INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

Hepatitis C virus (HCV) infections represent a serious health problem. A vaccine protective against HCV infection is not currently available, and antiviral treatments are ineffective in the majority of HCV-infected patients. Current estimates suggest that as many as 85% of HCV-infected individuals remain persistently infected, and chronic HCV infection is associated with cirrhosis and hepatocellular carcinoma (5, 6, 37).

HCV infection appears to persist despite the presence of virus-specific cytotoxic T lymphocytes (CTL) and circulating antibodies to HCV proteins (3, 12, 16). The HCV structural proteins include the capsid and two envelope glycoproteins, E1 and E2. Several hypervariable regions (HVR) are present within the envelope glycoproteins and may facilitate the maintenance of persistent infection (10, 15, 23, 25, 50). The most significant divergence has been observed in the first HVR (HVR-1) within E2. Since the HVR-1 may be a dominant neutralizing epitope (19), the presence within an individual of heterogeneous populations of virions, or quasispecies, may explain why HCV-specific CTL and antibodies are not sufficient to clear infection, since multiple variant genomes continuously escape neutralization (18). A greater understanding of the pathogenesis of HCV may facilitate the development of vaccines and antiviral treatments that are more-efficacious.

HCV pathogenesis is difficult to study, since small-animal models and conventional tissue culture systems have not been established. Currently, chimpanzees serve as the only animal model for HCV infection. The frequency of persistent infection in chimpanzees and humans appears to differ. Examination of the virological outcome in a large cohort of HCV-inoculated chimpanzees revealed that an unexpectedly high percentage of chimpanzees cleared the virus (61%) based on reverse transcriptase (RT)-PCR negativity (7). Since an antibody response elicited against the envelope protein has been proposed to be important for neutralization and clearance of the virus, we have examined HCV-inoculated animals for antibody reactivity to the envelope proteins and sequence variability in the envelope domain. The results revealed that (i) a low percentage of infected chimpanzees responded to E1 and E2, (ii) viral clearance did not appear to be associated with an antibody response to E1 or E2, and (iii) persistence did not appear to be due to immune escape of variants in the E1 and E2 regions. The significance of these findings to the chimpanzee animal model and their possible extrapolation to humans is discussed here.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Cloning and envelope proteins into baculovirus expression vectors. An E1 fragment representing nucleotides 915 to 1421 (amino acids [aa] 192 to 360) was amplified by PCR by using a previously described plasmid containing the E1 region of the HCV-1 strain (genotype 1a) (33). The E1 domains of HCV-1 and the Hutchinson strains are 98% homologous. The downstream primer for the E1 fragment spanned nucleotides 1404 to 1421 (aa 355 to 360, 5'-GAAGATCTTTAGTGGTGGTGGTGGTGGTGCGCTATGCCCGCCAGGAC-3') and contained nucleotide sequences encoding a 6-histidine tail, and a BglII site. The upstream primer (5'-TTCCCCGGGATACCAAGTGCGCAACTCC-3') contained a SmaI site and spanned nucleotides 915 to 932 (aa 192 to 196). The gel-purified E1 fragment was digested with SmaI and BglII and ligated into gel-purified SmaI- and BamHI-digested pPbac (Stratagene). The vector contained a translational start site in frame with a human placental alkaline phosphatase signal sequence (APSS) upstream of the SmaI site. The APSS directs polypeptides to the secretory pathway of insect cells. The selected clone was designated pE1-360.

Sf9 cells and baculovirus infections. Procedures involving the cultivation of insect cells and the production and use of recombinant baculoviruses were essentially as described (44). For the purification of E1 and E2, Sf9 cells were grown to a cell density of approximately 2 × 10⁶ to 3 × 10⁶ cells/ml in a 250-ml spinner culture as previously described (32). Cells were pelleted at 1,000 × g for 10 m in and resuspended in 25 ml of virus stock for 1 h at 27°C. After infection, 225 ml of Grace's medium supplemented with 2% fetal bovine serum and 0.1% Pluronic F-68 (JRH Biosciences) was added to the spinner of infected cells.

Purification of HCV recombinant envelope proteins. Sf9 spinner cultures infected with E1 and E2 recombinant baculoviruses were harvested at 72 h postinfection. The cells were washed three times with cold phosphate-buffered saline (PBS) and were extracted in 15 ml of EB buffer (50 mM Tris-HCl, pH 9.0; 100 mM NaCl, 1% Nonidet P-40) for 15 min, during which time the container was vortexed lightly every 5 min. The cell lysate was clarified at 31,000 × g for 20 min.

SDS-PAGE and Western blot analysis. Discontinuous sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed as previously described (28, 29). The gels were either silver stained according to the procedure of Wray et al. (51) or were further processed for Western blot analysis as previously described (30). For Western blot analysis, the gels were transferred to a Pall Biosupport Flourotrans polyvinylidene difluoride membrane (ICN Biomedical). The primary antibodies were rabbit antisera produced against synthetic peptides derived from capsid/E1 (aa 185 to 210) and E2 (aa 439 to 466) (33).

E1 and E2 ELISAs. Microtiter plates (flat-bottom, high-binding ELISA plates; Corning) were coated with antigen for 16 h at 4°C in borate-buffered saline (48.5 mM boric acid, 145 mM NaCl, 6.0 mM NaOH, 50 mM KCl [pH 8.2]). The volume of reagent added to the microtiter wells was 100 µl for all steps except for the blocking (200 µl) and washing (300 µl) steps. A titration assay was performed to determine the optimal coat concentrations (~5 ng/well for E1; ~100 ng/well for E2). Wells were blocked with PBS containing 5% nonfat dry milk (NFDM) and 0.05% Tween 20 (NFDM-PBS-Tween) at 37°C for 2 h and were washed three times with PBS-Tween. The primary antibody was diluted 1:200 in NFDM-PBS-Tween and was incubated with the wells for 2 h at 37°C. The wells were washed three times with PBS-Tween. Binding of primary antibodies was detected with horseradish peroxidase-conjugated goat anti-human immunoglobulin G secondary antibody (Southern Biotechnology Associates, Inc.). The conjugate was diluted in NFDM-PBS-Tween (1:1,000). Wells were incubated with the conjugate at 37°C for 1 h, washed three times in PBS-Tween, and incubated with the peroxidase substrate solution (0.54 mM 2,2'-azino-bis[3-ethylbenzthiazoline-6-sulfonic acid], 0.03% H₂O₂, 0.1 M citric acid; pH 4.0) at room temperature for 30 min. The color reaction was stopped with 10% SDS. The absorbance at 410 nm (reference wavelength, 490 nm) was read by using a Dynatech MR700 microtiter plate unit. The raw data were analyzed with a computer program in Microsoft Excel designed specifically for these assays by Mark Sharp (unpublished data). A human or chimpanzee pool of normal, anti-HCV negative sera was run in triplicate on every microtiter plate. The average optical density (OD) reading from the negative control wells was multiplied by 2.5 to establish a cutoff value. A negative value indicated that the average OD reading from a serum sample run in duplicate was below the cutoff value.

Analysis of sera for anti-HCV antibodies by ELISA and recombination immunoblot assay (RIBA). Chimpanzee and human sera were analyzed for anti-HCV antibody by using a second-generation ELISA (HCV 2.0 ELISA; Ortho Diagnostic Systems, Raritan, N.J.). Chimpanzee sera were also analyzed for HCV antibodies to individual HCV proteins by using HCV BLOT 3.0 (Genelabs Diagnostics). These assays were performed according to the manufacturer's instructions.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Purification of E1 and E2 and ELISA development. In order to evaluate the immune response of chimpanzees to E1 and E2, the proteins were expressed in insect cells by using recombinant baculovirus constructs. We have previously demonstrated that the humoral immune response to E1 and E2 was primarily to denaturation-sensitive epitopes (33). Thus, the carboxy-terminal hydrophobic domains of E1 and E2 were deleted to express these proteins in the soluble fraction of the cell to facilitate purification under nondenaturing conditions. The E1 (aa 192 to 360) and E2 (aa 364 to 715) constructs were engineered with truncations at the 3' end, followed by nucleotides encoding a 6-histidine tail. Signal sequences at the amino terminus of E1 and E2 were included to direct the protein into the secretory pathway. In addition, the E1 gene was cloned into a vector with a human placental alkaline phosphatase signal sequence at the amino terminus in an attempt to increase secretion into the culture medium. Approximately 60% of the E1 and E2 was present in the soluble fractions of the cells, and the proteins were glycosylated (33). Although approximately 20% of E2 was secreted into the medium, E1 was not secreted. E1 and E2 proteins were purified from Sf9 cell lysates under native conditions by using an agarose G. nivalis lectin I and a Talon metal affinity resin as described in Materials and Methods. Approximately 120 µg of E1 and 2.5 mg of E2 were purified per 2 × 10⁹ cells (1-liter spinner culture) based on BCA analysis. Western blot analysis revealed that the purification was not highly efficient, since a large fraction of E1 and E2 was present in the unbound fractions of both resins (data not shown). However, this purification scheme yielded proteins with very low reactivity to normal human sera as determined by ELISA. The purity of E1 and E2 was assessed by silver staining (Fig. 1) and was estimated to be approximately 80 to 95%.

View larger version (49K): [in this window] [in a new window]   FIG. 1.   SDS-PAGE analysis of purified E1 and E2 proteins. Western blot analysis of E1 (100 ng) and E2 (5.2 µg) was performed as described in Materials and Methods. Purified E1 (500 ng) and E2 (1.3 µg) were analyzed by SDS-PAGE and silver staining. Recombinant E1 and E2 proteins were purified from Sf9 cell lysates by using an agarose G. nivalis (snowdrop) lectin 1 and a Talon metal affinity resin as described in Materials and Methods.

Humoral immune response to envelope proteins in HCV-infected human cohorts. Prior to analysis of the chimpanzee cohort, human serum panels were examined to assess the performance of the assay and the frequency of E1 and E2 antibodies (anti-E1 and anti-E2) according to a widely used, commercially available test. Ortho-HCV 2.0-positive human serum samples from three different human cohorts were examined for anti-E1 and anti-E2 by ELISA (Table 1). Panel 1 was comprised of sera from intravenous drug users (IVDUs) from a cohort that has been previously described and in which more than 95% of the Ortho-HCV 2.0 positive sera are also positive as determined by RIBA and/or for HCV RNA (45). Anti-E1 and anti-E2 reactivity were observed in 58% (149 of 257) and 93% (239 of 257) of the serum samples, respectively. Therefore, these ELISAs appeared to be sensitive indicators of HCV infection. Panel 2 consisted of plasma rejected from the blood bank based on elevated alanine aminotransferase (ALT) values prior to 1992, when blood donations were not screened for anti-HCV. Only HCV 2.0-positive samples were included in this panel. Anti-E1 and anti-E2 were observed in 44 and 62% of the samples, respectively. The higher percentage of antibody reactivity in the IVDU panel compared to the blood donor panel may be due to a longer duration of infection, as well as to potential continuous inoculation with HCV in the IVDU cohort. Panel 3 represented individuals with recently acquired symptomatic non-A, non-B hepatitis (NANBH), indicating an acute infection, which was retrospectively screened to include only HCV 2.0-positive sera. Anti-E1 and anti-E2 reactivity were observed in 22% (4 of 18) and 33% (6 of 18) of the individuals, respectively. The low frequency of reactivity to E1 and E2 in this panel may be explained by the short duration of infection in the acute panel. Alternatively, the low frequency of reactivity may be an artifact due to the small sampling size.

Humoral immune response to envelope proteins in a cohort of HCV-inoculated chimpanzees. Serum samples from 46 HCV-inoculated chimpanzees were analyzed for the presence of anti-HCV antibody by HCV 2.0 ELISA and for the presence of HCV RNA by RT-PCR as previously described (7). The chimpanzees were exposed to HCV inocula during the past 2 to 19 years, with an average duration of 10.6 years since inoculation. Of the 46 chimpanzees, at least 14 had received the Hutchinson strain inoculum, at least 3 had received the HCV-1 inoculum, and 28 had received unknown inocula. Antibody to HCV proteins was detected in 48% (22 of 46) of the chimpanzees by the HCV 2.0 ELISA (Table 2). The majority of HCV-inoculated chimpanzees did not have detectable antibody as determined by ELISA against the envelope proteins. Anti-E1 and anti-E2 reactivity were observed in 22% (10 of 46) and 15% (7 of 46) of the HCV-inoculated chimpanzees, respectively. Of the HCV 2.0-positive animals, only 18% (4 of 22) and 32% (7 of 22) were anti-E1 and anti-E2 positive, respectively (Tables 1 and 2). None of the HCV 2.0-negative animals had antibody to E2, although 25% were anti-E1 positive (Table 2). Therefore, only HCV 2.0-positive animals had antibody to E2, while the anti-E1 response was similar for both HCV 2.0-positive and -negative animals.

Humoral immune responses to the envelope proteins in viremic chimpanzees. Although HCV infections after inoculation had been confirmed in all 46 of the animals in this study (7), HCV RNA was detected in recently collected serum samples from only 18 (39%) of the chimpanzees (Table 2). Of the viremic chimpanzees, anti-E1 antibody was detected in 17% (3 of 18) and anti-E2 antibody was detected in 39% (7 of 18) (Table 2). All of the anti-E2 positive chimpanzees were viremic (7 of 7) (Table 2). In contrast, anti-E1 was detected in both viremic (17%) and nonviremic animals (25%). Antibody to E1 and E2 were simultaneously detected (E1⁺ E2⁺) in 17% of the viremic chimpanzees (Table 3). Antibody reactivity to E1 but not to E2 (E1⁺ E2) was not observed in the viremic animals, and an antibody response elicited against E2 in the absence of anti-E1 antibody (E1 E2⁺ was observed in 22% of the viremic animals (Table 3). The majority of viremic chimpanzees lacked a detectable humoral immune response to both E1 and E2 (61% E1 E2) (Table 3).

Humoral immune responses to the envelope proteins in nonviremic chimpanzees. Viral clearance had apparently occurred in 28 chimpanzees (61%), since HCV RNA was not detected in the serum samples collected in 1995 (Table 2). Convalescence was supported by the observation that most RT-PCR-negative chimpanzees were also anti-HCV-negative by ELISA (Ab PCR; 86%, 24 of 28) and that nested PCR followed by Southern hybridization failed to reveal HCV RNA in serum from antibody-positive but RT-PCR negative (Ab⁺ PCR) chimpanzees. Anti-E1 was detected in 25% (7 of 28) of the HCV RNA-negative chimpanzees, while none of these animals had a detectable anti-E2 response (Table 2). The frequency of anti-E1 in HCV 2.0- and HCV RNA-negative chimpanzees (Ab PCR) was also 25% (6 of 24). Of the convalescent chimpanzees, 25% were E1⁺ E2, none were E1⁺ E2⁺ or E1 E2⁺, and the majority were E1 E2 (75%) (Table 3).

Longitudinal evaluation of humoral immune responses to the envelope proteins. An anti-E2 antibody response appeared to correlate with persistent HCV infection rather than convalescence in the cross-sectional study (Table 2). To determine whether anti-E2 was elicited prior to convalescence but declined after clearance of viremia, a longitudinal study of the humoral immune response in six HCV-infected chimpanzees was performed. The Hutchinson strain of HCV was serially passaged into six chimpanzees as described previously. Viral clearance was observed in four animals within 5 months postinoculation, with either gradual or rapid loss of anti-HCV antibody based on the HCV 2.0 ELISA. A chronic carrier profile characterized by persistent HCV RNA was observed in two animals. One of these chimpanzees was RT-PCR positive and antibody negative for 5 years. This animal would not have been detected with the standard assay employed by blood banks and thus represents a "silent carrier". Serial serum samples from these six chimpanzees were examined for anti-E1 and anti-E2 by ELISA. Anti-E1 was not detected in the serial serum samples from any of the chimpanzees. An anti-E2 response was not observed in the serial bleeds of the convalescent animals and was detected in only one viremic animal, x174 (Fig. 2). Although this animal was HCV 2.0 positive by 3 months postinoculation, the anti-E2 response was not detected until 7 years postinoculation, though it remained positive thereafter. Antibodies to the HVR-1 and to E1 were not detected in serum samples from x174. Although the number of animals examined was small, the results of the longitudinal study agreed with those of the cross-sectional study in that convalescent animals did not have detectable levels of antibody to E2 and suggested that the low reactivity to E1 and E2 observed in the cross-sectional study was not due to the sampling of a single time point.

View larger version (12K): [in this window] [in a new window]   FIG. 2.   Antibody response to E1, E2, and the E2 HVR-1. Serial bleeds obtained up to 9 years postinoculation with HCV from chimpanzee x174 were examined by ELISA for reactivity to native, purified E1 and E2, and a synthetic peptide to the HVR-1 as described in Materials and Methods.

Antibody responses to E2 and sequence changes in the E2 HVR-1. Several studies have suggested that viral persistence may be associated with the emergence of viruses that vary in the HVR-1 that are able to escape neutralizing antibody. To examine the role of an antibody response to the HVR-1 in chimpanzees, serial serum samples from the six chimpanzees inoculated with the Hutchinson strain of HCV were examined for anti-HVR-1 antibody. An ELISA was developed by using a synthetic peptide homologous to the HVR-1 of the Hutchinson strain of HCV. An anti-HVR-1 antibody response was not observed in these chimpanzees, irrespective of whether they convalesced or progressed to chronic infection. Although none of the chimpanzees were positive in the HVR-1 ELISA, all were positive for anti-HCV at some time point during the analysis and four demonstrated antibodies in a different peptide-based ELISA to the NS4A region (data not shown). The lack of anti-HVR-1 antibody was confirmed by an immunoprecipitation assay with a GST-HVR-1 fusion protein (data not shown).

View larger version (23K): [in this window] [in a new window]   FIG. 3.   Alignment of amino acid sequences of the E2 HVR-1. The amino acid sequences of the HVR-1s from the acute and chronic phases from 10 chimpanzees are aligned with the amino acid sequence of the HVR from the Hutchinson strain of HCV. Residues identical to the Hutchinson sequence are indicated by a period. Residues differing between the acute and chronic samples are shown in boldface. The clinical records indicate that animals x81, x174, x196, and x62 received the Hutchinson inoculum; animals x204 and x258 received a common inoculum; animals x341 and x342 received a common inoculum, and it is suspected that animal x304 received this inoculum as well. The inoculum for animal x130 was not indicated.

3

3

View larger version (18K): [in this window] [in a new window]   FIG. 4.   Relationship of immune responses to envelope proteins and nucleotide changes in the HVR-1. Temporally spaced bleeds from 10 HCV-infected chimpanzees were examined, 5 of which had an antibody response to E2. The HVR-1 sequence was amplified from each serum sample and was cloned and sequenced as described in Materials and Methods. The sera were examined by ELISA for antibody to E1 and E2. Positive responses are indicated by a "+" at the top of the bars. The bars represent the OD values from the E1 and E2 ELISAs of paired sera, with the earliest and latest bleeds represented by the bars to the left and right, respectively. The chimpanzee numbers, the duration of time between temporally spaced sera, and the number of nucleotide and amino acid changes are noted at the bottom.

Evaluation of E1 and E2 sequences for variability outside of E2 HVR-1. Immune escape of variants in the HVR-1 has been suggested to contribute to persistent HCV infection in humans. However, persistent infection was observed in three chimpanzees even in the absence of amino acid changes in the HVR-1, and changes were minimal in the other seven animals (Table 4). To determine whether variable regions exist in chimpanzees, other than the HVR-1 observed in humans, the E1 and E2 genes were sequenced from paired sera collected from x174 and x304 (Table 5). Serum samples were chosen to represent the greatest possible duration of time between samples, which was 7.9 years for x174 and 7.4 years for x304. Anti-E2 was detected in the most recent serum sample from x174 and from both samples collected from x304. Although anti-E2 only recently emerged in x174 (7 years postinoculation), it was present for approximately 1 year prior to the collection of the serum sample used for the sequencing of E1 and E2. Although both animals had anti-E2, no major sequence changes in the E1 and E2 regions were observed. The rates of variation in the HVR-1, E1, and E2 regions sequenced from x174 were 1.4 × 10³, 2.2 × 10³, and 1.6 × 10³, respectively (Table 5). Similarly, the rates of variation in the HVR-1, E1, and E2 regions sequenced from x304 were 0, 2.1 × 10³, and 2.4 × 10³, respectively. Similar rates of variation were observed in the E1 and E2 regions, despite the presence of antibody exclusively to E2. Although the position of amino acid changes varied between x174 and x304, in both animals, 4 aa changes were observed within E1 and 4 aa changes were observed within E2. Interestingly, the percentage of nucleotide changes that resulted in amino acid changes was higher in E1 (40 to 44%) than in E2 (21 to 29%) (Table 5). Since the chimpanzees had detectable anti-E2 but not anti-E1, immunoselection did not appear to be driving the nonsilent nucleotide changes observed in E1. However, the E1 ELISA may not have been adequate for the detection of neutralizing antibody that may have driven immunoselection. Perhaps neutralizing antibody only recognizes E1 in the form of E1-E2 heterodimers.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

We have examined HCV-inoculated chimpanzees for antibody reactivity to the envelope proteins and sequence variability in the envelope domain. The results revealed that (i) a low percentage of HCV-inoculated chimpanzees had antibody to E1 and E2, (ii) convalescence did not appear to be associated with an antibody response to E1 or E2, and (iii) viral persistence did not appear to be due to the antibody escape of variants in the E1 and E2 regions.

Prior to examining serum samples from HCV-inoculated chimpanzees for antibody to the envelope proteins, the validity of the E1 and E2 ELISAs was assessed by testing three panels of HCV 2.0-positive human sera (Table 1). The three panels were representative of individuals with acute NANBH, blood donors with elevated ALT levels, and IVDUs. The low percentages of anti-E1 (22%) and anti-E2 (33%) reactivity observed in the acute serum panel may be due to insufficient time to mount an immune response to the envelope proteins. Nishihara and co-workers observed a similar percentage of anti-E2 (29%) in an acute panel (38). Panel 2 consisted of serum samples from HCV 2.0-positive blood donors with elevated ALT levels. Antibody to E1 and E2 was observed in 44 and 62% of the serum samples, respectively. Although the presence of HCV RNA was not assessed in our HCV 2.0-positive panel of blood donors, a higher percentage of anti-E2 positive samples may have been observed in this panel if only viremic individuals were tested. We observed that 58 and 93% of HCV 2.0-positive sera from the IVDU panel were anti-E1 and anti-E2 positive, respectively (Table 1). The higher percentage of antibody reactivity in the IVDU panel compared to the other panels may be due to a greater proportion of long-term infections. Alternatively, a greater breadth of immunoreactivity may be noted in IVDUs with ongoing and repeated exposure to HCV.

In our human serum panels, lower reactivity to E1 (22 to 58%) than to E2 (33 to 93%) was observed. A study of truncated HCV glycoproteins revealed that E2 appeared to be required for the correct folding of E1 (35). Therefore, the lower antibody reactivity to E1 compared to E2 may be explained if conformational differences due to the misfolding of E1 hinder antibody recognition. The percentage of responders to E2 in the IVDU panel in our ELISA was similar to the percentages of E2 responders in chronically infected individuals reported in several earlier studies (range = 53 to 100%) (11, 13, 21, 24, 34, 38, 47, 54). Therefore, our assay appeared to be a sensitive indicator of HCV infection.

A low percentage of the 46 HCV-inoculated chimpanzees responded to E1 (22%) and E2 (15%) but, in contrast to the human panels, E1 reactivity was higher than E2 reactivity. The higher prevalence of anti-E1 versus anti-E2 in our chimpanzee cohort may be explained by the high percentage of convalescent animals in the cohort. The lack of detectable anti-E1 and anti-E2 in the majority of HCV-inoculated chimpanzees did not appear to be due to a deficient humoral immune response to HCV, since even within the chimpanzees that were positive by the HCV 2.0 ELISA, only 18 and 32% of these animals were anti-E1 and anti-E2 antibody positive, respectively (Table 2). Although it is possible that the negative serological responses in some chimpanzees may have been due to differences between strains or genotypes used for inoculation and the antigens in the serologic assays, antibody to E2 was detected in 93% of a human serum panel. This suggests that the E2 ELISA was probably sensitive to a wide variety of strains, since humans are exposed to highly heterogeneous inocula. Previous studies have suggested that humoral immune responses to HCV structural proteins are observed less frequently in chimpanzees than in humans for reasons not understood (2, 14, 22, 31, 36, 48, 49).

Viral clearance was not associated with an antibody response to E1 or E2 in the cross-sectional study. Chimpanzees that cleared HCV lacked an E2 antibody response but had approximately the same reactivity to E1 as the persistently infected animals (Table 2). E2 antibody is consistently detected in chronically infected humans (11, 13, 21, 24, 34, 38, 47, 54), and several studies suggests that E2 antibody is a marker for active HCV replication (11, 20, 54). Consistent with data from human studies, E2 antibody was detected exclusively in viremic chimpanzees (Table 3). In the cross-sectional study, it could not be determined if nonviremic chimpanzees elicited an E2 antibody response that declined after viral clearance or if antibody to E2 was never produced.

In the longitudinal study, antibodies to E1, E2, and the HVR-1 were not detected in serial serum samples from five chimpanzees (four nonviremic, one viremic). Antibody to E2, but not to E1 or to the HVR-1, was detected in one persistently infected animal (x174), but only after 7 years postinoculation (Fig. 2). If the data from the longitudinal study can be extrapolated to the 46 chimpanzees, it is possible that antibodies to E1, E2, and the HVR-1 were never elicited in the majority of chimpanzees and that the antibodies did not simply decline after viral clearance. Several studies have suggested that neutralizing antibodies target the envelope proteins (14, 19, 43) and that an early antibody response to the HVR-1 is associated with convalescence (4, 56). However, additional immune factors (e.g., cytotoxic T lymphocytes and cytokines) may be involved in viral clearance, since antibody to E2 and to the HVR-1 was not detected in serial serum samples from the four chimpanzees that cleared the virus in the longitudinal study.

Several studies have suggested that viral persistence may be associated with the continuous emergence of viruses variant in the HVR-1 that are able to escape neutralizing antibody. However, the prevalence of immune escape variants is often difficult to assess, since minor variation in quasispecies due to random genetic drift may be interpreted as immunoselection. Parameters such as viral load and the presence of complex quasispecies in the inoculum may affect the rate of variation. Increased viral replication, as reflected in higher viral loads, should lead to higher rates of variation due to increased errors by RNA polymerase. A minor variant in the quasispecies may emerge as a predominant strain if it replicates more efficiently or due to T-cell immune escape variants at any position in the polyprotein. Since these changes will be in linkage with variations in the HVR-1, it would appear as if a new HVR-1 was selected by immune escape mechanisms.

Mechanisms in addition to the emergence of escape variants in the HVR-1 must play a role in maintaining persistent infection. Strong selective pressure did not appear to be driving the emergence of variants in our study. We observed very few amino acid changes in the HVR-1 during the time between collection of temporally spaced bleeds in 10 chimpanzees, and the percentages of silent and nonsilent nucleotide changes were similar (Table 4, Fig. 3). Additionally, antibody to E2 or the HVR-1 did not coincide with the number of nucleotide and amino acid variations in the HVR-1 (Table 4, Fig. 3 and 4). In some studies of persistently infected humans and chimpanzees, lower variation in the HVR-1 has also been observed and is sometimes associated with the absence of an antibody response to the HVR-1, near-normal ALT levels, and an improved response to interferon treatment (1, 8, 17, 26, 27, 39, 41, 46, 52, 53, 55). Since the majority of chimpanzees also have near-normal ALT levels and lack antibodies to E2, the lower variation in the HVR-1 was not surprising (7, 31). Possibly, the chimpanzees may represent the healthier, more asymptomatic population of HCV-infected humans with normal ALT levels.

The entire E1 and E2 genes were sequenced from temporally spaced sera collected from two animals to determine whether additional variable regions within the envelope regions, other than the E2 HVR-1 region, could contribute to persistence in chimpanzees. Major sequence changes in the E1 and E2 genes were not observed, although both animals had antibody to E2 in the chronic phase. The rates of variation in the E1 region from x174 and x304 were similar to the rates observed by Okamoto and coworkers and by Ogata and coworkers (9, 40, 42). In contrast to these studies, lower rates of variation were observed here in the E2 regions. The emergence of variants due to a humoral immune response was not clearly apparent. Although antibody to E2 was detectable for 7.4 years in x304 and for only 1 year in x174, the rates of variation and the percentage of nucleotide changes that resulted in amino acid changes were similar.

The chimpanzee animal model will most likely be used in assessing experimental therapies and vaccines. Recombinant envelope proteins have already been used in one vaccine study (14) and may be used in future studies. Therefore, a better understanding of the humoral and cellular immune response to the envelope proteins and the genetic drift of HCV in this animal model may aid in the interpretation of such important studies.