INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Recently, a new virus related to flaviviruses, the hepatitis G virus (HGV), also called GBV-C virus, was identified by recombinant DNA techniques and was initially related to posttransfusion hepatitis (13, 25). HGV infection prevalence among blood donors from the United States has been estimated to be 1% (8), while the prevalence among high-risk groups, coinfected with HCV, can reach 15 to 20% (1, 13, 25). Parenteral exposure has been documented to be the main mode of transmission for HGV, and vertical transmission has already been reported (28). It is not known, however, if other transmission routes for HGV might be operating.

Most HGV epidemiological studies have been conducted in developed countries; information on HGV infection in developing countries, and particularly in isolated communities is scarce. In a previous study, we analyzed a cohort of hemodialysis patients in Venezuela with high risk for parenterally transmitted hepatitis viruses, and a high prevalence and incidence of HCV infection was found (21). Another population in Venezuela with a high prevalence of hepatitis viruses are Amerindians, among whom a high HBV prevalence with delta superinfection has already been documented (7). These two groups differ in the degree of access to medical care; the hemodialysis patients are subjected to multiple contacts and are susceptible to nosocomial and iatrogenic transmission, while for West Amerindians medical assistance is scarce. Thus, it was of interest to evaluate HGV infection among these two groups. On the other hand, the hepatotropic nature of HGV and its real pathogenic role is controversial (1-3, 8, 10, 14, 29), and additional studies are needed to evaluate the relationship of HGV infection to liver disease of unknown etiology. The aim of this study was to analyze HGV infection among hemodialysis patients, West Amerindians, and patients with non-A, non-B, non-C hepatitis (NABCH).

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Patients. Sera from hemodialysis patients (n = 64 [representing 25% of the cohort]) from four units in Caracas, Venezuela, were studied. A high prevalence of HCV antibodies (71%) was found in a previous study (21). The subset of sera chosen for the evaluation was almost equally distributed among the four units and was previously tested for HCV RNA. Among them, 38 were positive for HCV antibodies and/or HCV RNA and 24 were positive for hepatitis B surface antigen. Follow-up sera from HGV-positive patients were collected 1 and 2 years later. Sera from West Yukpa Amerindians (n = 80), a population with a high rate of HBV infection but negative for HCV antibodies (tested by at least two different second- and third-generation assays) (4), were also studied. Sera from 11 patients requesting routine PCR diagnosis for HCV RNA who were negative for any HAV, HBV, or HCV serological markers were also tested. Eight of these patients, three of whom had evidence of chronic liver disease upon biopsy, had had elevated alanine aminotransferase (ALT) levels for more than 6 months. The other three patients presented with transient ALT elevation.

PCR. HGV infection was determined by reverse transcription-PCR with random primers for cDNA priming, according to previously reported procedures for HCV (21). Nested PCR was performed with primers 874, 877, 875 and 876 from a highly conserved region of the putative NS5 protein (9). Negative controls were added in each step of the procedure (extraction, cDNA synthesis, and PCR). A sample was considered positive when found repeatedly positive after amplification of newly extracted material. The envelope region of the HGV genome was also assayed for amplification, with primers YK1090 (5'-GGT CAC GCC CCT TTG ACT AC-3') and YK1093 (5'-AGC CGA GGC CCC ACG CCG CAC C-3') for the first round of PCR and primers YK1091 (5'-GTT GAC TTG GCA GAC CTG CTC-3') and YK1092 (5'-AGC CGA GGC CCC ACG CCG CAC C-3') for the second round.

SSCP. For single-strand conformation polymorphism (SSCP), PCR-amplified products (1 to 2 µl) were incubated in 90% formamide for 10 min at 95°C and analyzed by 7% polyacrylamide gel electrophoresis (with a 5% stacking gel) at 160 V for 3 h. Gels were stained with silver nitrate (19).

Restriction analysis. For restriction analysis, NS5 amplified products were digested with 10 units of restriction enzyme BstUI, HinfI, ScrFI (New England Biolabs, Beverly, Mass.), or MvaI (Amersham, Cleveland, Ohio), as described previously (22).

Sequence analysis and cloning. Purified PCR fragments were sequenced by the dye terminator labeling method (ABI PRISM dye terminator cycle sequencing ready reaction kit; Perkin-Elmer, Foster City, Calif.) with a model 373 DNA sequencer (Applied Biosystems, Foster City, Calif.). Both strands of DNA were sequenced. NS5 PCR-amplified products (S2 and S2" specimens) were cloned with the pmos Blue kit (Amersham). White colonies (in which the cloned DNA is inserted within the -galactosidase gene) were directly amplified to detect the cloned fragments. These amplified fragments were digested with 10 units of ScrFI, and two clones of each variant were also sequenced with the Sequenase PCR product sequencing kit (Amersham).

Nucleotide sequence accession numbers. Nucleotide sequence data have been deposited into the GenBank database under accession no. U69615 to U69621 and U97577 to U97579.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

For the detection of HGV infection, reverse transcription-nested PCR was performed for 155 individuals. A 402-bp fragment corresponding to an NS5 conserved region was amplified in the sera from nine patients (Table 1). None of the HGV-positive patients had a history of blood transfusion when HGV RNA was first detected in their serum. A total of 64 hemodialysis patients attending four units in Caracas were tested. HGV RNA was found in 9% of patients attending two different units (A and D [Table 1]). Four of six HGV-positive hemodialysis patients were coinfected with HBV and/or HCV, and all viral genomes were found circulating in the respective sera. Only one of the other two hemodialysis patients who were HCV and HBV negative had high ALT values in one of the two serum samples taken 1 and 2 years later. This was the only patient who had elevated ALT values. No association between HGV RNA positivity and the presence of any HBV or HCV marker was found (data not shown). Additionally, 2 of the 80 West Amerindian sera tested were found to be positive for HGV RNA, as well as one NABCH patient exhibiting a transient elevation of ALT (Table 1).

On follow-up, HGV RNA was found in the sera of all of the hemodialysis patients and of patient S2. SSCP analysis showed differences among all of the PCR-amplified fragments from sera from different patients. The same patterns were observed on follow-up of hemodialysis patients, while a difference was observed in the PCR-amplified products of patient S2 2 years later (Fig. 1). In some SSCP patterns, a slight heterogeneity was observed (patient A6 [Fig. 1]).

View larger version (37K): [in this window] [in a new window]   FIG. 1.   SSCP analysis of HGV amplified products. x', serum 1 year later; x", serum 21 months or 2 years later.

The sequences of the amplified NS5 regions of the Venezuelan isolates had 85 to 90% identity with other reported sequences (Fig. 2). Only 0 to 5 substitutions in the analyzed 291-nt fragment were found in follow-up sera taken over a period of 2 years. Moreover, no change was found in the envelope region of the amplified product of one hemodialysis patient (A6; accession no. U97577) over a period of 2 years. The stability of the hemodialysis patient isolates over time is compatible with the SSCP patterns observed (Fig. 1). Significantly higher divergence (88 to 90% identity) was observed among different hemodialysis patients, even from the same unit, than for a single patient observed over 2 years (99 to 100% identity) (Fig. 2). NS5 sequences of HGV circulating in the two Amerindians were more closely related to each other (98%) than any others (Fig. 2). At the amino acid level, most of the nucleotide changes corresponded to silent substitutions. Changes were observed among all of the HGV isolates at six different positions, most of these substitutions being conservative (data not shown).

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Phylogenetic tree derived from the sequences of NS5 PCR-amplified regions from HGV circulating in infected patients, based on the analysis of 291 nucleotides (nt) (7315 to 7605) by the neighbor-joining method with CLUSTAL V. x', serum 1 year later; x", serum 21 months or 2 years later. Other reported sequences with different geographic origins were also included: PNF2161 (HGV prototype sequence; accession no. U45966), GBV-C (U36380), Iw (D87255), and C294 (U75356).

The one exception to the observed stability was the follow-up HGV isolate from patient S2 taken at 21 months, in which 33 mutations were found in the 291-nt fragment (Fig. 2 and 3A). Predictions of restriction enzyme sites based on sequence analysis indicated restriction patterns for the major variant circulating in S2 that were different from the one circulating in S2" (Fig. 3B). When S2 and S2" amplified specimens were cloned, all of the 16 clones derived from the S2 PCR product exhibited restriction patterns characteristic of the S2 major variant, while all of the 16 clones derived from the S2" amplified product bore restriction patterns characteristic of the S2" variant (Fig. 3C). Two of the 16 clones from the S2 PCR product and 2 of the 16 clones from the S2" PCR product were sequenced, confirming the restriction analysis (data not shown). In the digestion patterns of the S2 PCR product, with four different enzymes, faint bands were observed in each case, corresponding to the restriction pattern of the S2" variant. The restriction patterns from use of MvaI and ScrFI are shown in Fig. 3C. These results suggest that the minor S2" variant was present in the S2 specimen but at a low concentration. In contrast, no evidence of an S2 major variant could be obtained in S2", either by restriction analysis (Fig. 3C) or by cloning.

View larger version (39K): [in this window] [in a new window]   FIG. 3.   (A) Sequences of the NS5 PCR-amplified regions from patient S2 and S2" isolates, showing the restriction sites for BstUI, HinfI, MvaI, and ScrFI. (B) Expected restriction patterns for each enzyme in both isolates. nt, nucleotides. (C) Restriction enzyme analysis of HGV isolates from S2 and follow-up samples (S2"). PCR-amplified fragments from the NS5 regions (S2 or S2") (serum) and PCR-amplified fragments of the clones derived from these amplified fragments (clone) were digested with MvaI or ScrFI. One of the 16 clones (bearing the same restriction pattern) derived from each specimen is shown. The S2" pattern was also observed as a minor variant in S2 by using BstUI or HinfI (data not shown).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

HGV active infection was demonstrated among high-risk groups in Venezuela. This work corroborates recent studies which have shown that hemodialysis patients are also at risk for acquisition of HGV (15, 27). The HGV prevalence found in this particular cohort of hemodialysis patients was, however, lower than those previously reported, and significantly lower than that observed in this group, for HCV infection or HBV serological markers (21). The actual prevalence of HGV infection in these Venezuelan populations might be higher than that detected in this study, as only one region was targeted for PCR and since testing in the 5' noncoding region and in the NS5 region of HGV has been recently recommended to ensure appropriate sensitivity (23). However, the primers used for amplification of the NS5 region have proven to be highly reliable for HGV detection by PCR (9). Moreover, it has been recently shown that PCR in the NS5 region can be as sensitive as (6) and even more sensitive than (24) other regions of the HGV genome, suggesting that this region is in fact optimal for HGV RNA detection without further analysis of other regions. Another possible explanation is that since only active infection was detected in this study, some of the hemodialysis patients might have already cleared the virus from the circulation (24). Recent studies which have detected anti-HGV envelope antibodies have shown that a higher prevalence of antibodies against HGV than presence of active viremia can be observed among a specific population (26).

None of the HGV-positive patients had a history of blood transfusion when HGV was first detected in their sera, suggesting that other parenteral procedures may be responsible for the transmission of HGV. In this particular cohort of hemodialysis patients, the high prevalence and incidence of HCV infection described previously (21) suggest that nosocomial transmission may have played a role in the dissemination of HCV among these patients and might also be a mechanism to explain the dissemination of HGV. However, we could not find molecular evidence of nosocomial transmission among the patients studied. Analysis of more patients would be needed to determine if nosocomial transmission of HGV was occurring in this group. Some evidence of patient-to-patient transmission has already been observed by others in hemodialysis patients (15, 27).

HGV infection did not seem to induce significant ALT elevation in the patients tested. Although hepatitis viruses in general do not induce ALT elevations in hemodialysis patients (21), this observation is in agreement with a previous observation that HGV infection is not always correlated with ALT elevations (13), which has even raised the question about its real pathogenic role (1-3). More biochemical and histological studies are needed to address more properly the question of the pathogenicity of HGV.

In the patients followed up, HGV was associated with long-lasting viremia, as previously described (13, 15, 25). Significant sequence variations did not occur over a 2-year period for any of the hemodialysis patients, but they did occur in the NABCH patient. SSCP proved to be a valuable tool to detect these changes. The significant variation observed in this isolate after 21 months suggests the selection of a minor variant instead of the accumulation of multiple substitutions. Indeed, restriction enzyme analysis suggested a mixed infection in patient S2, with clearance of the dominant variant in the follow-up sample. Mixed infections of HGV have already been suggested (17). Selection of a minor variant during the evolution of HCV infection has been described (11). In addition, mixed HCV infection seems to lead to clearance of one of the infecting strains, and viral interference has been suggested as one of the possible mechanisms mediating this effect (12). No additional follow-up specimens from patient S2 were available to determine whether the phenomenon observed in S2 was due to clearance of a major infecting strain or intermittence of two strains.

We speculate that the two variants present in patient S2 may also differ in other, more variable regions of the HGV genome, like the envelope region, and that some degree of immune pressure of the host may have selected the minor S2" variant. Indeed, the S2 E2 region was successfully amplified with the nested primers used, while S2" was not (data not shown), suggesting some genetic differences between the two isolates in this region also and confirming the apparent complete clearance of the S2 variant in the follow-up serum. A correlation between the time of clearance of HGV RNA and the appearance of anti-E2 antibodies in the infected host has been shown recently (20, 26). It is not known, however, to what extent this putative immune clearance might be effective only against a homotypic strain, like HCV, or against a heterologous strain as well. In contrast, no change was found over a period of 2 years in the envelope regions of HGV isolates from one hemodialysis patient. A significantly higher stability over time of the HCV envelope region has been found in virus circulating in immunocompromised patients than in immunocompetent patients (5, 18). Sequence analysis of more envelope regions is needed to answer these questions for HGV-infected patients. By SSCP, some heterogeneity was also observed in PCR-amplified products from hemodialysis patients, suggesting that even in these immunocompromised patients, variants other than the dominant one might be circulating despite the apparent higher stability of HGV than HCV (9, 15).

As stated before, West and South Venezuelan Amerindians have high HBV prevalence, while HCV infection seems to be absent among West Amerindians, probably because of the lack of parenterally associated medical care (4). HGV, however, appeared to be present in this population. Thus, other mechanisms might be occurring that allow the penetration of HGV in this community. Ritual percutaneous practices have been cited for the transmission of HBV and HDV in these ethnic groups (7). Further studies of more Amerindian HGV-positive specimens are needed to analyze their 5' noncoding regions (16, 17) and to evaluate the genetic variability of HGV in these isolated communities.