Potential conflict of interest: Nothing to report.
Tenofovir is a new effective treatment option for patients with chronic hepatitis B, but could be potentially hampered by mutations in the hepatitis B virus (HBV) polymerase conferring drug resistance. Drug resistance may occur preferentially if long-term administration is required, for example, in patients with hepatitis B e antigen (HBeAg)-negative HBV infection bearing precore (PC) and basal core promoter (BCP) mutations. The rtA194T polymerase mutation has been found in HBV/HIV coinfected patients during tenofovir treatment and may be associated with tenofovir resistance. We generated replication-competent HBV constructs harboring rtA194T alone or in addition to lamivudine (LAM) resistance (rt180M + rtM204V), PC mutations, and BCP mutations and assessed their replicative capacity after transient transfection in human hepatoma cells. The rtA194T polymerase mutation alone or in conjunction with LAM resistance reduced the replication efficiency as compared with wild-type (WT) HBV. In contrast, combination of rtA194T (± LAM resistance) with HBeAg-negative PC or BCP mutants increased the replication capacity of the drug-resistant polymerase mutants, thereby restoring the viral replication to similar levels as WT clones. Clones harboring rtA194T showed partial resistance to tenofovir in vitro and also to LAM but remained susceptible to telbivudine and entecavir. Conclusion: The rtA194T polymerase mutation is associated with partial tenofovir drug resistance and negatively impacts replication competence of HBV constructs. Viral replication, however, can be restored to WT levels, if these polymerase mutations occur together with precore or basic core promoter substitutions as found in HBeAg-negative hepatitis B. Patients with HBeAg-negative chronic HBV infection may therefore be at particular risk when developing drug resistance to tenofovir. Telbivudine or entecavir should be considered as effective alternative treatment options for these patients. (HEPATOLOGY 2009.)
Tenofovir disoproxil fumarate, a nucleotide analog that is structurally closely related to adefovir, has recently been approved for the treatment of chronic hepatitis B. Even though initially developed and approved for the treatment of human immunodeficiency virus (HIV), it has potent antiviral activity against wild-type (WT) or lamivudine-resistant hepatitis B virus (HBV) strains both in vitro and in clinical trials.1–4 Its effects on HBV DNA suppression appear similar to entecavir or telbivudine and superior to lamivudine or adefovir,5, 6 and it remains largely active even against adefovir-resistant or entecavir-resistant mutations.7–9 The combination of potent viral suppression and sustained efficiency against drug-resistant mutants renders tenofovir an attractive candidate for rescue treatment in chronically HBV-infected patients.2, 10–12 Occurrence of tenofovir resistance, conversely, was very rarely observed in large trials of HBV-infected patients.4–6, 12 Up to now, only the rtA194T nucleotide exchange in the HBV polymerase, which was identified in two HBV-HIV–co-infected patients, has been associated with tenofovir drug resistance in vitro and in vivo.13
In this study, we assessed the replication capacity of HBV strains with the rtA194T polymerase mutation to provide a molecular rationale for handling the emergence of this mutation. We also tested its replication efficacy in conjunction with lamivudine (LAM) resistance (rtA194T + rtL180M/rtM204V), because tenofovir is often used in pretreated and LAM-resistant patients.2 Because the long-term use of nucleotide analogs with the risk of selecting drug-resistant mutants is particularly common in patients with chronic hepatitis B e antigen (HBeAg)-negative HBV infection, we further addressed whether the most frequent mutations associated with HBeAg negativity, precore (PC) and basal core promoter (BCP) substitutions, affect the replication of rtA194T HBV constructs. We and others have previously reported that polymerase mutants associated with LAM resistance show higher replication levels if coexistent PC or BCP mutations are present.14, 15 In addition, although the rtA194T polymerase mutation was selected in individual HBV-HIV–co-infected patients during tenofovir administration,13 whether the rtA194T mutation truly confers resistance against tenofovir has remained controversial, because in vitro phenotypic assays showed variable results between different laboratories.13, 16 We therefore readdressed the sensitivity against tenofovir in HBV constructs carrying the rtA194T mutation in vitro, and we also tested alternative antivirals such as lamivudine, telbivudine, and entecavir with these mutants.
β-Gal, beta-galactosidase; BCP, basal core promoter; DMEM, Dulbecco's modified Eagle medium; EC50, 50% efficient concentration; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HIV, human immunodeficiency virus; LAM, lamivudine; PC, precore; PCR, polymerase chain reaction; WT, wild-type.
Materials and Methods
Generation of Replication-Competent HBV Vectors.
Replication-competent HBV plasmids (genotype A, subtype adw2) carrying the 1.28-fold of the complete HBV genome were used in this study.17 The 1.28-fold HBV replication-competent plasmids containing rtA194T and rtA194T/rtL180M/rtM204V substitutions within the polymerase gene of HBV were generated previously.13 To introduce the PC mutant (G1896A/C1858T) and the BCP double mutations (A1762T/G1764A) to the rtA194T vectors, the EcoRI-NcoI fragment was cut out from each of the rtA194T vectors and ligated into the same region within the WT-PC and WT-BCP 1.28-fold HBV replication-competent plasmids, respectively. The WT-PC and WT-BCP 1.28-fold HBV replication-competent plasmids were described previously.14 All constructs were confirmed by digestion analysis followed by direct sequencing of the whole genome (Applied Biosystems, Foster City, CA).18
Cell Culture and Transfection.
Huh-7 cells were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum and antibiotics and then transiently transfected by the vectors (5 μg per 6-cm dish) using the calcium phosphate DNA precipitation method.14 Transfection efficiency was assessed by cotransfection of 0.5 μg porcine cytomegalovirus beta-galactosidase (β-Gal) vector (Stratagene, La Jolla, CA) and measurement of β-Gal activity from cell lysates as described.19 Transfected cells were harvested after 5 days. Cell lysates from three 6-cm dishes were pooled for further analysis and treated as one experiment. All experiments were performed three to five times.
Antiviral Drugs and Susceptibility Assay.
Tenofovir disoproxil fumarate [9-(2-[phosphonomethoxy]propyl)adenine) was provided by Gilead Sciences, telbivudine by Novartis, and entecavir by Bristol-Myers Squibb. LAM was purchased from Moravek Biochemicals, Brea, CA. For dose finding, different concentrations of antiviral agents (0, 0.1, 1, 10, 30, 50, and 200 μmol) were added into the cell culture medium; medium containing drugs was changed every other day. The lowest 50% efficient concentrations (EC50) were calculated, and the fold of resistance toward antivirals such as tenofovir was calculated by dividing mutant EC50 by WT EC50. For the other antivirals, efficient inhibitory concentrations for lamivudine (10 μmol), telbivudine (50 μmol), and entecavir (1 μmol) were used.
Isolation of Progeny HBV DNA and Dot Blot Assay.
HBV progeny DNA was quantified to assess intracellular HBV replication as previously described.14 In brief, cells were lysed 5 days after transfection, and HBV capsids were immunoprecipitated by polyclonal rabbit anti-hepatitis B core antibody (Dako, Carpinteria, CA) and protein A agarose (Roche Diagnostics, Mannheim, Germany). Contaminating HBV plasmids were removed using deoxyribonuclease/ribonuclease digestion (Promega, Madison, WI). HBV progeny DNA was isolated after digestion of capsids with proteinase K-sodium dodecyl sulfate (Roche) followed by phenol-chloroform extraction and ethanol precipitation. HBV progeny DNA was blotted onto Hybond-N+ membrane (Amersham Biosciences, Buckinghamshire, UK) using the MINIFOLD I dot blot machine (Schleicher & Schuell, Dassel, Germany) and subsequently detected using 3.2-kb α[32P]-radiolabeled HBV full genome fragment. Data were quantified by the Scion Image software and normalized to total protein content of the cell lysate and transfection efficiency.
Northern Blot Analysis.
Total RNA was extracted using RNeasy kit (Qiagen, Hilden, Germany). Northern blot analysis was performed as described,14 using 25 μg RNA per sample and a 3.2-kb α32P radiolabeled HBV probe for detection of all HBV mRNAs. The density of pregenomic and hepatitis B surface RNA were measured by the Scion Image software. Data were normalized to β-Gal transfection efficiency and to the density of the 28S/18S ribosomal RNA as a loading control.
HBV Viral Load.
Five days after transfection, released HBV virions were harvested from the cell culture supernatant using polyethylene glycol precipitation (Sigma, St. Louis, MO), followed by eliminating contaminant HBV vectors (deoxyribonuclease/ribonuclease treatment), digestion by proteinase K–sodium dodecyl sulfate, extraction by phenol-chloroform, and precipitation with alcohol. To perform real-time polymerase chain reaction (PCR), a 98-bp fragment from a highly conserved region of HBV S gene was amplified using SYBR Green qPCR SuperMix (Invitrogen, Carlsbad, CA). A standard curve was prepared using 10-fold serial dilutions of the HBV WT vector, and viral load was quantified by the ABI 5700 real-time PCR machine (Applied Biosystems). After quantification of HBV DNA copy numbers, the viral load was normalized to the β-Gal transfection efficiency.
HBV Proteins in Supernatant.
Hepatitis B surface antigen and HBeAg were measured in the supernatant by commercially available kits (Architect, Abbott Laboratories, Chicago, IL). The values were normalized to the transfection efficiency.
Monitoring Contamination by HBV Vectors.
To ensure that there is no contamination by remaining HBV plasmids during the different assays, a pair of primers was designed to amplify the pBluescript (pBS) backbone of the HBV vectors. The primer sequences were pBS-forward 5′-gat gct ttt ctg tga ctg gtg ag-3′ and pBS-reverse 5′-cgt ttt cca atg atg agc act t-3′. All extracted DNA from progeny assays (before and after drug susceptibility assays) as well as extracted DNA from viral load assays were tested for DNA vector contamination by real-time PCR. No contamination was detected in the experiments (data not shown).
Results are reported as mean ± standard deviation. The t test was used for comparisons between the groups, and P < 0.05 was considered statistically significant.
The rtA194T Polymerase Mutation Reduces Viral Replication Efficacy.
The rtA194T nucleotide exchange in the HBV polymerase has been identified in HBV-HIV–co-infected patients and associated with tenofovir drug resistance in vitro and in vivo.13 However, up to now, occurrence of tenofovir resistance was very rarely observed in large trials of HBV-infected patients.4–6 We therefore assessed the viral replication efficiency at a molecular level using replication-competent HBV constructs with either WT polymerase, the rtA194T mutation, or the rtA194T mutation in combination with the LAM-resistance mutations rtL180M and rtM204V (Fig. 1A). After transient transfection of human hepatoma cells, the transcriptional activity of these constructs was unchanged (Fig. 1B). However, constructs with the rtA194T mutation, alone or in conjunction with LAM resistance, had significantly reduced intracellular replication efficacies, as evidenced by reduced intracellular progeny DNA levels (Fig. 1C). This corresponded to reduced numbers of released virions found in the supernatant of transfected cells (Fig. 2A). The rtA194T mutation did not affect synthesis or secretion of the hepatitis B surface antigen and HBeAg (Fig. 2B). These data show that the rtA194T mutation impairs viral replication directly at a level of reduced polymerase activity.
Basal Core Promoter and Precore Mutations Enhance Replication Efficacy of rtA194T Mutants.
Reduced polymerase activity has been reported for several drug-resistance mutations, including LAM or entecavir resistance.4, 20 However, we and others have previously reported that LAM-resistant polymerase mutants show improved replication efficacies, if coexistent PC or BCP mutations are present.14, 15 Because the prevalence of HBeAg-negative chronic HBV infection with these PC or BCP mutations is increasing worldwide,4 we generated HBV constructs combining the polymerase (A194T ± LAM resistance) and PC/BCP mutations (Fig. 1A). Although again no significant differences were detected on a transcriptional level (Fig. 1B), the intracellular viral replication, measured as HBV progeny DNA formation, was significantly increased in the presence of BCP mutations for the rtA194T mutants (Fig. 1C). The BCP-rtA194T clone even restored its intracellular replication to the WT level. For PC mutants, significantly higher progeny DNA was found for the rtA194T+rtL180M/rtM204V construct (Fig. 1C). However, combined PC-polymerase mutants did not fully restore the intracellular replication capacity to the WT rate.
Furthermore, the presence of the PC or the BCP mutations was able to significantly enhance the release of mature HBV virions of rtA194T and rtA194T+rtL180M/rtM204V constructs (Fig. 2A). Five days after transient transfection, no differences in HBV copy numbers could be detected in the supernatant of the constructs with combined mutations (PC/BCP + polymerase) as compared with HBV WT. Of course, HBV vectors with the PC mutations failed to secrete HBeAg, whereas the BCP mutants had significantly reduced HBeAg concentrations in the supernatant as compared with HBV WT (Fig. 2B).
rtA194T Mutants Are Resistant to Tenofovir But Remain Susceptible to Telbivudine and Entecavir.
It has remained controversial whether the rtA194T mutation truly confers resistance against tenofovir, because in vitro phenotypical assays showed variable results between different laboratories.13, 16 Therefore, we tested the ability of WT and rtA194T-containing constructs to replicate in the presence of increasing concentrations of tenofovir by quantifying HBV Progeny DNA. For WT, 22.7 μmol of tenofovir reduced the replicative level to 50% (Fig. 3A). The presence of rtA194T increased the EC50 to above 110 μmol, irrespective of other concomitant mutations such as LAM resistance or PC/BCP. This corresponded to a fivefold to sixfold decreased susceptibility to tenofovir (Fig. 3B). LAM-resistant clones (rtL180M/rtM204V) without rtA194T remained susceptible to tenofovir (not shown).
We next tested the potency of other antivirals to reduce HBV replication in WT or rtA194T constructs. LAM did not impair the replication level of rtA194T clones at a concentration that efficiently inhibited WT HBV (Fig. 3C), whereas all rtA194T constructs remained susceptible to telbivudine (Fig. 3D) and entecavir (Fig. 3E). These data demonstrate that the rtA194T HBV polymerase mutation confers partial resistance against tenofovir and lamivudine, but can still be efficiently inhibited by telbivudine or entecavir.
The potent antiviral activity against WT and LAM-resistant strains as well as the low incidence of resistance make tenofovir an appealing drug for the treatment of chronic HBV infections, especially in pretreated patients, or for long-term administration.2, 6, 10, 11 However, tenofovir resistance has been described in two HBV-HIV–co-infected patients and linked to the rtA194T mutation in the HBV polymerase. We first addressed the impact of the rtA194T mutation on the HBV replication efficacy. As evidenced by reduced formation of intracellular HBV–progeny DNA and reduced release of virions into the supernatant, the rtA194T polymerase mutation, alone or in conjunction with LAM resistance, impaired viral replication. This is in general accordance with other polymerase mutants, for example, conferring resistance against LAM or entecavir, in which the polymerase activity is impaired in mutated viral strains, most likely because of sterical alterations of the reverse transcriptase motif.4, 20–22
The impaired replication efficacies of rtA194T mutants may possibly also explain the low incidence in (tenofovir-treated) patients. The HBV polymerase is a reverse transcriptase and lacks proof-reading function, so that mutant viral genomes are frequently found.4 However, it either requires endogenous (for example, replication efficiency) or exogenous (for example, antivirals) pressures to select mutated quasispecies. The favorable kinetics of tenofovir with a strong initial HBV DNA suppression may therefore prevent resistant strains with a reduced replicative capacity from taking over the majority of the virus population in a given patient taking relatively high doses of a potent inhibitor such as tenofovir. Longer observation periods in larger cohorts of patients that are treated with the recently approved tenofovir will ultimately reveal the true relevance of the rtA194T mutation in clinical practice.
It is important to note that also two or more mutations in different regions can occur in the HBV genome.4 HBeAg-negative chronic hepatitis B related to PC or BCP mutations has become the predominant clinical presentation in patients with HBeAg-negative chronic HBV infection in many parts of the world.23, 24 Patients with HBeAg-negative chronic HBV infection generally require long-term administration of nucleoside or nucleotide analogs, rendering them at high risk for developing drug resistance.25 We therefore also tested the replication efficacy of HBV constructs with the rtA194T (±LAM resistance) and coexistent PC or BCP mutations. Interestingly, PC mutations increased the intracellular progeny DNA levels of rtA194T mutants and restored the HBV copy number in the supernatant to WT levels, whereas the BCP mutations even showed a stronger effect enhancing the intracellular replication and a similar effect on virion release of the rtA194T constructs. Of course, BCP mutations diminished and PC abolished secretion of HBeAg in the respective strains.
These results are in full agreement with findings from LAM-resistant mutants, in which also PC and BCP mutations enhance the replication of the drug-resistant polymerase mutants.14, 15 For the PC mutants, two mechanisms are believed to facilitate HBV replication. First, the precore protein has been reported to inhibit HBV DNA synthesis, and point mutations that abolish proper precore protein expression enhance progeny DNA expression.26–29 Secondly, the epsilon (ϵ) structure involved in viral packaging and initiation of reverse transcription is usually stabilized by the PC mutations at its lower stem because of the A-T base pairing. The ϵ structure is a 5′-end-stem-loop of the viral pregenomic RNA where the HBV polymerase binds and which is essential for the encapsidation of the HBV pregenomic RNA by triggering assembly of core proteins.30, 31 The BCP hotspot mutations A1762T/G1764A, conversely, may increase HBV replication by altered transcription factor binding to the promoter, mutations in the overlapping HBx protein, and changes in the ratio of precore to pregenomic RNA.32–36
The rtA194T was selected in two HIV-HBV–co-infected patients during tenofovir administration.13 In the initial identification of this mutation, the susceptibility of rtA194T and rtA194T+rtL180M/rtM204V to tenofovir was assessed in transiently transfected HepG2 cells in vitro, revealing a 7.6 (rtA194T) and greater than 10-fold (rtA194T+rtL180M/rtM204V) increase in the 50% inhibitory concentrations.13 These findings have been discussed controversially, because a different research group from the manufacturer of tenofovir, Gilead Sciences, reported no resistance to tenofovir by rtA194T clones (fold resistance between 1.5 and 2.4), using also a HepG2 cell system.16 We set up an independent series of experiments comparing all six rtA194T-containing HBV constructs (±LAM-resistance, ±PC/BCP) to WT. The EC50 of tenofovir that we obtained for WT was slightly higher in transiently transfected Huh7 cells than in the prior assays.13, 16 However, these types of EC50 variations between Huh7 and HepG2 cells have already been observed previously with other nucleoside or nucleotide analogs, without affecting the ranking of antiviral potency.37, 38 Constructs with rtA194T, irrespective of additional mutations, consistently demonstrated a fivefold to sevenfold increase in the EC50 values compared with WT. By definition, a fivefold to 10-fold increase in the EC50 confers partial resistance.39 Therefore, our data now clearly demonstrate partial resistance to tenofovir in the presence of rtA194T polymerase mutations. In a three-dimensional model of the HBV reverse transcriptase, introduction of the threonine substitution at position 194 created a side branch, possibly changing the deoxyribonucleotide triphosphate-binding pocket of the beta-sheet and suggesting sterical alterations as the basis for tenofovir resistance.13 However, the fold-resistance in our experiments is lower than initially described in a slightly different experimental model. The fact that rtA194T confers only partial tenofovir resistance together with the reduced replicative capacity of these polymerase mutants possibly explains the low incidence of rtA194T mutations in tenofovir-treated HBV-infected patients in clinical practice.40
Furthermore, which antiviral rescue therapy may be administered in case of selection of tenofovir drug-resistant mutations has not been evaluated.13, 16, 40 Our data suggest that rtA194T reduces susceptibility not only to tenofovir, but also to lamivudine, even without the LAM-resistance mutations rtL180M/rtM204V. This might be related to the sterical changes in the beta-sheet of the HBV reverse transcriptase caused by the threonine substitution.13 However, telbivudine and entecavir remain active against rtA194T mutants, even with concomitant PC and BCP mutations. These data thereby indicate that patients developing tenofovir drug resistance can be effectively treated by telbivudine or entecavir.
Taken together, the HBV rtA194T polymerase mutation reduces viral replication and confers partial resistance to tenofovir. The clinical impact of this mutant, however, is unknown and remains to be elucidated by long-term observation of large cohorts of patients treated with tenofovir. Patients with HBeAg-negative hepatitis B may be at particular risk to select rtA194T mutations, because PC and BCP mutations enhance the reduced replicative capacity of rtA194T mutants. Patients with HBeAg-negative hepatitis B may benefit from close virological monitoring when treated with tenofovir to early detect drug-resistant viral strains. Telbivudine and entecavir are effective treatment options for rtA194T HBV polymerase mutants.
The authors thank Aline Müller for excellent technical assistance.