Resistance Studies of a Dithiazol Analogue, DBPR110, as a Potential Hepatitis C Virus NS5A Inhibitor in Replicon Systems

Abstract

Hepatitis C virus (HCV), a member of the Flaviviridae family, affects approximately 3% of the world's population and is becoming the leading cause of liver disease in the world. Therefore, the development of novel or more effective treatment strategies to treat chronic HCV infection is urgently needed. In our previous study, we identified a potential HCV NS5A inhibitor, BP008. After further systemic optimization, we discovered a more potent HCV inhibitor, DBPR110. DBPR110 reduced the reporter expression of the HCV1b replicon with a 50% effective concentration (EC₅₀) and a selective index value of 3.9 ± 0.9 pM and >12,800,000, respectively. DBPR110 reduced HCV2a replicon activity with an EC₅₀ and a selective index value of 228.8 ± 98.4 pM and >173,130, respectively. Sequencing analyses of several individual clones derived from the DBPR110-resistant RNAs purified from cells harboring genotype 1b and 2a HCV replicons revealed that amino acid substitutions mainly within the N-terminal region (domain I) of NS5A were associated with decreased inhibitor susceptibility. P58L/T and Y93H/N in genotype 1b and T24A, P58L, and Y93H in the genotype 2a replicon were the key substitutions for resistance selection. In the 1b replicon, V153M, M202L, and M265V play a compensatory role in replication and drug resistance. Moreover, DBPR110 displayed synergistic effects with alpha interferon (IFN-α), an NS3 protease inhibitor, and an NS5B polymerase inhibitor. In summary, our results present an effective small-molecule inhibitor, DBPR110, that potentially targets HCV NS5A. DBPR110 could be part of a more effective therapeutic strategy for HCV in the future.

INTRODUCTION

Hepatitis C virus (HCV) is a small enveloped RNA virus that affects nearly 170 million individuals worldwide, making it a leading cause of hepatitis C and liver disease (1). HCV infection is responsible for the development of severe chronic liver disease and cirrhosis and associated complications, including liver failure, portal hypertension, and hepatocellular carcinoma (2). The main goals of chronic HCV therapy are to eradicate the virus and prevent these potentially life-threatening complications. The mainstays of chronic HCV therapy are PEGylated alpha interferon (IFN-α) and ribavirin, but these compounds are poorly tolerated and may eventually lead to a suboptimal response rate and a high incidence of adverse effects, including flu-like symptoms, depression, and anemia (3, 4). The chances of sustained viral clearance are only 40 to 50% for genotype 1 infection, which is the predominant genotype in worldwide populations. Therefore, the development of specific antiviral therapies for hepatitis C with improved efficacy and better tolerance is a major public health objective that is urgently important.

HCV is a positive-strand RNA virus that has been classified as the sole member of the Hepacivirus genus within the Flaviviridae family. The HCV genome consists of a single strand of RNA that is approximately 9.6 kb long, with a large open reading frame encoding a polyprotein of approximately 3,010 amino acids. The viral polyprotein is cleaved cotranslationally and posttranslationally by both cellular and viral proteases to yield more than 10 different viral proteins. Among these viral proteins are the structural proteins C, E1, E2, and p7, which serve as the components of the mature virus particle and are required for viral assembly, and the nonstructural proteins NS2, NS3, NS4A, NS4B, NS5A, and NS5B, which are involved in membrane-associated RNA replication, viral assembly, and release (5–8). HCV NS3 is a bifunctional protein with an amino-terminal domain that has serine protease activity and a carboxy-terminal domain that shows helicase/NTPase activity (9–11). The small hydrophobic protein NS4A serves as a cofactor for the NS3 protease and helicase activities. The association of NS4A with the NS3 protease domain is essential for enzymatic function, stability, and anchoring to the cellular membranes (12, 13). NS4B is an integral membrane protein that plays a direct role in the remodeling of host cell membranes for the formation of the membranous web, which is presumably responsible for HCV replication complex assembly (14, 15). NS5A is a large hydrophobic phosphoprotein that plays an important role in HCV RNA replication (16) and is essential for virion morphogenesis (17). Structurally, NS5A is composed of three domains and an amphipathic α-helix that promotes membrane association (16–20). The amino terminus of NS5A contains a zinc and RNA binding motif (20, 21). Mutations that alter either the zinc binding or membrane association of NS5A result in the complete inhibition of RNA replication (22–24). In HCV replicon cells, the inhibition of NS5A-targeting molecules promotes a relocalization of the NS5A protein from the endoplasmic reticulum to lipid droplets and suppresses the formation of functional replication complex formation (25). Clinically, when an NS5A inhibitor is combined with polyethylene glycol (PEG)-IFN and ribavirin, the inhibition of NS5A has been associated with a significant decrease in HCV RNA and an enhanced, sustained virologic response (26, 27). NS5B is an RNA-dependent RNA polymerase (28, 29). Previous studies have indicated that the NS3, NS4A, NS4B, NS5A, and NS5B proteins form the HCV replicase complex and that all members play an essential role in HCV replication (30–32).

To date, there is still no vaccine to prevent or to cure HCV infection. Thus, the development of a more effective treatment for HCV infection will be crucial for drug discovery efforts. In the past, viral enzymes have been the most advanced targets for drug development. NS3/4A protease inhibitors and NS5B polymerase inhibitors have garnered the most attention as drug targets, with several candidates recently showing great promise in clinical trials (33–35). In 2011, drugs targeting the HCV NS3/4A protease, telaprevir and boceprevir, were approved, representing a major step toward improving therapeutic response and decreasing treatment time (36, 37). However, the promising development of inhibitors of HCV NS5A showed that nonenzyme HCV viral proteins can also be good direct antiviral agent (DAA) targets (26). Recently, using a cell-based replicon screen, a new class of anti-HCV compound was identified that appears to target NS5A (38). Antiviral drugs have often been shown to cause the emergence of drug-resistant mutations that lead to a low sustained virological response rate to drug treatment in patients. Therefore, exploring new mechanisms for anti-HCV drugs and the design of combination regimens is highly desirable.

In our previous study, we identified an effective HCV NS5A inhibitor, BP008, which is a thiazole analogue (39). We subsequently developed a dithiazole analogue, DBPR110. Here, we describe the profile of DBPR110, which exhibits a picomolar 50% effective concentration (EC₅₀) toward genotype 1b and 2a HCV replicons. Our data reveal that this small-molecule inhibitor may target the domain I region of the genotype 1b and 2a HCV NS5A proteins and confirmed that the function of NS5A in replication is a good target for regulation by small molecules. Moreover, the combination study using this inhibitor supports the continued exploration of DBPR110 as a component of a new combination therapeutic strategy for the treatment of chronic HCV infection.

MATERIALS AND METHODS

Escherichia coli and yeast strains.

Frozen, competent E. coli strain C41, a derivative of BL21(DE3) (40), was purchased from OverExpress, Inc. Standard yeast medium and transformation methods were used (41). Saccharomyces cerevisiae YPH857 was purchased from ATCC. The genotype of YPH857 is MATα ade2-101 lys2-801 ura3-52 trp1-Δ63 HIS5 CAN1 his3-Δ200 leu2-Δ1 cyh2. Competent yeast cells were prepared using the lithium acetate procedure (41).

Cell culture and HCV inhibitors.

Huh-7.5 cells and their derivative HCV replicon cell lines were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco/BRL) that was supplemented with 100 U/ml penicillin-streptomycin (Gibco/BRL), 0.1 mM nonessential amino acid (NEAA; Gibco/BRL), and 10% heat-inactivated fetal bovine serum (FBS) at 37°C in 5% CO₂. The HCV replicon cell lines were isolated from colonies as described by Lohman et al. (32). The culture medium for the replicon cells was additionally supplemented with 0.25 to 0.5 mg/ml of G418, unless specified otherwise. Compound DBPR110 was synthesized at the Institute of Biotechnology and Pharmaceutical Research at the National Health Research Institutes in Taiwan. The synthesis details of DBPR110 will be reported elsewhere. The structure of the NS5A inhibitor, DBPR110, is shown in Fig. 1. 2′-C-Methyl-adenosine (2′CMA) (42) and VX-950 (43) were purchased from Carbosynth (Berkshire, United Kingdom) and Acme Biosciences (Belmont, CA), respectively. The compounds were stored at −20°C as 10 to 500 mM dimethyl sulfoxide (DMSO) stock solutions until the assay. IFN-α was purchased from Calbiochem (La Jolla, CA) and stored at −80°C.

Fig 1.

Chemical structure of HCV NS5A inhibitors. (A) Structure of thiazole analogue BP008. (B) Structure of dithiazole analogue DBPR110. (C) Structure of diimidazol analogue BMS-790052.

Inhibitory assay for HCV replicons.

Cells were seeded at 1 × 10⁴ (high-throughput screening assay) or 1 × 10⁵ (regular assay) cells/well in 96- or 12-well plates, respectively, and incubated for 4 h. The medium was then aspirated and replaced with 0.1 (96-well plates) or 1 (12-well plates) ml of complete medium containing a single compound or combinations of compounds in serial concentration(s). The plates with compounds were incubated for 72 h and then assayed for luciferase expression (Promega). The EC₅₀ of each compound was determined independently and used to determine the range of concentrations used for the combination experiments. All data are presented as the means ± standard deviations (SD) from three independent experiments. The selectivity index (SI) was calculated as the ratio of the 50% cytotoxic concentration (CC₅₀) to the EC₅₀.

Cytotoxicity assay.

The sensitivity of the cell lines to inhibitors was examined using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, Huh-7.5 cells were plated at a density of 1 × 10⁵ cells per well in 12-well plates containing 1 ml of culture medium for 4 h. Serial dilutions of the compounds or DMSO (negative control) were added, and the plates were incubated for an additional 72 h. The MTT reagent was then added to each well, and the plates were incubated for 3 h at 37°C in a humidified 5% CO₂ atmosphere before being read at a wavelength of 563 nm using an enzyme-linked immunosorbent assay (ELISA) plate reader. All data are presented as the means ± SD from four independent experiments.

Small-molecule inhibition of HCV infectivity.

To quantify the inhibitory effect of DBPR110 on HCV particle formation, HCV replication in DBPR110-treated and untreated cells was quantified using a luciferase activity assay, as described previously (44, 45). In vitro transcribed RNA derived from a full-length HCV2a JFH1 infectious cDNA clone with the luciferase reporter gene was delivered to Huh-7.5 cells by electroporation. The cells were seeded at 1 × 10⁵ cells per well in 12-well plates and incubated for 4 h. The medium was then aspirated and replaced with 1 ml of complete medium containing DBPR110 in serial concentrations. The plates with compounds were incubated for 72 h, and the medium was then used to infect Huh-7.5 cells. Huh-7.5 cells were seeded in 12-well plates (1 × 10⁵ cells/well) in DMEM with 10% FBS for 24 h before infection. Then 1 ml of HCV cell culture (HCVcc)-containing supernatant per well was added to the Huh-7.5 cells. After 72 h of incubation at 37°C, the total cell lysate was assayed for luciferase expression (Promega).

Isolation of resistant replicons.

Selection of resistant replicon cells was performed by growing HCV genotype 1b Con1 and 2a JFH1 replicon cells in medium containing 0.2 or 200 nM and 60 nM or 1 μM DBPR110, respectively. Medium containing the compound was added to monolayers of HCV1b-neo replicon cells at ∼25% confluence in the presence of 0.2 to 0.4 mg/ml of G418. Replicon cells maintained in the presence of dimethyl sulfoxide (DMSO) were used as a control. After 40 days, total RNA was isolated from both the control replicon cells and homogeneous cell lines containing compound using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. The RNA was amplified by reverse transcription-PCR (RT-PCR). The PCR products of NS3-NS5B were gel purified and subcloned into the pRS-Luc-HCV1bRep vector to replace the parental NS3-NS5B by homologous recombination in yeast. Thirty-six colonies of plasmids were purified from the yeast cells and reamplified in E. coli strain C41 for DNA sequencing.

Construction of molecular clones containing resistance mutations.

To create point mutations derived from the resistant clones, the amino acid substitutions P58S, P58T, P58L, Y93H, Y93N, Y93C, V153M, M202L, and M265V were introduced into the phRlu-HCV1b plasmid; T24A, P58L, Y93N, and Y93H were introduced into the HCV2a plasmid either individually or in combination (primers are available upon request). The PCR products were gel purified and joined by overlapping PCR to form the fragments containing the following mutations for homologous recombination with linearized phRlu-HCV1b plasmids (digested with HpaI): V153M, M202L, M265V, and Y93N; V153M, M202L, and M265V; and Y93H, V153M, M202L, and M265V. The mutant replicon plasmids were purified from yeast cells and then reamplified and maintained in E. coli strain C41 strain. All constructs were sequenced to confirm the presence of the desired mutation and to ensure that there were no additional changes.

RNA transcription and transient replicon assay.

The RNA transcripts were synthesized in vitro using ScaI-digested DNAs and a T7 MegaScript transcription kit (Ambion) according to the manufacturer's directions. A transient replicon assay was performed to quantify the compound-mediated inhibition of viral translation (46). RNA transcripts were transfected into Huh-7.5 cells by electroporation, as described previously (47). Various concentrations of DBPR110 or the control medium were added to each well, and the cells were assayed to determine the luciferase activity at 4 h and 72 h posttransfection. The cells were lysed for luminometry, and a luciferase assay was performed by mixing 5 μl of lysate with 25 μl of Renilla Luciferase Assay Reagent (Promega). For quantification of the compound-mediated inhibition, the relative luciferase activity derived from the mock-treated cells was set to 100% (48).

Serum shift assay.

In the serum shift assay, the inhibitory activity of DBPR110 was determined using replicon 1b in the presence of 10, 20, 30, 40, and 50% fetal bovine serum or 10 and 40% of extracellular normal human serum. In the absence or presence of a serial dilution of DBPR110, the percentage of inhibition was determined by a 50% or 90% reduction in Renilla luciferase (Rluc) activity (EC₅₀ or EC₉₀, respectively) compared to the control after 72 h of incubation.

Inhibitor combination study.

Luciferase reporter-linked HCV replication assays were used to evaluate the potential use of DBPR110 in combination with IFN-α, an NS3 protease inhibitor (VX-950), and a nucleoside inhibitor of NS5B (2′CMA). For the combination index model, the cells were incubated for 72 h with serial dilutions of either IFN-α, VX-950, or 2′CMA and DBPR110 and a fixed ratio of inhibitors below their cytotoxic concentrations. CalcuSyn (Biosoft) was used to analyze the data obtained from the 72-h luciferase-based HCV replicon assay and quantify the differences between the observed effects and predicted ones. Compound interactions and concentration ratios were quantified using the approach described by Chou and Talalay (49). The degrees of synergistic and additive effects were evaluated using the median-effect principle with the combination index (CI) calculation (49). The combination indices (CIs) at the EC₅₀, 70% effective concentration (EC₇₀), and EC₉₀ were also determined (50). In total, six combinations were evaluated, with three to eight experiment replicates per condition. By convention, a CI of 0.9 was considered synergistic, a CI of >0.9 or <1.1 was considered additive, and a CI of >1.1 was deemed antagonistic.

Energy calculation.

The docking module implemented in the program Insight II from Accelrys, Inc. (San Diego, CA), was used to calculate the binding energy between the DBPR110 and the HCV NS5A variants. The hydrogen atoms were first added to the compounds and protein. The potentials for the DBPR110 and HCV NS5A variants were subsequently assigned by using the consistent force field (CFF). The parameters for the assignment of potentials using the CFF were set at the default values. The interaction energy, a combination of the van der Waals energy and electrostatic energy, between the DBPR110 and HCV NS5A variants was finally calculated using the docking module in the Insight II program.

Computational modeling.

We employed the Discovery Studio, version 2.1, program from Accelrys, Inc. (San Diego, CA), to build the computational models of the HCV NS5A protein. The three-dimensional structure of the parental HCV NS5A was used as a template to perform energy minimization. The force fields of the conformations were further verified using Chemistry at HARvard Macromolecular Mechanics (CHARMM), and the parameters used were set at the default values.

Statistical analysis.

The reported values are the averages of three independent measurements expressed as means ± standard deviations. The statistical significance of the difference between the means of the experimental groups was tested by the Student t test for unpaired data. A difference was considered statistically significant at a P value of <0.05 (Sigma Plot, version 10, software; Systat Software, San Jose, CA).

RESULTS

Identification of HCV inhibitors.

In our previous study, we discovered a specific and effective thiazole analogue inhibitor of HCV replication, BP008 (Fig. 1) that potentially targets HCV NS5A (39). BP008 formed the basis of an extensive series of chemical refinements that focused on improving antiviral potency. After defining symmetry as an important contributor to antiviral activity (38) and further systematic optimization, we identified a novel dithiazole analogue inhibitor of HCV replication, DBPR110 (Fig. 1), that demonstrated balanced EC₅₀s in the picomolar range for the HCV1b and -2a replicon cell lines. DBPR110 displayed improved potency against the genotype 1b and 2a replicons, as well as against the 2a infectious virus, all with calculated CC₅₀ values of over 50 μM and EC₅₀s of 3.9, 228.8, and 18.3 pM, respectively, as assessed by luciferase reporter activity (Table 1). DBPR110 displays an in vitro selective index (CC₅₀/EC₅₀) of over 12,800,000 for the HCV genotype 1b replicon, 173,130 for the genotype 2a replicon, and 720,461 for the 2a infectious virus. Moreover, the susceptibility of genotype 1b to DBPR110 was 74-fold greater than that of genotype 2a replicon cells. The other diimidazole analogue HCV inhibitor, BMS-790052 (Fig. 1), was shown to have comparable potency against HCV1b (EC₅₀ of 9 pM) and -2a replicon (EC₅₀ of 71 pM) activities (26). Analysis of the potency of DBPR110 by real-time PCR revealed similar effects (data not shown).

Table 1.

Potency of DBPR110 on HCV replicon cell line and virus particle formation

Virus type and cell line	DBPR110 potency
	EC50 (pM)a	EC90 (pM)a	CC50 (μM)	Selective index (CC50/EC50)
Genotype 1b, Con1	3.9 ± 0.9	8.2 ± 1.8	>50	>12,800,000
Genotype 2a, JFH1	228.8 ± 98.4	464.7 ± 96.6	>50	>173,130
Infectious HCV, genotype 2a, JFH1	18.3 ± 2.6	257.5 ± 50.2	>50	>720,461

^aEC values were determined based on luciferase reporter activity and are reported as means ± standard deviations relative to the parental cell line (n ≥ 3).

DBPR110 reduces the level of viral RNAs but not the translational efficiency of viral RNAs in cells transfected with HCV1b replicon RNAs.

To distinguish the inhibition of viral translation from the inhibition of RNA synthesis, we monitored the reduction rate of the reporter gene expression levels as an indicator of the inhibition of DBPR110. The HCV1b reporter replicon construct, pRS-Luc-HCV1bRep, was transcribed in vitro and transfected into Huh-7.5 cells. The luciferase activity was monitored several times over a period of 72 h posttransfection (Fig. 2), and a level of luciferase activity was sustained until 72 h posttransfection in the absence of DBPR110 (data not shown). The luciferase activity peaked within the first 8 h posttransfection and also after 72 h posttransfection, representing viral translation and RNA replication, respectively (51). We measured the luciferase activity at 4, 8, 24, 48, and 72 h posttransfection; DBPR110 had a minimal effect on the Rluc signals at 4 and 8 h posttransfection, but the signals were significantly reduced at 24, 48, and 72 h posttransfection, respectively (P < 0.001) (Fig. 2). In summary, our data demonstrated a significant suppression of viral RNA synthesis by DBPR110.

Fig 2.

Inhibition of HCV1b RNA replication by DBPR110. DBPR110 inhibits the HCV replication stages (24 to 72 h) rather than the viral translation stages (4 to 8 h). The HCV1b replicon was electroporated into Huh-7.5 cells, which were then maintained in the absence (DMSO) or presence of 10 or 100 pM DBPR110. Renilla luciferase (Rluc) activity was monitored at the indicated time points posttransfection. The numbers above the DBPR110-treated time points represent the percentages of luciferase signals relative to the DMSO-treated controls (100%), and the data are presented as the means ± SD of three independent experiments. ***, P < 0.001, compared with the DMSO group. Statistical significance was calculated by using an unpaired Student t test as described in Materials and Methods.

Isolation and characterization of genotype 1b replicons resistant to DBPR110.

Resistance to antiviral therapy is a major issue in managing HCV infection through DAA strategies (52, 53). To characterize the resistance profile of DBPR110, the cell clones that were resistant to DBPR110 were obtained by culturing HCV genotype 1b replicon cells in the presence of G418 and increasing concentrations of DBPR110 ranging from 50- to 50,000-fold the EC₅₀. The selection experiments revealed that the replication of the cognate replicons was resistant to inhibition by DBPR110 and that they displayed a loss of potency compared to the parental cell lines. Compared to the parental cells, the selected cells (showing DBPR110 resistance [DBPR110R]) were determined to be more than 14,000-fold more resistant, increasing from an EC₅₀ of 0.0039 nM for the parental cells to an EC₅₀ of more than 55 nM for the DBPR110R cells. The direct DNA sequencing of individual clones containing NS3-NS5B from genotype 1b resistant cells revealed multiple changes in the N terminus of NS5A (Table 2). P58L/T (20%), Y93N/H (73%), V153M (53%), M202L (47%), and M265V (40%) were the predominant mutations observed in 0.2 nM DBPR110-resistant clone selections (Table 2). In total, 100% of the cDNA clones isolated from the cells treated with 200 nM DBPR110 contained the mutations Y93N, V153M, M202L, and M265V (Table 2). None of these amino acid substitutions were observed in the NS5A cDNA clones isolated from the DMSO-treated control cells (data not shown). Substitutions at P58 and Y93 of NS5A are common mutations in HCV drug resistance studies (26, 38, 39, 54), signifying that these residues play an important role in the drug-resistant functions of HCV. We also checked if any frequent mutations occurred in the 5′ or 3′ untranslated region (UTR) and the other nonstructural regions of DBPR110-resistant HCV replicon cells. No frequent mutations were found outside the NS5A region.

Table 2.

Amino acid changes in genotype 1b HCV NS5A derived from cells resistant to 0.2 or 200 nM DBPR110

DBPR110 concn	Amino acid in NS5Aa	Residue in the indicated plasmida
		p1	p2	p3	p4	p5	p6	p7	p8	p9	p10	p11	p14	p15	p16	p17	p18	p19	p20	p21	p22
0.2 nM	P58	L															L			T
	Y93		N				N	H		N	H			N	H	H	N		H		H
	V153		M				M			M			M	M			M	M		M
	M202		L				L			L			L	L			L	L
	M265		V				V						V	V		I	V	V
200 nM	Y93	N	N	N	N	N	N	N	N	N	N	N
	V153	M	M	M	M	M	M	M	M	M	M	M
	M202	L	L	L	L	L	L	L	L	L	L	L
	M265	V	V	V	V	V	V	V	V	V	V	V

^ap, plasmid derived from DBPR110-resistant individual clone.

Validation of the genotype 1b mutations responsible for the resistance phenotype.

To determine the contributions of specific mutations to inhibitor sensitivity, the resistance phenotypes were further validated by engineering mutations into an HCV genotype 1b replicon that contained a luciferase reporter gene, which can be used to monitor replication in a transient reporter assay. The replication of the parental and mutant clone replicons was monitored over time in the presence or absence of DBPR110. Maximum replication efficiency for both the parental and mutant RNAs was determined to be 72 h posttransfection (data not shown). As shown in Table 3, the replication efficiencies of the P58S, P58T, P58L, Y93N, Y93H, and Y93C replicons were 42% ± 10%, 40% ± 15%, 19% ± 8%, 8% ± 3%, 8% ± 4%, and 9% ± 6% of the level of the parental replicon at 72 h, respectively. This result indicates that these resistant mutants had reduced fitness, with the amino acid substitutions Y93N/H/C showing the lowest replication capacity (Table 3). In a previous study by our laboratory and one by Fridell et al., substitutions at residue 93 also had a great impact on replication fitness (54). The replication efficiencies of V153M, M202L, and M265V were 70% ± 17%, 106% ± 37%, and 87% ± 23% of the level of the parental replicon, respectively, indicating that the V153M, M202L, and M265V mutants did not affect fitness (Table 3). Our data revealed that most of the DBPR110-resistant clones contained a combination of two or four amino acid substitutions at residues 58, 93, 153, 202, or 265 (Table 2). The complexity of the resistance pattern was verified by the analysis of individual cDNA clones. All of the 200 nM DBPR110-resistant clones contained the combination Y93N V153M M202L M265V (Table 2). Furthermore, to determine the phenotypes of the variants with linked mutations, replicons with the following representative combinations were tested in transient replication assays: V153M M202L M265V, Y93N V153M M202L M265V, and Y93H V153M M202L M265V. The Y93N V153M M202L M265V and Y93H V153M M202L M265V variants exhibited impaired replication capacities of 16 to 32% relative to the parental clone (Table 3). The individual amino acid substitutions P58S/T/L and Y93N/H/C exhibited different levels of resistance to DBPR110 with increasing EC₅₀s ranging from 25- to 2,547-fold above the parental control (Table 3). When Y93N was combined with V153M, M202L, and M265V on the same replicon, the effects on the inhibitor increased dramatically to give a 2,547-fold boost in resistance. On the other hand, V153M, M202L, and M265V identified in a single NS5A cDNA clone did not affect DBPR110 potency as a single mutation (Table 3), but the combination of Y93N, V153M, M202L, and M265V or the combination of Y93H, V153M, M202L, and M265V produced 18,217- or 5,824-fold resistance, respectively. This suggests that the primary conformation of NS5A, or of NS5A in the replication complex, is the predominant determinant for inhibitor sensitivity, while residues 58, 93, 153, 202, and 265 are the determinants for resistance selection in genotype 1b of HCV.

Table 3.

Effects of genotype 1b HCV NS5A amino acid substitutions on DBPR110 potency

Amino acid substitution(s)	Replication efficiency (%)a	EC50 (pM)a	Resistance at EC50 (n-fold)	EC90 (pM)a	Resistance at EC90 (n-fold)
None (parental)	100	1.5 ± 0.6	1	4.2 ± 2.1	1
P58S	42 ± 10	38 ± 14	25	64 ± 11	15
P58T	40 ± 15	243 ± 40	162	1303 ± 219	310
P58L	19 ± 8	564 ± 194	376	2731 ± 909	650
Y93N	8 ± 3	3,821 ± 1,677	2,547	13,305 ± 3,416	3,168
Y93H	8 ± 4	1,408 ± 293	939	7,337 ± 2,206	1,747
Y93C	9 ± 6	78 ± 40	52	177 ± 62	42
V153 M	70 ± 17	1.3 ± 0.5	1	4.1 ± 1.9	1
M202L	106 ± 37	2.1 ± 0.6	1	5.0 ± 1.4	1
M265V	87 ± 23	2.0 ± 0.9	1	5.1 ± 1.7	1
V153M M202L M265V	157 ± 52	1.1 ± 0.5	1	3.1 ± 1.1	1
Y93N V153M M202L M265V	16 ± 4	27,326 ± 12,349	18,217	98,912 ± 30,548	23,550
Y93H V153M M202L M265V	32 ± 10	8,736 ± 2,370	5,824	37,710 ± 6,970	8,979

^aMeans ± standard deviations determined from transient transfection assays (n ≥ 3).

Isolation and characterization of genotype 2a replicons resistant to DBPR110.

The cell clones that were resistant to DBPR110 were obtained by culturing HCV genotype 2a replicon cells in the presence of G418 and increasing concentrations of DBPR110 ranging from 60 to 1,000 nM. The selection experiments revealed that the replication of the cognate replicons was resistant to inhibition by DBPR110 and that they displayed a loss of potency compared to the parental cell lines. Direct DNA sequencing of the individual clones containing NS3-NS5B from genotype 2a resistant cells revealed multiple changes in the N terminus of NS5A (Table 4). T24A (50%) and P58L (50%) were the predominant mutations observed in the 60 nM DBPR110-resistant clone selections (Table 4). In total, 100% of the cDNA clones isolated from the cells treated with 1 μM DBPR110 contained only the mutation Y93H (Table 4). None of these amino acid substitutions were detected in the NS5A cDNA clones isolated from the DMSO-treated control cells (data not shown).

Table 4.

Amino acid changes in genotype 2a HCV NS5A derived from cells resistant to 60 nM or 1 μM DBPR110

DBPR110 concn	Amino acid in NS5Aa	Residue in the indicated plasmida
		p1	p2	p3	p4	p9	p6	p5	p8	p9
60 nM	T24	A	A					A	A
	P58			L	L	L	L
1 μM	Y93	H	H	H	H	H	H	H	H	H

^ap, plasmid derived from DBPR110-resistant individual clone.

Validation of genotype 2a mutations responsible for the resistance phenotype.

When tested in replicon transient assays, the T24A, P58L, and Y93N/H mutations reduced susceptibility to DBPR110. As shown in Table 5, the replication efficiencies of the T24A, P58L, Y93N, and Y93H replicons were 120% ± 12%, 154% ± 20%, 103% ± 28%, and 192% ± 13% of the parental replicon at 72 h, respectively. This result indicates that these resistant mutants did not have impaired fitness (Table 5). The individual amino acid substitutions T24A, P58L, Y93N, and Y93H exhibited different levels of resistance to DBPR110 with increasing EC₅₀s, ranging from 65- to 3,041-fold above the parental control (Table 5). The Y93H substitution had the greatest impact on susceptibility to DBPR110. This suggests that the primary conformation of NS5A is the predominant determinant for inhibitor sensitivity in genotype 2a, while residues 24, 58, and 93 are the determinants for resistance selection in genotype 2a of HCV.

Table 5.

Effects of genotype 2a HCV NS5A amino acid substitutions on DBPR110 potency

Amino acid substitution	EC50 (pM)a	Resistance at EC50 (n-fold)	EC90 (pM)a	Resistance at EC90 (n-fold)	Replication efficiency (%)a
None (parental)	250 ± 32	1	592 ± 70	1	100
T24A	16,245 ± 4,547	65	63,488 ± 8,467	107	120 ± 12
P58L	52,953 ± 8,045	212	89,348 ± 27,926	151	154 ± 20
Y93N	51,766 ± 6,307	207	85,243 ± 15,920	144	103 ± 28
Y93H	760,167 ± 175	3,041	>5,000,000	>8,446	192 ± 13

^aMeans ± standard deviations determined from transient transfection assays (n ≥ 3).

Protein binding activity of DBPR110.

To evaluate the effect of serum protein binding on DBPR110 activity, fetal bovine serum (FBS) and normal human serum (NHS) were added to determine the binding of DBPR110. Our results revealed that the EC₅₀s in the presence of 10, 20, 30, 40, and 50% FBS were 4.3 ± 0.8, 8.1 ± 1.6, 7.9 ± 0.9, 13.2 ± 1.7, and 21.5 ± 10 pM, respectively, and the EC₉₀s were 9.3 ± 3.4, 23.8 ± 11, 21.6 ± 17, 35.1 ± 7.4, and 41.9 ± 7.2 pM, respectively. The EC₅₀s in the presence of 10 and 40% NHS were 33.5 ± 0.4 and 210.9 ± 6.3 pM, respectively, and the EC₉₀s were 41.6 ± 1.3 and 588.1 ± 45.9 pM, respectively (Table 6). While the activity of DBPR110 at the higher serum concentrations was more favorable than that at lower levels, we observed that the EC₅₀ and EC₉₀ were increased 1.9- to 6.3-fold and 2.6- to 14.1-fold, respectively (Table 6). These results indicate that there is an apparent minor shift in the potency of DBPR110 in the presence of higher serum concentrations.

Table 6.

Effects of serum on the antiviral activity of DBPR110 in HCV1b replicon cell lines

Serum concn (%)b	DBPR110 potency
	EC50 (pM)a	Fold shift in EC50	EC90 (pM)a	Fold shift in EC90
FBS
10	4.3 ± 0.8	1.0	9.3 ± 3.4	1.0
20	8.1 ± 1.6	1.9	23.8 ± 11.0	2.6
30	7.9 ± 0.9	1.8	21.6 ± 17.0	2.3
40	13.2 ± 1.7	3.1	35.1 ± 7.4	3.8
50	21.5 ± 10.0	5.0	41.9 ± 7.2	4.5
NHS
10	33.5 ± 0.4	1.0	41.6 ± 1.3	1.0
40	210.9 ± 6.3	6.3	588.1 ± 45.9	14.1

^aMeans ± standard deviations determined from the parental cell line (n = 3).

^bFBS, fetal bovine serum; NHS, normal human serum.

In vitro combination studies with IFN-α, VX-950, and 2′CMA.

Standard care or single-agent therapies for viral infections often lead to the production of quasispecies, which increases the possibility of clinical drug resistance. Therefore, more effective and better-tolerated combination therapies to decrease the emergence of viral resistance are greatly needed. In order to evaluate the effect of DBPR110 used in combination with other HCV inhibitors, the inhibitory activity of pairwise combinations of IFN-α, VX-950, or 2′CMA with DBPR110 were analyzed using a genotype 1b replicon encoding a luciferase reporter gene. In this system, DBPR110 has a calculated EC₅₀ of 3.3 ± 0.8 pM, whereas IFN-α, VX-950, and 2′CMA have EC₅₀s of 35.1 ± 4.7 IU/ml, 301.6 ± 2.8 nM, and 167.3 ± 11.5 nM, respectively (Table 7). DBPR110 was mixed with different ratios of IFN-α, VX-950, or 2′CMA, and serial dilutions of each mixture were generated thereafter. The degree of inhibition for each drug combination was analyzed according to the median effect principle using the combination index calculation at 50%, 75%, and 90% (49). In three independent experiments, the combination of DBPR110 with IFN-α, VX-950, or 2′CMA produced synergistic effects at the 50%, 75%, and 90% effective doses (Table 8). No cytotoxicity was observed for DBPR110, IFN-α, VX-950, and 2′CMA at the concentrations used in these experiments. Moreover, DBPR110-resistant variants remained fully sensitive to IFN-α and small-molecule inhibitors of HCV NS3 protease and NS5B polymerase (Table 9).

Table 7.

Potency of DBPR110, IFN-α, VX-950, and 2′CMA on HCV1b replicon cell lines

Compound	EC50	EC90	CC50
DBPR110 (pM)	3.3 ± 0.8	7.4 ± 0.8	>50,000
IFN-α (IU/ml)	35.1 ± 4.7	327.0 ± 0.01	>2,000
VX-950 (nM)	301.6 ± 2.8	911.9 ± 75.4	>5,000
2p′CMA (nM)	167.3 ± 11.5	565.5 ± 47.7	>25,000

^aValues are means ± standard deviations determined from the HCV1b replicon cells (n ≥ 3).

Table 8.

DBPR110 in combination with IFN-α, VX-950, or 2′CMA yielded synergistic effects at the 50%, 75%, and 90% effective doses

Compound combined with DBPR110	DBPR110/compound ratio	CI value for:a			Influence
		ED50	ED75	ED90
IFN-α	1:1	0.50 ± 0 0.17	0.54 ± 0.19	0.58 ± 0.20	Synergistic
	2.5:1	0.57 ± 0.31	0.59 ± 0.31	0.61 ± 0.33	Synergistic
	1:2.5	0.45 ± 0.08	0.49 ± 0.09	0.54 ± 0.12	Synergistic
VX-950	1:1	0.43 ± 0.27	0.42 ± 0.18	0.43 ± 0.10	Synergistic
	2.5:1	0.67 ± 0.42	0.63 ± 0.33	0.60 ± 0.23	Synergistic
	1:2.5	0.34 ± 0.16	0.34 ± 0.11	0.34 ± 0.07	Synergistic
2′CMA	1:1	0.29 ± 0.14	0.31 ± 0.13	0.34 ± 0.13	Synergistic
	2.5:1	0.64 ± 0.13	0.60 ± 0.12	0.57 ± 0.12	Synergistic
	1:2.5	0.22 ± 0.04	0.25 ± 0.04	0.27 ± 0.05	Strongly synergistic

^aValues are means ± standard deviations determined from the HCV1b replicon cells (n ≥ 3).

Table 9.

DBPR110-resistant variants remained fully sensitive to IFN-α, VX-950, and 2′CMA

Amino acid substitution(s)	EC50 of the indicated inhibitora
	DBPR110 (pM)	IFN-α (IU/ml)	2′CMA (nM)	VX-950 (nM)
None (parental)	1.5 ± 0.6	3.9 ± 0.3	33.7 ± 6.9	127.3 ± 17.6
P58L	564 ± 194	3.2 ± 1.2	27.7 ± 9.8	256.5 ± 70.4
Y93N	3,821 ± 1,677	3.1 ± 1.2	28.9 ± 1.3	158.1 ± 33.3
Y93H	1,408 ± 293	3.7 ± 1.9	29.2 ± 7.4	159.4 ± 21.8
Y93N V153M M202L M265V	27,326 ± 12,349	2.5 ± 1.4	24.2 ± 5.7	142.1 ± 6.2
Y93H V153M M202L M265V	8,736 ± 2,370	2.3 ± 1.2	24.1 ± 5.9	158.2 ± 66.8

^aMeans ± standard deviations determined from transient transfection assays (n ≥ 3).

Structure biology studies.

HCV NS5A mutations can be associated with either altered drug-binding efficiency or drug resistance. Here, computational modeling was employed to give structural insights. The three-dimensional HCV NS5A structure was used (21), and the Discovery Studio, version 2.1, program (Accelrys, Inc.) was applied to build a model by mutating residues and performing energy minimization (Fig. 3 and Table 10). The DBPR110-associated mutation points, P58 and Y93, were mapped onto an HCV NS5A crystal structure of the DBPR110-NS5A protein complex. The results of modeling suggest that DBPR110 binds directly to the dimer interface of HCV NS5A (Fig. 3). We calculated the binding energy of DBPR110 in the HCV NS5A variants as a whole to gain a better insight into the role played by the DBPR110-resistant variants in the interactions with DBPR110 (Table 10). Parental NS5A and NS5A accompanied by V153M showed the most stable conformation with DBPR110, with −27.76 −29.06 kcal mol⁻¹ of binding energy (van der Waals energy and electrostatic energy), followed by P58L with −4.38 kcal mol⁻¹ and Y93H with 18.63 kcal mol⁻¹; Y93N showed the least stability, with 79.30 kcal mol⁻¹ of binding energy (Table 10). Thus, the mutation of these residues seems to affect affinity for the inhibitor, DBPR110.

Fig 3.

Docking of DBPR110 in the HCV NS5A crystal structure. An overall view of the NS5A-DBPR110 complex with the key in vitro resistance-conferring mutation positions highlighted. The figure was prepared with the PyMol program (Delano Scientific, LLC).

Table 10.

The EC₅₀ of DBPR110-resistant variants and the binding energy of DBPR110 to HCV NS5A

Amino acid substitution(s)	DBPR110 EC50 (pM)	Binding energy (kcal/mol)a
		VDW + electrostatic	VDW contribution	Electrostatic contribution
None (parental)	1.5	−26.79	−23.63	−3.16
V153 M	1.3	−29.06	−35.16	6.10
P58L	564	−4.38	−11.08	6.70
Y93H	1,408	18.63	21.14	−2.51
Y93N	3,821	79.30	87.63	−8.33

^aVDW, van der Waals.

DISCUSSION

The HCV NS5A inhibitor, BP008, with a thiazole core structure was identified via a cell-based high-throughput HCV replicon screen in our previous study (39). After further systematic optimization from studying its structure-activity relationship, we identified a more potent genotype 1b and 2a HCV replication inhibitor, DBPR110 (Fig. 1 and Table 1). The resistance selection and validation of the responsible mutations suggest that NS5A P58L/T and Y93H/N in the genotype 1b and T24A, P58L, and Y93H in the genotype 2a replicons are the major sites for the development of DBPR110 resistance, while in the 1b replicon, V153M, M202L, and M265V played a compensatory role in the processes of replication and drug resistance (Tables 3 and 5). In addition, DBPR110 displayed synergistic effects with IFN-α, an NS3 protease inhibitor, and an NS5B polymerase inhibitor, making it a promising candidate for HCV treatment. The docking of DBPR110 near P58 and Y93 showed some possible interactions between NS5A and DBPR110 (Fig. 3). To be more confident that this compound was specific for HCV, the antiviral selectivity of DBPR110 was measured by evaluating its activity against several other viruses. We found that DBPR110 did not demonstrate significant activity against dengue virus or Japanese encephalitis virus, suggesting that DBPR110 is highly selective for HCV (data not shown). We have demonstrated that DBPR110 is likely a highly potent inhibitor of HCV1b and -2a NS5A, a protein without any known enzymatic activity. This finding could be valuable as a potential therapeutic strategy for the treatment of HCV.

The predominant mutations associated with DBPR110 resistance were found at NS5A residues 58 (P58L/T) and 93 (Y93H/N) in HCV1b and T24A, P58L, and Y93H in the HCV2a replicon (Tables 2 and 4). The monomer thiazole inhibitor BP008 displayed a slightly different drug resistance pattern from that of dithiazole DBPR110. For example, the Q24L mutation found in BP008-resistant HCV1b replicon clones did not appear in the DBPR110-resistant HCV1b replicon clones (39). The degree of resistance conferred by the P58L, P58T, or P58S mutation to BP008 is similar. However, The P58L/T/S mutations conferred various degrees of drug resistance to DBPR110, with 376-, 162-, and 25-fold drug resistance conferred by P58L, P58T, and P58S, respectively. Similarly, the Y93H mutation in the HCV1b replicon was shown to confer strong resistance to both BP008 and DBPR110. The HCV1b Y93N/H mutant replicated at only 8% of the level of the parental type, indicating that its fitness is greatly reduced, although it affected DBPR110 potency significantly. In the genotype 2a HCV replicon, the resistant mutants did not show impaired fitness. In the transient replicon assays, the replicons containing these mutations showed substantially reduced susceptibility to DBPR110 (Tables 3 and 5). In particular, a change at Y93 has the strongest effect on the susceptibility of the HCV1b and -2a replicons to DBPR110.

The Y93H mutation within NS5A was discovered to be the key mutation for drug resistance within the HCV1b replicon variants selected from other NS5A inhibitors, such as BMS-824, BMS-858 (38), BMS-790052 (54), and BP008 (39). L31V and Q54L are the primary drug resistance mutations derived from BMS-824- and BMS-858-resistant HCV1b replicon cells (38), while L31F/V was identified as the primary resistance mutations in BMS-790052-resistant HCV1b replicon cells in addition to Y93H (54). In our previous study, Q24L/H, P58S/L/T, and Y93H were the key substitutions for BP008 resistance selection in HCV1b replicon cells (39). Interestingly, L31F/V or P32L mutation was not observed in the DBPR110-resistant HCV1b replicon cells selected in the presence of either a low (20 pM) or high (200,000 pM) concentration of DBPR110. The frequency of P58L, Y93H, and Y93N in the DBPR110-resistant HCV1b replicon clones (selected at 20 pM DBPR110) was found to be 13.3%, 40%, and 33.3%, respectively (data not shown). Y93N was the key mutation found within DBPR110-resistant HCV1b replicon cells at 200,000 pM of DBPR110 (Table 3). In contrast, the Y93H but not Y93N mutation was found in the BMS-790052-resistant HCV1b replicon cells selected at various concentrations of BMS-790052. The drug resistance profile of DBPR110 and the other NS5A inhibitors, such as BMS-824, BMS-858, BMS-790052, and BP008, indicated that Y93 is the common key drug resistance mutation site. In addition, DBPR110 and BMS-790052 have similar dimer chemical structures (Fig. 1) and comparable EC₅₀s against the HCV1b genotype (3.9 pM for DBPR110; 9 pM for BMS-790052) and the HCV2a genotype (228.8 pM for DBPR110; 71 pM for BMS-790052) (26) (Table 1). Overall, the results indicated that DBPR110 may have a mode of action similar to BMS-790052 based on the similar chemical structures and drug resistance profiles. Further investigation is needed to understand the differences in the molecular interaction between NS5A and inhibitors and the ability to overcome the drug resistance mutations within NS5A.

Viral sequence obtained from HCV patient isolates and deposited in the European HCV database was used for the sequence analysis. The naturally occurring variability among HCV1b sequences at residues P58, Y93, V153, M202, and M265 was analyzed to be 18%, 8%, 25%, 1%, and 5%, respectively, and the naturally occurring variability at both T24 and P58, was analyzed to be 5%, among HCV2a sequences (Table 11) (55). The lower variability at residue Y93, the predominant amino acid replacement associated with DBPR110 resistance in this study, within HCV1b patients' sequences may be related to the greatly reduced fitness induced by the mutation at Y93, e.g., Y93N. Interestingly, the amino acid residues Y93 and P58, but not Q24, are conserved between the HCV1b and -2a genomes, implying that residue T24 within the HCV2a genome may be responsible for the higher EC₅₀ of DBPR110 in the 2a genotype (228.8 pM) than in the 1b replicon (3.9 pM) (Table 1).

Table 11.

Prevalence of DBPR110 resistance mutations in the European HCV database

HCV genotype	Amino acid in NS5A	Prevalence in the database (%)		Frequency of variant(s) (%)
		Conserved	Variable
	1b	P58	82	18	H (11), S (5)
	Y93	92	8	H (7)
	V153	75	25	L (20), I (3)
	M202	99	1
	M265	95	5	T (1), L (1)
2a	T24	95	5	A (5)
	P58	95	5	S (5)
	Y93	100	0	None

The DBPR110 resistance and potency mapping study indicated that the N-terminal domain I of NS5A is the region primarily responsible for the DBPR110-mediated inhibition of HCV replicon activity (Tables 2 to 5). The structure of domain I was determined and shown to form a dimer via contacts near the N-terminal ends of the molecules that can adopt different conformations (20, 21). Y93, the major residue involved in DBPR110 resistance, is located near the dimer interface between two NS5A molecules in the structures proposed by Love et al. and Tellinghuisen et al. (20, 21). The location of the Y93N/H resistance substitution in domain I of NS5A suggests that the NS5A dimer interface may constitute the molecular surface recognized by NS5A-targeting compounds. Although the mode of action of NS5A-targeting molecules is still unclear, we propose a model to explain the molecular mechanism based upon findings from the present study. We conducted computer modeling to predict the interaction between DBPR110 and NS5A X-ray structure derived from either Tellinghuisen et al. (20) (Protein Data Bank code 1ZH1) or the other form of the NS5A structure (Protein Data Bank code 3FQQ) (21). For some unknown reason, docking experiments worked only with the 3FQQ structure. The results of our computer modeling suggest that DBPR110 binds directly to the dimer interface of HCV NS5A (Fig. 3). To determine if this NS5A inhibitor binds to the viral protein, we calculated the binding energy between DBPR110 and a key resistance point mutant in the NS5A region. The studies implied that, compared with parental and the V153M and P58L variants, the Y93H/N resistance substitution showed the least binding stability to DBPR110 (Table 10). The identification of Y93 as an important site for DBPR110 resistance suggests the possibility that NS5A dimer formation could be affected by the binding of an inhibitor. If transitioning between alternate dimer configurations is important for HCV replication, it is possible that DBPR110 could disrupt this transition, perhaps by binding across the dimer interface and stabilizing one conformation over the other. The binding of the inhibitor to NS5A may result in a disruption of specific protein-protein interactions and the abrogation of the cellular signal transduction involving NS5A, thus influencing HCV replication. Moreover, the in vitro binding experiment for the BMS-790052 analogue suggested that BMS-790052 inhibits HCV replication by directly binding to HCV NS5A (26). Because the BMS-790052 and DBPR110 resistance data share some degree of similarity with regard to the drug resistance mutations within the N terminus of NS5A, DBPR110 is likely to interfere with the role played by NS5A during the HCV replication cycle, possibly by binding to NS5A. The results of our study suggest that DBPR110 may bind across the dimer interface in the N terminus of NS5A (Table 10 and Fig. 3). Recent reports have indicated that BMS-790052 and other HCV NS5A inhibitors can induce alternations in the proper location of the HCV NS5A protein within the functional replication complexes (25, 56). It was hypothesized that an NS5A inhibitor could change the conformation of an NS5A dimer and that this change could be communicated allosterically to the NS5A molecules in an oligomeric complex. Further work, including cocrystallization studies, will be essential to better understand the interaction and molecular mechanisms of inhibition by DBPR110 and related inhibitors. Furthermore, NS5A possesses distinct cis- and trans-acting functions in HCV RNA replication, where the cis-acting function is part of a replication complex and the trans-acting function occurs outside the replication complex (57). BMS-790052 was shown to suppress these two distinct functions of NS5A in HCV RNA replication. Consequently, NS5A plays an important role in many steps or aspects of HCV viral replication, demonstrating that NS5A is a valuable target for the development of a therapeutic regimen. Continued research will aim to evaluate if DBPR110 inhibits HCV replication in a similar mechanism to other NS5A inhibitors.

Resistance to clinical antiviral chemotherapy has become a major issue in the management of patients with chronic viral infections. To achieve sustained viral responses, it is necessary to develop effective combination therapies, especially those targeting distinct HCV viral targets. For instance, IFN-α, which has known in vitro antiviral activities, is currently being used synergistically in combination therapy for chronic HCV infections (58). Our combination study using HCV1b replicon cells demonstrated synergistic antiviral actions between DBPR110 and IFN-α, VX-950, or 2′CMA (Table 8). More importantly, none of these combinations enhanced host cell toxicity, and no antagonistic antiviral effects were observed. In addition, DBPR110-resistant variants remained fully sensitive to IFN-α and small-molecule inhibitors of HCV NS3 protease and NS5B polymerase. Given that DBPR110 can yield a high level of synergy with “clinical standard care” inhibitors and inhibitors targeting the HCV NS3 protease or NS5B polymerase, incorporating DBPR110 into future anti-HCV cocktails can provide major advantages over single-drug therapy and represents an attractive paradigm for improving current virologic response rates. With a target distinct from those of the many HCV protease and polymerase inhibitors in various stages of clinical development, DBPR110 has the potential to be an important component in combination strategies for HCV therapy. In addition to drug resistance, another important issue with chemotherapy is the reduced therapeutic efficiency caused by a high proportion of drug binding to plasma or serum proteins. In our protein binding study, drug-serum protein binding had no apparent effect on the therapeutic efficiency of DBPR110 (Table 4). This indicates that DBPR110 is a good drug candidate for development as an HCV therapeutic regimen again.

In conclusion, the data presented in this study demonstrate the excellent anti-HCV activity of DBPR110. The in vitro combination of DBPR110 with IFN-α, an NS3 protease inhibitor, or an NS5B polymerase inhibitor led to a much stronger antiviral response than that observed with either agent alone. Importantly, the NS5A variants that represented high levels of resistance to DBPR110 remained fully sensitive to IFN-α and the small-molecule inhibitors of HCV protease and polymerase. Therefore, with its extraordinary potency, DBPR110 has the potential to become a novel therapeutic option for hepatitis C, and further evaluation of this compound in combination therapy is warranted.