Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Xu, Z. Articles by Di Bisceglie, A. M. Search for Related Content PubMed PubMed Citation Articles by Xu, Z. Articles by Di Bisceglie, A. M.

 Previous Article  |  Next Article 

Journal of Virology, October 2008, p. 9417-9424, Vol. 82, No. 19
0022-538X/08/$08.00+0     doi:10.1128/JVI.00896-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Comparative Analysis of Nearly Full-Length Hepatitis C Virus Quasispecies from Patients Experiencing Viral Breakthrough during Antiviral Therapy: Clustered Mutations in Three Functional Genes, E2, NS2, and NS5a

Zekuan Xu,^1,3 Xiaofeng Fan,^1,2* Yanjuan Xu,¹ and Adrian M. Di Bisceglie^1,2*

Division of Gastroenterology and Hepatology, Department of Internal Medicine, Saint Louis University School of Medicine, St. Louis, Missouri 63104,¹ Saint Louis University Liver Center, Saint Louis University School of Medicine, St. Louis, Missouri 63104,² Department of Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, China³

Received 29 April 2008/ Accepted 18 July 2008

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Viral breakthrough is a recognized response pattern to interferon-based antiviral therapy in patients with chronic hepatitis C virus (HCV) infection. The emergence of drug-resistant HCV quasispecies variants is assumed to be a major mechanism responsible for viral breakthrough. By using a long reverse transcription-PCR protocol recently developed in our lab, multiple nearly full-length HCV quasispecies variants were generated from 9.1-kb amplicons at both the baseline and breakthrough points in two patients experiencing viral breakthrough. Comparative analyses of consensus dominant quasispecies variants revealed that most mutations, occurring at the time of breakthrough, involved three functional viral genes, E2, NS2, and NS5a. Interestingly, similar mutation patterns were also observed in minor quasispecies variants at baseline. These three genes had the highest values of average amino acid complexity at the HCV 1a population level. No single amino acids were identified to be associated with viral breakthrough. Taken together, at the near-full-length HCV genome level, our data suggested that viral breakthrough might be attributed to the selection of minor quasispecies variants at the baseline with or without additional mutations during antiviral therapy. Furthermore, the pattern for mutation clustering indicated potential mutation linkage among E2, NS2, and NS5a due to structural or functional relatedness in HCV replication.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Hepatitis C virus (HCV), a single-stranded, positive-sense RNA virus within the Flaviviridae family, currently infects about 3% of the world's population. Upon HCV infection, up to 80% individuals will establish persistent infection with the potential to progress into end-stage liver disease, such as cirrhosis and hepatocellular carcinoma. Antiviral therapy is commonly used for patients with chronic HCV infection. Patients undergoing optimal antiviral therapy, 48-week peginterferon and ribavirin combination treatment, show a variety of therapeutic effects, including sustained virological response (SVR), nonresponse, breakthrough, and relapse (13). Current therapeutic regimens are associated with SVR rates of approximately 45% for HCV genotype 1 and 80% for genotypes 2 and 3 (13). Such remarkable differences of SVR rates among HCV genotypes suggest that HCV itself must be a critical determinant for antiviral therapy. Although variations in multiple HCV regions have been suggested to be associated with therapeutic resistance, such as the double-stranded RNA-dependent protein kinase-alpha subunit of eukaryotic initiation factor 2 phosphorylation homology domain (29), so-called alpha interferon sensitivity-determining region (9) for interferon, and HCV polymerase (NS5b) for ribavirin (32, 33), none of these studies reached a definitive conclusion (31). Both interferon and ribavirin have broad-spectrum antiviral activity via creation of nonspecific antiviral status rather than direct action with viral genomes. Consequently, to explore viral mechanisms mediating treatment resistance, it is necessary to perform any genetic studies at the full-length HCV genome level rather than focusing on small HCV domains, which may partially explain the plethora of controversial data with regard to the roles of HCV genetic variation in antiviral therapy (31).

By using long reverse transcription (RT)-PCR (LRP) and a cloning technique recently developed in our laboratory (12, 34), we have conducted comparative genetic analyses at the near-full-length HCV quasispecies (QS) level in two patients experiencing viral breakthrough during antiviral therapy in which HCV rebounded before the end of the therapeutic schedule (48 weeks). Compared to other response patterns, the breakthrough, in spite of its lower frequency, represents a distinct clinical situation in which viral factors may be a major explanation for viral reappearance. Thus, sequential genetic analysis at the near-full-length QS level will allow us to examine several debatable issues, including the existence of putative interferon-resistant domain(s) in the HCV genome.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patients. The two patients studied, LIV02 and LIV03, were enrolled in a study comparing the efficiency of peginterferon therapy with or without ribavirin (6). Both were infected with HCV genotype 1a and were treated with peginterferon alfa-2a and ribavirin for 48 weeks. After initiation of treatment, serum HCV RNA in both patients became undetectable and then reappeared at weeks 48 and 36, respectively, while the patients were still on treatment (Fig. 1). Serum samples, collected at multiple time points, were stored at –80°C. Samples collected at baseline, at the time of the breakthrough, and at follow-up were analyzed in this study (Fig. 1).

View larger version (18K): [in this window] [in a new window]

FIG. 1. Dynamics of HCV viral load during antiviral therapy in patients LIV02 (A) and LIV03 (B). The serum HCV RNA level was quantified at multiple time points using the Roche Amplicor HCV Monitor assay, version 2.0 (lower limit of quantification, 600 IU/ml). Serum samples used in this study were collected at the indicated time points. The area marked in gray shows the period where HCV RNA was undetectable.

Generation of HCV QS profiles by regular RT-PCR and cloning. The HCV QS profiles at the baseline were determined for these two patients in our previous study based on a 1.38-kb amplicon covering most of the HCV E1 and E2 domains (4). In this study, we have further evaluated viral heterogeneity at multiple time points by focusing on a 496-bp domain spanning the HCV HVR1 domain. The protocol was essentially the same as what we described previously except for the PCR primers (4). In brief, serum RNA was reverse transcribed with 200 U of Moloney murine leukemia virus reverse transcriptase (Promega) and reverse primer RR1, 5'-TSCGGA ARC ART CMG TGG GGC A-3' (nucleotides [nt] 2082 to 2103), followed by the first-round PCR with primers RR1 and CF11, 5'-CGG CGT GAA CTA TGC AAC AGG3' (nt 821 to 841). The second-round PCR was performed using primers RAF2, 5'-GTA CTG AAT TCA ACT GTT CAC CTT CTC TCC CA-3' (forward, nt 1207 to 1227), and 6AR1, 5'-ACT CGA AGC TTT CGG GAC AGC CTG AAG AGT TG-3' (reverse, nt 1682 to 1702). Primer numbering is according to HCV strain H77 (GenBank accession no. AF009606). Degenerate bases were matched with standard International Union of Pure and Applied Chemistry codes. The resulting PCR product, 496 bp in length without restriction sites, was first digested with EcoRI and HindIII and then gel purified by using a QIAEX II gel extraction kit (Qiagen), followed by ligation into EcoRI/HindIII-predigested pUC19 vector. Escherichia coli DH5 cells (Invitrogen) were used for transformation and recovery of recombinant clones. Approximately 15 clones for each sample were randomly picked, minicultured, and sequenced with the ABI Prism dye terminator cycle sequencing ready reaction kit using an ABI 373A automated sequencer (Applied Biosystems, Foster City, CA).

Generation of HCV QS profiles by long RT-PCR and cloning. Long RT-PCR was performed for the amplification of 9.1-kb near-full-length HCV genome in samples 4701, 5309, 4681, and 5159 (Fig. 1) as previously described (12). Briefly, RNA extracted from 280 µl of serum was reverse transcribed with RT matrix consisting of 1x SuperScript III buffer, 10 mM dithiothreitol, 1 µM QR2 (reverse primer), 2 mM deoxynucleoside triphosphates (Invitrogen), 20 U of RNasin RNase inhibitor (Promega), 200 U of SuperScript III, and 5 U of avian myeloblastosis virus reverse transcriptase (Promega). Five µl of the RT reaction mixture was applied for the first round of PCR with primers WF33 and QR2 under the use of 2 U of rTth XL DNA polymerase (Applied Biosystems). An aliquot of 2 µl of the first-round PCR product was used for the second-round amplification with primers WF5 and WR55. The 9.1-kb amplicon was digested with restriction enzymes PacI and FseI, followed by ligation into a pClone vector. The ligation product was electroporated into E. coli DH10B cells (Invitrogen). Positive recombinant clones were identified by either the digestion or partial sequencing of both ends of the insert.

After the confirmation of positive recombinant clones, all were partially sequenced with primer 1a-1031 (Table 1). HCV HVR1 QS profiles were derived from sequenced domains. Next, these HVR1 QS profiles were compared with those derived from the 496-bp amplicon, and both dominant and minor QS variants for each sample were determined. Three dominant clones with the same HVR1 sequence were selected for the full-length sequencing, which was done through a gene-walking approach with 17 primers (Table 1).

View this table: [in this window] [in a new window]

TABLE 1. Primers for sequencing nearly full-length HCV inserts in recombinant clonesa

Genetic analysis. Raw sequences were edited with the programs ClustalW (16) and BioEdit (14), for which HCV strain H77 (AF009606) served as the reference sequence. Primer sequences were removed prior to the genetic analysis. The mean genetic distance (d) was calculated with the Kimura two-parameter method (all sites) (18) in the Molecular Evolutionary Genetics Analysis software package (MEGA; version 4.0) (20). Phylogenetic trees were constructed using the neighbor-joining method (27) with a bootstrap test implanted in MEGA. We also retrieved 129 nonredundant full-length HCV genotype 1a sequences from the Los Alamos National Laboratories HCV database (19). Using ClustalW, all 129 full-length HCV genotype 1a sequences were aligned. The sequence alignment is available upon request. Next, genetic complexity was estimated at each amino acid position for this alignment by measuring Shannon entropy with the program BioEdit (14). Average genetic complexity for each HCV functional gene was then calculated.

Statistical tests. Student's t test was used to analyze differences between mean values for genetic parameters when data were normally distributed. Nonparametric tests were used to evaluate samples for which normal distributions were not present.

Nucleotide sequence accession numbers. A total of 12 nearly full-length HCV QS sequences generated in this study have been deposited in GenBank under accession numbers EU677247 through EU677258.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
LRP performance and cloning efficiency. Using an optimized LRP protocol, 9.1-kb viral fragments were amplified successfully from patient serum samples, further confirming the robustness of this technology (12, 34). In patient LIV02, the digestion of LRP product with PacI and FseI resulted in three fragments in agarose gel, suggesting the existence of restriction sites within this viral genome. We then replaced these restriction sites with AsiSI and XbaI for primers WF5 and WR55, respectively. A single band appeared in agarose gel after the digestion. The pClone vector was assembled with all these restriction sites not found in nine full-length HCV genotype 1a isolates (12). Thus, the combination of different restriction sites should cover the cloning of LRP product derived from most HCV genotype 1a isolates. The positive rate of recombinant clones varied among serum samples; thus, serum samples 5309 and 5159 had a near-100% efficiency, while only 35% of recombinant clones had correct inserts in sample 4681 (Fig. 2).

View larger version (80K): [in this window] [in a new window]

FIG. 2. Restriction maps of recombinant clones containing 9.1-kb HCV inserts. About 20 clones were randomly selected from each sample. Clones from patient LIV02 were digested with PacI only, which generated about 12-kb linear fragments in correct recombinants. Both PacI and FseI were used to digest clones from patient LIV03, which resulted in two fragments, a 9.1-kb HCV insert and a 3-kb pClone vector. Apparently, the rate for positive clones varied among samples, ranging from 35% (4681) to 100% (5159). M, 1-kb DNA ladder (Fisher).

Dynamics of HCV HVR1 QS over the therapeutic course. By cloning and sequencing recombinant clones derived from the 496-bp RT-PCR product, we determined HCV HVR1 QS profiles at multiple time points for each patient. Both patients showed similar evolutionary patterns as shown in phylogenetic trees (Fig. 3). At the time of breakthrough, the viral population was swept out by a single HCV HVR1 QS variant, which was replaced with new QS variants during the follow-up. Interestingly, after the withdrawal of antiviral therapy, HCV evolved back toward the viral population at baseline (Fig. 3). For both patients, HCV displayed an extremely homogeneous population at the breakthrough points, with genetic distances for HVR1 (d_HVR1) of 0.003 and 0.019 for samples 4701 and 4681, respectively. However, genetic heterogeneity was gradually recovered, and to even higher levels at follow-up points than at baseline, with d_HVR1 values of 0.158 versus 0.097, respectively, for samples 5661 and 4701 in patient LIV02 and 0.301 versus 0.194 for samples 5442 and 4681 in patient LIV03.

View larger version (12K): [in this window] [in a new window]

FIG. 3. Phylogenetic representation of HCV HVR1 quasispecies evolution in patients LIV02 (A) and LIV03 (B). The trees were constructed using the neighbor-joining method implanted in the program MEGA, version 3.1. The bootstrap analysis was done with 100 replicates and is shown at the major branch. Time points are shown as different symbols. Filled circles, baseline (day –7; samples 4681 and 4701); open squares, breakthrough points (samples 5309 and 5109); filled triangles, 4 weeks after drug withdrawal (samples 5388 and 5442); open triangles, 24 weeks after drug withdrawal (sample 5661). At the time of breakthrough, HCV displayed homogenous populations (shift 1), followed by an evolution toward a QS population at the baseline (shift 2). In sample 4681, we screened 160 clones using HVR1-specific PCR, which identified two clones (open diamonds) that were clustered into the same branch as breakthrough QS variants.

Comparison of HCV HVR1 QS derived from different amplicons. Samples 4701, 5309, 4681, and 5159 were used for the LRP procedure. For these samples, we compared HCV HVR1 QS profiles derived from either a 496-bp or 9.1-kb fragment. As shown in Fig. 4, considerable consistency was observed with regard to the recovery of dominant QS variants from both types of amplicons. Samples 5309 and 5159, collected at breakthrough points, showed essentially the same HVR1 QS profiles derived from either 496-bp or 9.1-kb fragments. Samples 4701 and 4681, collected at the baseline, had a similar HVR1 QS pattern, one dominant QS variant accompanied by minor QS variants (Fig. 4). While HVR1 QS profiles derived from the 496-bp and 9.1-kb fragments recovered different minor viral variants, the same dominant QS variants were identified (Fig. 4). Thus, the LRP procedure generated similar HCV HVR1 QS profiles as those derived from small fragments in these two patients.

View larger version (29K): [in this window] [in a new window]

FIG. 4. Comparison of QS profiles derived from both the 496-bp and 9.1-kb amplicons. Only HVR1 sequences are displayed at the amino acid level. Considerable consistency was observed with regard to the recovery of dominant QS variants from both types of amplicons. Dots indicate identity to the HVR1 amino acid sequence on the top line, which serves as the reference for the sequences in all boxes shown below. The dash indicates a coding failure due to a single-nucleotide deletion in the sequence alignment.

Sequential comparison of nearly full-length HCV QS derived at baseline and the breakthrough point. Full-length sequencing was applied to a total of 12 recombinant clones, including clones 7, 18, and 20 for sample 4701, clones 3, 6, and 11 for sample 5309, clones 6, 7, and 11 for sample 4681, and clones 2, 4, and 6 for sample 5159 (Fig. 4). Thus, for each sample, three dominant QS variants with the same HVR1 sequence were selected for fully sequencing. Genetic analysis showed minimized differences at the near-full-length HCV genome level among clones with the same HVR1 sequence. Average differences among three clones at both nucleotide (9,022 bp) and amino acid (2,939 bases) levels were 58.7 and 40 for sample 4701, 68 and 36.3 for sample 5309, 68 and 26.7 for sample 4681, and 38 and 25.3 for sample 5159, respectively. In contrast, when HVR1 was different, average differences at both the nucleotide and amino acid levels were significantly increased to 213.4 and 72.3 between samples 4701 and 5309 and 297.5 and 86.3 between samples 4681 and 5159 (P = 0.021 at the nucleotide level and P = 0.024 at the amino acid level). In the neighbor-joining tree reconstructed with a total of 12 nearly full-length QS variants, these QS variants displayed an HVR1-based clustering (Fig. 5). These findings further confirm our previous observation that HVR1 may serve as a genetic marker in terms of phylogenetic clustering of individual QS variants at the nearly full-length HCV genome level (34).

View larger version (12K): [in this window] [in a new window]

FIG. 5. Neighbor-joining tree constructed with 12 near-full-length QS variants. At the near-full-length HCV genome level, all QS variants showed HVR1-based clustering, suggesting a central role of HVR1 sequences in the reconstruction of phylogenetic relatedness. The bootstrap analysis was done with 500 replicates and is shown at each branch. HCV strain H77 was included as a reference.

Next, we performed sequential comparison of nearly full-length HCV QS variants derived at the baseline and breakthrough points. In spite of a very low nucleotide misincorporation rate related to the LRP procedure (34), it is possible that some mutations on HCV QS variants are artificial in nature. To minimize the potential interference from these artificial mutations, sequential comparison was conducted using a consensus dominant QS variant, which was inferred from three dominant HCV QS variants with the same HVR1 sequence for each sample. Consequently, 5309 and 4701 consensus QS variants had 159 (1.8%) and 35 (1.2%) differences at the nucleotide and amino acid levels, respectively. Compared to sample 4681, the sample 5159 consensus QS variant had 262 (2.9%) nucleotide mutations or 63 (2.1%) amino acid replacements. Variations in patient LIV02 seemed more narrow. Moreover, these mutations were not distributed evenly along the nearly entire HCV genome. For patient LIV02, there was no amino acid replacement in the HCV functional genes Core, p7, NS3, NS4a, and NS4b. HCV E1, E2 NS2, NS5a, and NS5b had 1, 25, 4, 4, and 1 amino acid substitution, respectively (Fig. 6). In patient LIV03, except for Core, p7, and NS4b, amino acid changes were identified in other domains, including E1 (9 changes), E2 (28 changes), NS2 (8 changes), NS3 (3 changes), NS4a (1 change), NS5a (11 changes), and NS5b (2 changes) (Fig. 6). Dependent on gene-based mutation rates, both patients displayed a similar pattern, with high mutation frequencies in E2, NS2, and NS5A (Fig. 6).

View larger version (17K): [in this window] [in a new window]

FIG. 6. Comparison of QS variants derived at the baseline and breakthrough points in patients LIV02 and LIV03. Each QS variant is shown as a straight line. Vertical lines indicate amino acid substitutions. All QS variants were drawn in proportion to the HCV genome structure shown at the top line. Con, consensus QS variant inferred from three dominant QS variants with the same HVR1 domain. Clones 5, 10, 13, and 18 were minor QS variants with distinct HVR1 sequences at the baseline. For these clones, only full-length NS2 and NS5a genes were sequenced. The E2, NS2, and NS5a sequences with potential covariance patterns are marked.

Since both HCV breakthrough QS variants had different HVR1 sequences from dominant QS variants at baseline, it was of interest to determine if the above mutation pattern could be seen in minor QS variants with distinct HVR1 domains. We therefore sequenced full-length NS2 and NS5A domains from minor QS variants, including clones 5 and 10 from sample 4701 and clones 13 and 18 from sample 4681 (Fig. 4). Sequence alignment revealed a comparable number of mutations in these two genes in sample 4701 (Fig. 6). For sample 4681, mutations in NS2 were reduced in minor QS variants, with two changes for clone 18 (P = 0.042, Fisher exact test) and one change for clone 13 (P = 0.017, Fisher exact test). However, both clones had increased amino acid substitutions in NS5a, with 17 and 23 changes for clones 13 and 18, respectively (Fig. 6). Finally, while most mutations were located near HVR1 in the HCV E2 gene, there was no consistent mutation pattern in NS2 or NS5a, such as clustering in a defined small region.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In this study, we performed a high-resolution genetic analysis at a near-full-length HCV QS level in two patients experiencing viral breakthrough during antiviral therapy. Based on the cloning and sequencing of the partial HCV E2 domain covering HVR1, we first determined dynamic changes of HCV QS over the entire therapeutic course and follow-up. At the time of breakthrough, a distinct QS variant swept through the viral population and was subsequently replaced by new variants after the withdrawal of antiviral therapy. Due to the lack of normalization of the HCV RNA amount during RT-PCR, it is possible that QS diversity may have been affected by viral titers. However, our previous study failed to detect a potential relationship between QS diversity and viral loads (11). Given an adequate number of clones and the use of fixated PCR primers, viral titers do not change QS structure and therefore would be expected to have a minimal role on our sequential QS analysis (10, 30). Overall, HCV showed an evolutionary pattern toward QS lineages at baseline (Fig. 3). Such QS dynamics are frequently observed in human immunodeficiency virus antiviral therapy in which drug-resistant variants assume a decreased replicative fitness and are rapidly replaced in a drug-free environment (1, 5, 15, 24). In line with these observations, it is reasonable to hypothesize that HCV QS variants sampled at the time of breakthrough represent the drug-resistant phenotype. By comparative analyses of near-full-length QS variants, we then explored the potential genetic characteristics that contributed to the drug-resistant nature.

Consensus dominant QS variants were used for the comparative analysis for two reasons. First, consistent with our previous observations, we found the number of mutations was significantly decreased among QS variants with the same HVR1 sequence. Second, our LRP protocol has a nucleotide misincorporation rate of about 2.17 x 10^–5 per cycle, which is equal to 22 mutations randomly scattered over a 9,022-bp amplicon (34). Thus, the use of consensus QS variants should minimize artificial mutations that may interfere with a genetic analysis. Each consensus sequence was inferred from three dominant QS variants. Sequential analyses showed that mutations in breakthrough QS variants were not randomly distributed. Most mutations involved three HCV genes, E2, NS2, and NS5a. Enomoto et al. found a similar mutation pattern through the comparative analysis at the full-length HCV genome level in three patients (all genotype 1b) with viral relapse on antiviral therapy (8). A putative interferon sensitivity-determining region, located in the C-terminal half of NS5a, was subsequently proposed (8, 9). However, Enomoto et al. used full-length HCV population sequences that were assembled from multiple overlapping RT-PCR products. Since all sequential comparisons, including Enomoto's study, occurred in samples with different HVR1 sequences, we then speculated whether such mutation patterns could be found in minor QS variants at baseline. Taking advantage of our LRP technique, we sequenced full-length NS2 and NS5a domains in minor QS variants with distinct HVR1 domains. It is interesting that minor clones from sample 4701 showed comparable numbers of mutations in both genes. In sample 4681, two minor clones had reduced mutations in NS2 while NS5a displayed more mutations (Fig. 6). However, within each individual gene, we didn't find any distinct mutation patterns, such as clustering in small domains.

The mutations clustering in E2, NS2, and NS5a may simply be a reflection of their high tolerance to amino acid substitutions. Indeed, E2, NS2, and NS5a are among the HCV functional genes that show higher values of average Shannon entropy at the HCV 1a population level (Fig. 7). It is well-documented that E2 is the most variable gene in the HCV genome. In particular, dynamic changes of HVR1 QS variants are frequently observed in various clinical settings (7). The NS2 is a "hot" domain for crossover points of natural HCV recombinants (17, 21). In the replicon system, adaptive mutations in NS5a are associated with enhanced replication (2). However, their respective capability to tolerate mutations cannot be used to explain the simultaneous existence of clustered mutations in these three genes from individual QS variants. Similar mutation patterns in different QS variants in these two patients suggest a lower probability of our observation being a random event. Consequently, a plausible explanation for our observation is a mutation linkage (covariance) among E2, NS2, and NS5a due to structural or functional relatedness in HCV replication, such as a possible epistatic interaction (3, 28). Near-full-length QS analysis in more patients is required to confirm this hypothesis. Finally, our observation of mutational clusterings in E2, NS2, and NS5a is not necessarily associated with viral breakthrough, since a single amino acid substitution is sometimes adequate for the phenotype alteration of a given gene.

View larger version (17K): [in this window] [in a new window]

FIG. 7. Average amino acid complexity of individual functional genes at the HCV genotype 1a population level. A total of 129 full-length HCV genotype 1a sequences were retrieved from the Los Alamos HCV Sequence Database and aligned. Genetic complexity was estimated at each amino acid position for the alignment by measuring Shannon entropy with the program BioEdit (10). The average genetic complexity for each HCV functional gene was then calculated. High complexity was observed in five HCV genes, three from structural domains (E1, E2, and p7) and two from nonstructural regions (NS2 and NS5a).

Neither breakthrough QS variant was found at baseline. For sample 4681, we screened 160 clones derived from the 9.1-kb amplicon using HVR1-specific primers. Positive amplification of the screening PCR product was obtained from only two clones that showed phylogenetic relatedness to the breakthrough QS variant (Fig. 3). Thus, breakthrough QS lineages, if present at baseline, must have a very low frequency. Since similar mutation patterns are observed in minor QS variants at the baseline, it is possible that viral breakthrough might simply be attributed to the selection of preexisting QS variants, which obtain a drug-resistant phenotype through additional mutations or recombination among QS variants (22). The recombination among QS variants is sometimes difficult to detect because of high sequence homology. In our study, recombination could not be excluded despite the lack of an apparent sign by bootscanning analysis (data not shown).

The emergence of distinct QS variants with diminished sensitivity to interferon is one mechanism to explain the appearance of viral breakthrough (23, 25, 26). Our data strengthen this assumption, based on evolutionary patterns of HVR1 QS variants over time. The QS analysis at the near-full-length HCV genome level indicates that selection rather than de novo generation is mainly responsible for emergence of the drug-resistant QS variants at breakthrough points. More importantly, mutation clustering in HCV E2, NS2, and NS5a has been found among different QS variants and is best explained by potential mutation linkage in these functional genes. Besides the contribution to our understanding of HCV genetic variation in antiviral therapy, such a mutation mode is also helpful for the improvement of HCV cell culture methods through the selection of appropriate QS variants generated by LRP.

 
* Corresponding author. Mailing address: Division of Gastroenterology and Hepatology, Department of Internal Medicine, Saint Louis University School of Medicine, 3635 Vista Avenue, St. Louis, MO 63110. Phone: (314) 977-7833. Fax: (314) 577-8125. E-mail for X. Fan: fanx{at}slu.edu. E-mail for A. M. Di Bisceglie: dibiscam{at}slu.edu

Published ahead of print on 30 July 2008.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1
1. Barbour, J. D., T. Wrin, R. M. Grant, J. N. Martin, M. R. Segal, C. J. Petropoulos, and S. G. Deeks. 2002. Evolution of phenotypic drug susceptibility and viral replication capacity during long-term virologic failure of protease inhibitor therapy in human immunodeficiency virus-infected adults. J. Virol. 76:11104-11112.[Abstract/Free Full Text]
2
2. Blight, K. J., A. A. Kolykhalov, and C. M. Rice. 2000. Efficient initiation of HCV RNA replication in cell culture. Science 290:1972-1974.[Abstract/Free Full Text]
3
3. Bonhoeffer, S., C. Chappey, N. T. Parkin, J. M. Whitcomb, and C. J. Petropoulos. 2004. Evidence for positive epistasis in HIV-1. Science 306:1547-1550.[Abstract/Free Full Text]
4
4. Chambers, T. J., X. Fan, D. A. Droll, E. Hembrador, T. Slater, M. W. Nickells, L. B. Dustin, and A. M. Di Bisceglie. 2005. Quasispecies heterogeneity within the E1/E2 region as a pretreatment variable during pegylated interferon therapy of chronic hepatitis C virus infection. J. Virol. 79:3071-3083.[Abstract/Free Full Text]
5
5. Charpentier, C., D. E. Dwyer, F. Mammano, D. Lecossier, F. Clavel, and A. J. Hance. 2004. Role of minority populations of human immunodeficiency virus type 1 in the evolution of viral resistance to protease inhibitors. J. Virol. 78:4234-4247.[Abstract/Free Full Text]
6
6. Di Bisceglie, A. M. 1998. Hepatitis C. Lancet 351:351-355.[CrossRef][Medline]
7
7. Domingo, E., and J. Gomez. 2007. Quasispecies and its impact on viral hepatitis. Virus Res. 127:131-150.[CrossRef][Medline]
8
8. Enomoto, N., I. Sakuma, Y. Asahina, M. Kurosaki, T. Murakami, C. Yamamoto, N. Izumi, F. Marumo, and C. Sato. 1995. Comparison of full-length sequences of interferon-sensitive and resistant hepatitis-C virus 1b-sensitivity to interferon is conferred by amino-acid substitutions in the NS5a region. J. Clin. Investig. 96:224-230.[Medline]
9
9. Enomoto, N., I. Sakuma, Y. Asahina, M. Kurosaki, T. Murakami, C. Yamamoto, Y. Ogura, N. Izumi, F. Marumo, and C. Sato. 1996. Mutations in the nonstructural protein 5A gene and response to interferon in patients with chronic hepatitis C virus 1b infection. N. Engl. J. Med. 334:77-81.[Abstract/Free Full Text]
10
10. Fan, X., A. C. Lyra, D. Tan, Y. Xu, and A. M. Di Bisceglie. 2001. Differential amplication of hypervariable region 1 of hepatitis C virus by partially mismatched primers. Biochem. Biophy. Res. Commun. 284:694-697.[CrossRef][Medline]
11
11. Fan, X., H. Solomon, J. E. Poulos, B. A. Neuschwander-Tetri, and A. M. Di Bisceglie. 1999. Comparison of genetic heterogeneity of hepatitis C viral RNA in liver tissue and serum. Am. J. Gastroenterol. 94:1347-1354.[CrossRef][Medline]
12
12. Fan, X., Y. Xu, and A. M. Di Biceglie. 2006. Efficient amplification and cloning of near full-length hepatitis C virus genome from clinical samples. Biochem. Biophys. Res. Commun. 346:1163-1172.[CrossRef][Medline]
13
13. Feld, J. J., and J. H. Hoofnagle. 2005. Mechanism of action of interferon and ribavirin in treatment of hepatitis C. Nature 436:967-972.[CrossRef][Medline]
14
14. Hall, T. A. 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41:95-98.
15
15. Hance, A. J., V. Lemiale, J. Izopet, D. Lecossier, V. Joly, P. Massip, F. Mammano, D. Descamps, F. Brun-Vezinet, and F. Clavel. 2001. Changes in human immunodeficiency virus type 1 populations after treatment interruption in patients failing antiretroviral therapy. J. Virol. 75:6410-6417.[Abstract/Free Full Text]
16
16. Higgins, D. G., and P. M. Sharp. 1988. CLUSTAL: a package for performing multiple sequence alignment on a microcomputer. Gene 73:237-244.[CrossRef][Medline]
17
17. Kalinina, O., H. Norder, S. Mukomolov, and L. O. Magnius. 2002. A natural intergenotypic recombinant of hepatitis C virus identified in St. Petersburg. J. Virol. 76:4034-4043.[Abstract/Free Full Text]
18
18. Kimura, M. 1980. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16:111-120.[CrossRef][Medline]
19
19. Kuiken, C., M. Mizokami, G. Deleage, K. Yusim, F. Penin, T. Shin-I, C. Charavay, N. Tao, D. Crisan, D. Grando, A. Dalwani, C. Geourjon, A. Agrawal, and C. Combet. 2006. Hepatitis C databases, principles and utility to researchers. Hepatology 43:1157-1165.[CrossRef][Medline]
20
20. Kumar, S., K. Tamura, and M. Nei. 2004. MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief. Bioinform. 5:150-163.[Abstract/Free Full Text]
21
21. Legrand-Abravanel, F., J. Claudinon, F. Nicot, M. Dubois, S. Chapuy-Regaud, K. Sandres-Saune, C. Pasquier, and J. Izopet. 2007. New natural intergenotypic (2/5) recombinant of hepatitis C virus. J. Virol. 81:4357-4362.[Abstract/Free Full Text]
22
22. Moreno, M. P., D. Casane, L. Lopez, and J. Cristina. 2006. Evidence of recombination in quasispecies populations of a hepatitis C virus patient undergoing anti-viral therapy. Virol. J. 3:87.[CrossRef][Medline]
23
23. Negro, F., M. Baldi, A. Mondardini, G. Leandro, M. Chaneac, P. Manzini, M. L. Abate, G. Verme, and F. Bonino. 1994. Continuous versus intermittent therapy for chronic hepatitis C with recombinant interferon alfa-2a. Gastroenterology 107:479-485.[Medline]
24
24. Nijhuis, M., R. Schuurman, D. de Jong, J. Erickson, E. Gustchina, J. Albert, P. Schipper, S. Gulnik, and C. A. B. Boucher. 1999. Increased fitness of drug resistant HIV-1 protease as a result of acquisition of compensatory mutations during suboptimal therapy. AIDS 13:2349-2359.[CrossRef][Medline]
25
25. Okada, S., Y. Akahane, H. Suzuki, H. Okamoto, and S. Mishiro, S. 1992. The degree of variability in the amino terminal region of the E2/NS1 protein of hepatitis C virus correlates with responsiveness to interferon therapy in viremic patients. Hepatology 16:619-624.[Medline]
26
26. Roffi, L., G. C. Mels, G. Antonelli, G. Bellati, F. Panizzuti, A. Piperno, M. Pozzi, D. Ravizza, G. Angeli, F. Dianzani, and G. Mancia. 1995. Breakthrough during recombinant interferon alfa therapy in patients with chronic hepatitis C virus infection: prevalence, etiology, and management. Hepatology 21:645-649.[Medline]
27
27. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406-425.[Abstract]
28
28. Shapiro, B., A. Rambaut, O. G. Pybus, and E. C. Holmes. 2006. A phylogenetic method for detecting positive epistasis in gene sequences and its application to RNA virus evolution. Mol. Biol. Evol. 23:1724-1730.[Abstract/Free Full Text]
29
29. Taylor, D. R., S. T. Shi, P. R. Romano, G. N. Barber, and M. M. Lai. 1999. Inhibition of the interferon-inducible protein kinase PKR by HCV E2 protein. Science 285:107-110.[Abstract/Free Full Text]
30
30. Torres-Puente, M., M. A. Bracho, N. Jimenez, I. Garcia-Robles, A. Moya, and F. Gonzalez-Candelas. 2003. Sampling and repeatability in the evaluation of hepatitis C virus genetic variability. J. Gen. Virol. 84:2343-2350.[Abstract/Free Full Text]
31
31. Wohnsland, A., W. P. Hofmann, and C. Sarrazin. 2007. Viral determinants of resistance to treatment in patients with hepatitis C. Clin. Microbiol. Rev. 20:23-38.[Abstract/Free Full Text]
32
32. Yoneyama, M., M. Kikuchi, T. Natsukawa, N. Shinobu, T. Imaizumi, M. Miyagishi, K. Taira, S. Akira, and T. Fujita. 2004. The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral responses. Nat. Immunol. 5:730-737.[CrossRef][Medline]
33
33. Young, K. C., K. L. Lindsay, K. J. Lee, W. C. Liu, J. W. He, S. L. Milstein, and M. M. Lai. 2003. Identification of a ribavirin-resistant NS5B mutation of hepatitis C virus during ribavirin monotherapy. Hepatology 38:869-878.[CrossRef][Medline]
34
34. Zhou, D., X. Fan, D. Tan, Y. Xu, J. E. Tavis, and A. M. Di Bisceglie. 2007. Separation of near full-length hepatitis C virus quasispecies variants from a complex population. J. Virol. Methods 141:220-224.[CrossRef][Medline]

---

Journal of Virology, October 2008, p. 9417-9424, Vol. 82, No. 19
0022-538X/08/$08.00+0     doi:10.1128/JVI.00896-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Xu, Z. Articles by Di Bisceglie, A. M. Search for Related Content PubMed PubMed Citation Articles by Xu, Z. Articles by Di Bisceglie, A. M.