 87% yield. Advancement of phosphate ester 13 to ddhCMP (15) required global deprotection: both the 4-N-benzoyl and fluorenylmethyl groups were readily removed using ammonia in methanol. Traceless deprotection of the TBDMS ether under mild acidic conditions gave ddhCMP (15) quantitatively, which was converted to its sodium salt by ion exchange.
ddhCDP (17) and ddhCTP (2) were both prepared from TBDMS-protected monophosphate 14 using the method reported by Hoard and Ott.17 Adoption of this strategy over direct di- or triphosphorylation of alcohol 3 was motivated by the scalability of the reaction18 as well as the simplicity of the reagent system, which would streamline purification. Activation of monophosphate 14 as the imidazolidate was performed using CDI, the excess of which was quenched using water. In situ treatment of the imidazolidate with the bis-(tributylammonium) salt of pyrophosphate gave triphosphate 18 in 70−88% yield. When MeOH was used as a CDI quenching agent instead of water, formation of a methyl carbamate at C4 of the cytosine nucleobase was observed. Triphosphate 18 exhibits good retention on C18 silica, a feature we attribute to the lipophilic TBDMS protecting group.19 This enabled rapid and scalable purification of the triphosphate by reverse-phase flash chromatography using an aqueous Bu3N-AcOH and MeOH ion-pairing eluent system, whereas oligophosphates usually require purification by strong ion-exchange chromatography.20 Traceless TBDMS cleavage was effected by treatment with Dowex resin in water, affording ddhCTP (2) in excellent yield. This synthesis was used to prepare approximately 1 g of ddhCTP (2) in a single pass and therefore provides ample access to this compound for biological evaluation. ddhCDP (17) was also readily accessed from monophosphate 14 by activation with CDI, treatment with orthophosphate, and Dowex-mediated TBDMS cleavage.

■ CONCLUSION

In conclusion, a synthetic route utilizing a 2′-O-TBDMS protected ddhC derivative has enabled facile synthesis of biologically relevant phosphates of ddhC (7). The benefits of the TBDMS protecting group are twofold: it provides a lipophilic handle that enables reverse-phase flash chromatographic purification of highly charged compounds that would otherwise require more intensive purification methods, and its traceless removal under mild conditions provides the deprotected targets in good purity. Taken together, these properties have facilitated a robust and scalable synthesis of ddhCTP (2), providing useful quantities of this antiviral metabolite and its prodrugs for biological studies.

■ REFERENCES

(1) WHO. Disease outbreaks. http://www.who.int/emergencies/ diseases/en/ (accessed Apr 19, 2020).
(2) (a) Liu, C.; Zhou, Q.; Li, Y.; Garner, L. V.; Watkins, S. P.; Carter, L. J.; Smoot, J.; Gregg, A. C.; Daniels, A. D.; Jervey, S.; Albaiu, D. Research and development on therapeutic agents and vaccines for COVID-19 and related human coronavirus diseases. ACS Cent. Sci. 2020, 6, 315−331. (b) Adalja, A.; Inglesby, T. Broad-spectrum antiviral agents: A crucial pandemic tool. Expert Rev. Anti-Infect. Ther. 2019, 17, 467−470. (c) Zhu, J.-D.; Meng, W.; Wang, X.-J.; Wang, H.C. R. Broad-spectrum antiviral agents. Front. Microbiol. 2015, 6, 517.
(3) (a) Safrin, S.; Cherrington, J.; Jaffe, H. S. Clinical uses of cidofovir. Rev. Med. Virol. 1997, 7, 145−156. (b) Gordon, C. J.; Tchesnokov, E. P.; Woolner, E.; Perry, J. K.; Feng, J. Y.; Porter, D. P.; Gotte, M. Remdesivir is a direct-acting antiviral that inhibits RNAdependent RNA polymerase from severe acute respiratory spectrum inhibitor of viral RNA polymerase. Proc. Jpn. Acad., Ser. B 2017, 93, 449−463.
(4) Gizzi, A. S.; Grove, T. L.; Arnold, J. J.; Jose, J.; Jangra, R. K.; Garforth, S. J.; Du, Q.; Cahill, S. M.; Dulyaninova, N. G.; Love, J. D.; Chandran, K.; Bresnick, A. R.; Cameron, C. E.; Almo, S. C. A naturally occurring antiviral ribonucleotide encoded by the human genome. Nature 2018, 558, 610−614.
(5) (a) Seo, J.-Y.; Yaneva, R.; Cresswell, P. Viperin: a multifunctional, interferon-inducible protein that regulates virus replication. Cell Host Microbe 2011, 10, 534−539. (b) Helbig, K. J.; Beard, M. R. The role of viperin in the innate antiviral response. J. Mol. Biol. 2014, 426, 1210−1219. (c) Lindqvist, R.; Överby, A. K. The role of viperin in antiflavivirus responses. DNA Cell Biol. 2018, 37, 725−730.
(6) Rivera-Serrano, E. E.; Gizzi, A. S.; Arnold, J. J.; Grove, T. L.; Almo, S. C.; Cameron, C. E. Viperin reveals its true function. Annu. Rev. Virol. 2020, 7, 421−446.
(7) Seifert, M.; Bera, S. C., van Nies, P., Kirchdoerfer, R. N., Shannon, A., Le, T.-T.-N., Grove, T. L., Papini, F. S., Arnold, J. J., Almo, S. C., Canard, B., Depken, M., Cameron, C. E., Dulin, D. Signatures and mechanisms of efficacious therapeutic ribonucleotides against SARS-CoV-2 revealed by analysis of its replicase using magnetic tweezers. bioRxiv 2020.08.06.240325; 2020 DOI: 10.1101/ 2020.08.06.240325.
(8) Dukhovny, A.; Shlomai, A.; Sklan, E. H. The antiviral protein viperin suppresses T7 promoter dependent RNA synthesis-possible implications for its antiviral activity. Sci. Rep. 2018, 8, 8100.
(9) (a) Petrová, M.; Budes̆ ínsky̌ , M.; Rosenberg, Í . Straightforward synthesis of 3′-deoxy-3′,4′-didehydronucleoside-5′-aldehydes via 2′,3′-O-orthoester group elimination: a simple route to 3′,4′didehydronucleosides. Tetrahedron Lett. 2010, 51, 6874−6876. (b) Petrová, M.; Budes̆ ínsky̌ ́, M.; Zborníková, E.; Fiedler, P.; Rosenberg, I. A Ferrier-type allylic rearrangement of 3′-deoxy-3′,4′didehydronucleosides mediated by DMF dimethyl acetal: Direct access to 4′-alkoxy-2′,3′-didehydro-2′,3′-dideoxynucleosides. Org. Lett. 2011, 13, 4200−4203.
(10) (a) Quan, L. G.; Cha, J. K. A convenient protecting group for aldehydes. Synlett 2001, 2001, 1925−1926. (b) Gimisis, T.; Arsenyan, P.; Georganakis, D.; Leondiadis, L. Tuning the reactivity of O-tertbutyldimethylsilylimidazolyl aminals towards organolithium reagents. Synlett 2003, 1451−1454.
(11) Abdel-Megied, A. E. S.; Hansen, P.; Pedersen, E. B.; Nielsen, C. M. New synthesis of 2′,3′-didehydro-2′,3′-dideoxynucleosides. Acta Chem. Scand. 1991, 45, 1060−1063.
(12) Ogilvie, K. K.; Schifman, A. L.; Penney, C. L. The synthesis of oligoribonucleotides. III. The use of silyl protecting groups in nucleoside and nucleotide chemistry. VIII. Can. J. Chem. 1979, 57, 2230−2238.
(13) Verheyden, J. P. H.; Moffatt, J. G. Halosugar nucleosides. II. Iodination of secondary hydroxyl groups of nucleosides with methyltriphenoxyphosphonium iodide. J. Org. Chem. 1970, 35, 2868−2877.
(14) Kubota, Y.; Nakamura, K.; Haraguchi, K.; Tanaka, H.Stereoselective preparation of 1-[5-O-(tert-butyldimethylsilyl)-2,3dideoxy-3-iodo-β-D-threo-pentofuranosyl]thymine from thymidine: An efficient entry to 3′,4′-unsaturated nucleoside. Synth. Commun. 2010, 40, 1758−1764.
(15) (a) Yoshikawa, M.; Kato, T.; Takenishi, T. A novel method for phosphorylation of nucleosides to 5′-nucleotides. Tetrahedron Lett. 1967, 8, 5065−5068. (b) Cheng, G.; Wei, H. 3′-Deoxy-3′,4′didehydronucleosides and application in anti-rabies virus, HIV virus, HCV virus, HBV virus and Zika virus. CN 108640959 A, October 12, 2018.
(16) Watanabe, Y.; Nakamura, T.; Mitsumoto, H. Protection of phosphate with the 9-fluorenylmethyl group. Synthesis of unsaturated-acyl phosphatidylinositol 4,5-bisphosphate. Tetrahedron Lett. 1997, 38, 7407−7410.
(17) Hoard, D. E.; Ott, D. G. Conversion of mono- and oligodeoxyribonucleotides to 5′-triphosphates. J. Am. Chem. Soc. 1965, 87, 1785−1788.
(18) Simon, E. S.; Grabowski, S.; Whitesides, G. M. Convenient syntheses of cytidine 5′-triphosphate, guanosine 5′-triphosphate, and uridine 5′-triphosphate and their use in the preparation of UDPglucose, UDP-glucuronic acid, and GDP-mannose. J. Org. Chem. 1990, 55, 1834−1841.
(19) Ohno, K.; Sugiyama, D.; Takeshita, L.; Kanamori, T.; Masaki, Y.; Sekine, M.; Seio, K. Synthesis of photocaged 6-O-(2-nitrobenzyl)guanosine and 4-O-(2-nitrobenzyl) uridine triphosphates for photocontrol of the RNA transcription reaction. Bioorg. Med. Chem. 2017, 25, 6007−6015.
(20) Murphy, P. J. Organophosphorus Reagents: A Practical Approach in Chemistry. The Practical Approach in Chemistry Series; Oxford University Press: Oxford, NY, 2004; pp 264−271.
(21) Hudson, H. R.; Powroznyk, L. Some new observations on an old reaction: disproportionation and the formation of P-O-P intermediates in the Michaelis-Arbuzov reaction of triaryl phosphites with alkyl halides. ARKIVOC 2004, IX, 19−33.
(22) Caron, J.; Lepeltier, E.; Reddy, H.; Lepetre-Mouelhi, S.; Wack,̂ S.; Bourgaux, C.; Couvreur, P.; Desmaële, D. Squalenoyl gemcitabine monophosphate: synthesis, characterisation of nanoassemblies and biological evaluation. Eur. J. Org. Chem. 2011, 14, 2615−2628.
(23) Kore, A. R.; Parmar, G. Convenient synthesis of nucleoside-5′diphosphates from the corresponding ribonucleoside-5′-phosphoroimidazole. Synth. Commun. 2006, 36, 3393−3399.
(24) Thillier, Y.; Sallamand, C.; Baraguey, C.; Vasseur, J.-J.; Debart, F. Solid-phase synthesis of oligonucleotide 5′-(α-P-thio)triphosphates and 5′-(α-P-thio)(β,γ-methylene)triphosphates. Eur. J. Org. Chem. 2015, 2015, 302−308.
(25) Kume, A.; Tanimura, H.; Nishiyama, S.; Sekine, M.; Hata, T. An improved procedure for the 2′,3′-O-methoxymethylenation of Nprotected ribonucleosides. Synthesis 1985, 1985, 408−409.