We have previously reported that CD4 antigenic variants driven by immune selection clustered within immune-dominant epitopes of the HCV nonstructural three (NS3) gene . Consistent to our data, a study showed HCV escapes CD8 T-cell immune response in HLA-B27+ patients through accumulated mutations within an CD8 immune-dominant epitope, viral protein RdRp which is bound by HLA-B27 molecule . In the present study we provide evidence that naturally occurring CD4 mutants of HCV act as antigenic variants that transfer a host protective peripheral Th1 immune response into an inhibitory Th3 or Tr1 response through mutiple mechanisms including: 1) induction of failed T cell activation and simultaneous induction of CD4+CD25- Th3 and CD4+IL-10+ Tr1 cells; 2) up-regulation of CD4+CD25+Th3 cells; and 3) direct inhibition of IL-10.
The T cell receptor (TCR) exhibits an exquisite specificity for its antigen as demonstrated by the significantly different biological outcomes induced by ligands that differ by only a single amino acid . VP 1 consisted of variant peptides with single or double amino acid substitutions located at the center of the core region of the Th1 peptide (see Table 1), and they can be either within or close to the TCR binding sites. Structural or conformational alteration of TCR binding sites of peptides can induce incomplete T cell activation that favors induction of anergy . Indeed, our data show that VP1 peptides dramatically decreased CD4 T cell proliferation (see Figure 1A). Furthermore, CD4 cells cultured with VP1 expressed fewer T cell activation, as determined by low level expression of the T cell activation markers of CD134 and CD45RBhigh (see Additional File 7, Figure S6), providing direct evidence of failed T cell activation by VP1 peptides. CD134 is expressed only on activated T cells, particularly activated CD4+ T cells . Interestingly, VP1 not only caused failed T-cell activation, but also at the same time up-regulated the regulatory cytokines IL-10 and TGF-β and either CD25- Th3 or Tr1 Tregs.
The amino acid substitutions in VP2 peptides are not likely involved in TCR binding directly since they are located outside the core region of the Th1 peptide (see Table 1). However, they may act as interface-disrupting residues which can actively disrupt the TCR binding ability . It is also possible that the amino acid substitutions in VP2 peptides induce conformational changes either in the TCR or in the HLA molecules on APCs. Minor changes at the TCR or APC level could have a significant impact on T cell responses, since specific T cell recognition is strictly determined . In support of this hypothesis, T cells pulsed with VP2 peptides failed to respond to the subsequent challenge with the Th1 peptide NS3358-375 (see Figure 1). This induction of anergy was not due to incomplete T cell activation since VP2 stimulated even higher levels of the T cell activation marker CD134 than wild type peptide (see Additional File 7, Figure S6A). Because up-regulation of CD134 expression by VP2 was accompanied by the highest levels of TGF-β and CD25+ Th3 Tregs, higher CD134 expression by VP2 might reflect activation of pathways related to CD25+ Treg differentiation instead of T effecter activation. These results suggest that VP2 peptides also work as antigenic variants through a distinct pathway.
Collectively, the immune evasion by antigenic variants has two features. First, the modulatory effect of antigenic variants on CD4 response is effective and extensive. As low as 0.1 μM of variant peptide and as many as 17 out of 18 (94%) of the antigenic variants can induce anergy (Figure 1 and unpublished data). Multiple inhibitory components such as IL-10 as well as Th3 and Tr1 cells were induced by these variants (see Figure 2, 3). Second, induction of a deviated immune response or antigen-inducible Tregs is favored over induction of a Th1 response. It seems that a specific Th1 response requires strict TCR engagement and even single amino acid alteration of a Th1 peptide can shift the direction of a Th1 response to an inhibitory direction . In contrast, induction of anergy or antigen-inducible Tregs does not require strict TCR engagement since extensive variants induced similar results of anergy and up-regulated Tregs although they may use different pathways . These features provide a potential mechanism by which a minority of circulating antigenic variants can effectively suppress a protective Th1 response to native viral epitopes. Together with the feature of high mutability of HCV, this hypothesis can explain why the majority of HCV infected patients progress to persistent infection.
Consistent with the studies by others , our current data show an elevated ratio of IL-10+ Tr1 cells in the early course (less than 5 years) of HCV infection (see Table 2). Moreover, the suppression of T cell proliferation by VP1 peptides was abrogated by anti-IL-10 antibody (see Figure 2B). We detected both CD25- and CD25+ Tr1 cells induced by VP1 as early as 1.5 years of HCV infection (see Additional File 3, Figure S2) whereas the majority of IL-10+ cells beyond 2.5 years of infection were predominantly CD25+.
Regardless the cytokine sources, VP1 consistently produced higher levels of IL-10 and CD25- Th3 cells, which directly contributed to the immune modulation by this group of variants. These data suggest that as antigenic variants such as VP1 evolved in the earlier course of HCV infection they initiate the regulatory mechanisms mediated by IL-10 and Th3 cells in order to escape immune pressure. Such VP1 variants up-regulate immunoregulatory IL-10 and Th3 cells that suppressed a protective Th1 response but did not influence the Th3 or Tr1 response. As a result of the existense of VP1, the balance of immune responses shifts to a Tr1 or Th3 profile as we observed in our previous  and present studies. This hypothesis is further supported by our data that: 1) the majority of variants as early as about one year into HCV infection consist of VP1 (see Table 1) ; and 2) patients infected with HCV for less than 5 years have significantly higher percentages of Tr1 and/or Th3 cells and higher serum levels of IL-10 and TGF-β (see Table 2 and Additional File 5, Figure S4).
TGF-β-secreting CD25+ and CD25- Th3 cells induced by antigenic variants occurred in early years of HCV infection. The immune inhibition induced by VP1 was abolished completely by deletion of TGF-β+ cells rather than by blocking or deletion of CD25+ cells, suggesting that this suppressive pathway by VP1 was not CD25-dependent but actually TGF-β-dependent. In contrast, the suppression induced by VP2 was CD25 and TGF-β dependent because either antibody to CD25 or deletion of CD25+ and TGF-β+ cells eliminated the suppressive effect. Accordingly, higher ratios of CD25-TGF-β+ Th3 cells were observed for VP1 and VP2 (see Figure 3D) and VP2 dramatically increased the ratio of CD25+TGF-β+ Th3 cells (see Figure 3B). In the presence of high levels of IL-2 and TGF-β produced by exposure to VP2 (see Figure 2G), CD25-TGF-β+ cells likely differentiated to become CD25+TGF-β+ cells . The fact that deletion of TGF-β+ cells instead of neutralization of TGF-β reversed the immune suppression by both VP1 and VP2 indicates that cell contact of TGF-β-bound cells instead of soluble TGF-β is necessary for the inhibitory effects. These data are consistent with the results reported by Carrier et al. Their data showed that the suppressive effect of TGF-β on T cell responses is due to the induction of Tregs instead of direct inhibition of TGF-β on T cell proliferation. The finding of higher ratio of CD25-TGF-β+ cells induced by VP1 supports the hypothesis that CD25- Th3 cells are the major source of TGF-β+ cells that play an important role in the immune evasion induced by VP1.
A recent study suggested that CD4 T cells function through other mechanisms rather than exerting immune selection in HCV infection . Consistently, the current study indicates possible inhibitory mechanisms used by CD4 T cells in chronic HCV infection. The immune modulation can be initiated or enhanced by naturally occurring antigenic variants of a CD4 epitope. This role of CD4 cells would in turn attenuate the functional efficiency of CD8 T cells that favors HCV persistence. It should be noted that data being obtained with PBMCs collected at one time point of an infection course would only represent "snapshots" of the immune response during entire HCV infection. Additional studies would be needed to precisely correlate the occurrence of antigenic variants with the loss or shift of HCV-specific T cell responses. The phenotypes of immune responses may vary somewhat among different subjects depending on the HCV time continuum and the subjects. Samples collected at different time points from any given subject might also induce variable results. However, the overall trend of immune inhibition induced by CD4 variants during the course of HCV infection was consistent.