Repeated Passage of Wild-Type Woodchuck Hepatitis Virus in Lymphoid Cells Does Not Generate Cell Type-Specific Variants or Alter Virus Infectivity

Abstract

Woodchuck hepatitis virus (WHV), which is closely related to human hepatitis B virus, infects the liver but also invariably establishes persistent infection in the lymphatic system. Although the dose of invading virus appears to be the main factor in determining whether WHV infection is restricted to the lymphatic system or also engages the liver, the nature of WHV lymphotropism remains unclear and a role for a specific lymphotropic variant was not excluded. The availability of woodchuck lymphocyte and hepatocyte cultures susceptible to WHV infection allows investigation of this issue in vitro. We hypothesized that repeated passage of wild-type WHV in lymphoid cells should lead to enrichment of a lymphotropic virus variant, if in fact such a variant exists. For this purpose, wild-type WHV with a homogeneous sequence was used as the inoculum, while lymphoid cells from a single healthy woodchuck donor and a normal woodchuck WCM-260 hepatocyte line served as infection targets. The serial passage of the wild-type virus repeated up to 13 times for both cell types did not lead to the emergence of cell type-specific WHV variants, as revealed by sequence analysis of the virus envelope and the core and X gene sequences. Moreover, the virus passaged in both cell types remained infectious for naive woodchucks, produced infection profiles that depended upon virus dose but not on virus cellular origin, and retained its initial DNA sequence. These results imply that WHV lymphotropism is a natural propensity of the wild-type virus and is not a consequence of infection with a viral variant.

Accumulated evidence indicates that hepadnaviruses are capable of replication not only in hepatocytes but also in cells of the immune system. In regard to symptomatic, chronic hepatitis B virus (HBV) infection, which is characterized by the continuous presence of HBV surface antigen (HBsAg) in serum and protracted liver necroinflammation, long-term persistence of virus has been shown in both the liver and the lymphatic system (6, 21, 26, 35, 42, 43, 45). However, liver and lymphoid cells have also been found to be the sites of HBV genome carriage in serum HBsAg-negative patients in whom acute hepatitis (AH) had apparently resolved completely when sensitive PCR-based assays were applied to detect the virus genome (3, 4, 30, 37, 40, 46). This occult form of HBV infection might be a source of infectious virus available for transmission to healthy individuals through blood or organ donations (1, 12, 21, 24, 27), as well as a potential cause of diseases of seemingly unknown etiology engaging the liver and the lymphatic system. In this regard, retrospective studies of hepatocellular carcinoma (HCC) of previously undetermined etiology showed low levels of HBV genomes in 63.5% of liver tissue samples tested (38). In some of those cases, integration of viral DNA into the host's genome was detected. This implies that trace amounts of HBV, identifiable by molecular assays of a sensitivity greater than those currently used in clinical laboratories, may retain pathogenic and oncogenic potential.

In a series of studies employing eastern North American woodchucks (Marmota monax) infected with woodchuck hepatitis virus (WHV) as the model of HBV infection and hepatitis B, we have shown the existence of two forms of persistent occult hepadnaviral carriage, designated primary occult infection (POI) and secondary occult infection (SOI) (9, 29, 31; reviewed in references 26 and 28). The SOI, continuing for life after the resolution of AH, was found to be associated with development of HCC in about 20% of the animals affected (31), whereas it remains undetermined whether POI can also be accompanied by the development of hepatoma. We have also documented that transmission of small amounts of virus from dams with SOI led to the establishment of POI in their offspring (8). Furthermore, this silent form of infection, unidentifiable by serum immunovirological markers, including antibodies to WHV core antigen (anti-WHc), can be experimentally induced in adult woodchucks by intravenous (i.v.) inoculation with WHV doses equal to or lower than 10³ virions (29). Most interestingly, the liver is not engaged in POI, at least initially, while cells of the lymphatic system, both those that circulate and those that reside in lymphatic organs, are infected (8, 29). These results implied that this extrahepatic form of WHV infection is due to invasion with a low dose of virus rather than a consequence of infection with an organ-specific (lymphotropic) viral variant. Nevertheless, the reason the development of WHV infection was restricted to the lymphatic system was not fully explained. In this context, it is necessary to emphasize that the replication of WHV in the cells of the woodchuck immune system has been thoughtfully documented by identifying the WHV genome and its replication intermediates and the expression of virus proteins, by enumerating lymphoid cells carrying intracellular WHV genomes, and by demonstrating infectivity and pathogenicity of the virus types residing in and produced by these cells (8, 23, 29, 31, 34; reviewed in reference 28).

In order to account for the possibility that a relatively minor subpopulation of WHV might be responsible for the infection of the lymphatic system, we serially passaged a wild-type WHV with a precisely defined genomic sequence in lymphoid cells in an attempt to enrich any potentially existing cell-specific virus variants. For this purpose, we developed an in vitro system whereby WHV could be serially passaged in normal woodchuck lymphoid cells or hepatocytes. The results showed that the ability of wild-type virus to infect cells of the lymphatic system is an inherent property and not due to the existence or emergence of a specific viral variant. As well, WHV retrieved after multiple serial passages in either lymphoid cells or hepatocytes remained infectious to virus-naive animals. The serological and molecular profiles of the induced infection correlated with the dose of virus administered but not with the cell type from which the virus originated.

MATERIALS AND METHODS

Animals.

Three woodchucks chronically infected with comparable serum WHV loads of ∼10¹⁰ virus genome equivalents (vge)/ml were used to isolate splenocytes and to prepare WHV for subsequent in vitro experiments. Thus, woodchuck WM.A, a chronic WHV carrier, was infected in the wild and showed sustained serum WHV surface antigen (WHsAg)-positive infection during the 30-month period before autopsy. The WHV inoculum prepared from this animal was used for the WHV passage experiments 1a and 1b (Table 1). Two other animals, designated WF.B and WM.C, were i.v. infected with WHV in our colony. Both animals developed classical, serum WHsAg-reactive chronic hepatitis (CH), which was confirmed by histological examination of serial liver biopsy specimens. Specifically, WF.B was infected with ∼10⁶ DNase-protected WHV vge, using well-defined WHV/tm3 inoculum (GenBank accession number AY334075) (29). The animal became serum WHsAg positive at 1 month postinfection (p.i.), and the antigen persisted for its lifetime. The WHV derived from splenocytes of WF.B was used for passage experiments 2a and 2b (Table 1). Animal WM.C developed chronic WHV infection after receiving an inoculation with WHV/tm2 inoculum containing ∼10¹⁰ DNase-protected vge (GenBank accession numbers AY628095 and AY628096) (9). WHV derived from the splenocytes of WM.C was used for multiple serial passage experiments 3 and 4 (Table 1). It is notable that the full WHV genome sequences from the liver, spleen, and serum of WM.C, whose pooled serum samples served as the WHV/tm3 inoculum and was used to infect WF.B, were determined to be that of the wild type and were entirely homogenous at all three locations. The GenBank accession numbers for complete WHV sequences from the liver, spleen, and serum of WM.C are AY334076, AY334077, and AY334075, respectively (29). Comparison of the complete WHV sequences from inocula WHV/tm2 and WHV/tm3 revealed that they differed only by the substitution of nine nucleotides. Seven of these alterations confer changes in the virus amino acid sequence, and three of the changes occurred in the X open reading frame (ORF) (i.e., A114T, T127I, and P141L); two occurred in the precore region of the core (C) ORF (i.e., T90A and Y125H); and two occurred in the polymerase gene (i.e., N290Y and T645M). Amino acid positions were enumerated according to their locations in the WHV/tm3 X, C, or P ORF. It is also notable that while there were single nucleotide differences between the WHV in the WHV/tm2 inoculum and the virus occurring in the liver and spleen of the animal which provided this inoculum, the virus isolates from the WHV/tm3 inoculum and from the liver and spleen of WM.C have identical sequences. This implies that the variation in the virus sequences became more concordant following passage of WHV/tm2 in animal WM.C.

TABLE 1.

WHV passage experiments performed in the course of this study

Experiment	Animal source of WHV inoculum	WHV cell target	Number of WHV passages
1a	WM.A	Lymphoid cells	3
1b	WM.A	Lymphoid cells	3
2a	WF.B	Lymphoid cells	3
2b	WF.B	Lymphoid cells	3
3-L	WM.C	Lymphoid cells	13
3-H	WM.C	Hepatocytes	13
4-L	WM.C	Lymphoid cells	12
4-H	WM.C	Hepatocytes	12

In addition, a healthy animal, WM.0, served as the source of peripheral blood mononuclear cells (PBMC) for in vitro WHV infection experiments, whereas two other healthy woodchucks, WF.1 and WF.2, were used to determine the in vivo infectivity of WHV recovered after multiple passages in lymphoid cells or hepatocytes (see below and Table 1). Randomly selected samples of sera and PBMC and liver biopsy samples collected from these woodchucks prior to the inoculation or the acquisition of PBMC were nonreactive for serological markers of WHV infection, i.e., WHsAg, anti-WHs, and WHV DNA, as determined by PCR/nucleic acid hybridization (NAH) assays, as previously described (29, 31).

Cells.

Splenocytes, composed mainly of lymphocytes, were isolated from WM.A, WF.B, and WM.C, using a metal 50-mesh cell dissociation sieve. After cells were separated on a Ficoll-Paque gradient (Pharmacia, Uppsala, Sweden) (18, 29), they were washed extensively with Hanks' balanced salt solution (HBSS; Invitrogen, Carlsberg, CA) and DNase/trypsin/DNase treated, as described previously in detail (34). Circulating lymphoid cells were isolated from the healthy animal WM.0 on a Ficoll-Paque density gradient, as reported previously (31). These cells served as targets for in vitro WHV infection in experiments 1, 2, 3-L, and 4-L (Table 1).

The woodchuck hepatocyte line WCM-260 was used as a control WHV infection target in experiments 3-H and 4-H (Table 1). This line was derived from primary hepatocytes isolated by collagenase microperfusion from a liver biopsy specimen of a healthy adult animal (11). WCM-260 cells were propagated in hepatocyte medium consisting of 80% Hepato-STIM with 10 μM dexamethasone, 10 ng/ml epidermal growth factor, 2 mM l-glutamine, 50 μg/ml penicillin, and 50 μg/ml streptomycin (all from Becton Dickinson, Rutherford, NJ) plus 20% (vol/vol) HepG2 cell culture supernatant, as described previously (7, 11).

Preparation of splenocyte-derived WHV.

Splenocytes from WM.A, WF.B, and WM.C were treated with DNase/trypsin/DNase (34) and washed three times in HBSS to remove any potentially attached WHV virions, WHV DNA fragments, or cellular debris carrying WHV DNA reactivity, and the cells were recovered by centrifugation. The cells were supplemented with 5 ml of hepatocyte culture medium at ∼3 × 10⁶ cells/ml and cultured in a 25-cm² tissue culture flask (Corning Costar Corp., Cambridge, MA). Supernatants obtained after a 72-h culture from WM.A and WF.B splenocytes were clarified by centrifugation at 4,000 rpm for 30 min in a 5810R Eppendorf centrifuge (Brinkmann Instruments Ltd., Mississauga, ON, Canada) and used as the inocula in experiments 1 and 2, each of which was performed in duplicate (Table 1). WHV from splenocytes of animal WM.C was prepared by collecting culture medium every 72 h during a 12-day culture. The supernatants were pooled and clarified by centrifugation, and WHV content was determined and used as the inoculum in experiments 3 and 4 (Table 1). Samples (1 ml) of each supernatant were preserved for WHV DNA analyses. In some instances, samples of the supernatants were also subjected to digestion with DNase to assess the presence of envelope-protected virions, as reported previously (34). Each splenocyte supernatant was assessed for the absence of residual cellular debris by phase-contrast microscopy. In addition, PCR using primers specific for human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) DNA (36), which were also found to amplify woodchuck GADPH DNA, was performed with DNA isolated from the splenocyte culture supernatant clarified as indicated. This testing gave negative results, confirming that no cellular debris were carried over.

Multiple passage of WHV in lymphoid cells and hepatocytes.

PBMC from WM.0 (∼1 × 10⁷) or WCM-260 hepatocytes (∼9 × 10⁵) were seeded 24 h prior to infection in 25-cm² tissue culture flasks in 5 ml of hepatocyte culture medium. The medium was removed and replaced with 4 ml of culture supernatant from WHV-infected splenocytes supplemented with 1 ml of fresh hepatocyte culture medium. These supernatants carried between 1 × 10⁵ and 3 × 10⁵ vge as determined by real-time PCR. The cells were incubated for 18 h at 37°C in a humidified 5% CO₂ atmosphere (Fig. 2). Then, the inoculum was removed, and the cells were washed and treated with DNase/trypsin/DNase to remove any potentially attached virions, free WHV DNA, or cellular debris carrying WHV DNA reactivity (34). The cells were washed, resuspended in 5 ml of fresh hepatocyte culture medium, and cultured for 48 h. Both adherent and nonadherent lymphoid cells were harvested, whereas hepatocytes were released by a brief treatment with trypsin-EDTA (Invitrogen). The recovered cells were DNase/trypsin/DNase treated and washed in HBSS, and DNA was extracted (Fig. 2). Final cell wash fluids were saved to evaluate the completeness of removal of extracellular virus material.

FIG. 2.

Schematic outline of the methodology used for serial passage of WHV in woodchuck lymphoid cell or hepatocyte cultures. X 3-13, passaged 3 to 13 times.

For virus passage, 4-ml volumes of the supernatants recovered from lymphoid cell or hepatocyte cultures were spun down at 4,000 rpm for 30 min to remove cellular debris and used for inoculation of either freshly prepared naive lymphoid cells or hepatocytes, while a 1-ml sample was preserved for analysis. In experiments 1 and 2 (Table 1), three consecutive passages of WHV were performed using culture supernatants from splenocytes of WM.A and WF.B for first-round (direct) infection. For multiple passages in experiments 3 and 4, WHV derived from splenocytes of WM.C was used. For each of the experiments, culture supernatants obtained from infected cells from the preceding passage were used to infect virus-naive lymphoid cells or hepatocytes in the next passage, as outlined in Fig. 2.

DNA extracted from 1-ml samples of randomly selected culture supernatants was tested for GAPDH DNA by PCR to assess for a possible carryover of cellular debris, as indicated above. The results showed that all supernatants tested were free of residual cell fragments.

Detection of WHV DNA, cccDNA, and mRNA.

WHV DNA content in lymphoid cells, hepatocytes, cell supernatants, and cell wash fluid was determined by direct and, if required, nested PCR, followed by NAH analysis of amplified products as described previously (29, 31). In brief, total DNA was extracted from 5 × 10⁶ to 1 × 10⁷ cells or from 200 μl of cell culture supernatants or sera by proteinase K digestion, phenol-chloroform extraction, and ethanol precipitation (31). For detection of WHV DNA, 1 μg of DNA from cells and DNA from 200 μl of cell culture supernatant or serum was amplified using primers specific for WHV C, surface (S), and X gene sequences. WHV covalently closed circular DNA (cccDNA) in cultured or control in vivo-infected cells was detected using 4 μg of total DNA treated with a single-strand-specific mung bean nuclease prior to PCR, with primers spanning the WHV nick region, as reported in detail previously (9, 23). For the detection of WHV mRNA, 2 μg of total RNA was DNase I (Sigma-Aldrich, St. Louis, MO) treated and then reverse transcribed under conditions reported previously (8, 9). The resulting cDNA was amplified by PCR with WHV C gene-specific primers. Each test RNA sample treated under the same conditions in the absence of reverse transcriptase (RT aliquot) served as a DNA contamination control. In addition, RNA samples extracted from PBMC of a healthy woodchuck and from an animal with serum WHsAg-positive chronic infection were used as negative and positive controls, respectively. RNA isolated from 400-μl samples of culture supernatants of WCM-260 and lymphoid cells infected in vitro with WHV were also tested. Amplicons were analyzed by agarose gel electrophoresis and, after they were blotted, were hybridized to complete recombinant WHV (rWHV) DNA as a probe to confirm the specificity of reactions and validity of controls. In some instances, real-time PCR using a Roche LightCycler (Roche Diagnostics, Mannheim, Germany) was employed to quantify the amount of WHV DNA in culture supernatants, as described previously (29).

Preparation of culture supernatants after WHV final passage.

In order to recover the entire amount of WHV produced after the final passage in either lymphoid cells (experiment 3-L) or WCM-260 cells (experiment 4-H), culture supernatants were ultracentrifuged at 200,000 × g for 24 h at 4°C in a Beckmann SW50.1 rotor. The pellet was resuspended in 500 μl of sterile phosphate-buffered saline (pH 7.4), and a 100-μl sample was used for DNA extraction. The remaining 400-μl portion of each of the two preparations was injected separately into two healthy woodchucks, as outlined below.

Inoculation of woodchucks with WHV recovered after repeated passages in lymphoid cells or hepatocytes.

Animals WF.1 and WF.2 were injected i.v. with WHV obtained from the final passages, after experiments 3-L and 4-H, respectively. The WHV DNA content was determined by real-time PCR. Thus, a total of 1.5 × 10⁴ vge was found in the lymphoid cell supernatant after experiment 3-L, and 8.8 × 10² vge was found in the supernatant from infected hepatocytes after experiment 4-H. Sera and PBMC samples were collected from both woodchucks before they were inoculated; samples were collected weekly for up to 10 weeks p.i. and then biweekly. Liver biopsy specimens were obtained prior to infection and at 6 weeks p.i. Animals were monitored for up to 7 months p.i. by testing for serological markers of WHV infection and for WHV DNA content in the sera, PBMC, and liver biopsy specimens. WHV DNA load was also determined for the lymphatic organs and for the hepatic tissue collected at autopsy.

WHV DNA sequencing.

For complete WHV genome sequence analysis, DNA derived from inocula and selected woodchuck sera, lymphoid cells, and tissue samples were subjected to PCR with back-to-back primers amplifying the whole WHV genomes (29, 31). When sequencing required it, fragments of the full-length WHV DNA were further amplified with WHV gene-specific primers, as reported elsewhere (8). In those cases where a partial sequence analysis of WHV genome was performed, WHV pre-S, S, C, and X gene fragments amplified by PCR were either cloned, using a TOPO TA cloning system (Invitrogen), and up to 10 individual clones were automatically sequenced or sequenced directly, using an fmol sequencing kit (Promega, Madison, WI). Thus, the pre-S/S and C gene sequences of WHV recovered in experiment 2b after passage three were directly sequenced, while the pre-S/S, C, and X gene sequences of virus recovered from the final lymphoid cell and hepatocyte culture supernatants obtained after experiments 3-L and 4-H, respectively, which were used to infect woodchucks WF.1 and WF.2, were obtained by clonal sequencing. In addition, the pre-S/S, C, and X gene PCR products of WHV from sera collected at autopsy from animals WF.1 and WF.2 were also subjected to direct sequence analysis and were further reevaluated by clonal sequencing. The resulting sequences were aligned with the sequence of the WHV/tm3 inoculum, which, as indicated above, was identical with those existing in the liver and spleen of WM.C, which provided this inoculum (29).

RESULTS

Infected in vivo splenocytes secrete WHV in culture.

Splenocytes obtained from woodchucks WM.A, WF.B, and WM.C, which were chronically infected with WNV and had comparable virus loads in serum, were treated with DNase/trypsin/DNase and washed extensively to eliminate potentially attached cell surface virions, virus DNA fragments, or WHV DNA-reactive cellular debris. The final cell wash fluids were found to be WHV DNA nonreactive by nested PCR/NAH assay (data not shown), while the splenocytes carried replicating WHV, which was confirmed by the detection of virus cccDNA (as illustrated for WM.C splenocytes in Fig. 5C). The supernatants collected after 72-h culture of these cells contained WHV DNA at the level of ∼3 × 10⁵ vge/ml, as determined by real-time PCR (29). To ensure comparable conditions for in vitro WHV infection of lymphoid cells and WCM-260 hepatocytes, both cell types were maintained in hepatocyte culture medium supplemented with 10 μM dexamethasone. This glucocorticoid has been previously shown to enhance the expression of WHV DNA in lymphoid cells by augmenting virus replication (23). We also found that naturally infected woodchuck splenocytes can be cultured for at least 12 days without altering their viability, extending our earlier observations indicating no loss in cell viability for up to 4 days under comparable culture conditions (23).

FIG. 5.

Detection of WHV DNA in cultured lymphoid cells and WCM-260 hepatocytes and in their culture supernatants collected during serial passage of WHV originating from WM.C splenocytes. DNA extracted from cultured cells (C) and their supernatants (S) were amplified with primers specific for either the WHV C, S, or X gene by direct PCR and, if required, nested PCR and then identified by NAH assay. WHV DNA signals detected after passages 1, 3, 5, and 12 in lymphoid cells (A) and in WCM-260 hepatocytes (B) are shown. For panels A and B, arrows indicate the molecular sizes of the amplicons generated with primer pairs specific for the WHV C gene (623-bp direct and 428-bp nested PCR products), the S gene (1,277-bp direct and 767-bp nested PCR products), or the X gene (386-bp direct and 192-bp nested PCR products). (C) Detection of WHV cccDNA in selected samples of lymphoid cells and WCM-260 hepatocytes obtained after the indicated passages (Pass) of WHV was performed, as described in Materials and Methods and in the legend to Fig. 4. Nested PCR products were analyzed by NAH to confirm the specificity of the 674-bp amplicons. Contamination controls consisted of water added to direct PCR (DW) and nested PCR (NW) mixtures instead of test DNA and of a mock (M) extract prepared and treated as test samples. Total DNA from the serum of a WHsAg-positive chronic WHV carrier served as a positive control in panels A and B, whereas DNA from liver tissue of the same animal served as a positive control in panel C.

For multipassage experiments 3 and 4 (Table 1), splenocytes from WM.C were used as a source of WHV. This virus was thoroughly characterized in our previous studies and was proven to be infective and oncogenic (29, 31). Furthermore, as indicated above, the full-length genomic sequences of WHV obtained from the serum, liver, and spleen of carrier WM.C were analyzed and found to be identical. To generate a WHV pool from cultured WM.C splenocytes, culture supernatants were collected every 72 h for 12 days. As shown in Fig. 1A, the amount of virus released by WM.C splenocytes progressively increased over the course of the culture, except at day 6. The centrifugation-clarified supernatant pool contained ∼3 × 10⁵ WHV vge/ml, as determined by real-time PCR. Importantly, the amount of WHV DNA in this pool remained unchanged after digestion with DNase/trypsin/DNase (Fig. 1B), indicating that the entire WHV DNA detected was encapsidated within intact virions.

FIG. 1.

Detection of WHV DNA in culture supernatants collected during a 12-day culture of splenocytes isolated from woodchuck WM.C with serum WHsAg-positive chronic WHV infection. (A) Splenocytes (1.7 × 10⁷) were treated with DNase/trypsin/DNase and cultured for 12 days, and their supernatants were collected at 3-day intervals. Serial 10-fold dilutions of complete rWHV DNA were used as quantification standards. (B) Effect of DNase digestion on WHV DNA in the pool of the supernatants collected during the 12-day culture of WM.C splenocytes. Samples of the pool were either treated (T) or not (NT) with DNase prior to DNA extraction to determine the presence of WHV genome in the envelope-protected virions. As a control for DNase digestion, 1 ng of rWHV DNA was resuspended in healthy woodchuck serum and was digested or not with DNase before DNA extraction. As the positive control, serum from a WHsAg-positive animal chronically infected with WHV (CH serum) was used. Contamination controls consisted of water (DW) added to the direct PCR instead of DNA and a mock (M)-extracted sample. All samples were amplified by direct PCR using the WHV C gene primers. The specificity of WHV DNA detection and validity of controls were confirmed by Southern blot hybridization. Positive samples showed the expected 623-bp amplicons. Numbers under the panel represent the relative densitometric units (DU) given by the hybridization signals after the PCR products were probed with rWHV DNA.

Splenocyte-derived WHV is infectious to naive lymphoid cells in vitro.

Figure 2 outlines the overall scheme used to serially transmit splenocyte-derived WHV in naive lymphoid cells. WCM-260 hepatocytes were used in parallel as a control, since it was established before that they are susceptible to WHV infection (23). As mentioned, we hypothesized that repeated passage of WHV in lymphoid cells, but not likely in hepatocytes, would enrich a lymphotropic variant, enabling its detection, if such a variant in fact exists.

Experiments 1 and 2 (Table 1) were conducted to establish the methodology for direct WHV infection and serial passage of virus in lymphoid cells. In these preliminary experiments, three sequential passages of virus were performed, and each was done in duplicate. Analysis of WHV DNA in the de novo-infected cells and their culture supernatants demonstrated not only that splenocytes secreted infectious virus but also that the virus released was serially transmitting infection from one lymphoid cell culture to the next, as illustrated for experiment 1a in Fig. 3. Estimated quantities of WHV produced in each passage were between 1 × 10³ and 5 × 10⁴ vge per 5 × 10⁶ cells, although virus levels in cells after passage 2 were usually lower or even apparently undetectable (Fig. 3). Despite this unexplained temporal decrease in WHV expression, it was evident that WHV can be serially passaged in naive woodchuck lymphoid cells without difficulty. Interestingly, such a transient decrease in intracellular hepadnavirus DNA level during the second passage was also found in serial passages of WHV or HBV in hepatocytes, as reported previously (16, 23).

FIG. 3.

Detection of WHV DNA in cultured lymphoid cells and their supernatants collected after three serial passages of WHV derived from splenocytes of woodchuck WM.A. Splenocytes (SPL; 2 × 10⁷) were treated with DNase/trypsin/DNase and cultured for 72 h. The resulting splenocyte supernatant (sSPL) was passaged to virus-naive woodchuck PBMC. Exposed cells and their culture supernatants were collected after a 72-h culture. DNA was extracted from 1 × 10⁶ cells or 200-μl supernatant samples and assayed for WHV DNA by nested PCR/NAH with WHV C gene-specific primers. Two 10-fold dilutions of complete rWHV DNA were used as quantification standards. Water added instead of test DNA to direct PCR (DW) and nested PCR (NW) and a mock (M) sample prepared and treated as test DNA were used as contamination controls. WHV C gene 428-bp amplicons were detected by NAH assay with rWHV DNA as a probe. Relative densitometric unit (DU) values given by the hybridization signals are indicated under the panel.

In addition, active replication of WHV in de novo-infected lymphoid cells and hepatocytes was confirmed by identification of WHV mRNA and cccDNA after the first passage of WHV released from WM.C splenocytes, as shown in Fig. 4. WHV RNA transcripts were detected in the infected cells, but not in their culture supernatants, and only after reverse transcription and not in nontranscribed RNA samples, excluding the possibility that WHV DNA might be a source of false-positive signals (Fig. 4A). In the same cells, WHV cccDNA was readily detected by a high-stringency PCR assay specifically detecting this replication intermediate (Fig. 4B).

FIG. 4.

Detection of WHV mRNA and cccDNA in WCM-260 hepatocytes and lymphoid cells exposed to WHV originating from WM.C splenocytes. (A) Identification of WHV mRNA in cultured hepatocytes and lymphoid cells after passage of WHV released by WM.C splenocytes. Total RNA extracted from cultured cells and 400-μl samples of their supernatants was DNase treated and reverse transcribed (RT +) or not (RT −) prior to PCR amplification with WHC C gene-specific primers. Nested PCR products were analyzed by NAH to validate the specificity of the 428-bp amplicons. (B) Detection of WHV cccDNA in the cells shown in panel A. Total DNA was digested with a single-strand-specific mung bean nuclease prior to amplification with primers spanning the nick region of the WHV genome. Nested PCR products were subjected to NAH to confirm the specificity of the 674-bp amplicons. Contamination controls shown in panels A and B consisted of water added to direct (DW) and/or nested (NW) amplification reaction mixtures instead of test cDNA or DNA and of a mock (M) sample prepared and treated as test samples. In addition, RNA and DNA extracted from PBMC of a virus-naive animal (H-PBMC) and a serum WHsAg-reactive chronic carrier (CH-PBMC) were included as negative and positive controls, respectively.

Furthermore, to recognize whether the three-round passage of wild-type WHV in lymphoid cells might lead to a rise of variations in the WHV genome, the virus pre-S/S and C gene fragments amplified from the cells recovered after the third passage of the virus produced by WF.B splenocytes (Table 1, experiment 2b), as well as the respective virus sequences amplified from the WHV/tm3 inoculum, which was used to infect WF.B, were directly sequenced and compared. This analysis showed the lack of nucleotide variations, suggesting that no WHV variant(s) had arisen.

Multiple serial passages in lymphoid cells or hepatocytes do not lead to the emergence of cell type-specific WHV variants.

Since three consecutive passages of WHV did not lead to the appearance of variants, the passage of virus in lymphoid cells or control hepatocytes was extended to 13 times. For this purpose, the same amounts of WHV (∼3 × 10⁵ vge) released by WM.C splenocytes were used as inocula for first-round (direct) infection of either lymphoid cells or hepatocytes (Table 1, experiments 3 and 4). It was found that the virus remained infectious in both cell types up to the end of the passage experiments. Figure 5 illustrates the detection of WHV DNA by PCR using primers specific for nonoverlapping regions of the virus C, S, and X genes after randomly chosen passages in lymphoid cells (Fig. 5A) or hepatocytes (Fig. 5B) and in corresponding culture supernatants. Analysis of WHV cccDNA in selected lymphoid cell and hepatocyte samples obtained from these passages confirmed that the transmitted virus was in fact reestablishing active replication (Fig. 5C).

Estimated levels of WHV DNA detected in de novo-infected cells were in the range of 5 × 10⁴ to 1 × 10⁵ vge, and these levels did not fluctuate meaningfully or increase over time regardless of the cell type. The amount of WHV released during each passage was between 10² and 10³ vge; however, in some passages, the levels decreased below the detection limit of the PCR/NAH assay. It is noteworthy that the supernatants which were found to be virus genome nonreactive were in fact WHV DNA positive when the pellets obtained after ultracentrifugation of their pools were reanalyzed by PCR/NAH. This showed that the virus was released but in very small quantities. Nonetheless, these amounts were sufficient to transmit infection and successfully establish replication in virus-naive lymphoid cells or hepatocytes.

The likelihood of the appearance or the enrichment of cell type-specific WHV variants was examined by determining sequences of the WHV pre-S/S and C gene amplicons generated from supernatants collected after the third passage in experiments 3-L and 4-H and those of the WHV pre-S/S, C, and X gene PCR products obtained from culture supernatants recovered after the final passages of virus in either lymphoid cells (experiment 3-L) or hepatocytes (experiment 4-H) (Table 1). In the first instance, amplicons of viral DNA from supernatants recovered after passage three were analyzed and compared with the respective sequences from the WHV/tm3 inoculum. This analysis showed no variations. Since it was considered that the emergence of a cell type-specific variant(s) might require a greater number of passages, the supernatants from the final passages from experiment 3-L (passage 13) and experiment 4-H (passage 12) were analyzed. Direct sequencing of the respective PCR products showed, again, no changes in the nucleotide sequence compared to those of the wild-type virus. To confirm this observation, further investigation included analysis of the sequences of multiple clones (usually 9 or 10 clones), which were generated from the PCR amplicons of the WHV pre-S/S, C, and X genes. Alignment of the cloned pre-S/S and C sequences with the WHV sequence of WHV/tm3 showed minor random nucleotide variations, which were unique to singular clones and were not shared across the clones derived from the particular amplicon (data not shown). However, a few single nucleotides in the X gene sequence, which were evident in all clones analyzed, varied from those of WHV/tm3. Thus, among the eight nucleotide substitutions detected in the 192-bp X gene fragment, three were nonsynonymous changes, i.e., A1596C, C1646A, and T1718C, while five substitutions resulted in amino acid changes, i.e., S34A, L45P, A47P, V72F, and E77D (amino acid positions were enumerated according to their locations in the WHV/tm3 X ORF). Nonetheless, these changes occurred regardless of whether the virus originated from lymphoid cells or hepatocytes, suggesting that they may reflect adaptive changes associated with virus survival under culture conditions. Interestingly, all these changes were reverted to the wild-type sequence upon passage of the virus in animals (see below). Taken together, these results revealed that no cell-specific variant had arisen during repeated passage of WHV in either lymphoid cells or hepatocytes.

In vivo infectivity of WHV recovered after serial passage in lymphoid cells or hepatocytes.

Since the splenocyte-derived virus was infectious to cultured lymphoid cells and hepatocytes, it was of interest to determine if the passaged virus also retained its in vivo infectious potential. For this purpose, the final supernatants obtained after passage of WHV from WM.C splenocytes in lymphoid cells (experiment 3-L) or hepatocytes (experiment 4-H) were concentrated by ultracentrifugation, and the recovered virus was injected into the healthy animals WF.1 and WF.2. Hence, WF.1 was inoculated i.v. with ∼1.5 × 10⁴ vge obtained after multiround passage in lymphoid cells and WF.2 with ∼8.8 × 10² vge recovered after passage in hepatocytes.

Figure 6 illustrates the profiles of serological markers of WHV infection and detection of WHV DNA in sequential serum and PBMC samples and in liver biopsy specimens collected over a 7-month p.i. follow-up with WF.1 and WF.2. Regardless of the source of WHV inoculum and the detection or not of WHsAg and anti-WHc in serum, WHV DNA appeared within 2 weeks after inoculation and persisted throughout the entire follow-up in the circulation, the PBMC, and the hepatic tissue in both animals. Noticeably, the virus passaged in lymphoid cells induced AH in WF.1, with detectable serum WHsAg and anti-WHc. On the other hand, WF.2, which was injected with WHV passaged in hepatocytes, displayed only molecular evidence of WHV infection. As we have previously reported, the difference between serologically evident disease in WF.1 and serologically silent infection in WF.2 can be explained by the differences in the amounts of WHV administered (29). The amount was about 100-fold greater for WF.1 than for WF.2. It was demonstrated previously that WHV doses equal to or greater than 10³ WHV vge cause serologically evident infection accompanied by AH, while doses lower than 10³ vge induce occult infection, progressing in the absence of detectable serum WHsAg, anti-WHc, or anti-WHs.

FIG. 6.

Serological and molecular profiles of WHV infection in woodchucks injected with virus derived from WM.C splenocytes, which was repeatedly passaged in cultured lymphoid cells or hepatocytes. The indicated amounts of WHV (vge) obtained after 13 passages in lymphoid cells (experiment 3-L) or 12 passages in hepatocytes (experiment 4-H) were injected i.v. at week 0 into woodchuck WF.1 or WF.2, respectively. Vertical bars show detection of WHsAg (light gray bar), anti-WHc (dark gray bar), and WHV DNA in sequential sera and PBMC samples. Estimated levels of WHV in sera are depicted as open bars for 10 to 10³ vge and as filled bars for >10³ vge per ml. For quantities of WHV DNA in PBMC, the estimated levels are depicted as open bars for 0.005 to 5 vge and as filled bars for >5 vge per 10⁴ cells. The amounts of WHV DNA detected in liver biopsy tissue and autopsy samples collected at the indicated time points (arrows) are shown as the estimated numbers of WHV vge per 10⁴ cells. Liver histology results are marked as follows: N, normal; AH, acute hepatitis; and MIN, minimal (residual) hepatitis. Lbx1, liver biopsy 1; Lbx2, liver biopsy 2; Ax, autopsy.

Direct sequence analysis of the WHV pre-S/S, C, and X amplicons derived from autopsy sera and PBMC of both WF.1 and WF.2 and an autopsy liver specimen from WF.1 showed sequences essentially identical with those detected in the culture supernatants obtained after final passages in experiments 3-L and 4-H and, thus, also identical with the sequence of the WHV/tm3 inoculum that was used to infect woodchuck WM.C (data not shown). There were only two nucleic acid changes noted in the X region in comparison to that of the WHV/tm3 sequence, both of which led to amino acid changes, i.e., A114T and T128I. These differences occurred in the virus recovered from sera and PBMC of WF.1 and WF.2 and from the liver of WF.1, and therefore, they did not support a cell type-specific origin for these WHV variations. Furthermore, clonal sequencing analysis of the WHV DNA C, pre-S/S, and X regions of the genomes derived from the same samples indicated that the wild-type WHV sequence clearly predominated. Thus, the X amino acid changes indicated above were detected only in less than half of the clones analyzed, indicating that the direct sequencing results, as expected, did not reflect the entire range of genomic variations existing in these samples. Additionally, a few random single nucleotide changes unique to individual clones were identified. Overall, the clones displaying wild-type WHV sequence identical with that of WHV/tm3 or WHV/tm2 prevailed in all samples analyzed (data not shown). These findings were consistent with the conclusion that the different patterns of WHV infection observed for WF.1 and WF.2 were related to the dose of the virus injected but not to a minor change in virus sequence, at least within the genomic fragments which were examined.

DISCUSSION

In this study, we have demonstrated that wild-type WHV can be serially transmitted in cultured naive lymphoid cells and hepatocytes and that multiple rounds of the virus passage in lymphoid cells did not lead to the emergence of cell type-specific viral variants. Furthermore, WHV remained infectious to WHV-naive woodchucks after repeated passage in either lymphoid cells or hepatocytes.

It has been shown that human B cell hybridomas infected with HBV could transmit the virus genome to normal lymphocytes, in coculture experiments (10). The B cells serving as the virus source were obtained from a patient with acute HBV infection. However, whether the HBV might have originated from within the B cells or adhered to their surface was not examined. Other subsets of lymphoid cells have also been found to be susceptible to HBV infection, although in many instances, no cell surface treatments to remove potentially adhering extracellular virions or evaluations of the expression of viral genome replicative intermediates were performed, making interpretation of some of the studies controversial (13, 15, 33, 41, 42, 45).

So far, no specific genomic variation has been linked to hepadnavirus lymphotropism. However, very few studies have aimed to investigate this issue. One of the studies showed that the HBV sequences in sera and PBMC samples collected during AH which lead to spontaneous recovery did not differ in the C gene promoter region (22). In the current work, the virus sequence analysis was focused on the determination of the WHV pre-S/S, C, and X gene sequences. This was because the virus passaged was recovered at relatively low levels, making it unfeasible to amplify its complete genome. On the other hand, we have routinely analyzed the virus pre-S sequence because this region is known to be most prone to variations (see below) and encodes virus envelope epitopes which are considered to be essential for virus-cell-specific recognition (18). Our extensive analysis of both the direct and the cloned WHV DNA amplicons showed no variations in the C or pre-S/S regions between the virus repeatedly passaged in lymphoid cells or hepatocytes, excluding the existence of lymphoid cell-specific WHV variants in the initial splenocyte-derived inocula and their rise during serial passage. The differences between the wild-type WHV X gene sequence and that found in the X region of the virus recovered from the culture supernatants were the same regardless of whether the supernatant was from WCM-260 cells or lymphoid cells, indicating that the differences were not cell specific but rather arose due to the adaptation of the virus to survive under in vitro conditions.

In the present study, we applied several methodological approaches which together allowed unbiased interpretation of the data in regard to the nature of WHV lymphotropism. Approaches included (i) detection of only intracellular WHV DNA sequences after cell surface enzymatic treatment to remove any potentially attached WHV nucleic acids and virions before DNA extraction; (ii) identification of WHV cccDNA in de novo-infected cells, confirming active virus replication; (iii) demonstration of the ability of the produced virus to transmit infection through many passages in these cells, rather than showing only the susceptibility of the lymphoid cells to virus; and (iv) proving that the virus passaged in cultured lymphoid cells remained infectious to healthy animals. Furthermore, the use of well-characterized, homogenous WHV as the inoculum, in which the complete virus genome sequence was shown to be identical with that in the spleen and liver of the donor, helped to conclude that no lymphoid cell-specific variants emerged during repeated passage of virus in lymphoid cells. Although we have analyzed the pre-S region of WHV sequences, which is most prone to nucleotide changes (17), as well as the C and X genomic regions, the existence of lymphoid cell-specific variants restricted to the remaining fragments of the genome located outside of the regions analyzed cannot be completely excluded.

The results from the present study corroborate data reported earlier which demonstrated that the lymphatic system-restricted WHV infection was unlikely to be the result of invasion with a WHV variant preferentially infecting lymphoid cells (23, 29). The current data support and expand these findings by showing that the amount of the virus but not its cellular origin determines the induction of hepatitis.

In one relevant study of self-limited HBV infection in humans, it has also been found that the virus residing in peripheral lymphoid cells isolated for 30 years after they were recovered from AH persisted without sequence variation (3). More recently, the study of HCC in patients with persistent occult HBV infection showed that the majority of the viral genomes detected in the liver do not contain mutations (38). Taken together, these results may suggest that low-level replication of hepadnaviruses persisting for an extended period of time may not be associated with the emergence of mutations. In turn, this raises a possibility that the genomic sequence of the persisting virus probably represents the best fit for its long-term survival.

Despite the findings indicating that a low hepadnavirus replication rate is not associated with the emergence of viral variants, there are other studies suggesting that variations in the hepadnavirus genome may occur during occult HBV infection. The hepadnavirus C gene sequence is considered to be relatively stable, and mutations in this gene have been rarely observed, especially after the resolution of AH. In contrast, the pre-S1 genomic region is relatively susceptible to sequence variations (17). The protein encoded by this region is thought to be involved in virus-host cell interactions, mediating virus binding to targeted hepatocytes and lymphoid cells. Recently, it has been postulated that mutations in the pre-S1 sequence are more prevalent in patients with occult HBV infection (32). Similarly, during phases of low viremia in patients with serum HBsAg-positive chronic hepatitis B, up to 37% of the HBV species identified in sera and liver samples contained pre-S mutations (14). This high frequency is greater than the variation identified in other HBV gene sequences (14). The above-cited studies raised a possibility that when the virus occurs at relatively low levels, pre-S1 variants may be more prevalent. In addition, it has been shown that HBV mutants with various types of in-frame deletions in the pre-S1 region were replication competent in vitro (17, 25). Since these pre-S1 gene deletions overlap the dispensable spacer domain of the polymerase protein, the generation of productive virus is possible, allowing these variants to accumulate to high levels in infected patients (25, 39, 44). Based on the above reasoning, we extensively analyzed the WHV pre-S/S regions to determine if specific variants had arisen or not during serial passage of WHV in lymphoid cells in vitro. As indicated, we did not find such evidence.

With respect to the hepadnavirus X gene sequence, very few data for variations in this genomic region, directly or indirectly relevant to occult long-term HBV persistence, have been reported. The main focus of the previous studies was on the mutations potentially associated with the development of fulminant hepatitis or HCC and on those located in the areas of the X gene encoding the basic core promoter and the core upstream regulatory sequences. Overall, the HBV X gene appears to be well conserved during chronic hepatitis B, and conservation of the X protein seems to be essential for the virus' ability to actively replicate (2, 19). On the other hand, mutations in the HBV X gene sequence have been documented in HCC (5, 20). We did not identify any conserved changes in the WHV X sequence analyzed.

In our study, no sequence variations were found in the pre-S, C, and X genomic regions in the virus recovered at autopsy from woodchucks WF.1 and WF.2, which were infected with WHV passaged in lymphoid cells or hepatocytes. The fact that we did see a small number of synonymous mutations in the X gene in some clones did not support the development of lymphoid cell-specific variation, since these changes were found in less than half of the clones analyzed and virus derived from animals infected with WHV passaged in either lymphoid cells or hepatocytes. It is notable that the WHV X gene changes detected in the supernatants from the WCM-20 and lymphoid cell cultures reverted to the wild-type WHV sequence in animals injected with these supernatants. Overall, our sequencing results corroborate the notion that lymphotropism is a natural propensity of WHV but not a result of the existence of a cell type-specific virus variant.

As we have previously shown, the pattern of WHV infection appears to be dependent on the amount of invading virus (29). In our previous study, doses above 10³ WHV vge caused serologically evident infection accompanied by AH, while those below this level induced primary silent infection, designated POI. The amounts of WHV recovered after multiple passages in lymphoid cells or hepatocytes in the current study were near this demarcating quantity. In this context, the development of AH in WF.1, injected with ∼1.5 × 10⁴ vge, and POI in WF.2, injected with ∼8.8 × 10² vge, are not unexpected findings.

A recent study employing the simian immunodeficiency virus (SIV) macaque model of human immunodeficiency virus showed the persistence of low levels of virus that evaded detection by conventional testing, i.e., 0.1 to 5.3 SIV DNA copies/10⁶ PBMC (47). The SIV envelope sequence remained homogeneous over a 6-year period in the lymphoid cells, and the animals remained free of evident illness for up to 10 years after inoculation. These data seem to be compatible with the results of our study and suggest that low levels of replicating virus do not normally permit the development of virus mutations and allow persistence of small amounts of virus which may evade detection by conventional methods.

Overall, the present study provides strong support for the concept that hepadnavirus lymphotropism is an inherent property of wild-type virus and that the dose of the virus, but not the existence or the rise during infection of a particular lymphotropic virus variant, predetermines the development of lymphatic system-restricted infection. A high degree of similarity between WHV and HBV may suggest that the same could be true for HBV infection.