Search tips
Search criteria 


Logo of jvirolPermissionsJournals.ASM.orgJournalJV ArticleJournal InfoAuthorsReviewers
J Virol. May 2001; 75(10): 4771–4779.
PMCID: PMC114232
Molecular Modeling and Biochemical Characterization Reveal the Mechanism of Hepatitis B Virus Polymerase Resistance to Lamivudine (3TC) and Emtricitabine (FTC)
Kalyan Das,1 Xiaofeng Xiong,2 Huiling Yang,2 Christopher E. Westland,2 Craig S. Gibbs,2 Stefan G. Sarafianos,1 and Edward Arnold1*
Center for Advanced Biotechnology and Medicine, Department of Chemistry, Rutgers University, Piscataway, New Jersey,1 and Gilead Sciences, Foster City, California2
*Corresponding author. Mailing address: CABM and Rutgers University, 679 Hoes Ln., Piscataway, NJ 08854. Phone: (732) 235-5323. Fax: (732) 235-5788. E-mail: arnold/at/
Received November 8, 2000; Accepted February 19, 2001.
Success in treating hepatitis B virus (HBV) infection with nucleoside analog drugs like lamivudine is limited by the emergence of drug-resistant viral strains upon prolonged therapy. The predominant lamivudine resistance mutations in HBV-infected patients are Met552IIe and Met552Val (Met552Ile/Val), frequently in association with a second mutation, Leu528Met. The effects of Leu528Met, Met552Ile, and Met552Val mutations on the binding of HBV polymerase inhibitors and the natural substrate dCTP were evaluated using an in vitro HBV polymerase assay. Susceptibility to lamivudine triphosphate (3TCTP), emtricitabine triphosphate (FTCTP), adefovir diphosphate, penciclovir triphosphate, and lobucavir triphosphate was assessed by determination of inhibition constants (Ki). Recognition of the natural substrate, dCTP, was assessed by determination of Km values. The results from the in vitro studies were as follows: (i) dCTP substrate binding was largely unaffected by the mutations, with Km changing moderately, only in a range of 0.6 to 2.6-fold; (ii) Kis for 3TCTP and FTCTP against Met552Ile/Val mutant HBV polymerases were increased 8- to 30-fold; and (iii) the Leu528Met mutation had a modest effect on direct binding of these β-l-oxathiolane ring-containing nucleotide analogs. A three-dimensional homology model of the catalytic core of HBV polymerase was constructed via extrapolation from retroviral reverse transcriptase structures. Molecular modeling studies using the HBV polymerase homology model suggested that steric hindrance between the mutant amino acid side chain and lamivudine or emtricitabine could account for the resistance phenotype. Specifically, steric conflict between the Cγ2-methyl group of Ile or Val at position 552 in HBV polymerase and the sulfur atom in the oxathiolane ring (common to both β-l-nucleoside analogs lamivudine and emtricitabine) is proposed to account for the resistance observed upon Met552Ile/Val mutation. The effects of the Leu528Met mutation, which also occurs near the HBV polymerase active site, appeared to be less direct, potentially involving rearrangement of the deoxynucleoside triphosphate-binding pocket residues. These modeling results suggest that nucleotide analogs that are β-d-enantiomers, that have the sulfur replaced by a smaller atom, or that have modified or acyclic ring systems may retain activity against lamivudine-resistant mutants, consistent with the observed susceptibility of these mutants to adefovir, lobucavir, and penciclovir in vitro and adefovir in vivo.
Hepatitis B virus (HBV) infection is among the top 10 viral infections, affecting an estimated 300 million people worldwide and over 1.5 million in the United States alone (10, 24). Chronic HBV infection can lead to cirrhosis, hepatocellular carcinoma, and liver failure. Treatment of chronically HBV-infected patients with alpha interferon (28) is limited by side effects, incomplete efficacy, restriction to patients with compensated disease, and the requirement for parenteral administration (8, 37). HBV, a hepadnavirus, replicates through an intermediate reverse transcription step carried out by the viral polymerase (19, 33), which is functionally and structurally related to human immunodeficiency virus (HIV) reverse transcriptase (RT). Some of the nucleoside analogs developed to treat HIV infection are highly potent against HBV infection (4, 5) at concentrations below cytotoxic thresholds. Treatment of chronically HBV-infected patients with nucleoside or nucleotide analogs (Fig. (Fig.1),1), like lamivudine (3TC), emtricitabine (FTC), famciclovir (the prodrug of penciclovir [PCV]), adefovir dipivoxil (ADV [also called PMEA]), and lobucavir (LBV), leads to significant decreases in serum virus levels (26). Treatment with the nucleoside or nucleotide analogs has shown immediate clinical benefits such as reduced viral load, suppression of progression of liver disease, and induction of immunological clearance or seroconversion (6, 18). Drug-resistant strains of HBV containing specific polymerase mutations emerge upon prolonged 3TC treatment (14, 23; H. Fontaine, V. Thiers, and S. Pol, Letter, Ann. Intern. Med. 131:716–717, 1999) and are the primary cause of treatment failure. Treatment of HBV-infected patients with 3TC in phase III clinical studies showed a sequential increase in appearance of genotypic resistance in HBV patients: 24% in the first year, 42% in the second year, 52% in the third year, and 67% in the fourth year (N. W. Y. Leung, C. L. Lai, J. Dienstag, G. Schiff, J. Heathcote, M. Atkins, C. Marr, and W. C. Maddrey, presented at the Management of Hepatitis B Meeting, 8 to 10 September 2000).
FIG. 1
FIG. 1
Chemical structures of dCTP, 3TCTP, FTCTP, ADVDP, LBVTP, and PCVTP.
As with other nucleotide polymerases, the triphosphates of the nucleotide substrates or their analog inhibitors are the catalytically active forms for polymerization by HBV polymerase, and the polymerization reaction has been shown to be Mg2+ ion dependent (34). Two of three catalytically essential aspartic acid residues are part of the highly conserved YMDD motif at the active site of HBV polymerase and its close viral relatives, including HIV type 1 (HIV-1) RT. The most common 3TC resistance mutations, Met552Ile and Met552Val (Met552Ile/Val), appear at the Met (M) position in the YMDD motif of the HBV polymerase, analogous to the lamivudine resistance mutations Met184Val/Ile of HIV-1 RT. In a departure from the pattern observed with HIV, 3TC-resistant HBV frequently contains a second polymerase mutation, Leu528Met. Met552Ile/Val mutations alone and in combination with the Leu528Met mutation confer a high degree of resistance to 3TC triphosphate (3TCTP) in vitro (Table (Table1).1). On the other hand, ADV has been reported to be active against 3TC-resistant HBV in vitro and in vivo (29, 30, 35). These data indicate complementary drug resistance profiles for 3TC and ADV against HBV, suggesting a potential advantage for combination therapy in treating chronic HBV infection where the emergence of resistance to either agent may be suppressed.
Inhibition of HBV polymerases containing prototypic 3TC resistance mutations
Knowledge of the structure of HBV polymerase would be valuable for understanding the molecular basis of many of its properties, including mechanisms of polymerization, inhibition, and drug resistance, and for interpretation of clinical and biochemical data. Attempts to determine the structure of HBV polymerase by various research groups have not yet been successful, as they have been limited by failure to obtain sufficient amounts of highly purified active protein.
The work presented here includes a molecular modeling study of HBV polymerase based on available retroviral RT structures. The validity of the model developed in the present study is supported by its ability to explain some of the key biochemical data. The inhibition potencies of 3TCTP, FTCTP, ADV diphosphate (ADVDP), PCVTP, and LBVTP were evaluated and compared with the Km for dCTP in in vitro enzyme assays for wild type HBV polymerase and a Leu528Met mutant, Met552Ile/Val mutants, and Leu528Met+Met552Ile/Val mutants. The results were analyzed at the atomic level using the modeled three-dimensional structure of HBV polymerase. dCTP, 3TCTP, FTCTP, and ADVDP were docked into the modeled enzyme so that the differential effects of Met552Ile/Val and Leu528Met mutations on different nucleotide analogs could be examined. Possible effects of these mutations on some other potent nucleotide inhibitors are addressed. Understanding of the roles of these drug resistance mutations might be helpful in achieving the broader goal of developing more effective antiviral strategies for the treatment of chronic hepatitis B.
Enzyme assay. (i) Inhibition of HBV polymerase.
Recombinant HBV polymerases were overexpressed and partially purified from insect cells as previously described (35). HBV polymerase activity was monitored by measurement of the incorporation of α-32P-labeled deoxynucleoside triphosphate (dNTP) into acid-precipitable products. Assays were performed in 40 μl of a solution containing 100 mM Tris (pH 7.5), 10 mM MgCl2, 0.6 U of RNasin/ml, 5% glycerol, 0.2 μg of activated calf thymus DNA/μl, 100 μM unlabeled dNTPs (e.g., dATP, dGTP, and dTTP), various concentrations of a α-32P-labeled dNTP (~500 Ci/mmol), and various concentrations of inhibitors. α-32P-labeled dATP was used for the determination of the inhibition constants for ADVDP α-32P-labeled dGTP was used for PCVTP and LBVTP, and α-32P-labeled dCTP was used for 3TCTP and FTCTP. HBV polymerase (5 μl, ~0.1 μg) was added to start the reaction. Aliquots (12 μl) were taken at various time points between 0 and 20 min and transferred onto 3MM paper disks. The paper disks were washed three times in 5% trichloroacetic acid plus 1% sodium pyrophosphate and once in 95% ethanol. The incorporated radioactivity was measured in a Beckman scintillation counter.
(ii) Enzyme kinetics.
Kinetic constants were determined by fitting the initial rates to Lineweaver-Burk plots based on the algorithms described by Cleland (3).
Molecular modeling. (i) Sequence alignments.
The protein segment from position 354 to 694 of the polypeptide chain translated from the HBV pol gene (Fig. (Fig.2)2) is responsible for the RT activity of HBV. A model was generated for amino acid residues 325 to 699 of the polypeptide chain, covering the entire polymerase/RT region. The amino acid sequence identity is significant among various HBV strains in the polymerase/RT region, but there is relatively low sequence homology with other viral RTs and polymerases. The nearest relatives of HBV polymerase, in terms of sequence homology, for which crystal structures are available are HIV-1 RT and murine leukemia virus (MuLV) RT, both with less than 25% sequence identity. Our sequence alignment (Fig. (Fig.2),2), however, indicates that the functionally important amino acid residues are highly conserved among the polymerases of HBV, HIV-1, and MuLV. The sequence alignments allowed us to derive a three-dimensional structural model for HBV polymerase from the known structures of HIV-1 RT and Moloney MuLV (MMLV) RT.
FIG. 2
FIG. 2
Schematic representation of the HBV pol gene, an HBV polymerase homology model (amino acids 325 to 699), and the HBV polymerase/HIV-1 RT sequence alignments used in constructing the model. The HBV polymerase is shown as a ribbon diagram (2) with the fingers (more ...)
(ii) Homology modeling, docking of substrates, and structure analysis.
Crystal structures of HIV-1 RT (7, 11, 12) and MuLV RT (9) were used as templates in the modeling of the HBV polymerase domain. Multiple initial models for the HBV polymerase were obtained using the amino acid sequence alignment-based three-dimensional structure-generating program MODELLER-4 (31) and using the crystal structures of HIV-1 RT (PDB codes: 1 RTD, 2HMI, and 1DLO) and of MULV RT (PDB code: 1MML) as templates. The protein conformation of the model obtained by using the HIV-1 RT-DNA-dNTP complex structure (1RTD) was used as the initial scaffold for the HBV polymerase model. The less conserved regions, insertions, and side chains were built by manual modeling using the computer program O (15) and its reference to databases of known main-chain conformations and preferred side-chain rotamers. The other three models, derived from the HIV-1 RT-DNA-Fab complex (2HMI), unliganded HIV-1 RT (1DLO), and MULV RT (1MML), were used as additional guides in building the molecular model of HBV polymerase. The secondary structure for the model constructed as described above agreed very well with a sequence-based secondary structure assignment for the region using the program Homologue (21). Buried side chains were manually oriented to have favorable interactions with each other. The final model was minimized using the molecular graphics and simulation program SYBYL, version 6.3 (Tripos, Inc.), and the quality of the geometrical parameters of the model was evaluated by PROCHECK (20); the overall G factor was −0.22, indicating that the molecular geometry is stereochemically reasonable. Amino acid residue Met552 was modeled as part of an unusual type II′ turn, as observed in other RT structures. The main-chain conformations for all other amino acids were within the favored regions of the Ramachandran plot. The drug resistance mutations Met552Ile/Val and Leu528Met were modeled so that (i) their side chains occupied positions that had minimal steric conflict with neighboring amino acids; (ii) their side-chain torsion angles fell within statistically favored ranges; and (iii) the side chain of Val/Ile552 had an orientation similar to that of Ile184 in the Met184Ile mutant HIV-1 RT-DNA structure (32). The individual dNTP substrates and analog inhibitors were initially modeled using as a guide the conformation of dTTP in the structure of the HIV-1 RT-DNA-dTTP complex (Fig. (Fig.3).3). After energy minimization, the substrates and nucleotide analog inhibitors were then docked into the active sites of the wild-type and mutant HBV polymerase models using the program SYBYL. The validity of the final model was further supported by the proximity of the positions of some of the important drug resistance mutation sites with respect to substrates (Table (Table2).2).
FIG. 3
FIG. 3
Electrostatic-potential surface diagrams of the modeled HBV polymerase (left) and of the HIV-1 RT-DNA-dNTP structure (right) plotted using the program GRASP (27). Regions in red and blue are charged negatively and positively, respectively. The locations (more ...)
Structurally conserved amino acid residues in HIV-1 RT and HBV polymerase
In vitro assay.
The effects of Leu528Met and Met552Val/Ile mutations on the binding of the natural substrate dCTP to HBV polymerase were evaluated by comparing the Km values for dCTP for the mutant enzymes to those for the wild-type enzyme. The Km values (Table (Table1)1) for the substrate dCTP were 0.64- to 2.6-fold relative to those for the mutants of HBV polymerase, indicating that the Leu528Met and Met552Val/Ile mutations do not significantly affect the binding of dCTP to HBV polymerase. In order to identify the desirable structural features for anti-HBV agents to be used to treat or prevent the emergence of 3TC resistance, five nucleotide analogs with different structural characteristics were tested for their sensitivities against 3TC-resistant HBV. ADVDP and PCVTP are acyclic nucleotide analogs, 3TCTP and FTCTP are l-configuration nucleotide analogs with a β-oxathiolane ring, and LBVTP bears a cyclobutyl replacement for the sugar moiety in the natural nucleotides. All of the structures are shown in Fig. Fig.1.1. Susceptibility to these inhibitors was assessed by determining Ki in in vitro polymerase assays using recombinant wild-type and mutant HBV polymerases.
The inhibition constants (Ki) for 3TCTP, FTCTP, ADVDP, PCVTP, and LBVTP against the Leu528Met mutant and Met552Val/Ile mutants are listed in Table Table1.1. Our in vitro enzyme assay results showed that a single mutation, M552V or M552I, in the YMDD motif caused significant resistance to 3TCTP and FTCTP, with the inhibition constants increased 8- to 30-fold compared to that for the wild-type HBV polymerase. The acyclic nucleotides ADVDP and PCVTP and the d-nucleotide LBVTP, however, remained active against all 3TC-resistant mutant enzymes, with the inhibition constants increased less than 3.1-fold (36). A moderate (2.6-fold) increase in Ki for 3TCTP and FTCTP against Leu528Met HBV polymerase is indicative of a minimal effect of the single Leu528Met mutation on the nucleosides. Similar observations were reported for 3TC resistance in various independent studies (17, 22).
Overview of the model.
The final model (amino acids 325 to 699) of HBV polymerase is shown in Fig. Fig.2.2. Like HIV-1 RT, the modeled HBV polymerase has fingers (325 to 403 and 469 to 519), palm (404 to 440 and 520 to 613), and thumb (614 to 699) subdomains. The catalytic triad residues Asp431, Asp553, and Asp554 of HBV polymerase correspond to Asp110, Asp185, and Asp186 in HIV-1 RT. Many of the key protein-DNA interactions and protein-dNTP interactions are conserved (Table (Table2)2) between the HIV-1 RT structure and the modeled HBV polymerase. It is intriguing that HBV polymerase probably contains an element analogous to the “primer grip” of HIV-1 RT (13), including residues Met598 and Gly599, which are equivalent to the conserved residues Met230 and Gly231 of HIV-1 RT. Some major differences between the HBV polymerase model and HIV-1 RT structure include four modeled disulfide bonds in HBV polymerase compared to none in HIV-1 RT, and a larger fingers region in HBV polymerase than in HIV-1 RT. The differences in the fingers region between HBV polymerase and HIV-1 RT may involve the different primers used by retroviral RTs and HBV polymerase. Differences in the palm and thumb regions of the HBV polymerase model and the HIV-1 RT structure are relatively small but significant. The DNA-binding cleft in the HBV polymerase model (Fig. (Fig.3)3) is well defined and more positively charged than the DNA-binding cleft in the HIV-1 RT-DNA-dTTP complex structure (12). The dNTP-binding region, between the palm and fingers subdomain, appears to be partially filled by additional amino acids in the HBV model, with the tip of its fingers touching the base of its thumb. This part of the HBV polymerase model corresponds to the β3-β4 region of the HIV-1 RT structure that contains some of the key HIV drug resistance mutation sites, where mutations can confer resistance to nucleoside drugs like zidovudine (AZT), dideoxyinosine (ddI), dideoxycytosine (ddC), and stavudine (d4T). These antiviral drugs are not very potent against HBV. Some of the HIV-1 RT mutations, conferring resistance to the above drugs, are the natural amino acids in the wild-type HBV polymerase. The nucleoside resistance mutations Asp67Asn and Leu74Val of HIV-1 RT correspond to Asn381 and Val391, respectively, of the HBV polymerase model. Two multidrug (AZT+d4T+ddI/ddC) HIV resistance mutations, Gln151Met and the insertion of three amino acids after Ser69, are found in wild-type HBV. Positions 151 and 69 of HIV-1 RT correspond to the positions of Met519 and Pro382, respectively, in the modeled HBV polymerase.
Positions of dNTP and nucleotide analog drugs.
In the modeled HBV polymerase, the relative positions of the α-, β-, and γ-phosphates of dCTP (and its analog inhibitors) with respect to the catalytic triad were assumed to occupy positions very similar to those of the dNTP in the crystal structure of the HIV-1 RT-DNA-dNTP complex (12). The sugar and the base moieties of the dCTP were oriented in their energy-minimized conformations, which are constrained to base-pair with the first DNA template overhang. The YMDD motif of the modeled enzyme interacts mostly with the sugar-phosphate portion of the docked dCTP. The Met552 side chain points towards the deoxyribose ring of dCTP. The position and orientation of this amino acid correspond to those of Met184 in HIV-1 RT. Leu528 of HBV polymerase, positionally equivalent to Phe160 of HIV-1 RT, is part of a helix, and its side chain points to the space between Met552 and Phe436. The aromatic ring of Phe436, positionally equivalent to Tyr115 in HIV-1 RT, stacks almost in parallel with the sugar ring of the substrate. Unlike Met552, Leu528 of HBV polymerase does not have close interactions with the dNTP substrate. Upon mutation, however, residue 528 has the potential to affect the binding of dNTP (or its analog inhibitor) by perturbing the side chains of surrounding amino acids, particularly of Phe436 and Met552.
Effects of the Met552Ile/Val mutation.
The Met552Ile/Val mutations in HBV polymerase, in both the presence and the absence of the Leu528Met mutation, conferred resistance to 3TC and FTC, as indicated by significant increases of Ki in in vitro polymerase assays (Table (Table11).
In our molecular modeling studies, the docked dCTP substrate was accommodated in a stereochemically feasible position and orientation (Fig. (Fig.4)4) in the wild-type HBV polymerase model. Residue Met552, which is part of the conserved YMDD motif in RTs, is adjacent to the bound nucleotide substrate. The accessible surface area (Fig. (Fig.4)4) of the YMDD region of HBV polymerase is complementary to the molecular surface of the dCTP. The Met552Ile/Val mutation limits the side-chain flexibility by introducing a branch, methyl group (Cγ2), to its Cβ atom. The most favorable conformation for a valine or an isoleucine at position 552 is with the Cγ2 atom pointing towards the bound dNTP. The side chain of Ile184 in the crystal structure of Met184Ile mutant HIV-1 RT-DNA (32), corresponding to position 552 of the HBV polymerase model, also had a similar conformation. Molecular modeling of the Met552Val mutation (Fig. (Fig.4)4) showed a decreased space between the protein and the substrate. Consistent with the small changes observed in kinetic constants, the Met552Ile/Val mutation does not appear to interfere significantly with the proposed binding of the dNTP in its catalytically favorable conformation, as shown in Fig. Fig.4.4.
FIG. 4
FIG. 4
The YMDD region of the modeled HBV polymerase with a docked dCTP substrate. Amino acids Met552 and Leu528 are mutated to confer resistance to 3TC and FTC. The orange molecular surface (left) corresponds to the deoxyribose of the docked dCTP. The green (more ...)
The nucleotide analog 3TCTP has an oxathiolane ring in a β-l configuration, replacing the β-d-deoxyribose ring of dCTP. Docking of 3TCTP into the active site of the wild-type HBV polymerase model, with its triphosphate and base oriented as in dCTP, showed (Fig. (Fig.5)5) that the sulfur in the oxathiolane ring points towards the site of mutation (position 552). As a consequence of the β-l configuration of the oxathiolane ring, which is inverted with respect to the β-d configuration of the deoxyribose ring of a natural substrate, the docked 3TCTP occupies a larger volume extending towards the side chain of Met552. The Met552Ile/Val mutation adds a methyl group at the Cγ2 position of the mutated amino acid, pointing toward the sulfur atom of the oxathiolane ring of 3TCTP (Fig. (Fig.5).5). Our molecular modeling studies suggest that steric hindrance between the Cγ2-methyl group of Ile/Val552 and the oxathiolane ring of 3TCTP may result from binding of 3TCTP to the Met552Ile/Val mutant HBV polymerase. A previous molecular modeling study of HBV polymerase (1), based on less detailed information about HIV-1 RT structure, concluded that the Met552Ile/Val mutation leads to decreased protein-inhibitor interactions. Subsequent biochemical and structural data, however, strongly support our proposed mechanism involving steric hindrance with the Cγ2-methyl group of Ile/Val552. This mechanism may also apply more broadly to other l-nucleoside analogs, like FTC, with anti-HBV activity.
FIG. 5
FIG. 5
Binding of 3TCTP to wild-type (left) and Met552Val mutant (right) HBV polymerase. Molecular modeling suggests that steric hindrance (right), between 3TCTP and the mutated amino acid, Val552, is the primary cause of 3TCTP resistance. This steric conflict (more ...)
Effects of Leu528Met mutation.
In the wild-type HBV polymerase model, Leu528 occupies a position between the side chains of Phe436 and Met552. Although the Leu528 side chain points toward the sugar ring of the docked dCTP, a shortest distance of about 4.5 Å between them suggests that interactions between dCTP and Leu528 in wild-type HBV polymerase are likely to be weak or indirect. The Leu528Met mutation introduces a longer, yet more flexible, side chain. As a consequence of this mutation, the side chain of Met528 may interact directly with the docked 3TCTP or FTCTP but its greater flexibility, unlike the Met552Ile/Val mutation, would disfavor steric conflict of Met528 with the nucleoside inhibitor. This hypothesis is in agreement with our results from in vitro studies on the effects of the Leu528Met mutation (Table (Table1)1) showing a modest increase in Ki for both 3TCTP and FTCTP of only 2.6-fold at the maximum.
A possible role of Leu528Met mutation would be a conformational perturbation of the dNTP-binding region, in particular the amino acids Phe436 and Met552. Phe436, whose HIV-1 RT equivalent is Tyr115, is positioned below and stacked with the sugar ring of dCTP or its analog inhibitors. As discussed above, Met552Ile/Val would introduce a rigid side chain in the vicinity of the 3TCTP oxathiolane ring. Our in vitro assays showed a higher degree of resistance of 3TCTP and FTCTP obtained with the double mutations Met552Ile/Val+Leu528Met than with the single Met552Ile/Val mutation (Table (Table1).1). A structural interpretation of this enhanced effect of the double mutation is the indirect involvement of Leu528Met mutation by reorienting the side chains of its surrounding amino acids, in particular of Phe436 and Ile/Val552. Such a rearrangement might also be responsible for compensating for the reduction of polymerase activity by the Met552Ile/Val mutation (22, 25).
Effects of Met552Ile/Val on other nucleoside inhibitors.
The nucleoside and nucleotide analog inhibitors ADV and PCV show complementary in vivo drug resistance profiles (16, 29, 35) with 3TC. An acyclic chain adds torsional flexibility to ADV and PCV compared to 3TC or FTC, both of which contain a five-membered oxathiolane ring (Fig. (Fig.1).1). In addition, the chain connecting the base and the α-phosphonate group is shorter in ADV than in the oxathiolane analogs. This disparity in length (and volume) is illustrated in comparisons of molecular models of wild-type and Met552Val+Leu528Met mutant HBV polymerases in complex with double-stranded DNA and ADVDP. Our molecular modeling studies predict that smaller acyclic nucleotide analogs can be accommodated more effectively than the bulkier oxathiolanes in a more constrained and “crowded” dNTP-binding pocket containing the Met552Val+Leu528Met mutations. This prediction is consistent with the resistance data that show that combinations of Met552Val/Ile and Leu528Met mutations confer only 0.8- to 2.3-fold resistance to ADVDP (35) and 0.9- to 1.8-fold resistance to LBVTP. Docking of LBVTP onto the modeled HBV polymerase fragment suggested that the interaction between the inhibitor and the mutating amino acids is qualitatively similar to that for ADVDP (Fig. (Fig.6).6). Furthermore, this model can explain why 3TC-resistant HBV mutants still retain susceptibility to ADV, which has thus far not been reported to select for resistance mutations in HBV (5, 29).
FIG. 6
FIG. 6
Both wild-type (left) and Met552Val+Leu528Met mutant (right) HBV polymerase appear to have no steric conflict with a docked ADVDP.
Summary and implications for drug design.
Our molecular modeling studies of HBV polymerase provide a plausible structural basis for the effects of Met552Ile/Val and Leu528Met mutations on the susceptibility of the enzyme to 3TCTP and FTCTP. Steric conflict between the β-branched mutant amino acid side chains and the sulfur atom of the β-l-oxathiolane ring of the inhibitor is proposed as a structural explanation for 3TC resistance by the Met552Ile/Val mutation in HBV polymerase. This explanation is in agreement with the proposed effects of the YMDD mutation Met184Ile in HIV-1 RT based on the comparison of the Met184Ile mutant HIV-1 RT-DNA (32) with wild-type HIV-1 RT-DNA-dTTP (12) and wild-type HIV-1 RT-DNA (7) structures. The Leu528Met mutation is proposed to have an indirect effect on substrate and inhibitor binding, potentially via rearrangement of its surrounding amino acids, particularly Phe436 and Met/Ile/Val552, although increased interaction between the side chain of Met528 and an incoming nucleotide cannot be ruled out. This mutation was reported to compensate for the decreased polymerization by Met552Val (22, 25). The presence of amino acids corresponding to drug resistance mutations in HIV-1 RT at the equivalent positions in the wild-type HBV polymerase model may explain the natural resistance of HBV to AZT and dideoxynucleoside inhibitors.
The structural explanation of the effects of the Met552Ile/Val mutation on inhibition by 3TCTP and FTCTP suggests that suitable modifications at the sugar ring of a dNTP analog could lead to the design of inhibitors with increased potency against the YMDD mutant strain. A similar suggestion was made for the design of HIV-1 RT inhibitors with reduced resistance due to a Met184Ile/Val mutation in the YMDD motif (32). Our studies predicted stereochemically feasible binding of ADVDP at the active sites of both wild-type and Met552Ile/Val mutant HBV polymerase models, which is consistent with earlier favorable reports of the potency of ADV against Met552Ile mutant HBV strains. Also, nucleoside analog inhibitors with a smaller sugar ring (e.g., LBV) or with different sugar ring conformational preferences might lead to development of additional nucleotide analogs that would be effective against Met552Ile/Val mutant HBV strains. Differences in the mode of binding of nucleotide inhibitors to the dNTP-binding pocket of the HBV polymerase, as predicted from the current modeling studies, may account for the complementary drug resistance profiles seen for different nucleotide analogs and support the concept that combination therapy against HBV may be more effective than monotherapy through mutual suppression of the emergence of drug-resistant variants.
We gratefully acknowledge Gilead Sciences for support of this work and an NIH MERIT award (R29 AI27690) from the National Institute of Allergy and Infectious Diseases to Edward Arnold for support of HIV-1 RT structural studies.
We thank Stephen Hughes for helpful discussions.
1. Allen M I, Deslauriers M, Andrews C W, Tipples G A, Walters K A, Tyrrell D L, Brown N, Condreay L D. Identification and characterization of mutations in hepatitis B virus resistant to lamivudine. Lamivudine Clinical Investigation Group. Hepatology. 1998;27:1670–1677. [PubMed]
2. Carson M. Ribbons 2 0. J Appl Crystallogr. 1991;24:958–961.
3. Cleland W W. Statistical analysis of enzyme kinetic data. Methods Enzymol. 1979;63:103–138. [PubMed]
4. Colacino J M, Staschke K A. The identification and development of antiviral agents for the treatment of chronic hepatitis B virus infection. Prog Drug Res. 1998;50:259–322. [PubMed]
5. De Clercq E. Perspectives for the treatment of hepatitis B virus infections. Int J Antimicrob Agents. 1999;12:81–95. [PubMed]
6. Dienstag J L, Perrillo R P, Schiff E R, Bartholomew M, Vicary C, Rubin M. A preliminary trial of lamivudine for chronic hepatitis B infection. N Engl J Med. 1995;333:1657–1661. [PubMed]
7. Ding J, Das K, Hsiou Y, Sarafianos S G, Clark A D, Jr, Jacobo-Molina A, Tantillo C, Hughes S H, Arnold E. Structure and functional implications of the polymerase active site region in a complex of HIV-1 RT with a double-stranded DNA template-primer and an antibody Fab fragment at 2.8 Å resolution. J Mol Biol. 1998;284:1095–1111. [PubMed]
8. Foster G R, Thomas H C. Effects of hepatitis B and hepatitis C virus replication on the actions of interferon. Antivir Res. 1994;24:131–136. [PubMed]
9. Georgiadis M M, Jessen S M, Ogata C M, Telesnitsky A, Goff S P, Hendrickson W A. Mechanistic implications from the structure of a catalytic fragment of Moloney murine leukemia virus reverse transcriptase. Structure. 1995;3:879–892. [PubMed]
10. Hoofnagle J H. Therapy of viral hepatitis. Digestion. 1998;59:563–578. [PubMed]
11. Hsiou Y, Ding J, Das K, Clark A D, Jr, Hughes S H, Arnold E. Structure of unliganded HIV-1 reverse transcriptase at 2.7 Å resolution: implications of conformational changes for polymerization and inhibition mechanisms. Structure. 1996;4:853–860. [PubMed]
12. Huang H, Chopra R, Verdine G L, Harrison S C. Structure of a covalently trapped catalytic complex of HIV-1 reverse transcriptase: implications for drug resistance. Science. 1998;282:1669–1675. [PubMed]
13. Jacobo-Molina A, Ding J, Nanni R G, Clark A D, Jr, Lu X, Tantillo C, Williams R L, Kamer G, Ferris A L, Clark P, Hizi A, Hughes S H, Arnold E. Crystal structure of human immunodeficiency virus type 1 reverse transcriptase complexed with double-stranded DNA at 3.0 A resolution shows bent DNA. Proc Natl Acad Sci USA. 1993;90:6320–6324. [PubMed]
14. Jardi R, Buti M, Rodriguez-Frias F, Cotrina M, Costa X, Pascual C, Esteban R, Guardia J. Rapid detection of lamivudine-resistant hepatitis B virus polymerase gene variants. J Virol Methods. 1999;83:181–187. [PubMed]
15. Kleywegt G J, Jones T A. Where freedom is given, liberties are taken. Structure. 1995;3:535–540. [PubMed]
16. Korba B E, Cote P, Hornbuckle W, Schinazi R, Gerin J L, Tennant B C. Enhanced antiviral benefit of combination therapy with lamivudine and famciclovir against WHV replication in chronic WHV carrier woodchucks. Antivir Res. 2000;45:19–32. [PubMed]
17. Ladner S K, Miller T J, Otto M J, King R W. The hepatitis B virus M539V polymerase variation responsible for 3TC resistance also confers cross-resistance to other nucleoside analogues. Antivir Chem Chemother. 1998;9:65–72. [PubMed]
18. Lai C L, Chien R N, Leung N W, Chang T T, Guan R, Tai D I, Ng K Y, Wu P C, Dent J C, Barber J, Stephenson S L, Gray D F. A one-year trial of lamivudine for chronic hepatitis B. Asia Hepatitis Lamivudine Study Group. N Engl J Med. 1998;339:61–68. [PubMed]
19. Lanford R E, Notvall L, Beames B. Nucleotide priming and reverse transcriptase activity of hepatitis B virus polymerase expressed in insect cells. J Virol. 1995;69:4431–4439. [PMC free article] [PubMed]
20. Laskowski R A, Rullmannn J A, MacArthur M W, Kaptein R, Thornton J M. AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR. J Biomol NMR. 1996;8:477–486. [PubMed]
21. Levin J M, Garnier J. Improvements in a secondary structure prediction method based on a search for local sequence homologies and its use as a model building tool. Biochim Biophys Acta. 1988;955:283–295. [PubMed]
22. Ling R, Harrison T J. Functional analysis of mutations conferring lamivudine resistance on hepatitis B virus. J Gen Virol. 1999;80:601–606. [PubMed]
23. Ling R, Mutimer D, Ahmed M, Boxall E H, Elias E, Dusheiko G M, Harrison T J. Selection of mutations in the hepatitis B virus polymerase during therapy of transplant recipients with lamivudine. Hepatology. 1996;24:711–713. [PubMed]
24. Malik A H, Lee W M. Chronic hepatitis B virus infection: treatment strategies for the next millennium. Ann Intern Med. 2000;132:723–731. [PubMed]
25. Melegari M, Scaglioni P P, Wands J R. Hepatitis B virus mutants associated with 3TC and famciclovir administration are replication defective. Hepatology. 1998;27:628–633. [PubMed]
26. Mutimer D. Hepatitis B virus antiviral drug resistance: from the laboratory to the patient. Antivir Ther. 1998;3:243–246. [PubMed]
27. Nicholls A, Sharp K A, Honig B. Protein folding and association: insights from the interfacial and thermodynamic properties of hydrocarbons. Proteins. 1991;11:281–296. [PubMed]
28. Niederau C, Heintges T, Lange S, Goldmann G, Niederau C M, Mohr L, Haussinger D. Long-term follow-up of HBeAg-positive patients treated with interferon alfa for chronic hepatitis B. N Engl J Med. 1996;334:1422–1427. [PubMed]
29. Ono-Nita S K, Kato N, Shiratori Y, Lan K H, Yoshida H, Carrilho F J, Omata M. Susceptibility of lamivudine-resistant hepatitis B virus to other reverse transcriptase inhibitors. J Clin Invest. 1999;103:1635–1640. [PMC free article] [PubMed]
30. Perrillo R, Schiff E, Yoshida E, Statler A, Hirsch K, Wright T, Gutfreund K, Lamy P, Murray A. Adefovir dipivoxil for the treatment of lamivudine-resistant hepatitis B mutants. Hepatology. 2000;32:129–134. [PubMed]
31. Sanchez R, Sali A. Evaluation of comparative protein structure modeling by MODELLER-3. Proteins. 1997;1997(Suppl.):50–58. [PubMed]
32. Sarafianos S G, Das K, Clark A D, Jr, Ding J, Boyer P L, Hughes S H, Arnold E. Lamivudine (3TC) resistance in HIV-1 reverse transcriptase involves steric hindrance with beta-branched amino acids. Proc Natl Acad Sci USA. 1999;96:10027–10032. [PubMed]
33. Summers J, Mason W S. Replication of the genome of a hepatitis B-like virus by reverse transcription of an RNA intermediate. Cell. 1982;29:403–415. [PubMed]
34. Urban M, McMillan D J, Canning G, Newell A, Brown E, Mills J S, Jupp R. In vitro activity of hepatitis B virus polymerase: requirement for distinct metal ions and the viral epsilon stem-loop. J Gen Virol. 1998;79:1121–1131. [PubMed]
35. Xiong X, Flores C, Yang H, Toole J J, Gibbs C S. Mutations in hepatitis B DNA polymerase associated with resistance to lamivudine do not confer resistance to adefovir in vitro. Hepatology. 1998;28:1669–1673. [PubMed]
36. Xiong X, Yang H, Westland C E, Zou R, Gibbs C S. In vitro evaluation of hepatitis B virus polymerase mutations associated with famciclovir resistance. Hepatology. 2000;31:219–224. [PubMed]
37. Zignego A L, Fontana R, Puliti S, Barbagli S, Monti M, Careccia G, Giannelli F, Giannini C, Buzzelli G, Brunetto M R, Bonino F, Gentilini P. Impaired response to alpha interferon in patients with an inapparent hepatitis B and hepatitis C virus coinfection. Arch Virol. 1997;142:535–544. [PubMed]
Articles from Journal of Virology are provided here courtesy of
American Society for Microbiology (ASM)