Antiviral Candidates for Treating Hepatitis E Virus Infection

Globally, hepatitis E virus (HEV) causes significant morbidity and mortality each year. Despite this burden, there are no specific antivirals available to treat HEV patients, and the only licensed vaccine is not available outside China.

ABSTRACT

Globally, hepatitis E virus (HEV) causes significant morbidity and mortality each year. Despite this burden, there are no specific antivirals available to treat HEV patients, and the only licensed vaccine is not available outside China. Ribavirin and alpha interferon are used to treat chronic HEV infections; however, severe side effects and treatment failure are commonly reported. Therefore, this study aimed to identify potential antivirals for further development to combat HEV infection. We selected 16 compounds from the nucleoside and nonnucleoside antiviral classes that range in developmental status from late preclinical to FDA approved and evaluated them as potential antivirals for HEV infection, using genotype 1 replicon luminescence studies and replicon RNA quantification. Two potent inhibitors of HEV replication included NITD008 (half-maximal effective concentration [EC₅₀], 0.03 μM; half-maximal cytotoxic concentration [CC₅₀], >100 μM) and GPC-N114 (EC₅₀, 1.07 μM, CC₅₀, >100 μM), and both drugs reduced replicon RNA levels in cell culture (>50% reduction with either 10 μM GPC-N114 or 2.50 μM NITD008). Furthermore, GPC-N114 and NITD008 were synergistic in combinational treatment (combination index, 0.4) against HEV replication, allowing for dose reduction indices of 20.42 and 8.82 at 50% inhibition, respectively. Sofosbuvir has previously exhibited mixed results against HEV as an antiviral, both in vitro and in a few clinical applications; however, in this study it was effective against the HEV genotype 1 replicon (EC₅₀, 1.97 μM; CC₅₀, >100 μM) and reduced replicon RNA levels (47.2% reduction at 10 μM). Together these studies indicate drug repurposing may be a promising pathway for development of antivirals against HEV infection.

INTRODUCTION

Hepatitis E virus (HEV) is a clinically important emerging virus, causing significant global morbidity and approximately 20 million infections, 44,000 deaths, and 3,000 stillbirths each year (1, 2). Despite this substantial disease burden, currently no specific antivirals exist for treating HEV infections, and the sole licensed vaccine (Hecolin) is only available in China (3). Ribavirin and pegylated alpha interferon (PEG-IFN-α) are often used as therapies for chronic HEV infections (4–6); however, treatment failure is common (7–9) and accompanied by severe side effects (6, 10). Therefore, further research is needed to identify safe and effective antivirals to treat HEV patients.

Recovery from acute HEV infection is often protracted, with typical hepatitis symptoms persisting for approximately 4 to 6 weeks (11). Fulminant hepatitis from HEV infection can prove fatal, with mortality rates reported between 0.5% and 4%, depending on the geographical region (12–14). Higher mortality rates of up to 27% have been reported in pregnant women (14, 15), with HEV infection being particularly severe in the third trimester and fulminant hepatitis often resulting in stillbirth and maternal death (16, 17). Although HEV infection is usually self-limiting, persistent infections have been frequently reported in immunosuppressed and immunocompromised patients (18–22).

HEV is classified within the Hepeviridae family, which is thought to have arisen from an ancient recombination event between viruses from different positive-sense RNA superfamilies (23). HEV comprises a single-stranded RNA genome, encapsidated within a 27- to 34-nm icosahedral, nonenveloped virion (24). The HEV genome is around 7.2 kb, with a 5′ cap and a 3′ polyadenylated tail (25, 26). The genome consists of three open reading frames (ORFs), with ORF1 encoding the nonstructural proteins and ORF2 the capsid protein, while ORF3 is a phosphoprotein thought to act as a viroporin to facilitate viral release from the host cell (27, 28).

Every stage of the HEV replication cycle can be exploited for antiviral design, as many of the nonstructural proteins are essential for viral replication, including the RNA-dependent RNA polymerase (RdRp), which makes them ideal antiviral targets. Therapies targeting the RdRp have proven highly successful, with several hepatitis C virus (HCV) antivirals being FDA approved for treatment (reviewed in reference 29).

The RdRp is highly conserved across all RNA viral families, forming the canonical protein structure resembling a closed right hand, with finger, palm, and thumb domains (30, 31). Two classes of antivirals are employed to target the RdRp, nucleoside analogs (NAs) and nonnucleoside inhibitors (NNIs).

HEV is a largely understudied virus, and antiviral development has been hampered by poor viral replication levels in cell culture systems (32) and difficulties in purifying the viral polymerase in its active form (25, 33). As such, several HEV replicons have been constructed from various infectious clones (34–38), which have allowed for effective preclinical screening of antiviral candidates (39–41).

As many antivirals have been successfully developed against other viruses, such as HCV, repurposing these compounds as potential therapies against emerging infections such as HEV should be considered. This study aimed to identify broad-spectrum antiviral candidates to combat HEV infection. All compounds examined in this study were previously developed against other viruses (Table 1) but have not been reported against HEV before, with the exception of sofosbuvir. We employed a subgenomic replicon approach to screen 16 compounds belonging to the NA or NNI classes of antivirals and identified three potent compounds that were profiled for dose-responsiveness and cytotoxicity, two of which were further examined for combinational synergism. The potent compounds identified in this work provide a promising platform for the development of antivirals to treat HEV infections.

TABLE 1.

Antiviral compounds examined in this study

Compound class/name	Chemical structure	Molecular mass (g/mol)	Original target virus	RdRp binding site	Developmental stage	Reference
Nonnucleoside inhibitors
Beclabuvir (BMS-791325)		659.8	Hepatitis C virus	Thumb I	Phase II clinical trials	81
Dasabuvir (ABT-333)		493.6	Hepatitis C virus	Palm I	FDA approved	82
Filibuvir (PF868554)		503.6	Hepatitis C virus	Thumb II	Halted after phase II trials	83
GPC-N114		436.8	Picornaviruses	RNA channel	Preclinical	46
JTK-109		638.1	Hepatitis C virus	Thumb I	Halted after phase II trials	84
Lomibuvir (VX-222)		445.6	Hepatitis C virus	Thumb II	Halted after phase II trials	85
Nesbuvir (HCV-796)		446.5	Hepatitis C virus	Palm II	Halted after phase II trials	86
Setrobuvir (ANA-598)		560.6	Hepatitis C virus	Palm I	Halted after phase II trials	87
Tegobuvir (GS-9190)		517.4	Hepatitis C virus	Palm β	Halted after phase II trials	88
TMC-647055		606.7	Hepatitis C virus	Thumb I	Halted after phase II trials	59
Triazavirin		228.2	Influenza	NDb	Preclinical	89
Nucleoside analogs
2’-C-Methylcytidinea (2CMC)		257.2	Hepatitis C virus	Active site	Halted after phase II clinical trials	90
7-deaza-2′-C-methyladenosine (7DMA)		280.3	Hepatitis C virus	Active site	Preclinical	91
Favipiravir (T705)		157.1	Influenza	Active site	FDA approved	92
NITD008		290.28	Dengue virus	Active site	Preclinical	45
Sofosbuvir (PSI-7977)		529.5	Hepatitis C virus	Active site	FDA approved	93
Other
Quercetagetin		318.23	Herpesviruses	NDb	Preclinical	94

^aPositive control compound.

^bND, not determined.

RESULTS

Polymerase-targeted antivirals are examined for anti-HEV activity.

Sixteen compounds (Table 1) were selected for examination of HEV replication inhibition and screened at a fixed concentration of 10 μM against the HEV human genotype 1 (G1) replicon pSK-HEV-2-Luc (Fig. 1A). The NA 2′-C-methylcytidine (2CMC) was used as a positive control to demonstrate inhibition of HEV replication (Fig. 1B). Antiviral efficacy was quantified by the relative luminescence of treated cells (test compounds at 10 μM) compared to that of mock-treated cells (vehicle control; 0.5%, vol/vol, dimethyl sulfoxide [DMSO]). Four antivirals, including NITD008, GPC-N114, sofosbuvir, and dasabuvir, demonstrated more than 50% inhibition at 10 μM (Fig. 1B). These four antivirals were selected for further examination to ascertain effective dose and cytotoxicity profiles, while the other 12 compounds did not reach the threshold of >50% inhibition at 10 μM and were dropped from further investigation in this study.

FIG 1.

Screening of broad-spectrum antivirals against HEV G1 replication. (A) A schematic representation of the HEV genome (top) is shown for comparison with the subgenomic replicon (bottom), with nonstructural proteins encoded by ORF1 and the structural proteins and phosphoproteins encoded by ORF2 and ORF3, respectively. Nucleotide positions, 5′ untranslated regions (UTR), and 3′ polyadenylated tails [poly(A)] are indicated. The HEV G1 subgenomic replicon pSK-HEV-2-Luc (bottom) (38) shows the disruption of the ORF2 capsid gene (nucleotides 5148 to 5816), with the firefly luciferase gene as described in reference 38. (B) An initial screen of 16 broad-spectrum antivirals (Table 1) for inhibitory activities against the human HEV G1 subgenomic replicon pSK-HEV-2-Luc through quantitation of luminescence is shown. All compounds were examined at a fixed concentration of 10 μM, and the percentages of mock-treated HEV replication (compound vehicle only, 0.5%, vol/vol, DMSO) for each compound are plotted. The black horizontal dotted line represents 100% HEV replication (0% inhibition), while the red dotted line represents 50% HEV replication (50% inhibition). The NA positive control 2CMC is used to demonstrate effective inhibition of HEV replication. Mean values ± SEM are shown.

Three antivirals exhibit dose-dependent inhibition of HEV replication without cytotoxic effects.

As NITD008, GPC-N114, sofosbuvir, and dasabuvir were identified as potential HEV antivirals in the initial screen (Fig. 1B), these compounds were further examined for their effects on cell viability and dose-dependent inhibition of HEV replicon replication.

NITD008 was examined over a concentration range of 0.02 to 2.50 μM, while GPC-N114, sofosbuvir, and dasabuvir were all examined at 0.16 to 25.00 μM (Fig. 2). HEV replication was assessed using relative luminescence production compared to that of mock-treated cells (0.5%, vol/vol, DMSO), while cytotoxicity was quantified using a metabolic activity conversion assay.

FIG 2.

Dose-response curves and cytotoxicity profiles of lead inhibitory compounds against HEV replication. The HEV inhibitory and cytotoxicity effects of four compounds identified in Fig. 1B are shown. Dose-response graphs were generated by quantification of luminescence (red bars, left y axis), and effects on cell viability were examined using a fluorescent resazurin-to-resorufin assay (blue lines, right y axis). The black dotted horizontal lines represent 50% inhibition. The EC₅₀ and CC₅₀ values are shown on the graphs and in Table 2 for compounds NITD008 (0.019 μM to 2.5 μM) (A), GPC-N114 (0.16 μM to 25.0 μM) (B), dasabuvir (0.16 μM to 25.0 μM) (C), and sofosbuvir (0.16 μM to 25.0 μM) (D). The synthesis of HEV replicon RNA was reduced by three broad-spectrum antivirals compared to the mock-treated control (0.5%, vol/vol, DMSO), as quantified by qRT-PCR, using primers to detect the HEV RdRp. Sofosbuvir and GPC-N114 were examined at 10 μM (E), and NITD008 was examined at concentrations of 0.04 to 2.50 μM (F). The NA 2CMC (10 μM) was used as a positive control, and RNA levels were normalized to the housekeeping gene β-actin, while the relative fold expression was calculated using the ΔΔC_T method. P < 0.001 (***); P < 0.01 (**); P < 0.05 (*). Mean values ± SEM are shown.

NITD008 was a potent inhibitor of HEV replication, with a half-maximal effective concentration (EC₅₀) of 0.03 μM (Fig. 2A and Table 2). The half-maximal cytotoxic concentration (CC₅₀) was not reached when examined up to 100 μM, indicating a therapeutic index for NITD008 of >3,333 (Table 2).

TABLE 2.

Potency and cytotoxicity of antiviral compounds with efficacy against the HEV G1 replicon

Test compound	EC50 (μM) (95% CIb )	CC50 (μM) (95% CIb )	Therapeutic index (CC50/EC50)
Dasabuvir	1.79 (1.38–2.32)	12.28 (11.13–13.59)	6.86
GPC-N114	1.07 (0.13–1.35)	>100	>93
NITD008	0.03 (0.02–0.04)	>100	>3,333
Sofosbuvir	1.97 (1.58–2.45)	>100	>51
2CMCa	3.04 (1.52–4.51)	>100	>33

^aPositive control compound.

^bCI, confidence intervals.

GPC-N114 was also an effective inhibitor of HEV in vitro, giving an EC₅₀ of 1.07 μM (Fig. 2B and Table 2). Huh7 cell viability started to drop at the higher compound concentrations (77.2% of mock-treated cells at 25 μM); however, the CC₅₀ was not reached with concentrations up to 100 μM, giving a therapeutic index of >93.

Dasabuvir demonstrated dose-responsive inhibition of HEV replication with an EC₅₀ of 1.79 μM (Fig. 2C and Table 2); however, evaluation of cell viability indicated that the compound was toxic to Huh7 cells at concentrations over 2.5 μM, with a CC₅₀ of 12.28 μM. This resulted in a poor therapeutic index of 6.86; therefore, dasabuvir was dropped from further investigation in this study.

Sofosbuvir efficacy against HEV G1 and G3 replicons in vitro has already been published previously, with an EC₅₀ of 1.20 μM against G3 and >10 μM against G1 (39, 42). These previously reported results for G1 HEV are in contrast to the findings in this study, as sofosbuvir inhibited the HEV G1 replicon in a dose-dependent fashion with an EC₅₀ of 1.97 μM (Fig. 2D and Table 2), demonstrating results similar to those for the G3 HEV replicon published previously (39). No cytotoxic effects on Huh7 cells were observed up to 100 μM, indicating a therapeutic index of >51 (Table 2).

The positive control 2CMC gave an EC₅₀ value of 3.04 μM (Table 2) against the HEV G1 replicon, compared to an EC₅₀ of 1.60 μM previously reported for the G3 replicon (43), with no effects on cell viability up to 100 μM (therapeutic index, >33) (Table 2).

HEV RNA levels are reduced by NITD008, GPC-N114, and sofosbuvir.

NITD008, GPC-N114, and sofosbuvir all exhibited dose-dependent inhibition of HEV replicon-derived luminescence production without cytotoxic effects (Fig. 2A, B, and D). As such, these three antivirals were examined further for their abilities to reduce HEV replicon RNA levels in vitro. The positive control 2CMC was assessed at 10 μM and compared to mock-treated RNA levels (0.5% DMSO) to demonstrate a reduction in HEV replicon RNA (Fig. 2E and F). 2CMC reduced RNA levels to 43.6% of that of the mock-treated samples, from 7.7 × 10⁵ replicon copies per well to 3.4 × 10⁵ copies per well (Fig. 2E and F). Sofosbuvir (10 μM) reduced HEV RNA levels to 38.5% of that of the mock-treated wells, from 7.7 × 10⁵ replicon copies per well to 3.0 × 10⁵ copies per well, while the NNI GPC-N114 (10 μM) also reduced HEV RNA levels to 42.9% of that of the mock-treated wells, down to 3.3 × 10⁵ copies per well (Fig. 2E).

NITD008 was the most potent inhibitor of HEV replication identified in Fig. 1 and 2 and therefore was investigated at a lower concentration range, from 0.04 to 2.50 μM. NITD008 reduced HEV replicon RNA levels in a dose-dependent fashion, from 7.7 × 10⁵ replicon copies per well in mock-treated samples to 5.3 × 10⁵ per well (69.5% of mock) at 0.04 μM, 4.3 × 10⁵ replicon copies per well (56.1% of mock) at 0.16 μM, and 2.8 × 10⁵ copies per well (36.7% of mock) at 2.50 μM (Fig. 2F).

NITD008 and GPC-N114 exhibited synergism in combination against HEV replication.

As GPC-N114 and NITD008 were the most potent HEV replication inhibitors identified in the replicon luciferase assays (Fig. 2A and B) and also effectively reduced HEV replicon RNA levels (Fig. 2E and F), these two antivirals were selected for combinational studies against HEV replication.

NITD008 is a broad-spectrum chain-terminating adenosine NA, initially developed as an antiviral against dengue virus (DENV), for which the reported EC₅₀ range is 0.64 to 1.60 μM (44, 45). In contrast, GPC-N114 is a picornavirus NNI with an EC₅₀ range of 0.13 to 5.44 μM, demonstrating broad-spectrum antiviral activity against multiple genera from within the Picornaviridae (46). GPC-N114 binds in the RNA channel of picornavirus polymerases (46), while NAs such as NITD008 bind in the polymerase active site for viruses such as DENV (45); therefore, they occupy different binding sites in their original target viruses. For these reasons we assessed combinational synergy against HEV replicon replication levels.

The effects of GPC-N114 (0.16 to 5.00 μM) and NITD008 (0.005 to 0.16 μM) were examined at a 32:1 ratio using the G1 HEV replicon. The Chou-Talalay method (47) was used to analyze the inhibition data, and Compusyn software was employed to generate the combinational isobologram (Fig. 3A). The mean of the combination indices at 50%, 75%, and 90% inhibition of HEV replicon replication was 0.4, indicating synergism of NITD008 and GPC-N114 in combination (Fig. 3A). This allowed for dose reduction indices at 50% inhibition of 20.42 for GPC-N114 and 8.82 for NITD008. The cytotoxic effects of the two drugs were also examined in combination, with no effect on Huh7 cell viability up to 100 μM (Fig. 3B).

FIG 3.

GPC-N114 and NITD008 exhibited synergistic inhibition of HEV replication. GPC-N114 and NITD008 were examined in combination against HEV replication in vitro as quantified by relative luminescence. GPC-N114 (0.16 to 5.00 μM) and NITD008 (0.005 to 0.16 μM) were evaluated in a 32:1 concentration ratio. (A) Isobologram of GPC-N114 and NITD008 in combination. Data were analyzed using the Chou-Talalay method (Compusyn software) with an average combination index of 0.4 over 50% (blue dots), 75% (red squares), and 90% inhibition (green triangles), indicating synergism. (B) Cytotoxicity effects of the two drugs in combination over the same concentration range as that described for panel A were assessed using a fluorescent resazurin-to-resorufin assay. Mean values ± SEM are shown.

DISCUSSION

HEV is a major cause of liver disease across the globe, resulting in significant morbidity and mortality (1, 2, 22). Despite this burden, HEV is a largely understudied virus, and as such, we currently lack specific therapies to combat HEV infection.

The development of safe and effective antivirals to combat HEV is needed to treat both chronically infected patients and also for acute infections, to reduce the lengthy recovery period, and to prevent fulminant hepatitis. Additionally, antivirals could be useful as a prophylactic measure in the case of outbreaks and epidemics, which are often reported in refugee camps (48, 49) and military troops (50–52). Furthermore, if safety and efficacy could be achieved without teratogenic side effects, then treating HEV-infected pregnant women to prevent the significant mortality rates for both mother and child may also be possible.

Current HEV treatment options include the reduction of immunosuppressants for HEV-infected organ transplant patients as the first choice of action, followed by courses of PEG-IFN-α and/or ribavirin (53). The reduction of immunosuppressants is effective in clearing the virus in around 30% of solid-organ transplant cases (22), leaving around 70% of patients requiring further action. PEG-IFN-α cannot be used in most organ donor recipients due to the risk of transplant rejection; therefore, ribavirin is prescribed in the majority of cases (9). Ribavirin monotherapy is usually required for at least 3 months and can result in significant and undesirable side effects, including severe anemia (4, 6). Furthermore, ribavirin and PEG-IFN-α are both contraindicated during pregnancy (54).

Clinical resistance to ribavirin resulting in treatment failure has been associated with several HEV G3 polymerase mutations, including Y1320H, K1383N, and G1634R, and an insertion into the HEV hypervariable region (8, 9, 55). The effects of these RdRp mutations have been studied in vitro using mutant replicons, cell culture of HEV isolates, and deep sequencing, and it was found that HEVs bearing Y1320H and G1634R mutations were still sensitive to ribavirin but were associated with enhanced replicative capacities, while the effects of K1383N mutations could not be elucidated in vitro (9). Further studies using deep sequencing of HEV G3 isolates from chronically infected patients revealed that viral heterogeneity was increased following ribavirin treatment and that the RdRp C-terminal mutation G1634R was particularly associated with ribavirin treatment failure (56).

As the current HEV therapies have several reported issues, ranging from drug resistance to severe side effects, further research is warranted to identify safe and efficacious antivirals to treat HEV patients.

G1 HEV infection results in the highest number of fetal and maternal mortalities, while G3 causes the majority of chronic infections worldwide (22). In addition, G1, G4, and G7 human HEVs have also caused chronic infections in immunocompromised and immunosuppressed patients (18–21), so effective, broad-spectrum therapeutics to combat all human HEV genotypes would be beneficial.

This study evaluated 16 antivirals (Table 1), previously developed against other viruses, for inhibition of HEV replicon replication in vitro. Four of the 16 antivirals examined inhibited the HEV G1 replicon with more than 50% inhibition at 10 μM (Fig. 1B). Of these four antivirals, three exhibited potent, dose-dependent, and nontoxic antiviral activity against HEV replicon replication in Huh7 cells (Table 2), with EC₅₀ values of 0.03 μM for NITD008 (Fig. 2A), 1.10 μM for GPC-N114 (Fig. 2B), and 1.97 μM for sofosbuvir (Fig. 2D). Moreover, the same three antivirals effectively reduced HEV replicon RNA levels (Fig. 2E and F), albeit not as efficiently as was observed in the luciferase assays. However, discrepancies in antiviral efficacy examined by RNA quantification compared to that of luciferase assays are widely reported (57–61), and the trend of inhibition of HEV replication was the same across the two methodologies.

Two HCV NAs, 2CMC and sofosbuvir, have already been evaluated for antiviral efficacy against HEV (43). The HCV developmental drug 2CMC is a chain-terminating NA and has been shown to inhibit the G3 HEV replicon in vitro, with an EC₅₀ of 1.6 μM (43). However, development of the oral prodrug for 2CMC (valopicitabine) was halted following reports of undesirable side effects, and further development of this NA as an antiviral is now unlikely.

The HCV NA sofosbuvir has also been previously evaluated against HEV G1 and G3 replicons in vitro, with inconsistent reports of antiviral efficacy (39, 42). One study reported efficacy of sofosbuvir against the HEV G3 replicon with an EC₅₀ of 1.2 μM, but it was unexpectedly ineffective against the G1 replicon at concentrations of up to 10 μM (39). However, another study reported sofosbuvir to be ineffective against both G1 and G3 replicons in vitro, even at concentrations as high as 10 μM (42). In both these studies where sofosbuvir had no effect on G1 HEV, the Sar55/S17/luc replicon, which harbors an insertion from the human ribosomal S17 protein within the hypervariable region, was utilized (37, 62). This S17 insertion dramatically enhances replication levels (37) and, as such, likely renders it less susceptible to sofosbuvir inhibition than wild-type viruses and other replicons which lack this insertion, as previously observed with G1634R-associated ribavirin failure (63, 64).

In contrast to previous reports, in the present study using the pSK-HEV2-Luc replicon, we show that sofosbuvir can inhibit HEV G1 replication in vitro, and further investigation as a potential HEV antiviral may be warranted if known mutations associated with enhanced replication are absent. The discrepancies between the observed effects of sofosbuvir against HEV G1 in this study compared to data from previous studies may also be attributed with the use of a stable replicon cell line versus the transient replicon. However, previous studies using both stable and transient HEV replicons to assess the antiviral effects of IFN-α or ribavirin have revealed very similar results, indicating that these systems are comparable for screening potential antivirals (4, 39). Additionally, sofosbuvir may demonstrate variable antiviral efficacy across the different HEV genotypes, and further preclinical work would be required to ascertain its cross-genotypic activities.

Clinically, sofosbuvir has been used in a few individual cases with or without ribavirin to treat HEV G3-infected patients with variable success (63, 65–68), ranging from a reduction of HEV RNA to undetectable levels (63, 66) to treatment failure or relapse following treatment (65, 67, 68). These mixed reports of efficacy have raised questions around the pursuit of sofosbuvir as an HEV antiviral (69), indicating that further work is required to ascertain its suitability as an anti-HEV therapeutic. Clinical treatment failure of sofosbuvir has been attributed to the HEV phylogenetic subtype, patient immune status (63), and presence of HEV RdRp mutations known to confer increased replication levels that cause ribavirin treatment failure, particularly G1634R (63, 64).

In contrast to sofosbuvir, the antiviral efficacy of GPC-N114 has not been evaluated against HEV replication before. Under preclinical development as a picornavirus antiviral, GPC-N114 has broad-spectrum activity against multiple viruses from the Picornaviridae, with potency ranging from 0.13 μM against human enterovirus 71 to 5.44 μM against mengovirus (46). It has been reported to affect the viral RNA-template duplex binding to the picornavirus RdRp template channel, thereby inhibiting replication (46).

Similarly, NITD008 has also not been previously evaluated as an HEV antiviral. A broad-spectrum NA, NITD008, effectively inhibits the replication of a number of viruses, including the Norwalk (human norovirus) replicon (unpublished data), enterovirus 71 (70, 71), and several flaviviruses, such as DENV (44, 45), HCV (72), tick-borne encephalitis (73), Zika virus (57), and Japanese encephalitis virus (74). Reported potency ranges between 8.7 nM against HCV G2a (72) and 3.31 μM in the tick-borne flavivirus Alkhurma hemorrhagic fever virus (73).

During development of NITD008 against DENV, toxic side effects were observed after 2 weeks of in vivo treatment in rats and dogs (45). This reported NITD008 toxicity might not be clinically significant for a short-term therapeutic course to treat human HEV infection; however, reduced treatment doses may in turn reduce the risk of toxic side effects. As such, NITD008 and GPC-N114 were evaluated together for combinational synergy.

NITD008 and GPC-N114 were synergistic in combinational treatment against HEV G1 replicon-derived luminescence levels (mean combination index of 0.4) (Fig. 3A). This suggests that the two compounds occupy different HEV binding sites, as they do in their original target viruses, and allowed for dose reductions to achieve the same antiviral effect in vitro. Combinational therapy is of great importance to overcome viral resistance to direct-acting antivirals in addition to allowing dose reductions of drugs to reduce side effects. Combinations of at least two therapeutics have been used to successfully treat many HCV and HIV patients (reviewed in references 75–77) and should be considered for HEV antivirals to safeguard against viral evolution that can confer drug resistance.

In this study, we have identified two novel HEV antiviral candidates, NITD008 and GPC-N114, that demonstrated potent antiviral activity and combinational synergy against G1 HEV in vitro. These compounds could provide useful scaffolds for further antiviral development against HEV infection. Additionally, we have shown that in the absence of known polymerase mutations that confer increased replication levels, sofosbuvir demonstrates antiviral efficacy against G1 HEV replicon replication.

Further preclinical evaluation is required for these compounds as potential HEV antivirals before clinical assessment is undertaken. First, it would be prudent to test NITD008, GPC-N114, and sofosbuvir against other human-infecting HEV genotypes (G3, G4, and G7) using replicons and live virus in culture, where available, to ascertain the cross-genotypic activities of these compounds. Second, structure-activity relationship studies for these antivirals may reveal structural moieties that confer antiviral activity, allowing for the development of more potent and less toxic derivatives, particularly in the case of NITD008 as a potential HEV therapeutic. Third, longer-term treatment could be evaluated in cell culture to ascertain toxicity over time and to reveal if any resistance mutations arise. Finally, further assessment of the potential synergy between NITD008 and GPC-N114 against HEV in humanized mice or another animal model, such as swine, may be able to confirm the potential to reduce the required dose for the same nontoxic antiviral effects as those we observed in this study. These compounds represent promising candidates for further HEV antiviral development to combat this pervasive virus.

MATERIALS AND METHODS

Test compounds.

Compounds examined in this study included dasabuvir, sofosbuvir, and favipiravir (MedChemExpress, Monmouth Junction, NJ), GPC-N114 (kind gift from Gerhard Pürstinger [46], formerly of University of Innsbruck, Innsbruck, Austria), JTK-109 (Dalton Pharma Services, Toronto, Canada), lomibuvir (Selleckchem, Houston, TX), nesbuvir and tegobuvir (Haoyuan Chemexpress, Shanghai, China), NITD008 (collaboration with Subhash Vasudevan, Duke-NUS Graduate Medical School, Singapore), quercetagetin and 7-deaza-2’-C-methyladenosine (Santa Cruz Biotechnology, Dallas, TX), filibuvir and setrobuvir (Acme Biosciences, Palo Alto, CA), beclabuvir and TMC-647055 (Taizhou Crene Biotechnology, Zhejiang, China), triazavirin (Mcule, Palo Alto, CA), and 2’-C-methylcytidine (2CMC; Sigma-Aldrich, St. Louis, MO). All compounds were >95% pure and dissolved in 100% dimethyl sulfoxide (DMSO), and they were freshly diluted on the day of the experiment. Details of the 16 compounds examined in this study are shown in Table 1.

HEV replicon.

The HEV pSK-HEV-2-Luc replicon plasmid was a kind gift from Sue Emerson (National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD). The pSK-HEV-2-Luc replicon is based on a human genotype 1 (G1) HEV infectious clone (Sar55; GenBank accession number AF444002.1). The ORF2 capsid gene (nucleotides 5148 to 5816) is disrupted by a luciferase reporter gene (38).

Transcription of HEV replicon.

The HEV replicon plasmid was linearized with BglII and RNA synthesized using mMessage in vitro transcription kits (Ambion, Austin, TX). Unless stated otherwise, 100 μl capping transcription reaction mixtures contained approximately 7.5 μg linearized template, 10 μl of 10× transcription buffer, 50 μl of 2× capping deoxynucleoside triphosphate mix (containing 15 mM ATP, CTP, and UTP, 3 mM GTP, and 12 mM cap analog), 5 μl of 30 mM GTP, and 10 μl of the T7 RNA polymerase mix and were incubated for 2 h at 37°C. Reaction mixtures were then DNase treated for 15 min at 37°C and RNA purified using RNeasy kits (Qiagen, Hilden, Germany). RNA integrity was confirmed using agarose gel electrophoresis and quantified using spectrophotometry before transfection.

Cell culture.

The human hepatoma Huh7 cell line was a kind gift from Mark Douglas (Westmead Institute for Medical Research, Sydney, Australia). Cell culture was carried out as previously described (78). Briefly, Huh7 cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Life Technologies, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich), 1× GlutaMAX (Life Technologies), 10 mM HEPES buffer (ThermoFisher Scientific, Waltham, MA), and 100 U/ml penicillin-streptomycin (Life Technologies).

Transfection.

Huh7 cells were grown to 80% confluence in 96-well plates, and medium was changed 1 h prior to transfection. HEV RNA transcripts (75 ng/well) were chemically transfected into Huh7 cells using the TransIT-mRNA transfection kit (Mirus Bio LLC, Madison, WI) per the manufacturer’s instructions.

Cytotoxicity assays.

Compound cytotoxicity was assessed using monolayers of Huh7 cells, seeded into 96-well plates at a density of 5,000 cells/well. The next day, fresh medium (without antibiotics) was added 1 h prior to adding compounds. All compounds were freshly diluted in complete DMEM, added to cells, and then incubated for a total of 72 h. Cytotoxicity was quantified using a metabolic conversion assay (CellTiter-Blue; Promega, Madison, WI) per the manufacturer’s instructions, and fluorescence was measured on a FLUOstar Optima microplate reader (BMG Labtech, Ortenberg, Germany). CC₅₀ values were determined by nonlinear regression in GraphPad Prism, version 7.02. Mean values were calculated from quadruplicate data sets from at least two independent experiments.

Luciferase assays.

Antiviral activities of test compounds were examined by adding a fixed concentration (10 μM) or increasing concentrations of each drug to replicon-transfected cells: 0.02 to 2.50 μM for NITD008 and 0.16 to 25.00 μM for GPC-N114, dasabuvir, and sofosbuvir. The NA 2CMC was used as a positive control at 10 μM to demonstrate inhibition of replicon replication (43). Huh7 cells were seeded into 96-well plates at a density of 5,000 cells/well. The next day, fresh antibiotic-free DMEM was added to the cells 1 h prior to transfection. Compounds were freshly diluted in complete DMEM, added to the cells 4 h posttransfection, and incubated for 72 h. Mock-treated cells were incubated with 0.5% (vol/vol) DMSO, the compound vehicle. Replication of the pSK-HEV-2-Luc replicon was determined by luciferase-derived luminescence using Luciferase assay system kits (Promega) per the manufacturer’s instructions. Luminescence was measured on a FLUOstar Optima microplate reader (BMG Labtech). HEV replicon luminescence levels in treated cells were compared to those from mock-treated cells (0.5%, vol/vol, DMSO) to calculate the percentage of HEV replication. All compounds were also tested in the absence of the HEV replicon to ensure that they did not interfere with the luciferase signal. EC₅₀ values were determined by nonlinear regression in GraphPad Prism, version 7.02. Mean values were calculated from quadruplicate data sets from at least two independent experiments.

Combinational compound treatment.

Synergism was calculated using Compsyn software V1.0, which employs the Chou-Talalay method (47). GPC-N114 (concentration range, 0.16 to 5.00 μM) and NITD008 (concentration range, 0.005 to 0.16 μM) were examined alone and together at a 32:1 ratio against the HEV replicon, with luminescence quantified as described above. Data were generated from quadruplicate data sets.

RNA extraction.

Viral and cellular RNA was extracted from transfected Huh7 cell monolayers 72 h posttransfection with TRIzol LS (Invitrogen, Carlsbad, CA, USA) using phase separation per the manufacturer’s instructions. RNA was further purified using the RNeasy Minikit (Qiagen), which included DNA removal using RNase-free DNase (Qiagen). RNA was quantified using spectrophotometry, and RNA integrity was assessed by agarose gel electrophoresis.

HEV RNA level quantitation.

HEV replicon RNA levels were measured from transfected Huh7 cells by quantitative reverse transcriptase PCR (qRT-PCR). Briefly, cDNA was synthesized using a SuperScript VILO cDNA synthesis kit (Life Technologies). Replicon RNA was measured using an iTaq universal SYBR green supermix (Bio-Rad, Hercules, CA) as described previously (79). Samples were normalized to the housekeeping gene β-actin, and the fold change was analyzed using the ΔΔC_T method, as described previously (80). HEV-specific primers were used for quantitation targeting the region encoding RdRp and included the forward primer (HEV replicon Fwd, 5′-TGTCCTGATTGCTGGCTGTG-3′) and the reverse primer (HEV replicon Rev, 5′-GAGAAGAATTGGGGCCCTGG-3′). Mean values were calculated from triplicate data sets, and two independent experiments were performed. All statistical calculations were performed using GraphPad Prism software (v7.02). Data were analyzed using an unpaired t test: P > 0.05, P ≤ 0.05 (*), P ≤ 0.01 (**), and P ≤ 0.001 (***).