|Home | About | Journals | Submit | Contact Us | Français|
T cells play an important role in the control of hepatitis C virus (HCV) infection. We have previously demonstrated that the HCV core inhibits T-cell responses through interaction with gC1qR. We show here that core proteins from chronic and resolved HCV patients differ in sequence, gC1qR-binding ability, and T-cell inhibition. Specifically, chronic core isolates bind to gC1qR more efficiently and inhibit T-cell proliferation as well as gamma interferon (IFN-γ) production more profoundly than resolved core isolates. This inhibition is mediated by the disruption of STAT phosphorylation through the induction of SOCS molecules. Silencing either SOCS1 or SOCS3 by small interfering RNA dramatically augments the production of IFN-γ in T cells, thereby abrogating the inhibitory effect of core. Additionally, the ability of core proteins from patients with chronic infections to induce SOCS proteins and suppress STAT activation greatly exceeds that of core proteins from patients with resolved infections. These results suggest that the HCV core/gC1qR-induced T-cell dysfunction involves the induction of SOCS, a powerful inhibitor of cytokine signaling, which represents a novel mechanism by which a virus usurps the host machinery for persistence.
Hepatitis C virus (HCV) is remarkable in evading host immune surveillance, resulting in persistent infections in the majority of infected individuals that may progress to liver cirrhosis and hepatocellular carcinoma, thus becoming a leading cause of liver transplantation in the United States. Virus-mediated CD4+ and CD8+ T-cell dysfunction seems to play a pivotal role in the establishment of persistent HCV infection. In acute HCV infection, an early and sustained virus-specific T-cell response is critical for viral clearance (1, 20, 33). In contrast, chronic HCV patients display impaired virus-specific CD4+ and CD8+ T cells with lower proliferative and gamma interferon (IFN-γ)-producing capacities (13, 19, 35). This impaired T-cell function may contribute to increased susceptibility to secondary microbial pathogens, including viruses, bacteria, and parasites, during chronic HCV infection (4, 21). Despite extensive investigations of HCV pathogenesis, it still remains unclear why some HCV patients exhibit effective T-cell responses and clear the virus during acute infection, whereas others fail to do so and progress to chronic infection.
HCV core, the first protein to be synthesized upon viral infection, exhibits multiple functions, including the regulation of viral and cellular gene expression, induction of tumorigenesis, modulation of apoptosis, and suppression of host immunity (28). We have previously demonstrated that HCV core can inhibit T-cell proliferation through interaction with gC1qR on T lymphocytes (17, 36-38). This finding, in light of the observation that free core particles circulate in the bloodstreams of HCV-infected patients (24, 25), is particularly noteworthy in HCV pathogenesis since the binding of C1q, the natural ligand for gC1qR, to T lymphocytes leads to immunosuppression (8, 11). It is well known that C1q is the first molecule to be activated in the classical complement cascade and is involved in modulating both innate and adaptive immunity (12). Therefore, the interaction between HCV core and gC1qR might provide the virus with a direct means of immunomodulation through the utilization of gC1qR-mediated signals that dysregulate T-cell immunity (39). Indeed, gC1qR has been shown to interact with several viral and bacterial proteins in addition to HCV core, potentially providing these organisms with a “shared” mechanism of immune evasion (6, 7, 23, 26, 27, 34). However, the mechanism(s) of HCV core/gC1qR-induced T-cell dysfunction, as well as its role in HCV persistence, has yet to be elucidated.
We have previously shown, both in vivo and in vitro (18, 38), that HCV core inhibits T-cell function through the disruption of Th1 cytokine production, including that of IFN-γ, a crucial and potent component of the early host immune response against virus infection. Compelling evidence suggests that IFN-γ production and T-cell responses are negatively regulated by the suppressor of cytokine signaling (SOCS) family members SOCS1 and SOCS3 through inhibition of the Jak/STAT pathway (2, 5, 9, 30). HCV core/gC1qR-mediated T-cell dysfunction involving recruitment or induction of these negative regulators for cytokine signal transduction is an appealing idea, as this would provide a novel mechanism by which a virus usurps the host machinery for persistence.
Despite HCV core being well conserved among various genotypes, the core protein from genotype 1a is able to dampen the host immune response (37), but the expression of core protein from genotype 1b in transgenic mice did not elicit such an effect (22). Thus, studies of the immunomodulatory functions of different HCV core clinical isolates will be valuable for understanding whether there is an intrinsic link between the sequence of core and the dysregulation of T-cell function in resolved and chronic HCV infections. To address this issue, we characterized HCV core proteins from the peripheral blood of two groups of HCV patients who contracted the virus by needlestick exposure: one group of HCV patients failed to clear the virus, and the other group of HCV patients resolved the infection. We found that the HCV core isolated from HCV patients who subsequently cleared the virus differs genetically and functionally with regards to gC1qR binding and T-cell suppression from the core isolated from chronic HCV patients. Importantly, we also found that HCV core/gC1qR-mediated T-cell dysfunction involves the induction of SOCS1 and -3 gene expression, which may play a pivotal role in establishing HCV persistence.
Five health workers who contracted HCV by accidental needlestick were enlisted in this study. At baseline and at one to four weekly intervals after exposure, blood samples were drawn for serological, virological, and immunological analyses, and the alanine transferase (ALT) level of each subject was recorded. A summary of the outcomes and relevant clinical information of these patients is shown in Table Table11.
Human peripheral blood mononuclear cells (PBMC) were isolated from donors at Virginia Blood Services by Ficoll density centrifugation with Lympholyte-H (Cedarlane Labs, Hornby, Ontario, Canada). CD4+ and CD8+ T cells were purified from human PBMC by incubation with a fluorescein isothiocyanate (FITC)-anti-CD4 or FITC-anti-CD8 antibody (Ab), followed by positive isolation with anti-FITC magnetic beads (Miltenyi Biotec, Auburn, Calif.). Purified cells were cultured with complete RPMI 1640 as described previously (38).
HCV core mutants were generated according to the sequences of core isolated from patients with chronic and resolved HCV infections, using QuickChange site-directed mutagenesis kits (Stratagene, La Jolla, CA). Multiple point mutations were carried out in a stepwise sequential strategy. All mutants, in both the pCIneo:core and pQE:core plasmids, were confirmed by sequencing.
For glutathione S-transferase (GST) binding assays, GST and GST-gC1qR were expressed and purified in our laboratory (17). Core (HCV-H, genotype 1a) or various core mutants were labeled with [35S]methionine by use of an in vitro transcription/translation system (TNT; Promega Corp.). GST pull-down analysis was carried out as described previously (17). For core binding on T-cell surfaces, His6-core or His6-dihydrofolate reductase (His6-DHFR) was expressed and purified under native conditions in our laboratory (38). The preps showed one band of 25 kDa upon Coomassie blue staining and were endotoxin-free after polymyxin B-agarose adsorption. Assays of core binding to gC1qR on Molt-4 T cells or PBMC were carried out as described previously (37).
Various concentrations of His6-core (0.25, 0.5, 1, 2, 4, and 8 μg/ml) or DHFR were added to human PBMC (2 × 105 cells in 200 μl/well) stimulated with anti-CD3/28 antibodies (1 μg/ml; BD Pharmingen, San Diego, CA). [3H]thymidine incorporation was performed as described previously (38). To determine IFN-γ production, 1 × 106 PBMC were stimulated with anti-CD3/28 (1 μg/ml) in the presence of various concentrations of His6-core or DHFR for 24 h. Supernatants were harvested, and the production of IFN-γ was analyzed using an OptEIA human IFN-γ enzyme-linked immunosorbent assay (ELISA) kit (BD Pharmingen) under the conditions specified by the manufacturer.
Purified PBMC or CD4+ and CD8+ T cells (2 × 106) were activated with anti-CD3/28 antibodies in the presence or absence of His6-core for various times (0, 0.5, 1, 3, 6, 24, and 48 h). Cell lysates were prepared as previously described (36). A total of 80 μg of protein, as determined by bicinchoninic acid analysis (Pierce, Rockford, Ill.), was denatured with sample loading buffer at 100°C for 5 min and resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by semidry transfer (Amersham Pharmacia Biotech) to a Hybond-P membrane (Amersham Biosciences, Arlington Heights, Ill.). After being blocked with Blotto Tween 20 (10 mM Tris, 0.9% NaCl, 0.1% Tween 20, 5% nonfat dry milk) at room temperature for 1 h, the membrane was probed with rabbit polyclonal Abs (1:1,000; Cell Signaling Technology, Beverly, MA) to phospho-STAT1 (Tyr 701), total STAT1, phospho-STAT3 (Tyr 705), total STAT3, phospho-STAT6 (Tyr 641), and total STAT6 at 4°C overnight. After being washed with Tris-buffered saline-Tween and Tris-buffered saline, the membrane was incubated with a horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G secondary antibody (1:5,000) and subsequently developed by enhanced chemiluminescence (ECL-plus; Amersham Biosciences) on X-OMAT-LS X-ray film (Kodak, Rochester, NY).
Purified PBMC or CD4+ and CD8+ T cells (1 × 106) were treated with or without anti-CD3/28 in the presence or absence of His6-core for various time points, as indicated in the text. Total RNA was isolated by the TRIzol method (Life Technologies). A total of 1 μg RNA was treated with DNase to digest genomic DNA, and 0.27 μg RNA was then reverse transcribed using murine leukemia virus reverse transcriptase (RT) under conditions of 10 min at room temperature, 20 min at 42°C, 5 min at 99°C, and 5 min at 4°C. One microliter of cDNA generated in the RT reaction was added to each PCR mixture. PCRs were carried out using the following primer pairs: SOCS1 sense (5′-ATGGTAGCACACAACCAGGTG-3′) and antisense (5′-TCAAATCTGGAAGGGGAAGGA-3′), SOCS3 sense (5′-CTCAAGACCTTCAGCTCCAA-3′) and antisense (5′-TTCTCATAGGAGTCCAGGTG-3′), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) sense (5′-TGATGACATCAAGAAGGTGG-3′) and antisense (5′-TTACTCCTTGGAGGCCATGT-3′), and β-actin sense (5′-CGAGCGGGAAATCGTGCGTGACAT-3′) and antisense (5′-CGTCATACTCCTGCTTGCTGATCCACATCT-3′). The cycling conditions were 40 cycles of 95°C for 40 s, 58°C for 30 s, and 72°C for 40 s, followed by a single 10-min extension at 72°C. To control for genomic DNA contamination, equal amounts of cDNA from each sample were PCR amplified without RT. The resulting PCR products were separated in a 2% BioGel (Bio 101, Carlsbad, CA). To examine whether gC1qR mediates the core-induced induction of SOCS, a 1:10-diluted anti-gC1qR polyclonal antibody (PAb) or rabbit prebleed control serum was coincubated with cells treated with core, and SOCS mRNA expression was assessed as described above.
To generate double-stranded RNA templates for SOCS1 and SOCS3, the following two primer pairs containing the T7 RNA polymerase promoter (underlined) were synthesized: SOCS1 sense (5′-GCGTAATACGACTCACTATAGGGAGACCGTGCCCCGCGGTCCCGGCC-3′) and antisense (5′-GCGTAATACGACTCACTATAGGGAGATCAAATCTGGAAGGGGAAGGA-3′) and SOCS3 sense (5′-GCGTAATACGACTCACTATAGGGAGACTCAAGACCTTCAGCTCCAA-3′) and antisense (5′-GCGTAATACGACTCACTATAGGGAGATTCTCATAGGAGTCCAGGTG-3′). A green fluorescent protein (GFP) control plasmid and primers were provided in a Dicer siRNA generation kit (Gene Therapy System Inc., San Diego, CA). SOCS1, SOCS3, and GFP PCR products were generated as described above. After Gene Clean (Qbiogene Inc., Carlsbad, CA) treatment of the PCR products, double-stranded RNAs were transcribed from the SOCS1, SOCS3, and GFP DNA templates, and small interfering RNAs (siRNAs) were generated by using a recombinant Dicer enzyme (Gene Therapy System Inc., San Diego, CA).
For T-cell transfection, 2 μg of either SOCS1, SOCS3, or GFP siRNA in 50 μl of diluent was incubated with 10 μl of GeneSilencer reagent in 40 μl of diluent at room temperature for 30 min. The siRNA-GeneSilencer complex was gently added to 5 × 105 Jurkat T cells in 500 μl of serum-free lymphocyte medium (Life Technology). Four hours after transfection, another 500 μl of RPMI 1640 containing 20% fetal bovine serum was added, and the siRNA- or mock (phosphate-buffered saline-GeneSilencer)-transfected cells were treated with concanavalin A (ConA; 0.1 μg/ml) in the presence or absence of His6-core (5 μg/ml). GolgiStop (4 μl/6 ml of medium; BD Pharmingen) was added to the cultures 6 h before harvesting at various times, as indicated in the text. The cells were washed two times with fluorescence-activated cell sorting (FACS) buffer and fixed with 100 μl of CytoFix/Perm (BD Pharmingen) at 4°C for 20 min. After washing of the cells two times with 200 μl of PermWash, the amounts of intracellular IFN-γ production in the cells were assessed using allophycocyanin-conjugated anti-human IFN-γ (1 μl/50 μl PermWash; eBioscience) and then analyzed by flow cytometry. Another set of cells transfected with siRNA and treated with ConA and core, as described above, was used for total RNA extraction and RT-PCR for SOCS1/SOCS3 and gC1qR. The specificity of the SOCS-siRNA effect was verified by measuring the amount of GAPDH cDNA in each sample.
Levels of significance were determined by Student's t test, using the SPSS program. P values of <0.05 were considered significant, and those of <0.01 were considered very significant.
In order to investigate the role of HCV core/gC1qR-induced T-cell suppression in the outcome of viral infection, we cloned the HCV core genes from two groups of patients who were accidentally infected with HCV by needlestick exposure. The outcomes and relevant clinical information for these patients are shown in Table Table1.1. These clones are valuable for addressing this issue since some patients successfully cleared the virus and resolved the infection, either spontaneously (Table (Table1;1; MP079, 1b/R-1) or after IFN-α treatment (Table (Table1;1; JH010, 1b/R-2), while others failed to do so and progressed to chronic infection (Table (Table1)1) (MP022, 1a/C-1; MP036, 1a/C-2; and MP047, 1a/C-3). An ex vivo analysis of T-cell responses in these patients has been previously reported: while patients with resolved infections mounted T-cell responses, responses were impaired in chronic HCV patients (33).
Blood samples were collected during the early phase (Table (Table1)1) (41 to 76 days) of infection from both groups of patients. HCV core genes were amplified by RT-PCR using RNAs isolated from blood and subsequently cloned into a T-A vector. Twenty clones from each patient were selected for sequencing, and the genetic differences in core proteins from patients with chronic and resolved infections are shown in Fig. Fig.1.1. The HCV core proteins from two patients who cleared the virus and subsequently resolved the infection (Fig. 1, R-1 and R--2)2) showed the same sequence, but those from three patients who progressed to chronic infection differed (Fig. 1, C-1, C-2, and C--3).3). Strikingly, among five amino acids that were variable in HCV patients, four of them (R70Q, T75A, Y81H, and C91M) reside within the gC1qR binding region (17).
Since most of the amino acid differences between the core isolates from patients with chronic and resolved HCV infections reside within the gC1qR binding region, it is likely that these alterations influences core's ability to associate with gC1qR. To test this possibility, the interaction between core isolates and gC1qR was first examined in vitro with a GST-binding assay (17). As shown in Fig. Fig.2A,2A, similar amounts of radiolabeled core proteins exhibited various degrees of binding to gC1qR. Specifically, GST-gC1qR pulled down a larger amount of core proteins from chronic patients than that from patients with resolved infections, demonstrating a gC1qR binding ability of chronic core isolates of more than double that of resolved core isolates. No association between core protein and GST alone was observed (data not shown).
We next determined the ability of core proteins from patients with chronic and resolved HCV infections to bind to gC1qR on the surfaces of T cells. We first verified that the anti-core MAb recognized the various core isolates similarly by Western blotting, allowing us to utilize this MAb for comparisons of cell surface binding by FACS analysis. As shown in Fig. Fig.2B,2B, core proteins from three chronic patients bound Molt-4 T cells better than those from two patients with resolved infections. Using PBMC gated on lymphocyte populations, similar results were obtained in terms of the percentage of core-bound cells as well as the total mean fluorescence intensity (data not shown). Since the level of gC1qR expression on the surfaces of Molt-4 T cells or PBMC remains constant, the difference in core binding on these cells reflects varying abilities of core to associate with gC1qR. In addition, core binding on the surfaces of T cells was blocked by anti-gC1qR Ab treatment, but not by a control Ab (Fig. (Fig.2C),2C), suggesting a specific binding of core to gC1qR displayed on the T-cell surface.
Since the concentration of core protein secreted into the peripheral blood varies from patient to patient as well as during different phases of infection (24, 25), we determined whether various concentrations of core isolates showed similar dose-dependent association profiles with gC1qR on the surfaces of T cells. As shown in Fig. Fig.2D,2D, binding at core concentrations below 0.25 μg/ml was barely detected by FACS analysis. Measurable differences in the gC1qR binding abilities of two sources of core proteins were observed to be linear at concentrations between 0.5 and 2 μg/ml. The gC1qR binding profile for core from chronic HCV patients became a plateau when core concentrations increased above 4 μg/ml. In contrast, the gC1qR binding of core from patients with resolved infections still increased, so that it eventually reached the same level that was observed for core from chronic patients.
While effective T-cell responses are crucial for viral clearance during acute HCV infection, impaired T-cell functions are observed in patients with chronic HCV infections. Since core-induced T-cell suppression depends on the association with gC1qR displayed on the cell surface (37), the observed differential gC1qR binding abilities of the variant core proteins may result in a differential modulation of T-cell function. To test this possibility, we compared the abilities of HCV cores from two groups of patients to suppress T-cell responses in primary human T cells stimulated with anti-CD3/28. As shown in Fig. 3A, T-cell proliferation was inhibited by HCV core proteins from patients with both chronic and resolved infections compared to that of cells treated with anti-CD3/28 alone. However, the magnitude of the T-cell inhibitory effect was higher for core proteins from chronic HCV patients than that from patients with resolved infections (P < 0.05).
We next examined the effect of various doses of core proteins on the suppression of T-cell responses to anti-CD3/28 stimulation. In line with the observed dose-dependent core binding on the cell surface, T-cell inhibition was rarely detectable at low concentrations of core proteins (<0.25 to 0.5 μg/ml), but a significant and dose-dependent inhibitory effect was observed with 1 to 2 μg/ml of core (Fig. (Fig.3B).3B). Interestingly, while core concentrations of >4 μg/ml exhibited comparable cell surface binding in both resolved and chronic infections, only the core from chronic patients maintained a dose-dependent inhibition of T-cell function. In contrast, although the gC1qR binding ability of resolved core isolates to T cells appeared similar to that observed for chronic core isolates at higher concentrations, there was no corresponding increase in T-cell inhibition. These results suggest that the core from chronic patients not only has a greater affinity for gC1qR but also exerts stronger T-cell inhibitory signaling than that from patients with resolved infections.
We also examined the effect of various concentrations of core proteins on IFN-γ production from anti-CD3/28-stimulated PBMC by ELISA. As shown in Fig. Fig.3C,3C, IFN-γ production from activated T cells was inhibited by HCV core proteins from both chronic patients and patients with resolved infections compared to that in the DHFR control but was more profoundly impaired in T cells treated with core from chronic HCV patients than in those treated with core from resolved infections.
Core proteins from patients with chronic and resolved infections display differences in their amino acid sequences at five positions, including four (R70Q, T75A, Y81H, and C91M) within the gC1qR binding region. To determine specific residues critical for the observed disparity in gC1qR binding and T-cell inhibition between chronic and resolved core isolates, a series of core mutants were generated by substituting amino acids stepwise from those of chronic to those of resolved isolates by sequential site-directed mutagenesis. As shown in Fig. Fig.4A,4A, the core sequence listed in line a corresponds to the HCV-H strain and chronic isolate C-3, line b corresponds to C-2, line c corresponds to C-1, and line g corresponds to the resolved core isolates. Mutants d to f do not correspond to any core isolates but represent sequences between those of chronic core and resolved core isolates. These lab-generated mutants were used to express 35S-labeled core proteins for gC1qR binding assays.
As shown in Fig. Fig.4B,4B, the ability of core to interact with gC1qR was deduced as amino acids of core were changed from those of chronic to those of resolved isolates. In contrast, no difference was observed for the previously reported association of GST-Fas with [35S]Met-core mutants (14), suggesting that these amino acid substitutions within the core protein specifically affect its association with gC1qR (data not shown). Similar results were observed by analysis of the binding of core mutants to the surfaces of Molt-4 T cells by FACS analysis (Fig. (Fig.4C).4C). Interestingly, the ability of core to associate with gC1qR varied only slightly among the three chronic isolates (a, b, and c); however, it was significantly reduced when the amino acids were gradually switched from those of chronic to those of resolved isolates (a to g). This was especially true for the amino acid switches of R70Q (d) and C91M (g). It is notable that the substitution of V187I, which is outside the gC1qR binding region of core, did not affect core's gC1qR binding ability.
We next determined if the amino acid changes to the core protein affected its inhibitory function in T-cell proliferation. As shown in Fig. Fig.4D,4D, the ability of core to inhibit T-cell proliferation was diminished when the amino acids were switched from those of chronic to those of resolved isolates. Likewise, residue switches of R70Q and C91M in the core protein were more critical for its ability to inhibit T-cell proliferation. Thus, these amino acid changes within the core protein seem to be critical for gC1qR ligation as well as T-cell inhibitory signaling.
Since SOCS1 and SOCS3 are negative regulators of T-cell proliferation and IFN-γ production (5, 9), they might be involved in core/gC1qR-induced T-cell dysfunction. To examine this possibility, we treated PBMC or purified T cells with HCV core in the presence or absence of T-cell receptor (TCR) stimulation. RT-PCR analysis of PBMC stimulated with anti-CD3/28 for 24 h in the presence of core revealed increased expression of SOCS1 and SOCS3 (Fig. (Fig.5A).5A). Notably, cells treated with core alone in the absence of TCR activation also elicited SOCS expression (Fig. (Fig.5A).5A). GAPDH was not affected by the treatment. Repeated semiquantitative experiments using series of diluted RT reactions of T cells treated with core protein confirmed the induction of SOCS1 and SOCS3 but not of SOCS2 (data not shown). To determine the role of gC1qR in the core-induced induction of SOCS, anti-gC1qR or a control serum was added to the culture simultaneously with core, and SOCS expression was detected as described above. As shown in Fig. Fig.5B,5B, core-mediated inductions of SOCS1/3 were diminished by the addition of anti-gC1qR, but not the control Ab, suggesting that SOCS induction is specifically mediated by core/gC1qR ligation.
An analysis of the kinetics of SOCS induction revealed that the expression of SOCS1 was detectable at 3 h and was sustained for 24 h in CD4+ and CD8+ T cells following treatment with core protein (Fig. (Fig.5C).5C). In fact, the up-regulation of SOCS1 could be observed as early as 1 h in CD8+ T cells following core treatment (data not shown). The relatively early up-regulation of SOCS1 in CD8+ T cells by core is probably due to the higher level of gC1qR expressed on these cells, as we have previously reported (37), so that core elicited a more profound inhibition of CD8+ T cells than of CD4+ T cells. SOCS3 mRNA could be detected in both CD4+ and CD8+ naïve T cells (time zero), and the up-regulation of SOCS3 was observed at 3 h and persisted until 24 h after the core treatment. To compare the abilities of core proteins from patients with chronic and resolved HCV infections to induce SOCS expression, we incubated various core proteins with purified CD4+ and CD8+ T cells for 6 h and detected SOCS mRNA as described above. As shown in Fig. Fig.5D,5D, core proteins from three chronic HCV patients induced more SOCS1/3 expression in T cells than those from two patients with resolved HCV infections. It is notable that the increased induction of SOCS1/3 by cores from chronic patients compared to that by resolved cores correlates well with the disparate effects of these isolates on T-cell functions, as previously reported (33).
Because SOCS1 and -3 inhibit T-cell signaling through the prevention of Jak/STAT signaling (2), we next examined the effect of HCV core on anti-CD3/28-induced phosphorylation of STAT1/3 in T cells. As shown in Fig. Fig.6,6, there were low levels of both phosphorylated STAT1 (tyrosine 701) and phosphorylated STAT3 (tyrosine 705) in unstimulated T cells (Fig. (Fig.6A,6A, time zero). This might be due to the nature of the heterogeneous cell populations used in the experiment, which were purified from donors who had encountered various pathogens in their lifetime. STAT phosphorylation was observed as early as 30 min after TCR stimulation and was maximal between 3 and 6 h after stimulation (Fig. (Fig.6A,6A, left panel). The HCV core inhibited the phosphorylation of both STAT1 and STAT3 as early as 3 h after stimulation, and this inhibition persisted for at least 48 h after treatment. In contrast, HCV core did not affect the levels of total STAT1/3. A time course of STAT1/3 phosphorylation in activated CD4+ and CD8+ T cells treated with or without core showed a pattern of inhibition similar to that described for PBMC (data not shown). The kinetics of inhibition of STAT1/3 activation correlate with the time course of core/gC1qR-induced SOCS1/3 induction. However, HCV core did not affect the phosphorylation status of STAT6 (tyrosine 641; data not shown), suggesting that core/gC1qR ligation leads to the selective inhibition of STAT1/3 activation.
We next determined the inhibition of Jak/STAT signaling in T cells mediated by core proteins from patients with chronic and resolved HCV infections. PBMC were stimulated with anti-CD3/28 in the presence of various core isolates for 6 h, and the phosphorylation of STAT1 and STAT3 was detected as described above. As shown in Fig. Fig.6B,6B, cores from patients with both chronic and resolved HCV infections diminished the phosphorylation of STAT1/3 in a dose-dependent manner compared with the levels seen in cells treated with anti-CD3/28 alone. However, the core proteins from chronic HCV patients inhibited STAT1/3 phosphorylation more profoundly than those from patients with resolved infections, particularly at a high dose (4 μg/ml), where both showed the same ability to bind to T cells. Total STAT protein levels were not affected by any of the treatments.
Since tyrosine-phosphorylated STAT dimerizes and translocates from the cytoplasm to the nucleus to activate transcription, we also examined the effect of core on the TCR-activated DNA binding activity of STAT1 by electrophoretic mobility shift assays. The kinetic inhibition of STAT1 nuclear translocation by core in a dose-dependent manner was significant 3 h as well as 6 h after the treatment, and cores from chronic HCV patients inhibited STAT1 DNA binding more profoundly than cores from patients with resolved infections, although both treatments inhibited DNA binding of STAT1 compared to the levels seen in cells treated with anti-CD3/28 alone (data not shown). Taken together, these results indicate that HCV core/gC1qR-induced suppression of T-cell responses is mediated by inhibition of Jak/STAT signaling, presumably through the induction of SOCS expression in T cells.
To further determine the role of SOCS proteins in the inhibition of T cells by HCV core/gC1qR ligation, specific siRNAs were designed and used to silence SOCS1/3 expression in Jurkat T cells. The experiment for knocking down the expression of SOCS was performed in Jurkat cells due to the lower transfection efficiencies of siRNAs in primary human T cells. Briefly, following transient transfection of Jurkat T cells with either SOCS1, SOCS3, or GFP siRNA, T cells were stimulated with ConA for 24, 48, or 72 h in the presence or absence of core protein. The effect of HCV core/gC1qR ligation on T cells was determined by measuring intracellular IFN-γ production by FACS analysis.
As shown in Fig. Fig.7A,7A, HCV core inhibited IFN-γ production in T cells stimulated with ConA for 24 h. Importantly, the production of IFN-γ was boosted in ConA-stimulated T cells following transfection of the cells with either SOCS1 or SOCS3 siRNA. In contrast, GFP siRNA transfection did not significantly affect the inhibition of IFN-γ production by HCV core. This suggests that the induction of SOCS1 and SOCS3 by HCV core is involved in negative regulation of IFN-γ production in Jurkat T cells.
We next assessed HCV core-mediated inhibition of IFN-γ production 48 h and 72 h after ConA stimulation. To this end, the inhibition of IFN-γ by HCV core in mock- or siRNA-transfected cells was compared to that in the absence of core protein 24 h, 48 h, and 72 h after ConA stimulation (Fig. (Fig.7B).7B). Consistent with the 24-h ConA stimulation, the core-mediated inhibition of IFN-γ production was also observed in core-treated Jurkat cells 48 h and 72 h after ConA stimulation (Fig. (Fig.7B).7B). In addition, the percent inhibition of HCV core-mediated IFN-γ production was significantly reduced by SOCS1/3 siRNA transfection at various time points (24 h, 48 h, and 72 h) after ConA stimulation. However, the percent inhibition of IFN-γ production by HCV core was not affected by GFP siRNA transfection. This enforces the finding that SOCS1/3 induction is targeted by HCV core/gC1qR ligation to inhibit IFN-γ production and T-cell function.
The effect of SOCS1/3 siRNA is target gene specific, as transfection of the SOCS1 siRNA only affected the expression of SOCS1, and not that of SOCS3, and vice versa (Fig. (Fig.7C).7C). In addition to GAPDH, gC1qR expression was not affected by the transfection. This suggests that abolishment of the inhibitory effect of core in siRNA-transfected cells is not due to an alteration in gC1qR expression, which may potentially interfere with the core/gC1qR interaction, but because of the absence of SOCS expression in the cells.
We have previously demonstrated that HCV core inhibits T-cell responses through interaction with gC1qR (17, 36-38). However, the link between core/gC1qR-induced T-cell dysfunction and HCV persistence has not been established. In this study, we demonstrated that HCV cores isolated from patients who cleared and failed to clear the virus differ in sequence, gC1qR affinity, and T-cell inhibitory activity. Specifically, the core protein from chronic HCV patients binds to gC1qR with a greater affinity than that from patients with resolved infections, which in turn delivers a more profound inhibitory signal to T cells. This inhibition is mediated by the disruption of STAT1/3 phosphorylation through the induction of SOCS 1/3 expression. Since core from chronic HCV patients induced more SOCS than core from patients with resolved infections, it is likely that SOCS induction by core/gC1qR ligation plays a role in T-cell dysfunction and HCV persistence. This is supported by recent reports showing that SOCS1 expression is enhanced in the livers of chronic hepatitis C patients (16), while SOCS3 induction is associated with IFN-α antagonistic activity of the HCV core protein (5). Additionally, SOCS3 induction was observed in other infections, such as Listeria monocytogenes, Leishmania donovani, and herpes simplex virus type 1 infections, suggesting that the induction of SOCS may be a common mechanism of immune evasion shared by these pathogens (3, 31, 32, 40).
Intriguingly, the HCV core protein is secreted from infected cells and circulates in the bloodstream of infected patients, where it can interact with immune cells, including dendritic cells, macrophages, and T lymphocytes (24, 25). We have previously shown that extracellular core-mediated T-cell suppression is gC1qR specific and core dose dependent (37). Thus, the amount of free core protein circulating in the blood is crucial for eliciting the observed T-cell dysfunction. Given the vigorous HCV replication in the early phase of viral infection, it is likely that a large quantity of core protein is released from infected cells and circulates in the bloodstream of infected patients. In addition, the amount of free core protein is likely greater in the microenvironment of the liver, where viral replication occurs, than in the circulating blood of HCV-infected patients. Therefore, HCV core protein in the periphery of infected patients may not be sufficient for eliciting generalized immunosuppression but enough to reach the threshold locally in the liver for dysregulation of T cells in the early “window phase” (absence of anti-core Ab), thus establishing persistent HCV infection. Importantly, we found here that this dose-dependent effect on T-cell suppression was different between core proteins from patients with chronic and resolved HCV infections. Specifically, HCV core from chronic patients not only displays a higher gC1qR binding ability but also has a stronger inhibitory effect than core from patients with resolved infections. These results suggest that in addition to the concentration of core, the sequence of the core protein plays a role in suppressing T-cell functions to control viral infection.
The patients in this study were infected with HCV by accidental needlestick exposure. It remains unclear whether the dominant viral clone in these patients resulted from acute or chronic infection and how the virus exhibits divergent core sequences and outcomes of infection. One possibility is that individuals contracted viruses from acute-phase HCV patients and that this led to acute infection or chronic infection through viral mutations generated under host immune pressure. These mutations, which may alter the ability of the core protein to interact with gC1qR, may occur very quickly in vivo because of the very high virus replication rate (estimated at 1012 virions per day) and the error-prone HCV RNA-dependent RNA polymerase. Alternatively, given the high rate of persistent HCV infection, it is more likely that these HCV-infected individuals contracted virus from chronic HCV patients. How, then, does an individual who received the virus from a chronic source have an acute infection? It is well known that HCV exists as a closely related, though genetically diverse, population of quasispecies within an individual. All virions were once thought to be equally capable of interacting with host cells after inoculation, but a recent study of viral kinetics demonstrated that certain quasispecies variants have increased replicative fitness over others (15). The host-mediated selection of quasispecies can result in a minor subset of the donor's HCV variants being selected after transfusion, thus becoming the dominant strain of virus in the new host (29). This selectivity of the host for specific quasispecies variants may change, presumably via the complexity and composition of putative HCV receptors, including CD81, LDL, and CD209L (L-SIGN or DC-SIGN) (10).
In summary, this study reporting that SOCS induction by core/gC1qR ligation inhibits T-cell function represents a novel mechanism by which HCV takes advantage of the host machinery for persistence. To our knowledge, this is the first report describing the differential characteristics of the HCV core protein in chronic infections versus resolved infections and the first study showing SOCS induction by HCV core for persistence. It is notable that due to the small number of patients in this study, the three chronic HCV patients who failed to produce a significant T-cell response (33) revealed genotype 1a, while the two patients who cleared the virus had genotype 1b. Further studies are needed to investigate the dependency of genotypes regarding HCV core/gC1qR-mediated immune dysregulation. On the other hand, the genetic difference of host factors may also play a role in the outcome of viral infection. Specifically, the gC1qR expression level or polymorphism on T cells might play a role in T-cell dysfunction and the outcome of viral infection. We therefore conclude that the induction of SOCS by core/gC1qR interaction may play a pivotal role in T-cell dysfunction during the early phase of viral infection, which in turn contributes to HCV persistence. Intervention targeted at these specific interactions may provide a potential rationale for designing therapeutics to prevent persistent HCV infections.
We thank our colleagues for their constructive criticism and comments. We also greatly appreciate the outstanding technical support of Travis Lillard and Susan Landes.
This work was supported by an American Association for the Study of Liver Diseases/Schering Advanced Hepatology fellowship (to Z.Q.Y.) and by Public Health Service grants DK066754, DK063222, and U19 AZ 066328 (to Y.S.H.).