JOURNAL OF VIROLOGY,0022-538X/01/$04.00�0 DOI: 10.1128/JVI.75.23.11827–11833.2001

Dec. 2001, p. 11827–11833 Vol. 75, No. 23

Copyright © 2001, American Society for Microbiology. All Rights Reserved.

A Single Amino Acid in the Reverse Transcriptase Domain ofHepatitis B Virus Affects Virus Replication Efficiency

XU LIN,1 ZHENG-HONG YUAN,1 LI WU,1† JIAN-PING DING,2 AND YU-MEI WEN1*

Department of Molecular Virology, Medical Center of Fudan University, Shanghai 200032,1 and Shanghai Institute ofBiochemistry and Cell Biology, Chinese Academy of Sciences, Shanghai 200031,2 People’s Republic of China

Received 12 June 2001/Accepted 28 August 2001

To explore functional domains in the hepatitis B virus (HBV) polymerase, two naturally occurring HBVisolates (56 and 2-18) with 98.7% nucleic acid sequence homology but different replication efficiencies werestudied. After transfection into HepG2 cells, HBV DNA isolated from intracellular virus core particles wasmuch higher in 56-transfected cells than in cells transfected with 2-18. The structural basis for the differencein replication efficiency between these two isolates was studied by functional domain gene substitution. Thecomplete polymerase (P) gene and its gene segments coding for the terminal protein (TP), spacer (SP), reversetranscriptase (RT), and RNase H in 2-18 were separately replaced with their counterparts from 56 to constructfull-length chimeric genomes. Cell transfection analysis revealed that substitution of the complete P gene of2-18 with the P gene from 56 slightly enhanced viral replication. The only chimeric genome that regained thehigh replication efficiency of the original 56 isolate was the one with substitution of the RT gene of 2-18 withthat from 56. Within the RT region, amino acid differences between isolates 2-18 and 56 were located atpositions 617 (methionine versus leucine), 652 (serine versus proline), and 682 (valine versus leucine). Pointmutation identified amino acid 652 as being responsible for the difference in replication efficiency. Homologousmodeling studies of the HBV RT domain suggest that the mutation of residue 652 from proline to serine mightaffect the conformation of HBV RT which interacts with the template-primer, leading to impaired polymeraseactivity.

Hepatitis B virus (HBV) infection results in a broad spec-trum of clinical manifestations, including asymptomatic car-riage, acute self-limited hepatitis, chronic hepatitis, and fulmi-nant hepatitis. Variability in host immune response is mainlyresponsible for the diverse clinical manifestations, but viralmutations may also affect the clinical course of HBV infection(11). Enhanced replication efficiency of HBV variants couldlead to increased viral virulence associated with high virus loadin the host, which may influence the response to treatment.Despite containing a DNA genome, HBV replicates via a re-verse transcription process, using the polymerase encoded byits own genome (4). This polymerase, designated P, is com-posed of four domains (18). Starting from the N terminus, theyare (i) the terminal protein (TP), which becomes covalentlylinked to the negative-strand DNA to initiate reverse transcrip-tion; (ii) the spacer, which has not been associated with aspecific function and is tolerant to mutations; (iii) the reversetranscriptase (RT), which contains the YMDD consensus mo-tif that is also found in human immunodeficiency virus type 1(HIV-1) reverse transcriptase; and (iv) RNase H, which digeststhe RNA in RNA-DNA hybrids. Studies have shown that thereplication of the HBV genome is a complex process involvinga priming reaction, translocation of complementary sequences,

reverse transcription, and second-strand DNA synthesis (16,19, 20).

Two main experimental approaches have been used to in-vestigate the functions of HBV polymerase. One is the molec-ular approach, which consists of developing systems for directanalysis of full-length polymerase function in the absence ofviral replication and other viral proteins (2, 3, 13), andtranscomplementation studies between functional segments ofHBV polymerase (14, 15). The other is the clinical approach,which consists of structural and functional analysis of HBV Pgenes in various naturally occurring HBV strains from infectedpatients. A nonreplicative HBV mutant with a single aminoacid change in the TP domain of HBV polymerase has beenreported (5). Different types of HBV splice variants have beendescribed in HBV-infected patients under immunosuppressivetreatment (10), and HBV core internal deletion mutations thatinhibited the replication of wild-type virus have also been de-scribed (22).

In cells, replication of HBV not only requires a functionalactive polymerase, but also depends on host factors that caninteract with various elements in the viral genome and viralproteins. Therefore, the most authentic system in which tostudy HBV polymerase function would involve full-length ge-nome replication in liver-derived cells. The approach of am-plifying full-length HBV genomes from serum or tissues ofHBV-infected patients followed by hepatic cell line transfec-tion (10) provided an excellent platform for studying the func-tional diversity of HBV polymerase from naturally occurringstrains.

In a study of the biological properties of various HBV iso-lates, we have come across two isolates with high nucleotide

* Corresponding author. Mailing address: Department of MolecularVirology, Medical Center of Fudan University (formerly ShanghaiMedical University), 138 Yi Xue Yuan Road, Shanghai 200032, Peo-ple’s Republic of China. Phone: 86-21-64041900, ext. 2116/2523. Fax:86-21-64174578. E-mail: ymwen@shmu.edu.cn.

† Present address: HIV Drug Resistance Program, National CancerInstitute, Frederick, MD 21702.

11827

sequence homology but marked difference in their replicationefficiency (21). In the current study, we used gene segmentsubstitution to generate chimeric HBV genomes and identifiedan amino acid in the RT domain which affected the efficiencyof intracellular replication of HBV.

MATERIALS AND METHODS

Full-length HBV genome cloning. Six HBV DNA-positive serum or liver tissuesamples from chronic hepatitis B patients or hepatocellular carcinoma patientswere collected (14, 16, 2-18, 20, 56, and 97). Full-length HBV genomes wereseparately amplified by PCR, followed by cloning into vector pUC18 via the SstIrestriction site as described by Gunther et al. (10). These six HBV DNA genomeswere sequenced separately (GenBank accession numbers: 2-18, AF100308; 56,AF100309; 14, 16, 20, and 97 have been submitted and accession numbers arepending).

Cell culture and transfection. HepG2 cells were cultured in Dulbecco’s mod-ified Eagle’s medium with penicillin (100 U/ml) and streptomycin (100 �g/ml),supplemented with 10% fetal calf serum. Recombinant plasmid DNA used fortransfection was extracted and purified with the Qiagen Maxiprep kits, and HBVDNA was released from the recombinant plasmids by digestion with SapI at 1 Uof enzyme/�g of DNA for 16 h, followed by extraction and purification. HBVDNA (25 �g) from each clone was used to transfect HepG2 cells in 60-mm platesusing the calcium phosphate precipitation method as reported previously (21).Duplicate plates were used for all samples, and 10 �g of reporter plasmid DNAexpressing secreted alkaline phosphatase (SEAP) was cotransfected into eachculture as an internal control to normalize the transfection efficiency amongplates. Cells were collected separately 72 h after transfection, which has beenshown previously to be the appropriate time for monitoring replication efficiencyand expression of antigens by HBV isolates in this laboratory. Vector pUC18DNA was used as a mock transfection control. Transfection experiments weredone twice, repeated on separate days.

Purification of HBV DNA from intracellular core particles. As reported byGunther et al. (10), with some modifications, cells were washed twice with chilledphosphate-buffered saline buffer and lysed with 600 �l of lysing buffer (10 mMTris-HCl [pH 7.9], 1 mm EDTA, 1% NP-40, 8% sucrose). After centrifugation at12,000 � g for 2 min, cleared supernatants were collected, and 6 �l of 1 Mmagnesium acetate (MgOAc), 12 �l of 5-mg/ml DNase I, and 3 �l of 20-mg/mlRNase A were added, followed by incubation at 37°C for 30 min to digestremaining DNA and RNA. After a quick spin at 12,000 � g for 1 min, super-

natants were collected, 16 �l of 0.5 M EDTA and 130 �l of 35% polyethyleneglycol 8000 in 1.75 M NaCl were added, and the mixture was kept on ice for 1 hto precipitate core particles. After spinning at 9,000 � g for 5 min, pellets wereresuspended in 100 �l of buffer with DNase I (10 mM Tris-HCl [pH 7.9], 6 mMMgOAc, 0.1 �g of DNase I per �g) and incubated at 37°C for 10 min to ensuredigestion of any remaining DNA used for transfection. The resuspended coreparticles were digested with proteinase K at 1 mg/ml in 0.5% sodium dodecylsulfate at 37°C for 5 h. Nucleic acids were extracted with phenol-chloroform (1:1)once and chloroform once and precipitated with ethanol.

Southern blot and HBV DNA hybridization. The amount of HBV DNAextracted from intracellular cores from each plate of transfected cells was nor-malized using the expression of SEAP by enzyme-linked immunosorbent assay asan internal control. After electrophoresis in a 1.2% agarose gel, HBV core DNAwas blotted onto a Hybond N� nylon membrane. Twenty-five nanograms ofcloned full-length HBV DNA (adr subtype) was labeled by random primer with[32P]dCTP and used as the probe for hybridization. The signals were detected byautoradiography and scanned with densitometry for semiquantification of theintensity of signals.

Source of HBV DNA used for construction of chimeric genomes. Two full-length HBV genomes (56 and 2-18) which showed 98.7% nucleotide sequencehomology but high (no. 56) and low (no. 2-18) replication efficiencies were usedfor the construction of chimeric genomes.

Construction of chimeric genomes based on p2-18 with the polymerase gene orits functional domain fragments replaced with counterparts from p56. (i) Searchfor the boundaries of functional regions. The P gene sequences of 2-18 and 56were analyzed for DNA homology by Vector NTI Suite 5.5 AlignX software toidentify boundaries of the functional regions to be replaced. In order to detect asignificant change in the replication efficiency of the chimeric constructs, replace-ment of genes from the high-replication-efficiency isolate into the low-replica-tion-efficiency isolate was preferred. Therefore, p2-18 (a pUC18 backbone con-taining the full-length HBV DNA from strain 2-18 with low replicationefficiency) was used as the acceptor, while p56 with high replication efficiency wasused as the donor. For identification of the roles played by each of the fourfunctional regions of the P gene, putative segments coding for terminal protein,spacer, RT/DNA polymerase, and RNase H in 2-18 were individually replacedwith their counterparts from 56. The schematic diagram of the strategy used forreplacement is shown in Fig. 1.

After the search, because of the limited number of restriction sites available,we adopted fusion PCR combined with enzyme digestion to achieve gene re-placement in most cases.

FIG. 1. Functional region of the 2-18 polymerase gene (open bars), HBV region flanking the polymerase gene of 2-18 (checkered bars),functional region of the 56 polymerase gene (shaded bars), and HBV region flanking the polymerase gene of 56 (stippled bars). 2-18P56, 2-18SP56,2-18TP56, 2-18RNaseH56, and 2-18RT56 are 2-18 chimeric genomes in which the complete P gene or each of the functional regions of the P gene,namely, spacer, TP, RNase H, and RT, was replaced with either the complete P gene or each of its counterparts from 56. 56RT2-18 is a 56 HBVchimeric genome in which the RT region of the P gene was replaced with its 2-18 counterpart.

11828 LIN ET AL. J. VIROL.

(ii) Construction of chimeric genomes. To construct recombinant vector p2-18TP56, in which the TP region of the P gene of 2-18 was replaced with the TPregion from 56, two pairs of primers were used: primer P1 (pUC reverse uni-versal primer), 5�-AGCGGATAACAATTTCACACAGGA-3�, and primer P2,5�-AGACCACCAAATGCCCCTATC-3� (nucleotides [nt] 2299 to 2319); andprimer P5, 5�-GATAGGGGCATTTGGTGGTCT-3� (nt 2319 to 2299, comple-mentary to primer P2), and primer P6, 5�-CTGAGTTGGCTTTGAAGCAGG-3� (nt 2971 to 2948). Primers P1 and P5 were used to amplify the 5�-flankingregion of the P gene of p2-18 DNA, and primers P2 and P6 were used to amplifythe TP region (nt 2309 to 2843) of p56 DNA. The two amplified fragments wereannealed and fused by PCR with primers P1 and P6. The fused fragment wasdigested with EcoRI and BstEII and used to replace the EcoRI-BstEII fragmentof p2-18 (Fig. 2.).

For the construction of p2-18P56, in which the complete P gene of 2-18 wasreplaced with that of 56, the BstEII-RsrII (nt 1573 to 2817) fragment of the Pgene of p56 was recovered and ligated into p2-18P from which the BstEII-RsrIIfragment had been removed (Fig. 2).

For the construction of p2-18SP56, in which the spacer region of 2-18 wasreplaced with the spacer region of 56, the BstEII-SpeI (nt 2817 to 681) fragmentof the P gene of p2-18 was removed and replaced with its counterpart from p56(Fig. 2).

For the construction of p2-18RT56, in which the reverse transcriptase region ofthe P gene of 2-18 was replaced with the same region of 56, two sets of primerswere used: primer P3, 5�-CAAGGTATGTTGCCCGTTTGTCCTC-3� (nt 457 to481); primer P4, 5�-CCTTTACCCCGTTGCTC-3� (nt 1139 to 1158); primer P7,5�-GAGCAACGGGGTAAAGGTTC-3� (nt 1158 to 1139, complementary toprimer P4); and primer P8, 5�-GTTCACGGTGGTCTCCAT-3� (nt 1610 to

1627). Primers P4 and P8 were used to amplify the RNase H region of the P geneof p2-18 DNA, while primers P3 and P7 were used to amplify the RT region (nt854 to 1183) of p56 DNA. The two amplified fragments were annealed and fusedby PCR with primers P3 and P8. The fused fragment was digested with SpeI andRsrII and used to replace the SpeI-RsrII fragment of p2-18 (Fig. 2).

The procedure used to construct p2-18RNaseH56, in which the RNase Hregion of the P gene of 2-18 was replaced with that of 56, was similar to theprocedure for constructing p2-18RT56. However, primers P4 and P8 were used toamplify the RNase H region of the P gene of p2-18 DNA, and primers P3 and P7were used to amplify the RT region (nt 854 to 1183) of p56 DNA.

Construction of chimeric genomes based on p56 with the RT domain replacedwith its counterpart from p2-18. To confirm the critical role played by the RTregion of the P gene, the RT region of 56 was replaced with its counterpart from2-18. The procedure for construction of recombinant p56RT2-18, in which the RTregion of 56 was replaced with that of 2-18, was similar to that for p2-18RNaseH56.

Replacement of amino acid in RT region of 2-18 with the 56 counterpart.Based on the results obtained by substitution of the RT gene in p2-18 with thatfrom p56 (see Results below), three site-directed mutants were generated. Toconstruct recombinant p2-18617Met3Leu at amino acid 617, p2-18652Ser3Pro atamino acid 652, and p2-18682Val3Leu at amino acid 682, PCR-based site-directedmutagenesis was used to create the mutations at the corresponding nucleotides,nt 942, 1047, and 1137, respectively. The sequences of the primers used and theirpositions are shown in Table 1.

Homologous modeling studies of HBV RT. In order to predict the differencein replication efficiency between isolates 56 and 2-18 at the three-dimensionalstructural level, homologous modeling studies of HBV polymerase were based

FIG. 2. p2-18 and p56 are pUC18 recombinant plasmids containing either full-length HBV strain 2-18 or 56 cloned at the SstI site. Thepolymerase gene, ranging from nt 2309 to 1625, consists of four functional regions, TP (terminal protein), spacer, RT/DNA polymerase (DNAP),and RNase H. The 5� border of the TP region, 3� border of TP/5� border of spacer, 3� border of spacer/5� border of RT/DNA polymerase, 3� borderof RT/DNA polymerase/5� border of RNase H, and 3� border of the RNase H region are located at nt 2309, 2843, 135, 1168, and 1625, respectively,and are indicated by vertical bars. Gray boxes indicate consensus regions flanking the borders of functional regions, and the nucleotide positionsare shown. P1 is the pUC18 reverse universal primer, and P2 to P8 are primers located in the consensus regions. P1, P2, P5, and P6 were usedfor construction of the TP chimera of the P gene, whereas P3, P4, P7, and P8 were used for the construction of the RT/DNA polymerase and RNaseH chimeras. The sequences of these primers are given in the text.

VOL. 75, 2001 SINGLE AMINO ACID IN RT AFFECTS HBV REPLICATION 11829

on the work reported by Das et al. (7), but the sequence alignment between HBVRT and HIV RT was modified. A homologous model of HBV RT using HIV-1RT (13), Moloney murine leukemia virus RT (9), and hepatitis C virus polymer-ase (1, 6) as modeling templates was constructed, which covered amino acidresidues 319 to 700, comprising the entire RT domain of HBV polymerase. TheHBV RT was modeled in a conformation similar to that of other viral poly-merases or RTs, with the nucleic acid substrate-binding cleft defined by struc-tural elements from the fingers, palm, and thumb subdomains. The DNA tem-plate-primer was docked into the model based on the crystal structure of theHIV-1 RT/DNA binary complex (8) and HIV-1 RT/DNA/deoxynucleosidetriphosphate ternary complex.

RESULTS

Replication efficiency of six different HBV isolates. The rep-lication efficiency of the six HBV isolates was compared by thesemiquantification of scanned intensity of signals detected byautoradiography. Setting the value for the isolate with thelowest replication efficiency (2-18) at 1, the relative intensitiesof the other isolates, 14, 16, 20, 56, and 97, were 32, 32, 21, 32,and 41, respectively.

Because 56 and 2-18 showed 98.7% nucleotide sequencehomology (21), and restriction sites between these two isolateswere matched, these two isolates were selected to constructchimeric genomes for structural analysis of differences in rep-lication efficiency.

DNA homology and consensus boundary analysis of P genebetween 56 and 2-18. DNA homology analysis (Vector NTISuite 5.5 AlignX software) indicated that there were 15 pre-dicted amino acid differences between the polymerase genes of2-18 and 56 (Table 2).

Consensus regions for both isolates flanking the boundariesof all four functional regions of the polymerase gene are shownin Fig. 2. Three restriction sites within these consensus regionssuitable for cloning were BstEII (nt 2817), SpeI (nt 681), andRsrII (nt 1573). Based on this analysis, procedures to generatechimeric HBV genomes were designed, and five chimeric ge-nomes that replaced either the complete P gene or each of itsfunctional domain gene fragments were successfully con-structed and confirmed by restriction enzyme analysis and se-quencing.

Replication efficiency of 2-18 HBV chimeric genomes withfunctional domain genes replaced with 56 counterparts. Thehybridization signals of HBV DNA isolated from intracellularcore particles in transfected cells were markedly differentamong the five constructs used for transfection (Fig. 3). Ascontrols, 56 showed strong hybridization signals indicating ahigh replication efficiency, while with 2-18 the hybridization

signal could barely be seen, indicating a low replication effi-ciency. When the spacer region of the P gene or the completeP gene of 2-18 was replaced with that of 56, the chimericgenome showed slightly enhanced replication compared to theoriginal 2-18 (ratio of intensity, 5 versus 1). However, the HBVDNA hybridization signal was markedly increased in cellstransfected with the chimeric genome 2-18RT56, in which thegene fragment coding for the reverse transcriptase region of2-18 was replaced with the counterpart from 56 (ratio of in-tensity, 26 versus 1). Other chimeric genome constructs

FIG. 3. 2-18 and 56 are original isolates with low and high rep-licative efficiencies, respectively. 2-18P56, 2-18SP56, 2-18TP56,2-18RNaseH56, and 2-18RT56 are 2-18 HBV chimeric genomes inwhich the full-length P gene and functional regions of the P gene(spacer, TP, RNase H, and RT) were replaced with their counterpartsfrom 56. Intensity of hybridization signal was measured by densitom-etry, and the relative intensity of these samples is indicated at thebottom. Lane M, molecular size standards.

TABLE 1. Sequences and nucleotide positions of primers used forPCR-based site-directed mutagenesis

Primer Sequencea (5�33�) Nucleotidepositions

PMD CGCAAAATACCTATGGGAGTG 1602–1622PMU GTTCACGGTGGTCTCCATGCG 630–650PA617 AAAACACAG�TTGATTTTTTGTACAATATG 921–951PS617 AAATCAAACTGTGTTTTAGGAAACTTCCTG 939–969PA652 TAAAGCAGGATATCCACATTGCGTGAAAG 1027–1057PS652 GTGGATATCCTGCTTTAATGCCTTTATATG 1040–1070PA682 AGGTTCAGATACTGTTTACTTAGAAAGGC 1117–1147PS682 AACAGTATCTGAACCTTTACCCCGTTGCTC 1129–1159

a Mutated nucleotides are underlined.

TABLE 2. Differences in nucleotides and predicted amino acids ofHBV polymerase between isolates 2-18 and 56

P gene portionand amino

acid position

2-18 56

Amino acid Codon Amino acid Codona

TP74 W TGG G 2528GGG110 T ACT I A2637TT139 N AAT D 2723GAT

Spacer219 S AGT R AG2965A246 G CAG R C3044GG261 R CGC G 3089GGC270 S AGC N A3117AC272 S TCC A 3122GCC300 H CAT R C3207GT

RT617 M ATG L 942CTG652 S TCT P 1047CCT682 V TGT L T1137CT

RNase H720 V TGT L T1251CT805 L TTG F TT1508C809 S TCC T 1518ACC

a The position of the altered nucleotide is indicated as a superscript to the leftof the changed base.

11830 LIN ET AL. J. VIROL.

showed no enhanced replication in this study. These resultsindicate that the RT gene of 56 was critical for the high rep-lication efficiency of this isolate.

Replication efficiency of 56 HBV chimeric genome with RTreplaced with its 2-18 counterpart. Conversely, the replicationefficiency of 56RT2-18, in which the RT region of the P gene of56 was replaced with the 2-18 counterpart, was reduced com-pared to 56 (Fig. 4). The ratio of intensity of hybridizationsignals between 56RT2-18 and 56 was 1:3.5, as assayed by den-sitometry.

Replication efficiency of site-directed RT domain mutants.Among three site-directed mutants (2-18RT617, 2-18RT652,and 2-18RT682) in which each of the amino acids at the 617,652, and 682 position of the RT region of 2-18 was replacedwith its 56 counterpart and studied by cell transfection, only2-18RT652 showed enhanced replication efficiency, with an in-tensity of hybridization signal similar to that of 2-18RT56 (Fig.5).

Structural analysis of mutations in HBV RT domain. Com-puter-based structural analysis predicted that residue 617 waslocated at the loop connecting the palm subdomain and thethumb domain and would have no contact with any importantstructural element of the protein and the DNA substrate. Sim-ilarly, residue 682 was predicted to be located in the middle ofthe third �-helix of the subdomain and oriented toward theouter surface of the protein, which would not have significantimpact on the interaction between RT and DNA. In contrast,residue 652 was predicted to be located at the connecting loopbetween two �-helices. Alteration of residue 652 from prolineto serine might affect the precise conformation of the twohelices and their interactions with the DNA template-primer,which might impair the polymerase activity (Fig. 6).

DISCUSSION

Functional domains of the HBV P gene product have beenidentified by mutational analysis and transient expression incells by Radziwill et al. (18). Alignment of the amino acidsequences of the three enzymatic domains (TP, RT, and

RNase H) with similar enzymes from other retroviruses andhepadnaviruses revealed conserved and consensus regionswhich were considered important motifs. To date, all attemptsat expression and production of large amounts of functionallyactive forms of HBV polymerase in animal cells or prokaryoticcells have been unsuccessful. Therefore, detailed functionalanalysis of HBV P protein has been hampered.

Since HBV replication has a unique strategy and molecularmechanism, it is speculated that there may be specific func-tional domains in HBV P protein which are not shared byother viruses. To explore novel functional domains or residuesof HBV P protein, we used the chimeric genomic cell trans-fection analysis method. Among six full-length naturally occur-ring HBV isolates, we selected isolates 2-18 and 56 becausethey had highly homologous nucleotide sequences, making itpossible to perform the domain-swapping experiments. Studieswere designed to use only the HBV genome for cell transfec-tion, without any exogenous plasmid sequences which mightinterfere with the replication efficiency of isolates. In addition,substitutions of the P gene or the gene fragments coding forthe putative functional domains were all carried out betweentwo naturally occurring HBV genomes. Therefore, the resultsobtained in this cell transfection study would correlate with thereplication efficiency of the virus in vivo and could reveal pos-sible novel functional residues in HBV P protein.

In this study we have shown that when amino acid 652 waschanged from serine to proline, HBV replication efficiency wasenhanced, and replication efficiency was decreased by the con-verse change. In the four other isolates (14, 16, 20, and 97) thatwe studied, the amino acid at 652 was always proline. Hase-gawa et al. (12) have reported enhanced replication of an HBVmutant associated with an epidemic of fulminant hepatitis. Theamino acid at 652 in that strain of HBV was proline. A searchof GenBank for HBV RT sequences showed that the consen-sus residue at 652 was proline. The counterpart of 652 is also

FIG. 4. 56RT2-18 is an HBV chimeric genome in which the RTregion of the P gene was replaced with its 2-18 counterpart. Intensityof hybridization signal was measured by densitometry, and the relativeintensity of samples is indicated at the bottom. Lane M, molecular sizestandards.

FIG. 5. 2-18617Met3Leu, 2-18652Ser3Pro, and 2-18682Val3Leu are 2-18HBV chimeric genomes in which the nucleotide sequences coding foramino acids at positions 617, 652, and 682 within the RT region wereeach mutated to change the amino acid to the corresponding aminoacid of isolate 56. Intensity of hybridization signal was measured bydensitometry. The intensity of signal of 2-18RT652Ser3Pro was similar tothat of 2-18RT56, in which the RT region of 2-18 was replaced with its56 counterpart. Lane M, molecular size standards.

VOL. 75, 2001 SINGLE AMINO ACID IN RT AFFECTS HBV REPLICATION 11831

proline in HIV-1 RT. Therefore, one may consider 56 as rep-resenting a wild-type virus and 2-18 as a mutant with a substi-tution of proline by serine at amino acid 652. This substitutionresulted in decreased replication efficiency.

Interestingly, this substitution only regulated the relativereplication efficiency of the virus strain, in contrast to a muta-tion in TP, which abolished virus replication (5). It is intriguingthat a virus strain with decreased replication efficiency couldpersist during natural infection. Since host immune responsesare highly relevant in HBV infections, this virus strain couldpersist in a host with decreased immune responses. Unfortu-nately, it was not possible to study the immune responses in thepatient.

In this study, when the complete P gene of 56 replaced thecomplete P gene of 2-18, although enhanced replication wasobserved, it was not as marked as that when only the RT genefragment was replaced. It is tempting to speculate that theremight be replication-inhibitory functional domains in frag-ments of the HBV P gene which counteracted the replication-enhancing effect of amino acid 652 in the RT region. Besides,in this study, substitution of the spacer region from 56 for thatof 2-18 also slightly enhanced replication efficiency. Recently,Landford et al. (15) reported that a portion of the spacerdomain was required for both TP and RT function and resi-dues 300 to 334 were identified as the minimal portion of thespacer domain. A substitution was found at residue 300 in thespacer region between 2-18 and 56 (Table 2, from histidine toarginine), which could be associated with the slightly enhancedreplication efficiency.

The mechanism of the change in replication efficiencycaused by changing the residue at amino acid position 652 isnot clear. However, computer modeling suggested that residue652 was located at the connecting loop between two �-helicesof the thumb subdomain, equivalent to �-helices H and I inHIV-1 RT. In HIV, �-helices H and I interact extensively with

the DNA template-primer substrate (8). It was suggested thatthese two helices form part of a translocation track that con-trols and modulates the relative movements of polymerase.During the polymerization reaction, the template-primer hasto translocate after incorporation of each nucleotide. The poly-merase should have the flexibility to allow the translocation ofthe template-primer but still maintain the precise positioningof the template-primer relative to the polymerase active site. Itis speculated that in the HBV RT, proline has a rigid confor-mation with less flexibility, and this rigidity might be necessaryto constrain the relative conformation and/or positions of thetwo helices. Mutation from proline to serine at residue 652could disrupt this constraint, allowing more flexibility and thusaffecting the polymerase activity.

Although HBV is recognized as a noncytopathic virus, en-hanced viral replication could contribute to the pathogenesisof liver disease in several aspects. First, enhanced virus repli-cation can lead to an increased number of hepatocytes beinginfected, resulting in more severe hepatocellular injury medi-ated by host immune responses. Second, antiviral-drug-resis-tant mutants could be selected in higher frequency amongpatients with a high virus load. Third, a higher potential couldarise for emergence of viral mutants with altered biologicalcharacteristics and for higher incidences of random integrationof viral DNA into the host genome. In general, enhancedreplication has been associated with increased virulence withother viruses. Therefore, studies on the regulatory mechanismsof the replication efficiency of HBV are of importance.

The polymerase of HBV has been the essential target fordevelopment of anti-HBV drugs, and most studies have fo-cused on the common and conserved functional domains at theactive site of RT shared with other known viruses. Our findingof the replication-enhancing or -decreasing effect of aminoacid 652 in the RT region provides the clue that there could beother functional domains outside the A-E functional domains

FIG. 6. Ribbon diagram showing the three-dimensional model of the RT domain of HBV polymerase (residues 319 to 700). HBV RT iscomposed of three subdomains, fingers (319 to 400 and 451 to 515, in blue), palm (401 to 450 and 516 to 610, in red), and thumb (611 to 700, ingreen). A docked template-primer double-stranded DNA is shown as a ribbon, with the template in magenta and the primer in cyan. The threecatalytically essential aspartic acid residues, Asp429, Asp551, and Asp552, are shown with side chains. The locations of the mutations are indicatedwith gold balls.

11832 LIN ET AL. J. VIROL.

predicted by amino acid alignment (17). Novel functional do-mains revealed in naturally occurring HBV strains could betterresemble the situation in patients and thus could become newtargets for development of antiviral drugs. Whether changesfrom proline to other amino acids at 652 would also affect virusreplication is currently under study. Research aimed at eluci-dating the functional domains at and around this residue is alsobeing conducted.

ACKNOWLEDGMENTS

This work was supported by Chinese State Basic Research Founda-tion Grant G1999054105. X.L. is a Ph.D. student supported by theChinese Ministry of Education.

We are grateful to Christopher Burrell for reviewing the paper. Theexcellent technical help of Xin Yao and Zhang-Mei Ma is much ap-preciated.

REFERENCES

1. Ago, H., T. Adachi, A. Yoshida, M. Yamamoto, N. Habuka, K. Yatsunami,and M. Miyano. 1999. Crystal structure of the RNA-dependent RNA poly-merase of hepatitis C virus. Struct. Fold Des. 7:1417–1426.

2. Ayola, B., P. Kanda, and R. E. Lanford. 1993. High level expression andphosphorylation of hepatitis B virus polymerase in insect cells with recom-binant baculoviruses. Virology 194:370–373.

3. Bartenschlager, R., C. Kuhn, and H. Schaller. 1991. Expression of theP-protein of the human hepatitis B virus in a vaccinia virus system anddetection of the nucleocapsid-associated P-gene product by radiolabelling atnewly introduced phosphorylation sites. Nucleic Acids Res. 20:195–202.

4. Bavand, M., M. Feitelson, and O. Laub. 1989. The hepatitis B virus-associ-ated reverse transcriptase is encoded by viral pol gene. J. Virol. 63:1019–1021.

5. Blum, H. E., E. Galun, and T. J. Liang. 1991. Naturally occurring missensemutation in the polymerase gene terminating hepatitis B virus replication.J. Virol. 65:1836–1842.

6. Bressanelli, S., L. Tomei, A. Roussel, I. Incitti, R. L. Vitale, M. Mathieu, R.De Francesco, and F. A. Rey. 1999. Crystal structure of the RNA-dependentRNA polymerase of hepatitis C virus. Proc. Natl. Acad. Sci. USA 96:13034–13039.

7. Das, K., X. Xiong, H. Yang, C. E. Westland, C. S. Gibbs, S. G. Sarafianos,and E. Arnold. 2001. Molecular modeling and biochemical characterizationreveal the mechanism of hepatitis B virus polymerase resistance to lamivu-dine (3TC) and emtricitabine (FTC). J. Virol. 75:4771–4779.

8. Ding, J., K. Das, Y. Hsiou, S. G. Sarafianos, A. D. Clark, A. Jacobo-Molina,

C. Tantillo, S. H. Hughes, and E. Arnold. 1998. Structure and functionalimplications of the polymerase active site region in a complex of HIV-1 RTwith a double-stranded DNA template-primer and an antibody Fab fragmentat 2.8 Å resolution. J. Mol. Biol. 284:1095–1111.

9. Georgiadis, M. M., S. M. Jessen, C. M. Ogata, A. Telesnitsky, S. P. Goff, andW. A. Hendrickson. 1995. Mechanistic implications from the structure of acatalytic fragment of Moloney murine leukemia virus reverse transcriptase.Structure 3:879–892.

10. Gunther, S., B. C. Li, S. Miska, D. H. Kruger, H. Meisel, and H. Will. 1995.A novel method for efficient amplification of whole hepatitis B virus ge-nomes permits rapid functional analysis and reveals deletion mutants inimmunosuppressed patients. J. Virol. 69:5437–5444.

11. Gunther, S., L. Fisher, J. Pult, M. Sterneck, and H. Will. 1999. Naturallyoccurring variants of hepatitis B virus. Adv. Virus Res. 22:25–137.

12. Hasegawa, K., J. Huang, S. A. Rogers, H. E. Blum, and T. J. Liang. 1994.Enhanced replication of a hepatitis B virus mutant associated with an epi-demic of fulminant hepatitis. J. Virol. 68:1651–1659.

13. Kim, S. S., H. J. Shin, and H. M. Rho. 2000. Expression of stable hepatitisB viral polymerase associated with GRP94 in E.coli. Arch. Virol. 145:1305–1320.

14. Landford, R. E., L. Notvall, and B. Beames. 1995. Nucleotide priming andreverse transcriptase activity of hepatitis B virus polymerase expressed ininsect cells. J. Virol. 69:4431–4439.

15. Landford, R. E., Y. H. Kim, H. Lee, L. Notvall, and B. Beames. 1999.Mapping of the hepatitis B virus reverse transcriptase TP and RT domainsby transcomplementation for nucleotide priming and by protein-protein in-teraction. J. Virol. 73:1885–1893.

16. Miller, R. H., P. L. Marion, and W. S. Robinson. 1984. Hepatitis B viralDNA-RNA hybrid molecules in particles from infected liver are converted toviral DNA molecules during an endogenous DNA polymerase reaction.Virology 139:64–72.

17. Poch, O., I. Sauvager, M. Delarue, and N. Tordo. 1989. Identification of fourconserved motifs among the RNA-dependent polymerase encoding ele-ments. EMBO. J. 12:3867–3874.

18. Radziwill, G., W. Tucker, and H. Schaller. 1990. Mutational analysis of thehepatitis virus P gene product: domain structure and RNase H activity.J. Virol. 64:613–620.

19. Wang, G. H., and C. Seeger. 1992. The reverse transcriptase of hepatitis Bvirus acts as a protein primer for viral DNA synthesis. EMBO J. 71:663–670.

20. Will, H., W. Reiser, T. Weimer, E. Pfaff, M. Buscher, R. Sprengel, R. Cat-taneo, and H. Schaller. 1987. Replication strategy of human hepatitis B virus.J. Virol. 61:904–911.

21. Wu, L., Z. H. Yuan, L. F. He, W. L. Jiang, S. X. Qiu, and Y. M. Wen. 2000.Functional comparison of two full-length genome of hepatitis B virus iso-lated from two chronic hepatitis B virus patients. Chin. J. Virol. 16:16–22.

22. Yuan, T. T., M. H. Lin, D. S. Chen, and C. Shih. 1998. A defective interfer-ence-like phenomenon of human hepatitis B virus in chronic carriers. J. Vi-rol. 72:578–584.

VOL. 75, 2001 SINGLE AMINO ACID IN RT AFFECTS HBV REPLICATION 11833