Infection of common marmosets with hepatitis C virus/GB virus-B chimeras


  • Potential conflict of interest: Nothing to report.

  • Supported by the grants from National Basic Research Program of China (973 Program No. 2012CBA01305, 2010CB530204 and 2009CB522507), S&T Grand Special Program of China (No. 2009ZX10004-305), National Natural Science Foundation of China (No. 30972765 and 81071348), and Guangdong Provincial S&T Project (No. 2010B060500010).


The development of vaccination and novel therapy for hepatitis C virus (HCV) has been hampered by the lack of suitable small-animal models. GB virus B (GBV-B), closely related to HCV, causes viral hepatitis in common marmosets (Callithrix jacchue jacchus) and might represent an attractive surrogate model for HCV infection. However, differences exist between GBV-B and HCV in spite of a short genetic distance between the two viruses. Here we report common marmosets infected with two HCV/GBV-B chimeras containing HCV structural genes coding for either whole core and envelope proteins (CE1E2p7) or full envelope proteins (E1E2p7) substituted for the counterpart elements of GBV-B. Naïve animals intrahepatically injected with chimeric RNA transcripts or intravenously injected with sera from primary infected animals produced high levels of circulating infectious chimeric viruses and they developed chronic infection. Tacrolimus-treated marmosets inoculated with a CE1E2p7 chimera had higher viral loads and long-term persistent infection. A moderate elevation of serum aspartate aminotransferase (AST) levels was observed in parallel with viral replication. Chimeras recovered from liver samples revealed 1/958 adaptive viral mutations. Histopathological changes typical of viral hepatitis were observed in liver tissues from all types of HCV chimeras-infected marmosets. HCV core and E2 proteins were detected in liver tissues from infected animals by immunohistochemical staining. Fluctuations of chimeric virus replication in marmosets with spontaneous and sporadic viral clearance might be related to specific antibody and T-cell response to HCV proteins in vivo. Replication of CE1E2p7 chimera was observed in primary hepatocyte cultures by immunofluorescent staining in vitro. Conclusion: Infectious HCV chimeras causing chronic hepatitis in marmosets might constitute a small primate model suitable for evaluation of virus-cell interaction, vaccination, and antiviral therapy against HCV infection. (Hepatology 2014;59:789–802)


enzyme immunoassay


GB virus B


hepatitis C virus


monoclonal antibody

Hepatitis C virus (HCV) has a host range restricted to humans and chimpanzees and lacks suitable primate models,[1-3] limiting progress in developing HCV vaccines and antiviral drugs. The chimpanzee is a rare and expensive experimental animal whose scientific utilization raises an ethical dilemma. Marmosets (Callithrix jacchue jacchus) as well as tamarins (Saguinus oedipus) are small New World monkeys susceptible to GB virus B (GBV-B), a hepatotropic flaviviridae phylogenetically close to HCV that rarely infects small primates.[4, 5] GBV-B has been considered a surrogate for HCV in infected marmosets or tamarins.[6-9] Despite the short genetic distance between the two viruses, there are differences in structural proteins which affect the immune response. It has been hypothesized that chimeric HCV/GBV-B viruses might mimic HCV infection in small primates. Only chimeras integrating short pieces of HCV 5′-NCR, p7, or HVR1 sequences were previously constructed to investigate gene functions in vivo.[10-13] Such constructs, however, are unable to answer questions related to virus-host interaction, viral pathogenesis, and immune response.

The core, envelope glycoproteins E1E2, and p7 structural proteins are major components of HCV. We hypothesized that if HCV/GBV-B chimeras containing whole HCV structural protein genes (CE1E2p7) or intact envelope protein genes (E1E2p7) were infectious to marmosets, a major model to study the functions of HCV structural proteins in vivo would become available. Here the infectivity and pathogenicity of several chimeric constructs were investigated in marmosets.

Materials and Methods

Animal and Study Approval

The 13 common marmosets (Callithrix jacchue jacchus) used in this study were juvenile males with an average weight of 350 g. Prior to inoculation, all animals tested negative for GBV-B, HCV, and HBV by quantitative polymerase chain reaction (qPCR) or reverse-transcription (RT-qPCR).[7, 14, 15] All marmosets were imported from Tianjin Medical University and individually housed and bred at the Laboratory Animal Research Center of Nanfang Hospital, Guangzhou, China. The animals used in this study were approved by the authorities of Forest Bureaus, Tianjin, and Guangdong governments, respectively (Yuelinhu [2010] 90 and Yuelinhu [2011] 404). Ethical approval to use the marmosets as experimental animals was also obtained from Nanfang Hospital, Southern Medical University, Guangzhou (permit numbers: SYXK [Yue] 2010-0056). All animal care and procedures were in accordance with national and institutional policies for animal health and well-being. Marmoset surgery was performed under appropriate anesthesia and all efforts were made to minimize the suffering of animals.

Construction of Chimeric Genomes

HCV-CE1E2p7 and HCV-E1E2p7 chimeric genomes were constructed by substituting the GBV-B counterpart elements with HCV structural genes coding for either whole core and envelopes (CE1E2p7) or full-envelope proteins (E1E2p7) (see Supporting Materials and Methods).

In Vitro RNA Transcription

1.6 μg plasmid DNA was linearized with Sac I and in vitro transcribed in a volume of 20 μL using the T7 Megascript kit (Ambion, Applied Biosystems/Ambion, Austin, TX). The chimeric RNA was investigated with 5′ and 3′ terminus qPCR and RT-nested PCR to confirm the generated full length of RNA genome.

Intrahepatic Inoculation of Marmosets With Chimeric RNA

Marmosets were intrainoculated hepatically with 300 μg of chimeric RNA as described in the Supporting Materials and Methods. The course of the disease progression was monitored for up to 44 weeks during which EDTA blood samples (0.6-1.0 mL) were collected weekly through either the caudal or the femoral vein.

Passaging of HCV Chimeras in Marmosets by Intravenous Inoculation of Primary Chimera Containing Serum

Marmosets were intravenously infected through the femoral vein with 0.2 mL (104 genomic copies per animal) of primary HCV chimera containing serum (P0). After inoculation, ∼0.6 mL of blood sample (P1) was collected weekly.


Animals were treated orally twice a day with 100 g/kg FK506 (Tacrolimus). Three weeks after the initiation of FK506 treatment, at which time the level had reached 2.5-7.7 ng/mL in circulating blood, marmosets were intrahepatically or intravenously inoculated with HCV chimeras. FK506 treatment was maintained throughout the experiment.

Detection and Quantification of Chimeric Virus RNA

Amplification of chimeric RNA was conducted on plasma and liver samples by RT-qPCR using two sets of primers and probes targeting either HCV or GBV-B genome (Supporting Materials and Methods).[7, 14] The plasmid pGBB containing GBV-B genome was used as standard to establish genome equivalents for viral load quantification.[6] The sensitivity of RT-qPCR was 130 copies/mL for serum sample and 73 copies/300 mg for liver samples.

RT Nested-PCR and Sequencing for HCV Chimeras

Chimeric viruses were further identified by RT nested-PCR with five sets of primers (Supporting Table 1) to recognize the specific portions of HCV core (140 bp), E1 (165 bp) or GBV-B E1 (217 bp) gene sequences, and joint portions of GBV-B 5′NCR + HCV Core and E1 (750 bp) or HCV E2p7 + GBV-B NS2 (670 bp), respectively. Longer DNA fragments were cloned into the pMD-20 T vector. All PCR products and clones were sequenced commercially (Huada Gene, Shenzhen, China).

Sequencing of Full-Length Chimeric Genome From Liver of Infected Marmoset

The genome sequence of HCV-CE1E2p7 chimeric virus (P1) recovered from liver of immunosuppressed animal M17 was cloned and fully sequenced from overlapping PCR products obtained by RT-nested PCR (Supporting Materials and Methods).

Histopathological Grading

The surgical biopsies from infected and noninfected marmoset livers were investigated. A small section of the liver tissue was cut, immediately fixed in 10% buffered formalin solution before embedding in paraffin wax using standard procedures. The liver tissue was stained with hematoxylin and eosin (H&E), whereas collagen was stained with the Masson's Trichrome method. Blind samples were microscopically examined for histopathological changes. The histological activity grade and fibrosis stage were determined in 4 to 5 sections according to the modified HAI system (Kondell score).[16]

Immunohistochemical Staining (IHC)

Monoclonal antibodies (mAbs) specific to HCV core (C7-50, Abcam, Cambridge, UK), HCV E2 (3L19, provided by Ping Zhao, Department of Microbiology, Second Military Medical University, Shanghai, China), or GBV-B NS3 (2E12, generated in the laboratory) were used to examine the specific expression of HCV or GBV-B proteins in liver cells by IHC (Supporting Materials and Methods). A score was attributed according to the intensity of the nucleic or cytoplasmic staining and the extent of stained cells in 4 to 5 sections.[17]

Biochemical Test

The alanine aminotransferase (ALT) and aspartate aminotransferase (AST) level in the marmoset sera was measured in U/L with ALT or AST kits (Biosino Bio-technology and Science, Beijing, China) on a Hitachi 7170S fully automatic biochemical analyzer (Tokyo, Japan).

Anti-HCV Detection

The truncated HCV (Z14 strain, 1b) core (aa 1-121) and E2 (aa 384-661) expressed in Escherichia coli or Drosophila S2 were purified and used as antigens to detect antibodies to HCV in marmoset sera by the enzyme immunoassay (EIA). The anti-HCV reactivity was expressed as the signal to cutoff ratio (S/CO). The reactivity cutoff was calculated as mean +3 SD.

Synthetic Peptides and ELISpot Assay

Sixteen peptides were commercially synthesized (Chinese Peptide Company, Hangzhou, Zhejiang, China). Fifteen HCV E1E2p7 peptides were predicted to reside within the T-cell epitope-rich regions or epitopes. An unrelated Brucella BP26 peptide was used as a negative control. The HCV-specific T-cell response of peripheral blood mononuclear cells (PBMCs) from marmosets was detected by an ELISpot assay against individual peptides (Supporting Materials and Methods). The response of PBMCs stimulated with an unrelated peptide was considered background, and the response of PBMCs stimulated with phytohemagglutinin (PHA) served as a strong positive baseline. The number of spot-forming cells (SFC) was normalized to SFC/106 PBMCs.

Immunofluorescence Staining of Primary Hepatocyte Cultures

Primary hepatocytes were isolated from naïve marmoset using collagenase perfusion as described elsewhere.[13] Marmoset hepatocytes were seeded onto 24-well collagen-coated plates and transfected with either the CE1E2p7 chimera RNA or the GBV-B RNA by Lipofectamine 2000 (Invitrogen, Guangzhou, China). Rabbit antibody to albumin (Abcam) and mAbs to HCV core (C7-50) and GBV-B NS3 (2E12) were used as primary antibodies, whereas Alexa Fluor 594 (red) donkey antirabbit IgG and Alexa Fluor 488 (green) goat antimouse IgG (Invitrogen) were used as secondary antibody conjugates for detection of the HCV chimera-infected cells by immunofluorescence staining (IF).[13] Diamidinophenylindoldiacetate (DAPI) was added to stain cell nuclei.

Statistical Analysis

The data were analyzed using the statistical package SPSS v. 16.0. Correlation of viral RNA between GBV-B and HCV quantification systems was assessed by Spearman's rho nonparametric correlation analysis. Comparison of viremia between immunosuppressed and immunocompetent animals as well as the comparison of anti-HCV reactivity was determined using one-way analysis of variance (ANOVA). P < 0.05 was considered statistically significant.


Construction of Chimeric Viral Genomes Containing Full HCV Genes Encoding for Structural Proteins

The full-length genes of HCV core, envelope E1E2, and p7 (CE1E2p7, 2,427bp) and envelope E1E2 and p7 (E1E2p7, 1,854bp) complementary DNA (cDNA) were extracted from a wild-type strain of HCV genotype 1b obtained from an infected Chinese blood donor (Z14, accession number JN870282). Two HCV chimeric genomes of CE1E2p7 (length 9,630bp, accession number: KF285483) and E1E2p7 (length 9,525bp, accession number: KF285484) were constructed by replacing the counterpart elements of GBV-B with the corresponding HCV genes into an infectious GBV-B clone (pGBB) (Fig. 1). Fully functional RNA from chimeric genomes were synthesized in vitro using the T7 promoter at the 5′ terminus. The in vitro-produced RNA was diluted in phosphate-buffered saline (PBS) prior to the intrahepatic injections of marmosets.

Figure 1.

Schematic diagram of constructs for HCV chimeric genomes. Construction of HCV E1E2p7 and CE1E2p7 chimera genomes by replacing GBV-B (blank) counterpart elements with HCV genes (shadow) using overlapping PCR to join three fragments and restriction site ligation (Not I and Afl II).

Infectivity of HCV Chimeras to Common Marmosets

Primary HCV chimeras (P0) were produced and delivered in seven marmosets by intrahepatic injection with a single solution of HCV chimeric RNA (Supporting Table 2). The marmosets (M) were injected with the following RNA and controls: M3 and M15 with CE1E2p7 chimeric RNA, M1 and M18 with E1E2p7 chimeric RNA, M13 with HCV JFH-1 RNA, M4 with GBV-B RNA as positive control, M2 with PBS as negative control. Sera were collected from the marmosets at 1- to 2-week intervals, up to 44 weeks. Viremia was quantified with two quantitative qRT-PCR assays using different specific primers and probes targeting GBV-B or HCV RNA (Supporting Fig. 1). Viral RNA levels were analyzed with the GBV-B quantification system (Fig. 2A) and a significant correlation between GBV-B and HCV RNA was observed (P < 0.001) (Fig. 2B). No viral replication was observed in animals injected with HCV JFH-1 RNA or PBS controls (Fig. 2A, P0). Two profiles of chimera replication with viral load up to 105 copies/mL were observed in primary infection sera (P0) at 1 to 16 weeks and at 20 to 26 weeks postinjection, respectively. Such a profile was observed in animals M3 and M15 infected with CE1E2p7 chimera, M1 and M18 infected with E1E2p7 chimera, as well as in GBV-B-infected animals. HCV CE1E2p7 chimera from animal M15 was detected consistently and persisted up to 44 weeks with ∼105 copies/mL viral load, while viral RNA was not detected in four other animals after 26 weeks when sacrificed or dead (Fig. 2A).

Figure 2.

Detection and identification of HCV chimeras in sera from infected marmosets. (A) Primary (P0) and passaged (P1) HCV chimeras were detected by RT-qPCR in sera from marmosets and AST levels were indicated according to time-course. (B) Correlation of viral RNA between GBV-B and HCV quantification systems. (C) Amplicons of the predicted sizes (165 bp, 140 bp, and 217 bp) in duplicate or triplicate were produced by RT nested-PCR for E1E2p7 chimera-infected M1-P0 (9th week) and M10-P1 (4th day), CE1E2p7 chimeras infected M3-P0 (9th week) and M6-P1 (2nd week), and GBV-B infected M4-P0 (9th week). (D) DNA fragment (750 bp) for joint portions of GBV-B-5′ NCR + HCV-core, and E1 were amplified from CE1E2p7 chimera-infected M15-P0 (33rd week). (E) DNA fragments (670 bp) for joint portions of HCV-E2p7 to GBV-B-NS2 were amplified from CE1E2p7 chimera-infected M15-P0 (33rd week), M16-P0 (7th week), and M17-P1 (8th week). *Where serum was taken from infected marmosets for inoculation of naïve or FK506-treated marmosets. #Where serum sample was taken for identification of viral RNA by RT nested-PCR and sequencing. M and -ve indicate DNA molecular marker and negative control.

To test whether HCV chimeras infection transferred, four naïve animals were intravenously inoculated with P0 sera collected from marmosets with primary chimera infection and controls (Supporting Table 2). The sera were injected as follows: M6 with M3-CE1E2p7-P0, M10 with M1-E1E2p7-P0, positive control M9 with M4-GBV-B-P0, and negative control M12 with M2-PBS-P0 (Fig. 2A, P1). Viremia up to 2.4 × 105 copies/mL (P1) was subsequently detected at 1-7 weeks post-P0 serum injection. This was followed by spontaneous clearance for a further 2 weeks. A sporadic recurrence of viremia at low level (<1,000 copies/mL) was then observed between 10 and 29 weeks. The presence of HCV chimeras in the P0 and P1 sera indicated that HCV CE1E2p7 and E1E2p7 chimeras could infect naïve marmosets and persistently replicate when intrahepatically or intravenously administered.

The potential role of the host immune system to control infection with these chimeric constructs was examined in two animals immunosuppressed with orally administered FK506 (Tacrolimus) mixed with bananas. After 2-3 weeks when the circulating levels of 2.5-7.7 ng/mL of FK506 were reached, M16 was intrahepatically injected with CE1E2p7 chimeric RNA, whereas M17 was inoculated intravenously with M15-CE1E2p7-P0. Viremia was detected in these two FK506-immunosuppressed marmosets in every sample at significantly higher levels than observed in immunocompetent animals (P = 0.037); for example, M3-P0 and M6-P1, but similar to M15-P0.

The hepatic RNA load in these immunosuppressed animal M17-P1 was estimated to be 219GE/mg by GBV-B quantification system and 230GE/mg by the HCV system, respectively. The ALT levels remained mostly normal (<40U/L), whereas AST was elevated over 40U/L in sera of HCV chimera and GBV-B infected marmosets (Fig. 2A).

Identification of HCV Chimeras From Infected Marmosets

In order to determine whether viral RNAs detected by qRT-PCR in sera of infected marmosets corresponded to the expected HCV chimeras and GBV-B, DNA fragments from virus-containing sera P0 and P1 were amplified by RT nested-PCR using primer sets that specifically targeted portions of HCV, GBV-B, or contiguous portions of HCV and GBV-B, respectively. Amplicons of the predicted size were obtained for the HCV core corresponding to CE1E2p7 chimera (M3-P0 and M6-P1), the HCV E1 to E1E2p7 chimera (M1-P0 and M10-P1) as well as the GBV-B E1 to GBV-B (M4-P0) infected sera (Fig. 2C). Likewise, DNA fragments from contiguous portions of HCV-E2p7 to GBV-B-NS2 were amplified corresponding to CE1E2p7 chimeras (M15-P0, M16-P0, and M17-P1) (Fig. 2E). The PCR amplicons were verified to contain the predicted viral sequences by DNA sequence analysis.

The possible development of adaptive viral mutations occurring in the HCV chimera-infected marmosets was further investigated by DNA sequence analysis of viral clones of PCR amplicons (670 bp and 750 bp fragments) and compared to the parental HCV chimeric sequences. Consensus sequences from 3/4 PCR products were found to be identical with the HCV strain used, although some substitutions within viral quasispecies from CE1E2p7 chimeras occurred during various periods of infection, as was observed by comparing multiple clone sequences (Supporting Table 3). The substitution of T2466A/I674K was found in 6/9 clones, C642T/P66L in 3/14 clones, T1051C/S202P in 3/14 clones, and T2445C/L667P in 2/9 clones were found in the immunocompetent animal M15-P0 (Table 1). These changes might be associated with HCV chimera replication in marmosets. Substitutions within this region were likely to occur randomly in immunosuppressed animals M16-P0 and M17-P1. Viral load from M15-P0 was found at levels similar to those of immunosuppressed marmosets (Fig. 2A).

Table 1. Mutations in HCV CE1E2p7 Chimera Recovered From Serum and Liver Tissue of Immunocompetent or Immunosuppressed Marmosets

nt (aa)

RegionOriginal ConstructMutation
  1. Mutations listed in the table are detected in PCR consensus sequence with a substitution or an overlapping nucleotide or in at least two clones for each site. Dominant and minor sequences are shown in capital letters or lowercase letters, respectively. Nb, number of mutated/sequenced clones. In the sequenced 9,583 consensus nucleotides (nt 22 to 9604), the percentage of homology with the initial construct was 99.89%. The ratio of mutations (1/809 nts) in the structural protein genes (HCV) was similar to that (1/1022 nts) in GBV-B elements within HCV chimera.

Fragments (5′NCR-Core-E1, nt 352-1101) from serum of M15-P0 (33w)
422GBV-B 5′NCRTTC2/14
642 (66)HCV coreC (P)C (P)T (L)3/14
1051 (202)HCV E1T (S)T (S)C (P)3/14
Fragments (E2p7-NS2, nt 2289-2958) from serum of M15-P0 (33w)
2445 (667)HCV E2T (L)T (L)C (P)2/9
2466 (674)HCV E2T (I)T (I)/a (k)A (K)6/9
Fragments (E2p7-NS2, nt 2289-2958) from serum of M16-P0-FK506 (4d)
   No repeated mutation
Fragments (E2p7-NS2, nt 2289-2958) from serum of M17-P1-FK506 (8w)
   No repeated mutation
Full-length genome from liver tissue of M17-P1-FK506 (25w)
507 (21)HCV coreA (D)G (G)G (G)8/8
1956 (504)HCV E2T (V)G (G)G (G)6/6
2392 (649)HCV E2A (G)A (G)T (G)2/3
3460 (1005)GBV-B NS2A (A)A (A)G (A)2/10
4309 (1288)GBV-B NS3C (G)C (G)/t(g)T (G)4/8
4524 (1360)GBV-B NS3G (G)G (G)A (D)2/8
5037 (1531)GBV-B NS3G (R)G (R)/a (h)A (H)4/8
5273 (1610)GBV-B NS3C (R)C (R)/t (c)T (C)4/8
6267 (1941)GBV-B NS5AG (G)G (G)/t (v)T (V)5/9
8721 (2759)GBV-B NS5BA (Y)T (F)/a (y)T (F)4/8
Poly(U)GBV-B 3′NCR23 nts19 nts17 nts2/10
    18 nts3/10

Analysis of Mutations in HCV-CE1E2p7 Chimera From Liver of Infected Marmoset

A nearly complete genome sequence (nt 22-9604, accession number KF430633) was obtained from a liver tissue sample collected from immunosuppressed animal M17-P1 25 weeks postinfection. The entire open reading frame (ORF) of HCV-CE1E2p7 chimeric virus recovered was obtained from multiple clones (Supporting Table 4). Three mutations, A507G/D21G, T1956G/V504G, A8721T/Y2759F, were detected from PCR consensus sequence in core, E2, and NS5B, respectively. Ten mutations (1/958 nucleotides) were detected in at least two clones for each site and seven of them were nonsynonymous (Table 1). Mutations detected only in single clones are shown in Supporting Table 4.

Histopathological Observation of Liver Tissues From HCV Chimera-Infected Marmosets

At the end of the observation period, liver tissues of CE1E2p7 chimera-infected marmoset M3-P0 (44 weeks) as well as the CE1E2p7 chimera-infected FK506-treated M17-P1 (25 weeks), respectively, were found to contain pathological changes typical of hepatomega and brownish yellow particles when compared with liver of negative control M12-P1 (44 weeks). This suggests a fatty liver degeneration present in HCV chimera-infected animals (Fig. 3A). H&E staining and Masson's Trichrome histopathological examinations were conducted on the liver tissue collected at the timepoints shown in Fig. 3 and Table 2. Liver tissue from M2-P0 intrahepatically and M12-P1 intravenously injected with PBS, and M13-P0 intrahepatically injected with HCV JFH-1 RNA were used as corresponding negative controls. Figure 3B and Table 2 show that the negative control animal M12-P1 had a normal hepatic architecture while M2-P0 and M13-P0 presented slight edema and cell disarray. Compared to the control marmosets, lymphocyte infiltration, ground glass hepatocytes, eosinophilic cells, and fibrous expansion was observed in liver sections from primary (P0) and passage (P1) CE1E2p7 or E1E2p7 chimera-infected marmosets (Fig. 3C,D). Necrosis, inflammation, and fibrosis scores are listed in Table 2. The liver tissues from the CE1E2p7 chimera-infected immunosuppressed animal M16-P0 displayed serious cholestasis (Fig. 3C). These observations suggested that the marmosets were susceptible to HCV chimera infection and presented histopathological features indicative of acquired viral hepatitis, although viremia was occasionally cleared from infected marmosets during the experiments. A transmission electron microscopy examination revealed abnormal mitochondria in ultrathin sections of the liver tissue from CE1E2p7 chimera-infected M15-P0 (Fig. 3F). Most hepatocytes from the CE1E2p7 chimera-infected immunosuppressed M17-P1 animal carried multiple lipid droplets and lysosomes (Fig. 3F), which indicates that fatty liver degeneration and damage had occurred similar to Fig. 3A.

Table 2. Histopathological and Immunohistochemical Observations
Virus or ControlMarmosetTimepoint (weeks)Necroinflammatory GradeaFibrosis StagebIHC Stainingc
Anti-HCV coreAnti-HCV E2Anti-GBV-B NS3
Periportal +/− Bridging NecrosisIntralobular Degeneration and Focal NecrosisPortal InflammationTotalIntensityExtentIntensityExtentIntensityExtent
  1. a

    The histological status was determined by the modified HAI system (Kondell score), which grades necrosis and inflammation on a scale of 0 to 18 (periportal inflammation and necrosis, 0 to 10; lobular inflammation and necrosis, 0 to 4; portal inflammation, 0 to 4).

  2. b

    Fibrosis was scored as 0 to 6 (0, no fibrosis; 1 or 2, portal fibrosis; 3 or 4, bridging fibrosis; and 5 or 6, cirrhosis).

  3. c

    The score was given according to the intensity of the nucleic or cytoplasmic staining (no staining = 0, weak staining = 1, moderate staining = 2, strong staining = 3) and the extent of stained cells (0% = 0, 1-10% = 1, 11-50% = 2, 51-80% = 3, 81-100% = 4). NT, not tested.

Figure 3.

Histopathological observation of liver tissue of HCV chimera or GBV-B infected marmosets. (A) Representative pictures of livers from healthy and chimera-infected marmosets. (B-E) Microscopic features of liver tissues by H&E staining (left) and Masson's Trichrome method (right). Original magnification of liver tissue was ×200. (B) Negative controls of PBS or HCV JFH-1 RNA injected marmosets. (C) CE1E2p7 chimera-infected marmosets. (D) E1E2p7 chimera-infected marmosets. (E) GBV-B infected marmosets. (F) Transmission electron microscopy of the liver of CE1E2p7 chimera-infected marmosets. Small letters with arrows indicate histopathological features such as (a) lymphocytic infiltrates, (b) ballooning degeneration (edema), (c) ground glass liver cells, (d) eosinophilic cells, (e) cholestasis, (f) fibrous expansion, (g) abnormal mitochondria, (h) lipid droplet, and (i) increased lysosome.

Immunohistochemical Identification of HCV Chimeras Induced Hepatitis in Marmosets

Pathological changes in liver cells induced by HCV chimera were detected by IHC using mAbs to HCV core, E2, and GBV-B NS3. A score was created according to the intensity of the cytoplasmic staining and the extent of the stained cells (Table 2). PBS served as a negative control. As expected, the negative control animals showed no HCV or GBV-B specific staining (Fig. 4A, Table 2). The liver cells of positive controls GBV-B-infected M4-P0 and M9-P1 were clearly stained with anti-GBV-B NS3 but not with the antibodies to HCV core and E2 (Fig. 4B, Table 2). In the CE1E2p7 chimera-infected animals (M3-P0, M15-P0, M6-P1, M17-P1) the liver cells stained positive with all three specific mAbs (Fig. 4C, Table 2), while the liver tissue from FK506-treated M16-P0 hardly stained and differentiated from the severe cholestasis in the tissue (Fig. 3C). The liver tissue of animals infected with the E1E2p7 chimera (M1-P0, M18-P0, M10-P0) could be stained with the anti-HCV E2 and anti-GBV-B NS3 antigen (Fig. 4D, Table 2). These results suggest that HCV chimeras clearly replicated and were expressed in liver cells independently from viral particles released in peripheral blood from infected animals.

Figure 4.

Immunohistochemical staining of liver tissue from chimera-infected marmosets. Original magnification was ×400. (A) Negative control marmosets. (B) GBV-B infected marmosets. (C) CE1E2p7 chimera-infected marmosets. (D) E1E2p7 chimera-infected marmosets.

Immune Response to HCV Chimera Infection

The infected marmosets exhibited a distinct humoral and cellular immune response to the HCV chimera infection. Antibodies to HCV core and E2 were detected in serial sera from five and eight HCV chimera-infected marmosets, respectively. Levels of reactivity (S/CO) to HCV core in five CE1E2p7 chimera infections were significantly elevated in 1/5 animals (M16-P0-FK506) (P = 0.001) and moderately elevated in a further 2/5 animals (M3-P0 and M15-P0) relative to those observed in the PBS control animal (Fig. 5A). The same three animals presented with an elevation of anti-E2; however, only in two animals was this elevation significant (M15-P0 and M16-P0-FK506) (Fig. 5B). The E1E2p7 chimera infection did not, however, convincingly elicit an anti-E2 response in these three animals (Fig. 5B). The immunosuppression with FK506 apparently did not suppress the antibody response in M16-P0 and M17-P1 and was essentially unresponsive to both core and E2.

Figure 5.

Specific antibody and T-cell response in chimera-infected marmosets. (A) Reactivity (S/CO) of antibody to HCV core in CE1E2p7 chimera-infected marmosets and control animals of PBS, JFH1, and GBV-B. (B) Reactivity (S/CO) of antibody to HCV E2. S/CO ratio >1.0 is considered reactive. (C) Specific response of PBMCs from HCV chimera-infected marmosets was tested for secretion of IFN-γ by HCV peptide stimulation in an ELISpot assay. The lines indicated the strongly reactive baselines equal to responses of PBMCs stimulated by PHA from CE1E2p7 chimera-infected M15-P0, E1E2p7 chimera-infected M1-P0, and M10-P1, respectively.

Specific T-cell response to HCV epitopes was investigated using PBMCs collected from CE1E2p7 chimera-infected M15-P0 and E1E2p7 chimera-infected M1-P0 and M10-P1 animals. PBMCs were tested with the ELISpot assay for secretion of interferon-γ (IFN-γ) by stimulation with 15 predicted HCV T-epitope peptides (Supporting Table 5). The number of spot-forming cells (SFCs) per million PBMCs stimulated by an unrelated peptide or PHA was defined as background and strongly reactive baseline, respectively (Fig. 5C). M15-P0 responded to three peptides, whereas M1-P0 and M10-P1 responded to four and five peptides, respectively. The E1-282 and p7-779 peptides were reactive in both animals while other peptides induced dissociated response. The E2-695 and E2-725 peptides elicited a strong response only in the animal infected with CE1E2p7 chimera. These data suggested that HCV chimeras could stimulate specific and sensitive cellular immune responses in all chimera-infected marmosets.

Infectivity of HCV Chimera in Marmoset Primary Hepatocyte Culture

To determine whether CE1E2p7 chimera was infectious in vitro, primary hepatocyte cultures were transfected with CE1E2p7 chimeric RNA and GBV-B RNA, and examined by IF with mAbs directed against both HCV core and GBV-B NS3. Hepatocytes were identified by IFS with antibody to albumin and cell nuclei were stained by DAPI. Compared with naïve hepatocytes (Fig. 6A), CE1E2p7 chimeric RNA transfected hepatocytes were stained for HCV core (Fig. 6B), and GBV-B RNA transfected hepatocytes were positive for GBV-B NS3 (Fig. 6C), which indicates that HCV chimeras replicated and that viral proteins were expressed.

Figure 6.

Immunofluorescence staining for HCV or GBV-B antigens in infected primary hepatocyte cultures. Naïve hepatocytes (A), CE1E2p7 chimera-infected hepatocytes (B), and GBV-B infected hepatocytes (C) were stained in brightfield, in red with antibody to albumin and in green with mAb to HCV core or GBV-B NS3. Cell nuclei were stained in blue with DAPI. Original magnification was ×200.


This study describes the construction of two HCV/GBV-B chimeric viruses that contained full-length HCV genes (1,854-2,427 bp) encoding either the whole structural core and envelopes or all envelope proteins. Both the CE1E2p7 and E1E2p7 chimeras appear infectious to common marmosets and caused viral hepatitis. Several lines of evidence supported this conclusion.

First, substantial levels of viremia were found in two HCV chimera-infected marmosets and GBV-B controls during the initial 16 weeks postintrahepatic injection. In one of two animals (M15-P0) primarily infected with CE1E2p7 chimera, high levels of viremia persisted up to 44 weeks (Fig. 2A). Passaged infection with HCV chimeras presented viremia patterns similar to previously described GBV-B infections.[9, 13] An initial peak of viremia observed at 30 days (4-5 weeks) was followed in some animals by a second peak at 8-10 weeks, depending on the infectious dose.[9, 12, 13] This pattern resembles the fluctuating viremia observed in acute human HCV infection.[18] These HCV chimeras also appear infectious since, after primary infection, circulating chimeric viruses were able to infect naïve marmosets, although the viremia pattern was slightly different, possibly related to the infectious dose and the different route of infection (intravenous versus intrahepatic). An initial high viral load (>105 copies/mL) was detected 1-5 weeks postinoculation, and was cleared between weeks 4-7, and then remained detectable at a lower level up to 29th weeks postinfection (Fig. 2A, P1). In addition, both intrahepatic (M16) and intravenous (M17) injections in FK506-immunosuppressed animals developed consistently high viral load infections (Fig. 2A). In situations of both primary and passaged infections, the ALT level appeared to be a poor indicator of liver damage, as previously described,[19] while AST levels were moderately elevated in most infected marmosets. The presence of circulating HCV chimeras was shown by the simultaneous detection of GBV-B and HCV genome fragments. The viremia from the two chimeras was further authenticated by RT nested-PCR specific of regions from either the HCV core, E1, or adjacent regions of HCV and GBV-B gene sequences. As shown in Fig. 2B, two qRT-PCR systems were significantly correlated between HCV and GBV-B target amplification.

A second line of evidence was obtained by investigating the chimeric RNA in the liver tissue. Chimeric GBV-B and HCV RNA was detected in the M1-E1E2p7-P0 and M17-CE1E2p7-P1-FK506 animals. Furthermore, the sequence of the nearly-complete genome (nt 22-9604) was observed, including the entire ORF of M17-CE1E2p7-P1-FK506 was 99.89% homologous to the initial construct. This demonstrates a practically identical sequence identity between the injected and replicating viral RNA (Table 1). Only three mutations from the consensus sequence were observed in 50%-100% sequenced clones and seven mutations in 2-4 of eight clones are not unexpected after 25 weeks of HCV CE1E2p7 chimera replication.

Third, the histopathological evidence suggests viral hepatitis was induced. Typical but relatively discreet changes were observed in liver tissues, although clinical symptoms were not clearly recorded. Pathological examination of liver from all HCV chimera-infected marmosets revealed lymphocyte infiltration, severe ground glass degeneration, cholestasis, eosinophilic cells, fibrous expansion, hepatic edema, and cell disarray, and ultrastructural changes including abnormal mitochondrial, lipid droplets and increased numbers of lysosomes (Fig. 3, Table 2). Furthermore, the immunohistochemical staining indicated the presence of viral protein in pathologically modified liver tissues, suggesting that the infection with either of the two HCV chimeras was associated with production of HCV proteins in vivo (Fig. 4).

Lastly, indirect evidence was shown in the immunological parameters. Modest antibody response to HCV core and E2 proteins was detected in the HCV chimera-infected animals. A strong HCV-specific T-cell response from three representative marmosets were identified by ELISpot from a panel of 15 predicted HCV T-cell epitope peptides, of which 10 peptides triggered T-cell responses (Fig. 5). Among the reactive peptides, six were identified previously from the HCV database, and four are unique to this report (Supporting Table 5).

Previous studies with GBV-B chimeras integrating short sequences (<212 bp) of HCV 5′-NCR, p7 or HVR1 resulted in acute resolving infection in marmosets and tamarins mostly within 20 weeks,[8, 10-13] while GBV-B persistent infections were associated with viral molecular evolution.[20, 21] However, in this study HCV CE1E2p7 and E1E2p7 chimeric viruses remained persistently detectable up to 23-29 weeks and in one animal up to 44 weeks, suggesting that chronic infection had developed in these marmosets. A long-term HCV chimera replication with high viral load was more firmly established with oral administration of anti-T-cell drug FK506, as described previously.[22] Few high-frequency substitutions (T2466A/I674K in 67%, C642T/P66L and T1051C/S202P in both 21%) within multiple clones of HCV chimeras were detected from M15-P0 (Table 1), but it remains to be seen whether adaptive mutations for high viral load are involved in viral persistence in vivo.

Fluctuations of low-level HCV RNA has been described often in humans.[23] Compared with HCV clearance or persistence in humans,[24-26] sporadic spontaneous viral clearance of HCV chimeras might be associated with cellular immune response to viral proteins. Virus-specific T-cell response against viral epitopes was described in HCV infection.[27-29] In this study, three representative animals infected with HCV chimeras presented a strong T-cell response to HCV antigens. These data suggest that the HCV chimera-infected small primate model might prove useful in evaluating cellular immune response in HCV infection. Neutralizing antibodies were reported as contributing to the early control of viral replication, but might not be sufficient for long-term protective immunity to HCV infection.[30] Antibodies to HCV chimera core and E2 antigens were found to be weak or undetectable in the study. Whether or not HCV antibodies played a role in the clearance of viremia in HCV chimera-infected marmosets, or whether spontaneous clearance or persistence of chimeras in marmosets mimicked the outcome of human HCV infection associated with IL28B genetic variation in human,[31] requires further investigation.

The data presented in this report raise new questions regarding the mechanisms of interaction between HCV and marmoset host in terms of receptor, tropism, and replication. HCV is restricted to infecting humans and chimpanzees and rarely infects marmosets.[4, 5] Some important results indicated that the envelope glycoproteins E1 and E2 interacted with multiple host cell surface receptors.[32-34] If those human receptors are specific for HCV entry into the host cells, marmosets susceptible to HCV CE1E2p7 and E1E2p7 chimeric viruses might carry similar viral receptors. In this study, HCV CE1E2p7 chimera could infect marmosets, as well as primary marmoset hepatocyte culture (Figs. 2, 6), but could not infect human Huh7.5.1 cells (data not shown). Similarly, HCV JFH-1 could infect Huh 7.5.1 cells but not marmoset liver cells (Fig. 4A). Our data suggest that the interaction between cell receptors and viral envelope proteins may not tightly restrict HCV and GBV-B to infect humans or marmosets.

In conclusion, marmosets infected with replicating HCV chimeras carrying HCV whole structural proteins could provide a novel small primate model for investigation of a virus-host interaction, vaccination, and antiviral therapy in HCV infection.


We thank Dr. Jens Bukh (NIH, Bethesda, MA) for kindly providing the plasmid pGBB. We thank Dr. Ping Zhao (Department of Microbiology, Second Military Medical University, Shanghai, China) for kindly providing mAbs specific to HCV E2. We thank Dr. Qingling Zhang for histopathological analysis, Mr. Mourong Liu, Ms. Xianghui Wu, and Mr. Junling Zeng for assistance in animal surgery and blood sample collection. We thank Dr. Peter Gowland (BTS SRC Berne, Berne, Switzerland) for editing the article.