3-Deazaguanosine inhibits SARS-CoV-2 viral replication and reduces the risk of COVID-19 pneumonia in hamster

Summary

The COVID-19 pandemic highlighted the serious threat that coronaviruses have on public health. Because coronavirus continuously undergoes cross-species transmission, additional therapeutic agents and targets are urgently needed. Here, we show that a 3-deazapurine ribonucleoside, 3-Deazaguanosine (C³Guo, 2), has potent antiviral activity against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Unexpectedly, C³Guo (2) does not act as an inhibitor of RNA-dependent RNA polymerase (RdRp), which is the therapeutic target of two key nucleoside/nucleotide inhibitors approved for the treatment of COVID-19 (Remdesivir and Molnupiravir); instead, it seems to function by targeting the capping machinery of viral RNA. In hamsters infected with SARS-CoV-2, administration of 2 markedly reduced infectious viral titers, and prevented the development of COVID-19 pneumonia better than Molnupiravir. The potency of 2 against SARS-CoV-2 underscores its potential as an effective therapeutic agent for COVID-19 and future zoonotic coronavirus infections and raises the possibility of antiviral nucleoside analogs with alternative therapeutic targets to RdRp.

Graphical abstract

Highlights

• •3-Deazaguanosine (C³Guo, 2) suppresses SARS-CoV-2 replication in vitro and in vivo
• •C³Guo (2) prevents COVID-19 pneumonia better than Molnupiravir in hamster models
• •C³Guo (2) acts independently of RdRp, unlike Remdesivir and Molnupiravir
• •C³Guo (2) is a promising an anti-COVID-19 agent with a unique antiviral mechanism

Health sciences; Biological sciences; Immunology

Introduction

The COVID-19 pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has dominated world headlines since this positive-strand RNA virus emerged in late 2019. Within 2 years of the first outbreak, an astonishing 10 vaccines were developed and approved, leading to worldwide vaccination and significant success in controlling infection and exacerbation.¹ In addition to vaccines, numerous drug discovery programs also emerged with some success. The first United States Food and Drug Administration (FDA)-authorized drug to treat COVID-19 was Remdesivir (GS-5734) (Figure 1A), a nucleotide prodrug that targets the RNA-dependent RNA polymerase (RdRp) of SARS-CoV-2.^2,3 Remdesivir is a repositioned drug, originally evaluated in clinical trials for treatment of Ebola virus disease, and has broad-spectrum antiviral activity.⁴ Subsequently, Molnupiravir (EIDD-2801; also known as MK-4482) (Figure 1B), a nucleoside prodrug of N⁴-hydroxycytidine (NHC, EIDD-1931), was developed and approved as the first orally active COVID-19 drug.⁵ Molnupiravir also targets RdRp of SARS-CoV-2.^6,7 More recently, nirmatrelvir (PF-07321332), an inhibitor of the main protease (Mpro) of SARS-CoV-2, was authorized by the FDA in 2021.^8,9

Figure 1.

Chemical structures of nucleoside and nucleotide inhibitors (NIs) clinically approved for the treatment of COVID-19

Chemical structure of (A) Remdesivir and (B) Molnupiravir and its parental nucleoside, NHC.

Among the various drug targets in RNA viral infections, RdRp is a focus for almost all RNA viruses. Indeed, inhibition of RdRp has become an integrated approach to managing several viral infections, including SARS-CoV-2, hepatitis C, influenza, and dengue viruses, among others.^10,11 Nucleoside and nucleotide analog inhibitors (NIs) are an important class of RdRp inhibitors, as evidenced in the early application of Remdesivir and Molnupiravir during the COVID-19 pandemic.¹² After conversion to the corresponding 5′-triphosphate, NIs are incorporated into the growing viral RNA chain and inhibit viral replication mainly by one of two mechanisms: (1) chain termination and disruption of subsequent viral RNA replication or transcription; or (2) mutagenesis owing to mispairing with and/or substitution of natural nucleotides, leading to impaired RNA synthesis or structure, RNA-protein interactions, or protein functions. Of the two NIs approved as COVID-19 drugs, Remdesivir functions by the chain-termination mechanism; that is, the active 5′-triphosphate of Remdesivir is incorporated into the growing viral RNA product by RdRp, where it acts as a chain-terminator of further RNA elongation.¹³ The C1′-cyano group in the Remdesivir ribose moiety is critical for antiviral potency¹⁴: steric hindrance between this cyano group and the SARS-CoV-2 RdRp complex acts as a translocation barrier on the growing RNA strand, causing RdRp to stall (Figure 1A).^13,15 On the other hand, Molnupiravir mainly acts via mutagenesis. After incorporation in the viral RNA, its 4-hydroxyamino group forms two tautomers: an amino form that pairs with G and an imino form that pairs with A (Figure 1B).¹⁶ This tautomeric behavior increases the frequency of transition mutations (G-to-A and C-to-U) in the viral genes, leading to lethal mutagenesis.

In terms of the inhibition of DNA polymerase, strand elongation is greatly inhibited when 3-deaza-2′-deoxyadenosine 5′-triphosphate (dC³Ado-TP) lacking the minor groove electron pair is incorporated instead of dATP in primer extension reactions with Taq DNA polymerase.¹⁷ Similarly, the presence of 3-deaza-2′-deoxyguanosine (dC³Guo) at the primer terminus decreases the rate of dNTP incorporation by the Klenow fragment DNA polymerase.¹⁸ These observations suggest that, if metabolized to active 5′-triphosphates and incorporated into the growing DNA strand, dC³Ado and dC³Guo function as chain terminators for DNA polymerase. However, whether the corresponding 3-deazapurine ribonucleosides/ribonucleotides act as chain terminators for RdRp, has not been elucidated, despite their potential utility as anti-RNA viral agents.^19,20

In this study, therefore, we prepared 3-deazapurine ribonucleosides and explored their effects on RdRp of SARS-CoV-2. Although we found that 3-Deazaguanosine (C³Guo, 2) (Figure 2) has potent anti-SARS-CoV-2 activity, contrary to our expectations its triphosphate 2-TP did not terminate RNA elongation catalyzed by RdRp of SARS-CpV-2; instead, it seemed to function by targeting the capping machinery of viral RNA. In SARS-CoV-2-infected hamsters, administration of 2 significantly decreased SARS-CoV-2 viral titer, and blocked the development of COVID-19 pneumonia better than the existing clinical drug, Molnupiravir. The potency of 2 against SARS-CoV-2 underscores its potential as an effective therapeutic agent for COVID-19 and future zoonotic coronavirus infections and paves the way for antiviral nucleoside analogs with alternative therapeutic targets to RdRp.

Figure 2.

Anti-SARS-CoV-2 activity of 3-Deazapurine ribonucleosides 1–3

(A) Structures of 3-deazaadenosine (C³Ado, 1), 3-Deazaguanosine (C³Guo, 2), and 3-deazainosine (C³Ino, 3).

(B) CPE assays. Vero E6 cells pre-treated with the indicated compounds (30 μM) were infected with SARS-CoV-2 at a multiplicity of infection (MOI) of 0.001. On day 3 post-infection, cells were fixed and stained with crystal violet to assess CPE. Viable attached cells were stained purple; dead cells due to infection were not stained as they had detached from the wells. Representative data from three independent assays are shown. Mock, an uninfected control; DMSO, an infected control without any compound.

Results and discussion

C³Guo (2) has antiviral activity against SARS-CoV-2

Using our previously reported practical synthesis of 3-deazapurine nucleosides,^21,22 we first prepared three kinds of 3-deazapurine ribonucleosides, 3-deazaadenosine (C³Ado, 1), 3-Deazaguanosine (C³Guo, 2), and 3-deazainosine (C³Ino, 3) (Figure 2), and evaluated their inhibitory activity against SARS-CoV-2 by cytopathic effect (CPE) assay (Figures 2 and S1). In this assay, host cell viability, visualized by crystal violet staining, is surrogate readout for viral replication. Of the three 3-deazapurine ribonucleosides prepared, CPE assay clearly showed that guanosine analog 2 inhibited viral replication in Vero E6 cells, strongly suggesting that 2 has effective anti-SARS-CoV-2 activity. In contrast, the other two analogs had limited effect. Real-time quantitative reverse-transcription PCR (RT-qPCR) analysis of viral RNA further confirmed that 2 suppressed SARS-CoV-2 replication in a dose-dependent manner (Figure S2), with its half-maximal effective concentration (EC₅₀) was determined to be 1.14 μM. Remdesivir and NHC, the parental analog of Molnupiravir, were included as control compounds, yielding EC₅₀ values of 0.47 μM and 0.27 μM, respectively, both consistent with reported literature values (Table 1).^6,23 Additionally, the half-maximal cytotoxic concentration (CC₅₀) of 2 was sufficiently high at >200 μM for Vero E6 cells (Table 1).

Table 1.

Anti-SARS-CoV-2 activity and cytotoxicity of 3-deazapurine derivatives 1–6, 10, and 11

Compound	R1	R2	R3	R4	EC50 (μM) (95% CI)	CC50 (μM)c
3-deazaadenosine (C3Ado, 1)	–NH2	–H	–H	–OH	>30a	>200
3-Deazaguanosine (C3Guo, 2)	–OH	–NH2	–H	–OH	1.14b (0.44–3.34)	>200
3-deazainosine (C3Ino, 3)	–OH	–H	–H	–OH	>30a	>200
3-fluoro-3-Deazaguanosine (4)	–OH	–NH2	–F	–OH	>30a	>200
3-chloro-3-Deazaguanosine (5)	–OH	–NH2	–Cl	–OH	>30a	>200
3-deazadiaminopurine ribonucleoside (6)	–NH2	–NH2	–H	–OH	>30a	>200
3-deazaguanine (C3Gua, 10)	–	–	–	–	12.6b (3.51–161)	>200
2′-deoxy-3-Deazaguanosine (dC3Guo, 11)	–OH	–NH2	–H	–H	5.35b (1.62–32.7)	>200
Remdesivir (GS-5734)	–	–	–	–	0.47b (0.30–0.75)	>200
N4-hydroxy cytidine (NHC, EIDD-1931)	–	–	–	–	0.27b (0.14–0.59)	>200

^aAnti-SARS-CoV-2 activity was evaluated by CPE assay using SARS-CoV-2-infected Vero E6 cells.

^bAnti-SARS-CoV-2 activity was evaluated by qRT-PCR of viral RNA using SARS-CoV-2-infected Vero E6 cells. Each EC₅₀ value was calculated by non-linear curve fit analysis using GraphPad Prism software.

^cCytotoxicity was evaluated by WST-8 assay using uninfected Vero E6 cells.

To probe the structure-activity relationship, we evaluated the anti-SARS-CoV-2 activity of two C³Guo analogs with halogen atoms at the C3 position, 4²⁴ and 5,²⁴ that we previously developed. Substitution of the C3 hydrogen of 2 with fluoride (4) or chloride (5) led to a loss of anti-viral activity, suggesting the C3 hydrogen atom is critical for exerting an inhibitory effect (Table 1). Next, we prepared 3-deazadiaminopurine ribonucleoside (6) (Scheme 1), which might be converted to 2 by 6-deamination by endogenous adenosine deaminase, and evaluated its anti-SARS-CoV-2 activity (Table 1). Starting with 7,²¹ a common synthetic intermediate for 2, dehydration of 4-carboxamide group with p-toluenesulfonyl chloride (TsCl) afforded the di-cyano derivative 8. Subsequent treatment with methanolic ammonia under heating facilitated cyclization to give the protected 3-deazadiaminopurine derivative 9 in 77% yield. The desired 3-deazadiaminopurine ribonucleoside (6) was ultimately obtained when 9 was treated with triethylamine trihydrofluoride (Et₃N･3HF); however, the resulting compound 6 showed no inhibitory effects on SARS-CoV-2 (Table 1). It has been previously reported C³Ado (1) has resistance to adenosine deaminase.²⁵ In the case of 6, therefore, the substitution of N3 for C3 might similarly confer resistance to 6-deamination.

Scheme 1.

Chemical synthesis of 3-deazadiaminopurine ribonucleoside (6)

TBDMS, tert-butyldimethylsilyl; TsCl, p-toluenesulfonyl chloride.

For their antiviral effect, NIs generally rely on cellular kinases to undergo stepwise addition of phosphate groups to form the corresponding active nucleoside 5′-triphosphate.²⁶ Similarly, C³Guo (2) may also depend on cellular kinases for stepwise phosphorylation to give its active form, C³Guo (2) 5′-triphosphate (2-TP) (Figure S3A).²⁷ Additionally, it is possible that 2 would be first phosphorolysed to 3-deazaguanine (C³Gua, 10) and d-ribose 1-phosphate by purine nucleoside phosphorylase (Figure S3B). The resulting C³Gua (10) would be then converted into C³Guo (2) 5′-monophosphate (2-MP) via glycosylation with phosphoribosyl pyrophosphate catalyzed by hypoxanthine guanine phosphoribosyl transferase.²⁸ Finally, subsequent stepwise phosphorylation of 2-MP would yield an active 5′-triphosphate, 2-TP.

Based on this prediction, we evaluated the anti-SARS-CoV-2 activity of C³Gua (10)²⁹ and 2′-deoxy-3-Deazaguanosine (dC³Guo, 11) (Scheme 2),³⁰ which were chemically synthesized as follows. Starting with 12,³¹ dehydration with trifluoroimidazole (TFAI) followed by cyclization with aqueous Na₂CO₃ in EtOH under heating afforded the dC³Guo derivative 13. Initially, we tried deprotection of the triisopropylsilane (TIPS) groups on the sugar moiety of 13, but purification of the resulting compounds was difficult; therefore, we protected the exocyclic amino group of 13 with a dimethylaminomethylene group to give 14. The TIPS groups were then deprotected with tetrabutylammonium fluoride (TBAF), and the resulting 15 was gently heated in 1 N HCl to give 16. Treatment of 16 with methanolic ammonia ultimately afforded the desired C³Gua (10) in good yield. The same steps were followed to prepare dC³Guo (11), except that 15 was treated with ammonium hydroxide. By CPE assay in Vero E6 cells, the resulting C³Gua (10) and dC³Guo (11) both had anti-SARS-CoV-2 activity, as expected (data not shown). Furthermore, the EC₅₀ values of 10 and 11, based on measured viral titers in SARS-CoV-2 infected Vero E6 cells, were 12.6 μM and 5.35 μM, respectively. These findings suggested that 2-TP serves as the active species responsible for the antiviral activity of C³Guo (2).

Scheme 2.

Chemical synthesis of C³Gua (10) and dC³Guo (11)

TIPS, triisopropylsilyl; TFAI, trifluoroacethyl imidazole; THF, tetrahydrofuran; DMF-DMA, N,N-dimethylformamide dimethyl acetal; TBAF, tetrabutylammonium fluoride.

C³Guo-TP (2-TP), the predicted active metabolite of C³Guo (2), does not act as a chain terminator of RdRp, but significantly reduces 5′-capping of viral RNA

Next, to verify that C³Guo (2) exerts anti-SARS-CoV-2 activity as a chain terminator of RdRp, we synthesized C³Guo-TP (2-TP), which, as discussed above, is predicted to be the active metabolite of 2 (Scheme 3). Starting with 17,²² the exocyclic amino group was protected with a dimethylaminomethylene group to give 18. After the deprotection of tert-butyldimethylsilyl (TBDMS) group with TBAF, the resulting 19 was used as a substrate for triphosphate synthesis. Based on a previous method for producing 5′-triphosphate,³² 19 was treated with 2-chloro-4H-1,3,2-benzodioxaphosphorin-4-one, followed by bis(tri-n-butylammonium)pyrophosphate to give the cyclotriphosphite intermediate. This intermediate was then oxidized by 1% iodine in pyridine–H₂O, followed by successive treatment with NH₄OH and then 50% aq. trifluoroacetic acid (TFA) to give the desired product 2-TP in five steps and in 20% yield after purification on a DEAE Sephadex column.

Scheme 3.

Chemical synthesis of C³Guo (2) 5′-triphosphate (2-TP)

TBDMS, tert-butyldimethylsilyl; TFA, trifluoroacetic acid; TBAF, tetrabuthylammonium fluoride.

We tested 2-TP in an in vitro SARS-CoV-2 RdRp-dependent RNA synthesis assay (Figure 3A). In this assay, the reactions comprised an RNA template-product scaffold, recombinant SARS-CoV-2 RdRp complex, NTPs, and 2-TP. A fluorescent intercalating dye was then added after incubation to determine the levels of RNA synthesis based on fluorescence intensity. Unexpectedly, RNA synthesis by the recombinant SARS-CoV-2 RdRp complex was not inhibited by the addition of 2-TP, whereas it was reduced by 2′,3′-dideoxy GTP (ddGTP), a potential chain terminator of viral RNA synthesis³³ (Figure 3B). This observation suggested that 2-TP does not act as a chain terminator of SARS-CoV-2 viral RNA replication, contradicting our initial prediction that 3-deazapurine ribonucleosides might inhibit SARS-CoV-2 RdRp.

Figure 3.

An active metabolite of C³Guo (2), 2-TP, is not a chain terminator of SARS-CoV-2 RdRp

(A) Schematic of the SARS-CoV-2 RdRp-dependent RNA synthesis assay.

(B) Results of the SARS-CoV-2 RdRp-dependent RNA synthesis. The reactions comprised RNA template–product scaffold, recombinant SARS-CoV-2 RdRp complex, NTPs, and 10 μM ddGTP or 2-TP. After incubation at 34°C for 2 h, fluorescent intercalating dye was added. SARS-CoV-2 RdRp-dependent RNA synthesis levels were determined from fluorescence intensity (Ex, 485 ± 5 nm; Em, 535 ± 10 nm) mean. All experiments were performed in quadruplicate; data are presented as mean + SEM. Statistically significant differences against the control reaction were determined using one-way ANOVA followed by Dunnett’s multiple comparisons test: ns (not significant), ∗∗p ≤ 0.01. NC, negative control.

To further elucidate the mechanism underlying the anti-SARS-CoV-2 activity of C³Guo (2), we next considered the capping pathway as a potential target of the active metabolite 2-TP. To this end, we tested the effect of 2-TP on the capping reaction catalyzed by vaccinia capping enzyme (VCE), which has been well characterized (Figure S4A).^34,35,36 In the control reaction with model RNA fragment and natural GTP, 5′-capping and subsequent N7 methylation by VCE proceeded efficiently to give 5′-m7GpppG capped RNA (IVa), as determined by liquid chromatograph-mass spectrometry (LC-MS) (Figure S4B), while no obvious capped RNA (IIIa or IVa) was observed in the reaction without GTP (Figure S4C). When 2-TP was included in the reaction, it was incorporated into the 5′-cap structure, yielding small amounts of 5′-C³GpppA-capped RNA (IIIb) and 5′-m7C³GpppG-capped RNA (IVb) (Figure S4D). Furthermore, in the presence of both GTP and 2-TP, the capping reactions yielded no 5′-capped RNA (IIIa,b or IVa,b) and the RNA substrates (I) were degraded (Figure S4E). Although this is only one possibility because viruses have different capping machinery and various molecular recognition mechanisms,³⁷ these data suggest that C³Guo (2) exerts its antiviral activity by targeting the capping machinery of SARS-CoV-2 rather than by targeting RdRp.

C³Guo (2) reduces viral titer and inhibits COVID-19 pneumonia in hamsters infected with SARS-CoV-2

To further assess the effect of C³Guo (2) on SARS-CoV-2 infection in vivo, we initially conducted a repeated dose toxicity study in C57BL/6J mice. Intraperitoneal (IP) administration of C³Guo (2) to mice at a concentration of 12.5 mg/kg/day or 112.5 mg/kg/day for 5 consecutive days led to no abnormal body weight changes or clinical observations, indicating that 2 does not cause adverse effects in mammals (Figure S5). Next, we investigated the in vivo antiviral activity of 2 using a Syrian hamster model of COVID-19.^38,39 Hamsters were inoculated with SARS-CoV-2 (1 × 10² PFU) and, starting 2 h before infection, given daily IP doses of 2 at 110 mg/kg/day (qd). Control hamsters (0 mg/kg/day) received a daily dose of saline IP as the vehicle alone. Molnupiravir (220 mg/kg/day) was used as the primary comparator drug and given twice daily (bid), following a previously established protocol,^40,41 as it demonstrated superior anti-SARS-CoV-2 compared to Remdesivir in our in vitro evaluation, as shown in Table 1. Whereas uninfected animals gained weight, animals in the infected groups showed a marginal reduction in body weight regardless of treatment with C³Guo (2), Molnupiravir, or the vehicle (Figure 4A). At 2 days post-infection (dpi), when viral load reaches its peak, the titer in lungs of hamsters treated with antiviral drugs (2 or Molnupiravir) was lower than that in those of vehicle controls (Figure 4B). Notably, treatment with 2 (110 mg/kg/day) resulted in more than a 10⁵-fold lower infectious viral titer, comparable to the levels observed in uninfected controls.

Figure 4.

Intraperitoneal treatment of C³Guo (2) controls viral burden and disease in hamsters inoculated with SARS-CoV-2

(A) Body weight changes in uninfected hamsters and SARS-CoV-2-infected hamsters treated with vehicle, C³Guo (2), or Molnupiravir are shown.

(B) Viral titer of lung samples at 2 dpi.

(C and D) Lung histopathology at 5 dpi. (C) Representative data from four or more independent assays are shown. The area of inflammation was determined (black line, lower panels), and (D) the percentage of inflammation in the lung sections was calculated. Groups of uninfected (n = 4) and SARS-CoV-2-infected hamsters treated with vehicle (n = 8), C³Guo (2) (n = 4), or Molnupiravir (n = 4) were shown. Horizontal bars represent the mean value of each group.

Statistically significant differences against the vehicle control were determined using one-way ANOVA, followed by Dunnett’s multiple comparisons tests: ns (not significant), ∗∗p ≤ 0.01, and ∗∗∗p ≤ 0.001. IP, intraperitoneal.

Histopathological analysis of lung tissues using hematoxylin and eosin staining at 5 dpi showed that severe pneumonia was evident in the vehicle control group, with inflammation observed in 7.28 ± 3.05% of lung sections (Figures 4C and 4D). Conversely, inflammation in lungs from animals treated with C³Guo (2) was notably low (0.02 ± 0.02%) and comparable to that observed in the uninfected group (0.01 ± 0.01%); even though lung histopathology from the Molnupiravir-treated group showed mild pneumonia (0.24 ± 0.16%). Collectively, these findings demonstrate that IP treatment with C³Guo (2) significantly inhibits viral replication and prevents the development of pneumonia in SARS-CoV-2-infected hamsters.

We also tested the in vivo anti-SARS-CoV-2 activity of C³Guo (2) when administered by inhalation using a nebulizer, with a view to exploring its potential clinical application (Figure 5). To simplify the comparison of antiviral and pneumonia-preventive effects, the same doses of Molnupiravir and C³Guo (2) were used. Thus, nebulizers were filled with saline or 10 mg/mL of antiviral drug (2 or Molnupiravir) to a volume corresponding to 170 mg/kg/day; they were inhaled by the hamsters once a day (qd) until no aerosol was produced. Based on the residual drug in the nebulizer, the estimated aerosolized inhalation dose was approximately 110 mg/kg/day for either drug.

Figure 5.

Inhalation treatment of C³Guo (2) controls viral burden and disease in hamsters inoculated with SARS-CoV-2

(A) Body weight changes in uninfected hamsters and SARS-CoV-2-infected hamsters treated with vehicle, C³Guo (2), or Molnupiravir are shown.

(B) Viral titer of lung samples at 5 dpi.

(C and D) Lung histopathology at 5 dpi. (C) Representative data from four or more independent assays are shown. The area of inflammation was determined (black line, lower panels), and (D) the percentage of inflammation in the lung sections was calculated. Groups of uninfected (n = 3) and SARS-CoV-2-infected hamsters treated with vehicle (n = 9), C³Guo (2) (n = 7), or Molnupiravir (n = 4) were shown. Horizontal bars represent the mean value of each group.

Statistically significant differences against the vehicle control were determined using one-way ANOVA, followed by Dunnett’s multiple comparisons tests: ns (not significant), ∗p ≤ 0.05, and ∗∗∗p ≤ 0.001. INH, inhalation.

Similar levels of body weight loss associated with infection were observed in the vehicle-treated control group and the Molnupiravir-treated group (Figure 5A). In contrast, hamsters treated with C³Guo (2) showed a reduction in the body weight loss associated with infection. Furthermore, inhalation of 2 resulted in more than 10²-fold lower infectious viral titer in the lungs of hamsters at 5 dpi (Figure 5B). Whereas histopathological findings of viral pneumonia were pronounced in the vehicle group (11.65 ± 2.06%), C³Guo (2) significantly attenuated the lung pathology, reducing the proportion of inflammatory area per lung section to 0.98 ± 0.36% (Figure 5C). These results indicate that inhalation of C³Guo (2) largely prevented viral pneumonia, making it a promising option for COVID-19 treatment.

Conclusion

The present study showed that C³Guo (2) exerts antiviral activity against SARS-CoV-2. Our structure–activity relationship analysis suggested that 2 is intracellularly converted into an active metabolite, 2-TP. In vitro assays of SARS-CoV-2 RdRp-dependent RNA synthesis and the VCE capping reaction showed that 2-TP does not act as a chain terminator of RdRp, but it significantly reduces 5′-capping of viral RNA. In animals infected with SARS-CoV-2, administration of C³Guo (2) markedly reduced infectious viral titer to a greater extent than the existing clinical drug Molnupiravir, which was administered more frequently. Furthermore, the development of pneumonia caused by SARS-CoV-2 infection was inhibited by the administration of C³Guo (2). There have been several reports on the antiviral activity of 3-deazapurine nucleosides, but most of them concern C³Ado (1), which acts as an inhibitor of S-adenosylhomocysteine hydrolase.^19,42,43 The potency of C³Guo (2) against SARS-CoV-2 underscores its potential as an effective therapeutic agent for COVID-19 and other future zoonotic coronavirus infections and highlights the potential of nucleoside analogs as antiviral agents with alternative therapeutic targets to RdRp.

Limitations of the study

We showed that C³Guo (2) has potent antiviral activity against SARS-CoV-2, but its activity against other RNA viruses was not evaluated. Future studies should explore the breadth of activity of C³Guo (2). We further showed that the active metabolite of C³Guo (2), 2-TP, does not act as a chain terminator of SARS-CoV-2 RdRp, but instead seems to target the capping machinery of SARS-CoV-2. However, we did not provide direct evidence of 5′-capping inhibition by 2. Future experiments should investigate in detail the antiviral mechanism of C³Guo (2).

Resource availability

Lead contact

Further information and requests for resources should be directed to and will be fulfilled by Noriaki Minakawa (minakawa@tokushima-u.ac.jp).

Materials availability

Compounds in this study will be provided following the request to the lead contact.

Data and code availability

• •Data: All data reported in this paper will be shared by the lead contact upon reasonable request.
• •Code: This study did not generate original code.
• •Others: Any additional information regarding the data reported in this paper is available from the lead contact upon reasonable request.

Acknowledgments

We thank Ms. Ayako Matsuo for kindly providing technical assistance for the in vitro experiments. This work was financially supported, in part, by JSPS KAKENHI grant numbers 21H02606 (N.M.), 24K02149 (N.M.) and 22K06527 (N.S.-T.), the Uehara Memorial Foundation (N.M.), Takeda Science Foundation (T.K.), the Ichiro Kanehara Foundation (T.K.), Taiju Life Social Welfare Foundation (T.K.), YOKOYAMA Foundation for Clinical Pharmacology grant numbers YRY-2324 (T.K.), and the Naito Foundation (N.S.-T.). N.H., K.Y., M.O., A.M., and S.I. are grateful to the research program for development of the intelligent Tokushima artificial exosome (iTEX) at Tokushima University.

Author contributions

N.-S.T. and T. Koma: Contributed equally to study. conceptualization, N.-S.T. and N.M.; methodology, N.-S.T., T. Koma, M.N.; formal analysis, N.S.-T., T. Koma, T.Kondo, and K.T.; investigation, N.S.-T., T. Koma, N.H., K.Y., M.O., A.M., S.I., T. Kondo, N.D., and M.N.; data curation, N.S.-T.; writing – original draft, N.S.-T. and T. Koma; writing – review and editing, N.S.-T., T. Koma, M.N., and N.M.; visualization, N.S.-T. and T. Koma; supervision, N.S.-T., M.N., and N.M. All authors have given approval to the final version of the manuscript.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Bacterial and virus strains
SARS-CoV-2/Hu/DP/Kng/19-020	Kanagawa Prefectural Institute of Public Health	GenBank LC528232
Chemicals, peptides, and recombinant proteins
3-deazaadenosine (C3Ado, 1)	Hinotani et al.22	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
3-Deazaguanosine (C3Guo, 2)	Hinotani et al.22	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
3-deazainosine (C3Ino, 3)	Hinotani et al.22	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
3-fluoro-3-Deazaguanosine (4)	Minakawa et al.24	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
3-chloro-3-Deazaguanosine (5)	Minakawa et al.24	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
3-deazadiaminopurine ribonucleoside (6)	This study	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
3-deazaguanine (C3Gua, 10)	This study	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
2′-deoxy-3-Deazaguanosine (C3Guo, 2)	This study	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
N4-hydroxycytidine (NHC, EIDD-1931)	Paymode et al.44	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
Remdesivir (GS-5734)	Cayman Chemical Company	Cat#30354
Molnupiravir (EIDD-2801)	Merck & Co., Inc.	LAGEVRIO CapsulesⓇ
C3Guo-TP (2-TP)	This study	Prepared at the Faculty of Pharmaceutical Science, Tokushima University, and its chemical purity (>95%) was verified by UPLC-MS analysis (see the supplemental information)
ddGTP	BIOLOG Life Science Institute GmbH & Co. KG	Cat#D019-05
DMSO	FUJIFUILM Wako Pure Chemical Corporation	Cat#048-32811
Otsuka Normal Saline	Otsuka Pharmaceutical Co., Ltd.	https://www.otsukakj.jp/med_nutrition/dikj/menu1/000224.php
10% Formalin solution	FUJIFUILM Wako Pure Chemical Corporation	Cat#060-03845
Crystal violet	FUJIFUILM Wako Pure Chemical Corporation	Cat#031-04852
QIAamp Viral RNA Mini kit	Qiagen	Cat#52926
Fast SYBR Green RT-PCR kit	Qiagen	Cat#204156
MEGAshortscript	Thermo Fisher Scientific	Cat#AM1354
10% Phosphate-buffered formalin solution	FUJIFUILM Wako Pure Chemical Corporation	Cat#062-01661
Recombinant SARS-CoV-2 RdRp complex	ProFoldin	Cat#RDRP-100S2
Critical commercial assays
RNA polymerase assay kit	ProFoldin	Cat#S2RPA100K
ScriptCapTM m7G Capping System	CellScript	Cat# C-SCCS2250
Experimental models: Cell lines
Vero E6 cells	ATCC	Cat#CRL-1586
Experimental models: Organisms/strains
C57BL/6J (female, 6 weeks old)	CLEA Japan, Inc.	C57BL/6JJcl
Syrian hamster (male, 5 weeks old)	Japan SLC	Slc:Syrian
Oligonucleotides
5′-d(AAATTTTGGGGGACCAGGAAC)-3′	SARS-CoV-2 nucleocapsid	Shintaro et al.45 Chemically prepared.
5′-d(TGGCACCTGTGTAGGGTCAAC)-3′	SARS-CoV-2 nucleocapsid	Shintaro et al.45 Chemically prepared.
5′-r(gggcgaauuaa)-3′	Ohno et al.46	Prepared by in vitro transcription.
5′-d(CAGTGAATTGTAATACGACT CACTATAGGGC)-3′	Ohno et al.46	Chemically prepared.
5′-d(TTAATTCGCCCTATAGTGAG TCGTATTACAATTCACTG)-3′	Ohno et al.46	Chemically prepared.
Software and algorithms
ChemDraw Professional 21.0	PerkinElmer	https://www.perkinelmer.com/category/chemdraw
Delta 6.1.0	JEOL	https://nmrsupport.jeol.com/
GraphPad Prism 9	GraphPad	https://www.graphpad.com/
UNIFI	Waters	https://www.waters.com/waters/en_US/UNIFI-Scientific-Information-System/nav.htm?cid=134801648&lset=1&locale=en_US&changedCountry=Y
7500 Real-Time PCR Software v2.3	Applied Biosystems	https://www.thermofisher.com/jp/ja/home/technical-resources/software-downloads/applied-biosystems-7500-real-time-pcr-system.html
NDP View 2	Hamamatsu Photonics KK	https://www.hamamatsu.com/jp/ja.html
Other
NMR: FT-NMR AV400NEO	Bruker	https://www.bruker.com/en/products-and-solutions/mr/nmr.html
NMR: JNM-ECZL500R	JEOL	https://www.jeol.com/products/scientific/feature_nmr/
LRMS: ACQUITY QDa	Waters	https://www.waters.com/waters/en_US/ACQUITY-QDa-Mass-Detector-for-Chromatographic-Analysis/nav.htm?cid=134761404&lset=1&locale=en_US&changedCountry=Y
UPLC-MS: BioAccord	Waters	https://www.waters.com/waters/en_US/BioAccord-LC-MS-System-for-Biopharmaceuticals/nav.htm?cid=135005818&locale=en_US
QIAcube	Qiagen	https://www.qiagen.com/jp/listpages/ez1-cards/qiacube/
Tecan i-control	Tecan	https://www.tecan.com/
NanoDrop spectrophotometer	Thermo Fisher Scientific	https://www.thermofisher.com/jp/ja/home/industrial/spectroscopy-elemental-isotope-analysis/molecular-spectroscopy/uv-vis-spectrophotometry/instruments/nanodrop.html
Acquity Premier BEH C18 1.7-μm VanGuard column	Waters	ID#186009457
7500 Real Time PCR System	Applied Biosystems	https://www.thermofisher.com/order/catalog/product/4377354?SID=srch-srp-4377354
NanoZoomer S210	Hamamatsu Photonics KK	https://www.hamamatsu.com/jp/ja.html
Mesh Nebulizer NE-U100	OMRON Corporation	https://www.healthcare.omron.co.jp/product/ne/ne-u100.html

Experimental model and subject details

Cells, viruses, and test compounds

Compounds 1–19, 2-TP and NHC (EIDD-1931)⁴⁴ were synthesized at the Faculty of Pharmaceutical Science, Tokushima University, and their chemical purity (all >95%) was verified by ultra-high-performance liquid chromatography (UPLC) and HRMS analysis (see the supplemental information). Remdesivir (GS-5734) was purchased from Merck & Co., Inc. (Rahway, NJ, USA). Molnupiravir (EIDD-2801) was purchased as LAGEVRIO Capsules^Ⓡ 200 mg from Merck & Co., Inc. (Rahway, NJ, USA). For in vitro studies, all compounds were solubilized in DMSO (Kanto Chemical Co., Inc., Tokyo, Japan). For in vivo studies, all compounds were dissolved in Otsuka Normal Saline (Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan) or 50% DMSO for intraperitoneal administration. Vero E6 cells (ATCC: CRL-1586) were maintained in Eagle’s MEM (EMEM) containing 10% heat-inactivated fetal bovine serum (FBS) and 1% penicillin–streptomycin–glutamine (Thermo Fisher Scientific Inc., MA, USA). Vero E6 cells were inoculated with SARS-CoV-2 (SARS-CoV-2/Hu/DP/Kng/19–020; GenBank LC528232; kindly provided by Kanagawa Prefectural Institute of Public Health, Kanagawa, Japan). Two days after inoculation, culture supernatants were collected and filtered to make virus stocks.

Animal experiments

To assess the toxicity of C³Guo (2) in mammals, 6-week-old female C57BL/6J mice (CLEA Japan, Inc.) were given 12.5 mg/kg or 125 mg/kg of C³Guo (2) IP daily for 5 days and observed for body weight changes and clinical signs. The antiviral effects of C³Guo (2) were tested in vivo in a Syrian hamster model of SARS-CoV-2 infection.³⁸ Male hamsters (Slc:Syrian, 5 weeks old) obtained from Japan SLC (Shizuoka, Japan) were intranasally inoculated with 1 × 10² PFU of SARS-CoV-2. They were given 110 mg/kg/day of C³Guo (2) (qd) or 220 mg/kg of Molnupiravir (bid) by IP injection, starting 2 h before infection.

For daily inhalation administration, each hamster was placed in a polypropylene tube (diameter, 36 mm; length, 90 mm) connected to a nebulizer (NE-U100, Omron, Kyoto, Japan; mass median aerodynamic diameter of particles, ∼5 μm) such that only the head was exposed to aerosolized drug. A solution of 10 mg/mL of C³Guo (2), 10 mg/mL Molnupiravir, or saline was placed in the nebulizer at a volume corresponding to a 170 mg/kg dose, and hamsters were exposed until no more aerosol was produced. Approximately 0.5 mL (5 mg equivalent) remained in the nebulizer; thus, hamsters were exposed to an aerosolized volume of ∼110 mg/kg. Hamsters were observed for appearance.

Infectious viral titer in the lung was evaluated at 2 dpi (peak infection) and 5 dpi; histopathological analysis of the lungs was performed at 5 dpi after the animals were euthanized. The right lung of each hamster was weighed and homogenized with stainless steel beads (5.0 mm) in 1 mL of phosphate-buffered saline (300 rpm, 90 s) twice using a Shakeman 6 homogenizer (BMS Inc., Tokyo, Japan). After centrifugation (13,200g, 10 min 4°C), the supernatants were collected and used for plaque assay. The left lung was perfused with 10% phosphate-buffered formalin solution (FUJIFILM Wako Pure Chemical Corporation) for histopathological analysis. All animal experiments with SARS-CoV-2 were performed in animal biosafety level 3 (ABSL3) containment laboratories at Tokushima University. All experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Tokushima University (Approval Number: T2020-116).

Method details

Chemical synthesis and compound characterization

All reactions were carried out using oven-dried glassware and magnetic stirring under argon atmosphere unless otherwise stated. Analytical thin-layer chromatography (TLC) was performed on Merck Kieselgel F254 and visualized by UV light (254 nm). Column chromatography was performed using KANTO Chemical silica gel 60N (neutral). Physical data were measured as follows; nuclear magnetic resonance (NMR) spectra were recorded on FT-NMR AV400NEO (Bruker, MA, USA) or JNM-ECZL500R (JEOL, Kyoto, Japan). ¹H NMR spectra were recorded at 400 or 500 MHz, referenced to in CDCl₃ with tetramethylsilane (TMS) (0.00 ppm), DMSO-d₆ (2.50 ppm), and D₂O (4.79 ppm). ¹³C NMR spectra were recorded at 125 MHz, referenced to in CDCl₃ with TMS (0.00 ppm). ³¹P NMR spectrum was recorded at 202 MHz, referenced to in D₂O (0.00 ppm, phosphoric acid as an external reference). Chemical shifts are reported in parts per million (δ), and signals are expressed as s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), or br (broad). All exchangeable protons were detected by addition of D₂O. Mass spectra were measured on a ACQUITY QDa (for LRMS, quadrupole, Waters), and BioAccord ACQUITY RDa (for HRMS, TOF, Waters). The purity of assay compounds was analyzed using UPLC-UV/MS (BioAccord ACQUITY, Waters) with an Acquity Premier BEH C18 1.7-μm VanGuard column (Waters, 50 × 2.1 mm), eluted with appropriate solvent system.

Detailed synthetic procedures and compound characterization are provided in the supplemental information, along with NMR spectra were also available.

Abbreviations

Standard abbreviations for the protecting groups are followed by the IUPACIUB Commission on Biochemical Nomenclature.

Reagents

All the reagents and solvents used were commercially available and used without further purification.

Synthesis of 3-deazadiaminopurine nucleoside (6), related to Scheme 1.

1-[2,3,5-Tri-O-(tert-buthyldimethylsilyl)-β-D-ribofuranosyl]-5-cyanomethyl-1H-imidazole-4-carbonitrile (8). To a solution of 7²¹ (1.48 g, 2.37 mmol) in pyridine (25.0 mL) was added p-toluenesulfonyl chloride (TsCl) (2.71 g, 14.2 mmol), and the whole was stirred for 27 h under room temperature. The reaction mixture was quenched by the addition of ice, and the solvent was removed in vacuo. The residue was partitioned between AcOEt and saturated aqueous NaHCO₃. The separated organic layer was further washed with H₂O (twice), followed by brine, dried (Na₂SO₄) and concentrated in vacuo. The residue was purified by a silica gel column, eluted with hexane/AcOEt (9/1–2/1) to give 8 (1.40 g, 97%) as a pale brown oil. ESI-LRMS m/z 629 [M+Na]⁺; ESI-HRMS calcd for C₂₉H₅₄N₄O₄Si₃ [M + H]⁺ 607.3526, found 607.3526; ¹H NMR (CDCl₃, 500 MHz) δ 7.79 (1 H, s, H-2), 5.67 (1 H, d, J = 6.7 Hz, H-1′), 4.17–4.13 (4 H, m, H-2′, H-3′ and CH₂CN), 3.97–3.95 (1 H, H-4′, m), 3.94–3.93 (1 H, m, H-5′a), 3.80 (1 H, dd, J = 2.2, 11.7 Hz, H-5′b), 0.95, 0.94, and 0.82 (each s, each 9 H, t-Bu × 3), 0.14 (6 H, s, Me×2), 0.12, 0.11, −0.02, and −0.32 (each 3 H, each s, Me×4); ¹³C NMR (CDCl₃, 125 MHz) δ 138.51, 128.88, 115.86, 113.90, 113.26, 89.42, 87.94, 77.33, 72.70, 63.21, 26.02, 25.76, 25.67, 18.49, 18.01, 17.84, 13.76, −4.36, −4.53, −4.60, −5.43, −5.73.

4,6-Diamino-1-[2,3,5-tri-O-(tert-buthyldimethylsilyl)-β-D-ribofuranosyl]-1H-imidazo[4,5-c]pyridine (9). A solution of 8 (1.40 g, 2.31 mmol) in NH₃/MeOH (saturated at 0°C, 30.0 mL) was heated at 100°C for 12 h in a sealed stainless tube. The reaction mixture was concentrated in vacuo, the residue was purified by a silica gel column, eluted with MeOH in CHCl₃ (0%–6%), to give 9 (1.11 g, 77%) as a dark red solid. ESI-LRMS m/z 624 [M + H]⁺; ESI-HRMS calcd for C₂₉H₅₇N₅O₄Si₃ [M + H]⁺ 624.3791 found 624.3830; ¹H NMR (CDCl₃, 500 MHz) δ 7.83 (1 H, s, H-8), 5.99 (1 H, s, H-3), 5.66 (1 H, d, J = 7.1 Hz, H-1′), 5.23 (2 H, br s, exchangeable with D₂O, NH₂), 4.31 (1 H, dd, J = 4.5, 7.1 Hz, H-2′), 4.24 (2 H, br s, exchangeable with D₂O, NH₂), 4.19 (1 H, dd, J = 4.0, 4.5 Hz, H-3′), 4.10–4.08 (1 H, m, H-4′), 3.90 (1 H, dd, J = 2.9, 11.5 Hz, H-5′a), 3.81 (1 H, dd, J = 2.4, 11.5 Hz, H-5′b), 0.97, 0.95, and 0.78 (each 9 H, each s, t-Bu×3), 0.17, 0.16, 0.12, 0.11, −0.10, and −0.44 (each 3 H, each s, Me×6); ¹³C NMR (CDCl₃, 125 MHz) δ 152.38, 149.26, 141.77, 138.43, 121.43, 88.33, 86.62, 79.24, 76.16, 72.80, 63.34, 26.07, 25.84, 25.71, 18.53, 18.09, 17.81, −4.49, −4.53, −4.65, −5.30, −5.41, −5.50.

4,6-Diamino-1-β-D-ribofuranosyl-1H-imidazo[4,5-c]pyridine (3-deazadiaminopurine nucleoside, 6). To a solution of 9 (250 mg, 0.4 mmol) in CH₂Cl₂ (5.0 mL) was added triethylamine trihydrofluoride (228 μL, 1.4 mmol) at 0°C. After being stirred at 25 h under room temperature, the solvent was removed in vacuo. The residue was purified by a silica gel column, eluted with MeOH in CHCl₃ (0%–35%), to give 6 (61 mg, 54%) as a dark brown solid. The analytical and assay samples were recrystallized from MeOH–hexane. ESI-LRMS m/z 282 [M + H]⁺; ESI-HRMS calcd for [M + H]⁺ C₁₁H₁₅N₅O₄ 282.1197 found 282.1241; ¹H NMR (DMSO-d₆, 500 MHz) δ 7.91 (1 H, s, H-8), 5.78 (2 H, br s, exchangeable with D₂O, NH₂), 5.75 (1 H, s, H-3), 5.52 (1 H, d, J = 5.9 Hz, H-1′), 5.41 (1 H, d, J = 6.2 Hz, 2′-OH, exchangeable with D₂O), 5.16 (1 H, d, J = 5.0 Hz, exchangeable with D₂O, 3′-OH), 5.09 (2 H, br s, exchangeable with D₂O, NH₂), 5.00 (1 H, t, J = 5.3 Hz, exchangeable with D₂O, 5′-OH), 4.29–4.24 (1 H, m H-2′), 4.06–4.02 (1 H, m, H-3′), 3.91–3.86 (1 H, m, H-4′), 3.65–3.60 (1 H, m, H-5′a), 3.58–3.53 (1 H, m, H-5′b); ¹³C NMR (DMSO-d₆, 125 MHz) δ 154.60, 150.30, 141.24, 137.46, 120.16, 88.25, 85.06, 76.63, 73.35, 70.17, 61.39.

Chemical synthesis of 3-deazaguanine (C³Gua, 10) and 2′-deoxy-3-Deazaguanosine (dC³Guo, 11), related to Scheme 2.

6-Amino-1-(2-deoxy-3,5-bis-O-triisopropylsilyl-β-D-ribofuranosyl)-1H-imidazo[4,5-c]pyridin-4(5H)-one (13). To a solution of 12³¹ (6.03 g, 10.1 mmol) in THF (150 mL) was added 1-(trifluoroacetyl)imidazole (1.73 mL, 15.2 mmol), and heated to reflux for 4 h. After the reaction mixture was concentrated in vacuo, the resulting residue was dissolved in EtOH (80.0 mL) and added to 5% aqueous Na₂CO₃ (40.0 mL). After being heated to reflux for 4 h, the reaction mixture concentrated in vacuo. The residue was partitioned between CHCl₃ and H₂O. The separated organic layer was further washed with H₂O (twice) and followed by brine, dried (Na₂SO₄), and concentrated in vacuo. The residue was purified by a silica gel column, eluted with MeOH in CHCl₃ (0%–15%), to give 13 (5.45 g, 93%) as a dark green foam. ESI-LRMS m/z 601 [M+Na]⁺; ESI-HRMS calcd for [M + H]⁺ C₂₉H₅₄N₄O₄Si₂ 579.3756 found 579.3770; ¹H NMR (CDCl₃, 400 MHz) δ 12.74 (1 H, br s, exchangeable with D₂O, NH), 7.78 (1 H, s, H-8), 6.00 (1 H, dd, J = 5.5, 8.3 Hz, H-1′), 5.53 (1 H, s, H-3), 4.85 (2 H, br s, exchangeable with D₂O, NH₂), 4.74–4.69 (1 H, m, H-3′), 4.11–4.05 (1 H, m, H-4′), 3.87 (1 H, dd, J = 3.3, 11.0 Hz, H-5′a), 3.81 (1 H, dd, J = 4.5, 11.0 Hz, H-5′b), 2.44 (1 H, ddd, J = 5.2, 8.3, 13.1 Hz, H-2′a), 2.37 (1 H, ddd, J = 2.1, 5.5, 13.1 Hz, H-2′b), 1.16–1.05 (42 H, m, TIPS); ¹³C NMR (CDCl₃, 125 MHz) δ 158.60, 147.30, 142.92, 136.52, 124.03, 88.42, 84.89, 73.01, 72.75, 63.63, 41.64, 18.01, 18.01, 17.98, 12.08, 11.89.

6-[{(Dimethylamino)methylene}amino]-1-(2-deoxy-3,5-bis-O-triisopropylsilyl-β-D-ribofuranosyl)-1H-imidazo[4,5-c]pyridin-4(5H)-one (14). To a solution of 13 (3.47 g, 6.00 mmol) in DMF (60.0 mL) was added N,N-dimethylformamide dimethyl acetal (4.00 mL, 30.0 mmol), and the whole was stirred for 5.5 h at room temperature. The solvent was removed in vacuo and the residue was partitioned between CHCl₃ and H₂O. The separated organic layer was further washed with H₂O (twice) and followed by brine, dried (Na₂SO₄) and concentrated in vacuo. The residue was purified by a silica gel column, eluted with MeOH in CHCl₃ (0%–20%) to give 14 (3.83 g, quant) as a dark brown foam. ESI-LRMS m/z 656 [M+Na]⁺; ESI-HRMS calcd for C₃₂H₅₉N₅O₄Si₂ 634.4178 found 634.4184; ¹H NMR (CDCl₃, 400 MHz) δ 8.91 (1 H, br s, exchangeable with D₂O, NH), 7.85 (1 H, s, H-8), 7.82 (1 H, s, CH), 6.06 (1 H, dd, J = 5.5, 8.3 Hz, H-1′), 5.77 (1 H, s, H-3), 4.75–4.71 (1 H, m, H-3′), 4.12–4.07 (1 H, m, H-4′), 3.86 (1 H, dd, J = 3.4, 11.0 Hz, H-5′a), 3.79 (1 H, dd, J = 4.8, 11.0 Hz, H-5′b), 3.12 and 3.04 (each 3 H, each s, Me×2), 2.50 (1 H, ddd, J = 5.2, 8.3, 13.1 Hz, H-2′a), 2.40 (1 H, ddd, J = 2.2, 5.5, 13.1 Hz, H-2′b), 1.21–1.04 (42 H, m, TIPS); ¹³C NMR (CDCl₃, 125 MHz) δ 158.21, 154.32, 150.51, 141.43, 136.87, 127.83, 88.57, 84.83, 78.28, 72.80, 63.67, 41.61, 40.66, 34.74, 18.01, 17.96, 12.08, 11.87.

6-[{(Dimethylamino)methylene}amino]-1-(2-deoxy-β-D-ribofuranosyl)-1H-imidazo[4,5-c]pyridin-4(5H)-one (15). To a solution of 14 (3.73 g, 5.88 mmol) in THF (60.0 mL) was added TBAF (1.0 M THF solution, 14.7 mL, 14.7 mmol) at 0°C. After being stirred for 50 min at room temperature, the reaction mixture was concentrated in vacuo. The residue was crystallized from MeOH to give 15 (1.56 g, 83%) as white crystals. ESI-LRMS m/z 344 [M+Na]⁺; ESI-HRMS calcd for [M + H]⁺ C₁₄H₁₉N₅O₄ 322.1510 found 322.1529; ¹H NMR (DMSO-d₆, 400 MHz) δ 10.59 (1 H, br s, exchangeable with D₂O, NH), 8.04 (1 H, s, H-8), 8.00 (1 H, s, CH), 6.10 (1 H, dd, J = 6.0, 7.3 Hz, H-1′), 6.07 (1 H, s, H-3), 5.28 (1 H, d, J = 3.8 Hz, exchangeable with D₂O, 3′-OH), 4.99 (1 H, t, J = 5.2 Hz, exchangeable with D₂O, 5′-OH), 4.39–4.32 (1 H, m, H-3′), 3.87–3.81 (1 H, m, H-4′), 3.59–3.48 (2 H, m, H-5′a and H-5′b), 3.06 (3 H, s, CH₃), 2.94 (3 H, s, CH₃), 2.48–2.44 (1 H, m, H-2′a), 2.25 (1 H, ddd, H-2’b, J = 3.1, 6.0, 13.4 Hz); ¹³C NMR (DMSO-d₆, 125 MHz) δ 157.38, 155.05, 150.99, 140.94, 137.62, 126.92, 87.65, 84.28, 77.05, 70.52, 61.51, 34.15.

6-[{(Dimethylamino)methylene}amino]-1H-imidazo[4,5-c]pyridin-4(5H)-one (16). A solution of 15 (1.56 g, 4.85 mmol) in 1.0 N HCl (50 mL) was stirred at 50°C for 4 h. After being cooled to room temperature, the reaction mixture was neutralized by the addition of 1.0 N NaOH at 0°C, and the solvent was removed in vacuo. The residue was purified by a silica gel column, eluted with MeOH in CHCl₃ (0%–30%) to give 16 (498 mg, 50%) as a pale pink solid. ESI-LRMS m/z 206 [M + H]⁺; ESI-HRMS calcd for [M + H]⁺ C₉H₁₁N₅O 206.1036 found 206.1036; ¹H NMR (DMSO-d₆, 400 MHz) δ 12.90 (0.4 H, br s, exchangeable with D₂O, NH), 12.24 (0.6 H, br s, exchangeable with D₂O, NH), 10.66 (0.4 H, br s, exchangeable with D₂O, NH), 10.43 (0.6 H, br s, exchangeable with D₂O, NH), 7.99–7.74 (2 H, m, H-8 and CH), 5.94 (0.4 H, s, H-3), 5.79 (0.6 H, s, H-3), 3.04 and 2.92 (each 3 H, each s, Me×2).

6-Amino-1H-Imidazo[4,5-c]pyridin-4(5H)-one (C³Gua, 10).²⁹ A solution of 16 (74.0 mg, 0.36 mmol) in NH₃/MeOH–H₂O (3:1) (saturated at 0°C, 8.0 mL) was stirred at room temperature for 43 h. The solvent was removed in vacuo. The resulting residue containing 10 was diluted in potassium phosphate buffer (1.0 mL, c = 10 mmol/L, pH 8.0) and purified on a C18 cartridge column (YMC Dispo SPE C18), eluted with 5% MeOH to give C³Gua (10) (41.9 mg, 77%) as a pale pink solid. ESI-LRMS m/z 173 [M+Na]⁺; ESI-HRMS calcd for [M + H]⁺ C₆H₆N₄O 151.0614 found 151.0648; ¹H NMR (DMSO-d₆, 400 MHz) δ 12.68 (0.2 H, br s, exchangeable with D₂O, NH), 11.89 (0.8 H, br s, exchangeable with D₂O, NH), 10.36 (0.2 H, br s, exchangeable with D₂O, NH), 10.15 (0.8 H, br s, exchangeable with D₂O, NH), 7.85 (0.2 H, s, H-8), 7.61 (0.8 H, s, H-8), 5.52–5.18 (2 H, m, exchangeable with D₂O, NH₂), 5.52–5.18 (1 H, m, H-3).

6-Amino-1-(2-deoxy-β-D-ribofuranosyl)-1H-imidazo[4,5-c]pyridin-4(5H)-one (dC³Guo, 11).³⁰ A solution of 15 (104 mg, 0.32 mmol) in 28% NH₄OH (5.0 mL) was allowed to stand at room temperature for 7 days. The reaction mixture was concentrated in vacuo, the resulting solid was crystallized from MeOH–hexane to give 11 (21 mg, 24%) as white crystals. ESI-LRMS m/z 267 [M + H]⁺; ¹H NMR (CDCl₃, 500 MHz) δ 10.28 (1 H, s, exchangeable with D₂O, NH), 7.86 (1 H, s, H-8), 5.93 (1 H, t, J = 6.9 Hz, H-1′), 5.56 (2 H, br s, exchangeable with D₂O, NH₂), 5.43 (1 H, s, H-3), 5.31 (1 H, d, J = 4.3 Hz, exchangeable with D₂O, 3′-OH), 4.90 (1 H, t, J = 5.4 Hz, exchangeable with D₂O, 5′-OH), 4.33–4.26 (1 H, m, H-3′), 3.83–3.76 (1 H, m, H-4′), 3.55–3.34 (2 H, m, H-5′a, H-5′b), 2.48–2.41 (1 H, m, H-2′a), 2.21 (1 H, ddd, J = 3.3, 6.9, 13.4 Hz, H-2′b).

Synthesis of 3-Deazaguanosine 5′-triphosphate (2-TP), related to Scheme 3.

6-[{(Dimethylamino)methylene}amino]-1-(5-O-tert-butyldimethylsilyl-2,3-O-isopropylidene-β-D-ribofuranosyl)-1H-imidazo[4,5-c]pyridin-4(5H)-one (18). To a solution of 17²² (273 mg, 0.60 mmol) in DMF (6.0 mL) was added N,N-dimethylformamide dimethyl acetal (358 μL, 3.0 mmol), and the whole was stirred for 4 h at room temperature. The solvent was removed in vacuo and the residue was partitioned between CHCl₃ and H₂O. The separated organic layer was further washed with H₂O (twice), followed by brine. The separated organic layer was dried (Na₂SO₄) and concentrated in vacuo. The residue was purified by a silica gel column, eluted with MeOH in CHCl₃ (0%–10%), to give 18 (288 mg, 98%) as a white foam. ESI-LRMS m/z 492 [M + H]⁺; ESI-HRMS calcd for [M + H]⁺ C₂₃H₃₇N₅O₅Si [M + H]⁺ 492.2637, found 492.2641; ¹H NMR (CDCl₃, 500 MHz) δ 9.05 (1 H, br s, exchangeable with D₂O, NH), 7.86 (1 H, s, H-8), 7.81 (1 H, s, CH), 5.81 (1 H, d, J = 2.9 Hz, H-1′), 5.78 (1 H, s, H-3), 4.87 (1 H, dd, J = 2.9, 6.2 Hz, H-2′), 4.84 (1 H, dd, J = 2.9, 6.2 Hz, H-3′), 4.42–4.39 (1 H, m, H-4′), 3.85 (1 H, dd, J = 3.3, 11.2 Hz, H-5′a), 3.79 (1 H, dd, J = 3.3, 11.2 Hz, H-5′b), 3.12 and 3.04 (each 3 H, each s, NMe×2), 1.62 and 1.39 (each 3 H, each s, Me×2), 0.85 (9 H, s, t-Bu), 0.05 and 0.04 (each 3 H, each s, SiMe×2); ¹³C NMR (CDCl₃, 125 MHz) δ 157.91, 154.25, 150.45, 140.50, 136.99, 128.03, 114.32, 92.38, 86.56, 85.46, 81.21, 77.79, 63.35, 40.69, 34.78, 27.32, 25.89, 25.35, 18.34, −5.41, −5.61.

6-[{(Dimethylamino)methylene}amino]-1-(2,3-O-isopropylidene-β-D-ribofuranosyl)-1H-imidazo[4,5-c]pyridin-4(5H)-one (19). To a solution of 18 (263 mg, 0.53 mmol) in THF (11.0 mL) was added 1M TBAF in THF (640 μL, 0.64 mmol) at 0°C, and the whole was stirred for 4 h at room temperature. The solvent was removed in vacuo. The residue was purified by a silica gel column, eluted with MeOH in CHCl₃ (0%–25%), to give 19 (187 mg, 93%) as a white solid. ESI-LRMS m/z 400 [M+Na]⁺; ESI-HRMS calcd for C₁₇H₂₃N₅O₅ [M + H]⁺ 378.1772, found 378.1788; ¹H NMR (DMSO-d₆, 500 MHz) δ 10.64 (1 H, br s, exchangeable with D₂O), 8.04 (1 H, s, H-8), 7.97 (1 H, s, CH), 6.04 (1 H, s, H-3), 5.90 (1 H, d, J = 3.4 Hz, H-1′), 5.14–5.12 (1 H, m, H-2′), 5.14–5.12 (1 H, m, exchangeable with D₂O, OH), 4.93 (1 H, dd, J = 2.8, 6.3 Hz, H-3′), 4.16–4.13 (1 H, m, H-4′), 3.54–3.45 (2 H, m, H-5′a, H-5′b), 3.06 and 2.91 (each 3 H, each s, NMe×2), 1.55 and 1.33 (each 3 H, Me×2); ¹³C NMR (DMSO-d₆, 125 MHz) δ 157.37, 155.12, 151.15, 140.95, 137.83, 127.02, 113.56, 89.83, 85.53, 83.26, 80.76, 77.39, 61.13, 34.17, 27.00, 25.21, 13.55.

6-Amino-1-(β-D-ribofuranosyl)-1H-imidazo[4,5-c]pyridin-4(5H)-one 5′-triphosphate (2-TP). To a solution of 19 (150 mg, 0.4 mmol) in pyridine (0.4 mL) and 1,4-dioxane (1.4 mL) was added an 1.0 M solution of 2-chloro-4H-1,2,3-dioxaphosphorin-4-one in 1,4-dioxane (0.44 μL, 0.44 mmol). After 15 min, a 0.5 M solution of bis(tri-n-butylammonium)pyrophosphate in DMF (1.2 mL, 0.6 mmol) and tri-n-butylamine (0.40 mL, 1.7 mmol) were quickly added, and the reaction mixture was stirred for 10 min. A solution of 1% iodine in pyridine/water (98/2, v/v) (ca. 8 mL) was then added. After 5 min, the excess iodine was decomposed by adding 5% aqueous solution of Na₂S₂O₃ (ca. 5 mL), and the reaction mixture was stirred for 5 min. The reaction mixture was concentrated in vacuo, and ammonium hydroxide (28%, 5 mL) was added to the residue. After 20 h, the solution was concentrated in vacuo, and the residue was added 50% trifluoro acetic acid (TFA) (10 mL) at 0°C, and the whole was stirred for 2 h at room temperature. The solvent was removed in vacuo and co-evaporated with EtOH–H₂O (1:1), and the resulting residue was diluted in water (300 mL). The solution was applied to a DEAE Sephadex column (2.1 × 20 cm), which was eluted with a linear gradient of 800 mL each of water and 0.8 M triethylammonium bicarbonate (TEAB) buffer. Fractions containing 2-TP were concentrated in vacuo, and the residue was co-evaporated with EtOH. The residue was dissolved in water (5 mL), which was applied to a column of DIAION PK 212 (H⁺ form) and then DIAION WK 20 (Na⁺ form), and the fractions containing the desired triphosphates were concentrated in vacuo to give 2-TP as a tri-sodium salt (47 mg, 20%) as a brown solid. ESI-LRMS m/z 520 [M–H]^–; ESI-HRMS calcd for C₁₁H₁₇N₄O₁₄P₃ [M–H]^– 520.9881, found 520.9908; ¹H-NMR (D₂O, 500 MHz) δ 7.95 (1 H, s, H-8), 5.75 (1 H, d, J = 7.5 Hz, H-1′), 4.60 (1 H, dd, J = 5.6, 7.5 Hz, H-2′), 4.53 (1 H, dd, J = 2.4, 5.6 Hz, H-3′), 4.29–4.26 (2 H, m, H-4′, H-5′a), 4.15–4.11 (1 H, m, H-5′b); ³¹P-NMR (D₂O, 202 MHz) δ −5.08 (d, J = 20 Hz), −10.17 (d, J = 20 Hz), −20.84 (t, J = 20 Hz).

Antiviral assays in vitro

Vero E6 cells were seeded at a density of 3×10⁴ cells per well in a 96-well plate. On the next day, culture media were replaced with 2% FBS-EMEM containing the indicated concentration of the compounds and preincubated at 37°C for 1 h. SARS-CoV-2 was then added at a multiplicity of infection (MOI) of 0.001, and the cells were incubated again at 37°C for 1 h. Virus-containing medium was then replaced with the compound-containing medium and the cultures were incubated at 37°C for 3 days. On day 3 post-infection, culture supernatants were collected to measure viral copy number, and the cells in the wells were fixed by 10% formalin solution (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) containing 0.5% crystal violet (FUJIFILM Wako Pure Chemical Corporation) to assess cytopathic effects (CPE assay). Viral titer was determined by the plaque assay method as described previously.⁴⁷ All in vitro experiments handling SARS-CoV-2 were conducted at biosafety level 3 (BSL3) in the Tokushima University.

Quantification of SARS-CoV-2 viral RNA genome by qRT-PCR

Viral RNA in the collected supernatants was automatically extracted by using a QIAamp Viral RNA Mini kit (Qiagen, Hilden, Germany) and QIAcube system (Qiagen). The copy number of each viral RNA sample was measured using a primer pair specific for the SARS-CoV-2 N gene [5′-d(AAATTTTGGGGGACCAGGAAC)-3′ and 5′-d(TGGCACCTGTGTAGGGTCAAC)-3′]⁴⁵ and a Fast SYBR Green RT-PCR kit (Qiagen) in accordance with the manufacturer’s instructions.

SARS-CoV-2 RdRp-dependent RNA synthesis assay

Recombinant SARS-CoV-2 RdRp complex (RDRP-100S2) and an RNA polymerase assay kit (S2RPA100K) were purchased from ProFoldin (Hudson, MA, USA). RNA synthesis assays were performed in a 50-μL volume with or without 10 μM 2-TP or ddGTP (BIOLOG Life Science Institute GmbH & Co. KG, Bremen, Germany) in accordance with the manufacturer’s instructions. Reactions were incubated at 34°C for 2 h and then stopped by adding fluorescence dye (150 μL). Fluorescence intensity (Ex = 485 nm, Em = 535 nm) was measured by using a plate reader (Tecan i-control, Tecan, Männedorf, Switzerland).

Capping analysis

The VCE capping reaction was performed as reported by Ohno et al.⁴⁶ To evaluate RNA capping efficiency, a short RNA [5′-r(gggcgaauuaa)-3′] with 5′-triphosphate was prepared by in vitro transcription. For the template, two oligo DNAs [5′-d(CAGTGAATTGTAATACGACTCACTATAGGGC)-3′, and 5′-d(TTAATTCGCCCTATAGTGAGTCGTATTACAATTCACTG)-3′] were annealed by heating and cooling in hybridization buffer (10 mM Tris–Cl, 100 mM NaCl, pH 7.6). Using the template and MEGAshortscript (Thermo Fisher Scientific), transcription was carried at 42°C overnight. After treatment with DNase, the transcript was purified by denaturing PAGE (8.0 M urea, 0.5×TBE buffer, 20% polyacrylamide gel). Eluted RNA was extracted with phenol/chloroform and precipitated with ethanol. The concentration of recovered RNA was measured by using a NanoDrop spectrophotometer (Thermo Fisher Scientific). Capping reactions were then carried out by using the ScriptCap m⁷G Capping System (C-SCCS2250, CellScript, WI, USA) at a scale of 1/5 (20 μL) in accordance with the manufacturer’s protocol. For 200 pmol of RNA, 20 nmol of GTP or 2-TP and 8 units of Script-Cap Capping Enzyme (VCE) were incubated in ScriptCap Capping Buffer with 0.1 mM SAM and ScriptGuard RNase Inhibitor at 37°C. After 2 h, an aliquot of sample was analyzed by UPLC-MS (BioAccord, Waters) using an Acquity Premier BEH C18 1.7-μm VanGuard column (Waters, 50 × 2.1 mm), eluted with linear gradient of 2.5%–25% MeOH containing 20 mM Et₃N and 40 mM 1,1,1,3,3,3-hexafluoropropan-2-ol at a flow rate of 0.3 mL/min for 20 min.

Histopathology

Lung tissues were fixed in 10% phosphate-buffered formalin (FUJIFILM Wako Pure Chemical Corporation) and routinely embedded in paraffin, sectioned, and stained with hematoxylin and eosin (HE), The slides were scanned by NanoZoomer S210 (Hamamatsu Photonics KK, Shizuoka, Japan), and the degree of pneumonia was assessed by measuring the percentage area of inflammation within the total area of each section using NDP View 2 software (Hamamatsu Photonics KK).

Quantification and statistical analysis

Prism version 9.5.1 (GraphPad, MA, USA) was used for data analysis. The data in bar graphs are presented as mean +SEM or ±SEM. One-way ANOVA followed by Dunnett’s multiple comparisons test was performed to determine statistical significance among groups. A p value of <0.05 was considered statistically significant: ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, and ^∗∗∗∗p < 0.0001.

Published: March 1, 2025