Potential conflict of interest: F.Z. has received research support from and is a consultant for Gilead. He is also a consultant for BMS, Schering, and Roche. C.T. has received research support from and is a consultant for Gilead.
Mutations within the hepatitis B virus (HBV) polymerase gene conferring drug-resistance are selected during prolonged lamivudine (3TC) or adefovir dipivoxil (ADV) treatment. Because there is no other approved drug against HBV, treatments with 3TC or ADV are used either sequentially or in addition, depending on treatment response or failure. Considering the use of de novo or add-on 3TC+ADV bitherapy, we investigated the possibility of the emergence of an HBV strain harboring polymerase mutations conferring resistance to both 3TC (rtL180M+M204V) and ADV (rtN236T). We constructed the L180M+M204V+N236T mutant and determined its replication capacity and its susceptibility to different nucleos(t)ide analogs in transiently transfected hepatoma cell lines. The triple mutant replicates its genome in vitro, but less efficiently than either the wild-type (wt) HBV or L180M+M204V and N236T mutants. Phenotypic assays indicated that the L180M+M204V+N236T mutant is resistant to pyrimidine analogs (3TC, -FTC, β-L-FD4C, L-FMAU). Compared with wt HBV, this mutant displays a 6-fold decreased susceptibility to ADV and entecavir and a 4-fold decreased susceptibility to tenofovir. Interferon alfa inhibited equally the replication of wt and L180M+M204V+N236T HBV. In conclusion, the combination of rtL180M+M204V and rtN236T mutations impairs HBV replication and confers resistance to both 3TC and ADV in vitro. These results suggest that the emergence of the triple mutant may be delayed and associated with viral resistance in patients treated with 3TC+ADV. However, other nucleos(t)ide analogs in development showed an antiviral activity against this multiresistant strain in vitro. This provides a rationale for the clinical evaluation of de novo combination therapies. (HEPATOLOGY 2005;41:1391-1398.)
The main goals of the treatment of chronic hepatitis B virus (HBV) infection are the sustained suppression of HBV replication and a remission of liver disease. An immunomodulator, interferon alfa (IFN-α), and two synthetic nucleos(t)ide analogs, lamivudine (3TC) and adefovir dipivoxil (ADV), are currently approved for the treatment of chronic hepatitis B. With respect to IFN-α, the overall response rate to this drug is less than 40%, and IFN-α therapy is associated with a number of adverse effects.1 3TC, a nucleoside analog of L-deoxycitydine, and ADV, a phosphonate nucleotide analog of adenosine monophosphate, target the HBV reverse transcriptase (RT) activity, thus inhibiting viral replication. They are both powerful antivirals, and monotherapy with 3TC or ADV results in significant improvement in virological, biochemical, and histological status in most of the patients.2–5
However, long-term 3TC or ADV monotherapy leads to the emergence of drug-resistant HBV strains. 3TC-resistance increases at a rate of approximately 20% of patients per year, to reach approximately 70% of treated patients after 4 years.6, 7 Mutations conferring resistance to 3TC have been mapped in the conserved YMDD motif within the C domain of the viral RT (M204I/V).8–10 They are frequently associated with compensatory mutations in the conserved B domain (V173L, L180M) that restore partially the replicative capacity of YMDD mutant strains in vitro.11–13 Mutations associated with 3TC-resistance induce a reduction in the susceptibility to 3TC9 and may be associated with a progression of liver disease in a significant number of patients.6
ADV, the latest antiviral approved for the treatment of chronic hepatitis B, is used successfully for the treatment of patients with 3TC-resistant chronic hepatitis B either in monotherapy14 or in combination with 3TC.14, 15 This agrees with in vitro studies that showed that 3TC-resistant mutations do not confer resistance to ADV.16–18 ADV-resistance is less frequent and more belated compared with 3TC. Two recent clinical studies reported cases of ADV-resistance associated with the selection of the mutation N236T within the D domain of HBV RT after 2 years of ADV monotherapy.19, 20 This mutation induces a reduction in the susceptibility to ADV by 4.4-20 to 23-fold19 compared with wild-type (wt) HBV in vitro, but retains susceptibility to 3TC.
Several other synthetic nucleos(t)ide analogs are in clinical development. Emtricitabine (-FTC), a close analog of 3TC, has been recently approved for the treatment of HIV infection and is currently in phase III clinical trials for the treatment of HBV infection,21, 22 whereas clevudine (L-FMAU), a thymidine analog, has entered phase II clinical trials.23 Elvucitabine (β-L-FD4C), another L-deoxycytidine analog, is also a potent antiviral against HBV and hepadnaviruses in animal models.24–26 However, 3TC-resistance mutations confer cross-resistance to these three pyrimidine analogs with the exception of the M204V mutant that retains susceptibility to L-FMAU and β-L-FD4C.12, 27 Entecavir (ETV), a deoxyguanosine analog, inhibits specifically the replication of wt and 3TC-resistant HBV in vitro and in vivo.12, 28, 29 ETV has recently been approved in the United States.30 Tenofovir (PMPA), a nucleotide analog closely related to ADV, has already obtained approval for the treatment of HIV infection22 and inhibits efficiently the replication of 3TC-resistant M204V HBV mutant in vitro.31 Moreover, promising results were reported with respect to the efficacy of a 3TC+PMPA combination to reduce HBV DNA levels in HIV/HBV coinfected patients carrying wt or 3TC-resistant HBV.32–34 Two options are available to treat patients who develop a 3TC resistance.14, 15 3TC can be switched to ADV with the risk of developing ADV-resistance over the long-term, or ADV can be added to the ongoing 3TC monotherapy to delay further resistance, because there is no cross-resistance between these drugs.19, 20 However, a 3TC+ADV bitherapy could select for HBV mutants resistant to both drugs. In this study, we determined the phenotypical characteristics of an HBV laboratory strain carrying mutations conferring resistance to both 3TC (L180M+M204V) and ADV (N236T). The combination of L180M+M204V+N236T mutations was assembled in an HBV replication-competent plasmid. Its effects on the replication of the HBV genome and the susceptibility to nucleos(t)ide analogs (3TC, -FTC, L-FMAU, β-L-FD4C, ETV, ADV, PMPA) and to IFN-α were assessed in transiently transfected hepatoma cell lines.
HBV, hepatitis B virus; IFN-α, interferon alfa; 3TC, lamivudine; ADV, adefovir dipivoxil; RT, reverse transcriptase; wt, wild-type; L-FMAU, clevudine; β-L-FD4C, elvucitabine; ETV, entecavir; PMPA, tenofovir; HBsAg, hepatitis B surface antigen; IC50, 50% inhibitory concentration of viral replication in cell culture; CC50, 50% inhibitory concentration of cellular growth in cell culture; dNTP, deoxynucleotide triphosphate.
Materials and Methods
3TC, -FTC, L-FMAU, ADV, and PMPA were provided by Gilead Sciences (Foster City, CA). ETV was obtained from Moravek (Brea, CA), and IFN-α 2b from Schering Plough (Kenilworth, NJ). β-L-FD4C was a generous gift from Dr. Y.C. Cheng (Yale University, New Haven, CT).
Construction of N236T and L180M+M204V +N236T Laboratory Strains.
Plasmids pCMV-HBV containing wt or L180M+M204V HBV genome (genotype D, subtype ayw)35 were engineered with the QuickChange kit (Stratagene, La Jolla, CA) to install the N236T mutation. All constructs were sequenced to confirm this modification. The different EcoRI-NcoI restriction fragments encompassing the B, C, and D domains were transferred into the plasmid pTriEXModHBV containing 1.1 genome unit of the same wt HBV strain (D, ayw). This expression vector enables the production of pregenomic RNA under the control of the chicken beta actin promoter and therefore triggers HBV DNA synthesis.36 Three pTriEXModHBV plasmids with L180M+M204V, N236T, or L180M+M204V+N236T mutations were thus obtained.
Viral DNA Synthesis Capacity of HBV Mutants.
Transient transfection of plasmids containing wt or mutated HBV genome into Huh7 and HepG2 cell lines were performed using Fugene6 transfection reagent (Roche Diagnostics, Mannheim, Germany) according to the manufacturer protocol. Transfected cells were grown for 1 week, and culture medium was changed every 4 days. Intracellular encapsidated HBV DNA was purified and submitted to Southern blot analysis as previously described.35–37 Viral DNA synthesis levels were determined after phosphor-imager analysis of Southern blot. Transfection efficiency was monitored by quantifying the amount of hepatitis B surface antigen (HBsAg) secreted into the cell culture supernatant at the end of the experiment with the Monolisa HBsAg Ultra kit (Bio-Rad, Hercules, CA).
Drug Susceptibility Assay.
Huh7 cells were transfected as described above with wt or mutant HBV constructs. Treatment with drugs or combination of drugs started 60 hours after transfection and was renewed every other day for 5 days. To screen which drugs in development display an antiviral activity against the triple-mutant replication, transfected cells were treated with a single-drug concentration (-FTC: 15 μmol/L; L-FMAU: 100 μmol/L; β-L-FD4C: 1.25 μmol/L; PMPA: 100 μmol/L; ETV: 5 μmol/L; IFN-α: 103 IU/mL). To determine inhibitory concentrations (IC50), transfected cells were treated with 0, 0.5, 15, and 100 μmol/L 3TC, PMEA, or 3TC+PMEA; 0, 1.25, 2.5, 5, and 10 μmol/L ETV; or 0, 12.5, 25, 50, 100, and 200 μmol/L PMPA. For hydroxyurea (Sigma, St. Louis, MO) susceptibility assay, HEK 293 cells were transfected as previously described and treated with 0 or 0.1 mmol/L hydroxyurea for 2 days.11 At the end of treatment with drugs, intracellular encapsidated HBV DNA was extracted and submitted to Southern blot analysis, which was quantified after a phosphor-imager analysis. For each drug, the IC50 was determined by a 50% decrease in the amount of intracellular viral DNA detected in treated cells at the end of the treatment compared with untreated cells.35 Analysis of cellular toxicity was performed as previously described.35
Association of 3TC- and ADV-Resistant Mutations Impairs HBV Genome Replication.
First, we asked whether an HBV polymerase mutant gene carrying mutations conferring resistance to both 3TC and ADV was replication competent. Huh7 and HepG2 cells were transiently transfected with plasmids containing wt, L180M+M204V, N236T, or L180M+M204V+N236T HBV genome. One week after transfection, DNA from intracellular core particles were extracted and submitted to Southern blot analysis. Relative replication capacities of the three HBV mutants were determined after quantification by phosphor-imager analysis of Southern blots and standardized according to the amount of HBsAg secreted in the culture medium (data not shown). The level of replication of the wt HBV was taken as standard of replication (Fig. 1B). The relative replication capacities of the L180M+M204V mutant in Huh7 and HepG2 cells, 83.7% ± 11.3% and 22.9% ± 2.8%, respectively, were similar to those previously measured in these cell lines.12, 38 The replication efficiency of the N236T mutant was not significantly different in Huh7 (104% ± 10%) and HepG2 cells (87.3% ± 12%). With respect to the L180M+M204V+N236T mutant, the relative level of viral DNA synthesis was 27.8% ± 2.1% in Huh7 cells and 6.8% ± 1.8% in HepG2 cells. Thus, the triple mutant, like the L180M+M204V mutant, replicates its genome approximately 4-fold less efficiently in HepG2 cells than in Huh7 cells. Differences in the cellular concentration of the deoxynucleotide triphosphates (dNTPs) pool may explain this discrepancy in the viral DNA synthesis levels. Indeed, Melegari et al.11 previously observed that YMDD-mutants, but not wt HBV, displayed a reduced capacity of replication under conditions of intracellular dNTPs depletion produced by hydroxyurea in HEK 293 cells. To address this point, the effect of 0.1 mmol/L hydroxyurea on the replication of wt, L180M+M204V, N236T, or L180M+M204V+N236T HBV genomes was determined in transiently transfected HEK 293 cells. The viral DNA synthesis of wt HBV and N236T mutant were both inhibited by less than 30% by hydroxyurea administration (Fig. 2). However, the replication of L180M+M204V and L180M+M204V+N236T mutants were both inhibited by more than 70%, suggesting that the mutant polymerases harboring mutations in the YMDD motif are more sensitive to intracellular dNTPs depletion than HBV strains with a wt YMDD motif.
Susceptibility of the L180M+M204V+N236T Mutant to 3TC and ADV.
Because the replication of the triple mutant is very low in HepG2 cells, we assessed the susceptibility of this mutant to antiviral drugs in Huh7 cells. We first determined the triple-mutant susceptibility to 3TC and ADV, individually or in combination, and compared it with wt, N236T, and L180M+M204V HBV counterparts. As previously observed, ADV and 3TC were efficient against the replication of the L180M+M204V (ADV IC50 = 15.5 ± 1.8 μmol/L) and N236T (3TC IC50 = 2.65 ± 0.52 μmol/L) mutants, respectively (Table 1). The L180M+M204V mutant was resistant to 3TC (fold resistance > 40), and the N236T mutant was approximately 3-fold less susceptible to ADV than wt HBV (IC50 = 50.3 ± 11 μmol/L and 15.8 ± 1.9 μmol/L, respectively). The L180M+M204V+N236T mutant was as resistant to 3TC as the L180M+M204V mutant (fold resistance > 40), and it presented a 6.3-fold decreased susceptibility to ADV (IC50 = 100 ± 20 μmol/L) compared with wt HBV. Thus, the combination of drug resistance mutations did not restore susceptibilty to one or the other drug. As a consequence, the triple mutant was resistant to the treatment combining 3TC and ADV, showing an IC50 (100 ± 20 μmol/L) similar to the IC50 for ADV that is more efficient on the triple mutant replication than 3TC. Similarly, the response of the N236T and L180M+M204V mutants to the 3TC+ADV combination reflected their susceptibility to these 2 drugs (Table 1). With respect to wt HBV, the IC50 for the bitherapy (1.7 ± 0.36 μmol/L) and 3TC (2.48 ± 0.67 μmol/L) were not significantly different, 3TC being more efficient at lower concentrations than ADV on wt HBV.
Table 1. In Vitro Susceptibility of the L180M+M204V+N236T Laboratory Strain to 3TC, ADV, and 3TC+ADV Bitherapy
Values represent the mean of 3 independent experiments, each performed in duplicate.
Fold resistance = (mutant IC50)/(wt IC50).
In the combination therapy experiment, both drugs were administrated at the same concentrations.
2.48 ± 0.67
15.8 ± 1.9
1.7 ± 0.36
2.65 ± 0.52
50.3 ± 11
2.68 ± 0.35
15.5 ± 1.8
18.75 ± 5.7
100 ± 20
100 ± 20
Susceptibility of the L180M+M204V+N236T Mutant to Different Nucleos(t)ide Analogs in Development.
Because the L180M+M204V+N236T mutant was resistant to both 3TC and ADV, we next determined its susceptibility to several nucleos(t)ide analogs in development (-FTC, β-L-FD4C, L-FMAU, ETV, and PMPA). Huh7 cells transfected with plasmids containing wt, N236T, L180M+M204V or L180M+M204V+N236T genomes were treated for 5 days with each compound at a high single concentration. All the nucleos(t)ide analogs reduced the viral DNA synthesis level of wt HBV and N236T mutant by more than 70% and 50% respectively, which confirmed the antiviral potency of these drugs (Table 2). The levels of viral genome replication of the L180M+M204V and L180M+M204V+N236T mutants were decreased by less than 20% in presence of 15 μmol/L -FTC, 1.25 μmol/L β-L-FD4C, or 100 μmol/L L-FMAU. However, the replication of the triple mutant was reduced by 48.9% with 5 μmol/L ETV and by 70.8% with 100 μmol/L PMPA. The IC50s of these 2 drugs were thus determined by treating transiently transfected Huh7 with increasing drug concentrations ranging from 0 to 10 μmol/L for ETV (data not shown) and from 0 to 200 μmol/L for PMPA (Fig. 3). With ETV, the IC50s for the wt and N236T HBV strains were 0.8 ± 0.1 μmol/L and 0.7 μmol/L, respectively (Table 3). The L180M+M204V and L180M+M204V+N236T mutants both displayed an approximatively 6-fold resistance to ETV (IC50s = 5 ± 0.25 μmol/L and 5 ± 0.7 μmol/L, respectively) compared with wt HBV. With PMPA, a 3.4-fold resistance compared with wt HBV was observed for the L180M+M204V mutant (IC50 = 35.2 ± 5.1 μmol/L), a 4.5-fold resistance for the N236T mutant (IC50 = 46 ± 6 μmol/L) and 4.4-fold resistance for the L180M+M204V+N236T mutant (45.5 ± 6.1 μmol/L). The efficiency of IFN-α 2b against the replication of the triple mutant was also tested. It reduced equally the levels of viral DNA synthesis of the wt HBV and the 3 mutants by approximately 50% (Table 2).
Table 2. In Vitro Efficacy of Antivirals and IFN-α to Inhibit the Genome Replication of HBV Carrying 3TC-, ADV-, or 3TC+ADV-Resistance Mutations
Inhibition of the Replication Compared With No Drug (%)
Values represent the mean of at least 3 independent experiments, each performed in duplicate.
Fold resistance = (mutant IC50)/(wt IC50).
Wild type (wt)
10.3 ± 1.3
0.8 ± 0.1
46 ± 6
35.2 ± 5.1
5 ± 0.25
45.5 ± 6.1
5 ± 0.7
The emergence of HBV strains resistant to antiviral treatments is a major clinical concern. One alternative to delay the occurrence of drug-resistant mutants is to combine drugs that do not share the same resistance profile. We therefore asked whether an HBV mutant with a polymerase gene harboring both 3TC- and ADV-resistance mutations would have the capacity to escape the combination of 3TC+ADV given as a de novo or an add-on therapy for 3TC resistance. We constructed this mutant and investigated its replicative capacity and susceptibility to drugs in hepatoma cell lines.
Compared with wt, L180M+M204V, and N236T laboratory strains, the L180M+M204V+N236T mutant exhibited the lowest level of replication in vitro. This effect was more pronounced in HepG2 than in Huh7 cell lines. Previous studies showed that polymerases harboring mutations in the YMDD motif are less active than wt polymerase12, 39 and require higher dNTP concentrations to reach maximum efficiency.38 Under conditions of dNTPS depletion in transiently transfected HEK 293 cells, we observed that the replication of L180M+M204V+N236T and L180M+M204V mutants are both significantly reduced compared with wt and N236T HBV. Thus, the concentration of dNTPs within HBV-producing cells may be involved in the decreased replication efficiency of the L180M+M204V polymerase mutant, as suggested previously by Melegari et al.11 With respect to the N236T mutant, the replication of its genome is inhibited by less than 30% when the dNTPs pool is depleted, as well as the replication of wt HBV. Thus, variations in the intracellular dNTPs pool concentration affected weakly the replication of ADV-resistant mutant, which may explain the similarity of its replicative capacity in Huh7 and HepG2 cells. However, the addition of the N236T mutation seemed to increase the impairment of the polymerase processivity of the L180M+M204V mutant, the triple mutant replicating its genome approximately threefold less efficiently than the double mutant. According to the HBV RT model based on homology with HIV RT, the N236 amino acid is adjacent to the active site of the RT, and the N236T mutation may change the conformation of the dNTPs binding site, thus conferring resistance to ADV.40 The N236T mutation might also affect the binding of dNTP and thus impair their affinity for HBV RT harboring mutations in the YMDD motif. The low genome replication capacity of the L180M+M204V+N236T mutant may have different consequences in terms of viral fitness, such as reduced virion production (data not shown) or decreased infectivity. The lower replication level of the L180M+M204V mutant may result in a lower virion production.41 Until now, studies of the fitness of HBV drug-resistant strains have been hampered by the lack of easy-to-handle tissue culture or animal models for infectivity studies. It would be interesting to see whether the low viral genome replicative capacity of the L180M+M204V+N236T mutant may result in a significant decrease in virus infectivity, which could therefore result in delayed emergence of such mutant during combination therapy in vivo.
According to our results, combining 3TC- and ADV-resistance mutations did not restore some level of susceptibility to nucleos(t)ide analogs. This differs from what was observed with HIV, where the 3TC-resistant M184V mutation, the counterpart of HBV M204V, restores some susceptibility to HIV strains carrying thymidine analog mutations.42 With respect to HBV, the addition of L180M+M204V mutations to N236T mutation seemed to increase the level of resistance to ADV, although 3TC-resistant strains are sensitive to this drug in vitro and in vivo.12, 14, 15, 16–18 Indeed, the triple mutant was 2-fold less susceptible to ADV than the N236T mutant (Table 1). At the molecular level, the steric hindrance attributable to the L180M+M204V mutations43 may worsen the binding of the ADV in a dNTPs binding site already altered by the N236T mutation40 and thus decrease the susceptibility to this drug. These in vitro findings suggest that, in vivo, the selection of the triple mutant could induce viral resistance during a prolonged 3TC+ADV combination therapy.
With respect to the susceptibility of the triple mutant to other drugs, the L180M+M204V mutations confer resistance to L-pyrimidine analogs, such as -FTC, L-FMAU, and β-L-FD4C,12, 27 and probably accounted for the absence of response of the triple mutant to these drugs. The 3TC-resistance mutations also decrease the susceptibility to ETV in vitro,12, 28 which explains that the triple mutant was 6.25-fold less sensitive to ETV than wt HBV. Finally, L180M+M204V+N236T and N236T mutants were 4.5-fold less susceptible to PMPA than wt HBV. The mechanism of resistance conferred by the N236T mutation probably accounts for this decreased antiviral effect of PMPA, because this drug and ADV share common structural features.
Phenotypical analysis of the L180M+M204V+N236T mutant also indicated that IFN-α inhibit equally its replication and those of L180M+M204V, N236T and wt laboratory strains in vitro (Table 3). This was foreseeable because IFN-α exhibits at least two HBV-specific antiviral activities independent of the viral polymerase sequence: one reduces the levels of core protein and replicative intermediates, and the other leads to posttranscriptional degradation of HBV RNA.44
The therapeutic potency of a drug depends on different factors, including its antiviral capacity and its toxicity. Consequently, the selectivity index (IC50/CC50) of a drug may prevent the use of optimal dose to reach a more potent antiviral effect. As triple mutant susceptibilities to ADV and ETV are decreased by approximately 6-fold in vitro, the treatment of patients infected with this HBV strain may require higher doses of drugs to overcome this decreased susceptibility. However, increasing the ADV daily dose from 10 to 30 mg leads to renal adverse effects,5 which precludes its use in the treatment of this multiresistant strain infection. Clinical studies showed that ETV is safe and inhibits viral replication more potently than 3TC when administered at 0.5 mg/day.45, 46 Although L180M+M204V mutant displays a decreased susceptibility to ETV in vitro,12, 28 ETV used at a 1-mg dose was shown to decrease viral load in patients infected with 3TC-resistant HBV strains29 and may be efficient against infection by the triple mutant because its susceptibility to ETV in vitro is similar to the L180M+M204V mutant counterpart (Table 3). However, viral strains with a decreased ETV susceptibility have been observed and correspond to the selection of complex mutants harboring specific polymerase mutations in addition to 3TC-resistant ones.29 PMPA is also interesting for the treatment of patients who may be infected with 3TC+ADV-resistant HBV strain. Indeed, PMPA was slightly more potent than ADV and ETV to inhibit the triple mutant replication in vitro and, by contrast to ADV, PMPA is not associated with renal side effects at doses that suppress significantly the viral load in patients.32 Moreover, PMPA is efficient for the treatment of HIV/HBV co-infected patients infected with 3TC-resistant HBV strains.32–34
In conclusion, the combination of ADV- and 3TC-resistance mutations impaired the replication of the HBV genome in hepatoma cell lines, which suggests that the emergence of an HBV strain harboring these two mutations may be delayed compared with either 3TC- or ADV-resistant HBV strains in patients treated with a 3TC+ADV bitherapy. Our in vitro findings also showed that this combination of mutations conferred resistance to both 3TC and ADV. Consequently, nonoptimal strategies based on the sequential use of nucleos(t)ide analogs could lead to the selection of multiple drug-resistant strains. Interestingly, the triple mutant retains susceptibility to IFN-α, ETV, and PMPA in vitro, suggesting that these drugs could be evaluated to rescue resistance in patients infected with this multidrug–resistant HBV strain. These in vitro data also provide useful information for the rational evaluation of de novo combination therapies in clinical trials.47