Background. Infection with hepatitis C virus (HCV) is characterized by impairment of viral effector T cell responses and a high propensity for viral persistence. Previous studies have demonstrated that chronic HCV infection is associated with an increased frequency of regulatory T (Treg) cells, compared with that in persons whose infection resolved and in healthy persons. However, all patients in prior analyses had exposures in the distant past, precluding the ability to determine whether Treg cells play a causal role in establishing persistence during the earliest stages of infection or whether they are expanded because of viral persistence.
Methods. For the first time, we longitudinally analyzed Treg cells in patients with acute HCV infection (n = 27). We used a multiparameter approach, including fluorescence-activated cell sorting analysis of cell-surface and intracellular antigens, coculture experiments with highly purified CD4+CD25high regulatory and CD4+CD25− responder cell populations, and multiplex analysis of secreted cytokines.
Results. Forkhead transcription factor 3 (FoxP3) expression and Treg cell suppression were greater in patients with acuteHCVinfection than in healthy control subjects but were not different at the first time point among patients who subsequently developed persistence or resolved HCV infection spontaneously; however, 6 months later, the resolution of disease was associated with a relative loss of functional suppression.
Conclusions. Collectively, these data indicate that patients with acute HCV infection who develop chronicity versus spontaneous resolution exhibit temporal changes in Treg cell function. It is possible that repetitive viral antigenic stimulation alters the function of Treg cells over time.
Considerable data demonstrate that CD4+CD25high regulatory T (Treg) cells constitute an important component of the normal healthy immune response , and they have been implicated in a range of disease states, including autoimmune diseases, allergy, cancer, allograft rejection, and infectious diseases . The majority of Treg cells constitutively express CD25 (the interleukin [IL]-2 receptor α chain); the CD4+CD25high Treg cell population comprises ∼2%-5% of peripheral CD4+ T cells in humans. Treg cells do not proliferate in response to antigenic stimulation and can potentially suppress the proliferation of other CD4+ or CD8+ T cells induced by polyclonal or antigen-specific stimuli . Their depletion leads to the spontaneous development of various autoimmune diseases in rodents  and enhances immune responses against a number of infectious microbes in vitro .
Because the hallmarks of infection with hepatitis C virus (HCV) include a very high likelihood of viral persistence  as well as impaired HCV-specific effector T cell responses [6-8], recent studies have focused on the role played by T cell counteractive processes. Studies have revealed an increased frequency of Treg cells in the peripheral blood of patients with chronic HCV infection, compared with that in persons whose HCV infection spontaneously resolved or in healthy control subjects [9, 10]. Furthermore, because Treg cells have been shown to suppress the proliferation of HCV-specific cytotoxic T lymphocytes (CTLs), it has been suggested that they play a central role in viral persistence and may be a target of immunotherapy for chronic HCV infection [11, 12]. However, all reports have focused on patients with chronic or resolved infection after an exposure in the distant past and thus have not been able to determine whether Treg cells play a role in establishing persistence during the early stages of infection or whether they are expanded because of viral persistence. This is the first investigation to address these critically important questions by characterizing the proportional frequency, phenotype, and function of Treg cells by use of a multiparameter approach in human subjects with acute HCV infection who either developed viral persistence or experienced spontaneous resolution of the infection. We found that, compared with healthy subjects, patients with acute HCV infection demonstrated a higher frequency of circulating forkhead transcription factor 3 (FoxP3)- expressing Treg cells, irrespective of whether they ultimately developed spontaneous resolution or persistence. To characterize the suppressive function of Treg cells (i.e., their interaction with responders), highly purified CD4+CD25high regulatory and CD4+CD25− responder cell populations were cocultured at various ratios and with different concentrations of anti-CD3 and anti-CD28 antibodies. Patients with acute HCV infection demonstrated greater ex vivo suppression than did healthy control subjects. Examination of contact-dependent suppression demonstrated that, although Treg cells at the earliest time point were functionally similar in both groups of patients with acute HCV infection, 6 months later, the Treg cells derived from patients with chronicity were significantly more suppressive than those from patients with spontaneous resolution. In selected patients, comixing experiments with effector cells from a constant time point (baseline) and Treg cells from different time points confirmed the temporal change in the function of Treg cells during acute HCV infection. Taken together, these results suggest that, although differences in the function of Treg cells at the earliest stages of acute HCV infection do not determine virologic outcome— that is, persistence versus recovery—the development of chronicity may lead to expansion or maintenance of Treg cells that regulate effector immune responses.
Study population. The study group comprised patients with acute HCV infection who were recruited from multiple sites. The study protocol was approved by all appropriate institutional review boards. Acute HCV infection was diagnosed based on the following: HCV antibody seroconversion in a subject with previously negative HCV testing results; a positive test result for HCV antibody in a subject with new-onset risk factors and an alanine aminotransferase (ALT) level>10-fold higher than normal; or HCV RNA positivity with HCV antibody negativity. Twenty-seven treatment-naive patients (13 male and 14 females) were recruited for the present study. The majority of patients were white (93%). Spontaneous viral resolution and chronicity were defined as the absence or presence of HCV RNA at 6 months after enrollment with at least 2 (range, 2–5) viral determinations.
Duration of infection describes the time from the estimated date of HCV acquisition to enrollment. For subjects with known iatrogenic exposures, acquisition date was defined as the date of exposure; these included 2 health care workers infected by needlestick and 1 person with iatrogenic infection, confirmed by molecular sequencing for acquisition from a cadaveric tissue donor. For subjects who lacked known iatrogenic exposures but who reported symptoms, the acquisition date was defined as 6 weeks before the onset of symptoms. Among the remaining participants, the midpoint between the last negative antibody test result and the earliest of the first positive antibody test or positive HCV RNA test result was used as the date of acquisition. Twenty-two (81%) of the patients were enrolled within 6 months of infection, and all were enrolled within 8 months.
Sample preparation and storage. Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll (Amersham Biosciences) density gradient centrifugation and cryopreserved for subsequent analyses. PPT tubes (Becton Dickinson) were used to isolate plasma from whole blood at months 0, 2, 4, 6, 9, and 12, which was frozen and later thawed for viral load and genotype or serotype (Abbott Murex Laboratories) testing. HCV genotyping (line probe assay) and viral level determination (HCV RNA 3.0 bDNA; lower limit, 615 copies/mL) were performed by Bayer Reference Testing Laboratory.
Flow cytometry analysis of cell-surface and intracellular antigens. Monoclonal antibodies specific for CD3, CD4, CD8, CD25, CD69, CD62-L, and CD152 (CTL-associated antigen [CTLA]-4) were obtained from BD Biosciences. Antibody to lymphocyte activation gene (LAG)-3 was supplied by Alexis Biochemicals. Cryopreserved PBMCs (2.5 × 105) were stained for expression of cell-surface antigens at 4°C in the dark for 30 min, washed twice in 2 mL of PBS containing 1% bovine serum albumin and 0.01% sodium azide (FACS wash), and fixed in 200 μL of 1% paraformaldehyde. Intracellular staining for CTLA-4 was performed after membrane permeabilization of fixed cells with 0.02% saponin (Sigma). The anti-human FoxP3 staining set from eBioscience was used for FoxP3 analysis in accordance with the manufacturer's instructions. Isotype-matched control antibodies were used to determine background levels of staining. Four-color multiparameter flow cytometry was performed using a FACSCalibur instrument (BD Biosciences) compensated with single fluorochromes and CellQuest software (version 3.0; BD Biosciences).
Proliferation/Treg cell suppression assay. For cell purification, CD4+CD25high Treg cells were separated from PBMCs by use of a CD4+CD25+ Regulatory T cell Isolation Kit (Miltenyi Biotech), in accordance with the manufacturer's protocol. Briefly, CD4++ cells were negatively selected by magnetic bead separation and further separated into CD4+CD25high Treg cells and CD4+CD25low/- effector cells. Purity of the isolated cell populations was determined by flow cytometry. If cell yield allowed, an aliquot of isolated cell fractions was retained for polymerase chain reaction (PCR) analysis of FoxP3 mRNA.
For the proliferation assay, a modification of a method that has been described elsewhere 13, 14] was used. All cells were cultured in 96-well round-bottom plates (Costar) with RPMI 1640 supplemented with 5 mmol/L HEPES, 2 nmol/L l-glutamine, gentamicin, and 0.5 mmol/L sodium pyruvate (all from Gibco BRL, Invitrogen) and 1% heat-inactivated human serum. CD4+CD25− effector cells (2 × 104) were cultured alone or in combination with CD4+CD25high Treg cells in triplicate. For all patients, cells were cocultured at effector to Treg cell ratios of 1:0, 1:2, 1:0.2, and 0:2 (and, for selected patients, at the additional ratios of 1:1, 1:0.5, 1:0.3, and 0:1) in wells containing either plate-bound anti-CD3 (2 μg/mL) and plate-bound anti-CD28 (2 μg/mL) or plate-bound anti-CD3 (0.1 μg/mL) and soluble anti-CD28 (1 μg/mL) with 1 × 104 irradiated autologous PBMCs as accessory cells . On day 5 of culture, 50–100 μL of supernatant was removed from each well for cytokine determination before the addition of 0.5 μCi of 3Hthymidine (Amersham Biosciences) to each well and incubated for an additional 16–18 h. Cells were harvested onto glass fiber filters, and proliferation (the uptake of labeled thymidine) was measured by liquid scintillation. Percentage of suppression was calculated using the following formula:
where CE is the counts per minute for effector cells only and CE:Treg is the counts per minute for a given effector to Treg cell ratio.
Multiplex cytokine analysis. On day 5 of the proliferationassay coculture of Treg cells with responder cells, before the addition of tritiated thymidine, 50–100 μL of supernatant was removed from each replicate well and then pooled and frozen for later cytokine/chemokine analysis (for more information, see appendix A, which appears only in the electronic edition of the Journal).
Statistical analysis. Descriptive statistics, including means, SDs, and frequencies, were determined for the 2 groups of patients with acute HCV infection (resolved or chronic) and for healthy subjects. Differences between groups and over time were tested using median analysis (Wilcoxon rank sum and Mann-Whitney U test) with JMP statistical software (version 6.0; SAS). Statistical significance was assumed when the resultant P value was <.05.
Table 1 demonstrates the demographic, virologic, and clinical features of the 27 study patients with acute HCV infection; 15 patients ultimately developed chronic infection, and 12 had infections that resolved spontaneously. Figure 1 shows the longitudinal results of serum ALT and HCV RNA levels in patients with acuteHCVinfection according to whether they developed a chronic infection or whether it resolved spontaneously. In these 2 different virologic outcome groups, the median and mean age (38 years) was the same, and the distributions of risk factors and serum ALT levels at baseline (enrollment) were comparable. The estimated time between acquisition of HCV infection and enrollment was equivalent in the 2 groups (median, 129 days for those who developed chronic infections and 133 days for those whose infections resolved spontaneously); the time from presumed exposure to enrollment was also comparable in the 2 groups. However, a significantly higher proportion of patients whose infections resolved spontaneously were women (10/12) versus those who became chronically infected (4/15) (P = .006).
Frequency and phenotype of CD4+CD25high regulatory T cells in acute HCV infection. As has been described previously [2, 14], human peripheral blood contains a heterogeneous population of CD4+CD25+ T cells that express either high levels of CD25 (exhibiting regulatory function) or moderate levels of CD25. By use of FACS analysis, the frequency of Treg cells was enumerated directly ex vivo from patients with acute HCV infection. As shown in figure 2, there was no significant difference in the proportions of T cells that were CD4+CD25high Treg cells at baseline (median, 4.79% vs. 5.54% of CD4+ T cells, for chronic vs. resolved, respectively). The median frequency of Treg cells was statistically lower in healthy control subjects, compared with that in both acute HCV groups (P < .005). The frequency of CD4+CD25high Treg cells did not vary over time when assessed either by grouping into early (enrollment), medium (6 months), or later (12 month) time points (figure 2B) or when patients were examined individually over time (figure 3). There was no correlation between Treg cell frequency and viral level (figure 4).
CD4+CD25high Treg cells have been characterized by the absence of the early activation marker CD69 and the constitutive expression of CD62L [1, 14], intracellular expression of CTLA-4 , FoxP3 (the most specific marker for Treg cells) [16, 17], and LAG-3 . As shown in table 2, there was no significant difference in the expression level of these markers at baseline (n = 27) or over time in the 8 patients with chronic infection and the 9 patients whose infection resolved who were studied longitudinally. CTLA-4 and CD62L were relatively down-regulated in the patients with acute HCV infection, compared with healthy control subjects.
FoxP3 protein and mRNA expression. FoxP3 has been shown to be the transcription factor responsible for the development and suppressive function of Treg cells . We used FACS analysis for detection of intracellular FoxP3 protein and a quantitative TaqMan PCR assay (see appendix A) to detect human FoxP3 mRNA. As expected [13, 19], FoxP3 protein and mRNA were selectively expressed in bead-purified CD4+CD25+ cells compared with CD4+CD25− responder cells (figures 5 and 6). As shown in table 3 and figure 6A and 6B, the vast majority of CD25high Treg cells stained intracellularly for FoxP3; the protein and mRNA data matched closely. Moreover, there was a statistically significant relationship between the levels of CD4+CD25high Treg cells and FoxP3+ cells (figure 7). Because it has been recently appreciated that FoxP3+CD25low/- CD4+ T cells can be as equally suppressive as FoxP3+CD25high Treg cells , thus indicating that FoxP3 is a more reliable marker of Treg cells than CD25, we compared its expression in the different CD25 subsets. As shown in table 3, FoxP3 expression was statistically higher (by percentage and median fluorescent intensity) in the CD25low subset in patients with acuteHCVinfection, compared with that in healthy control subjects. Although there was no statistical difference within the HCV groups (likely because of the relatively wide distribution of values within the acute-chronic group and the limited sample size), at 12 months, there was a trend toward higher FoxP3 expression in the chronic versus resolved group (P = .08). Moreover, the percentage of total CD4+ T cells that expressed FoxP3 was statistically higher in patients with HCV infection than in heathy control subjects (figure 6C).
Variable suppressive function of CD4+CD25high Treg cells over time according to virologic outcome. To examine the function of Treg cells in the acute HCV model, we cocultured highly purified autologous populations at various ratios and with different concentrations of anti-CD3 and anti-CD28 antibodies as described in Methods and elsewhere [13, 14]. Examples of this approach for a representative patient are shown with proliferation (figure 8A) and suppression (figure 8B) results. Figure 8C summarizes the functional suppressive data for the study cohort (at a 1:0.2 responder to Treg cell ratio). These results show that Treg cells derived from patients who ultimately develop chronic infection and from those whose infections resolve spontaneously display comparable levels of suppressive activity at the earliest time point after acuteHCVinfection, in accordance with the FoxP3 data shown in figure 6C. However, 6 months later, Treg cells from patients with resolved infection, although similar in frequency, demonstrate significantly lower inhibition of responder CD4+CD25− T cell proliferation than do those from patients with chronic infection. As shown in figure 8C, in subjects whose HCV infection resolved, the relative loss of functional suppression over time (0 vs. 6 months) was statistically significant (P = .016). At the higher concentration of Treg cells (1:2 ratio), similar patterns were noted—on average, 96% versus 88% suppression for the chronic versus the resolved group at 6 months (P = .026); the percentage of suppression for the healthy control subjects at this ratio was 64% (P < .0001). Although there was not a statistically significant difference in Treg cell frequency between the 2 patient groups, these data indicate that acute resolving infection is associated with a temporal decrease in Treg cell suppressive function, whereas acute infection that becomes persistent is associated with long-term maintenance of Treg cells that regulate effector immune responses. This pattern over time in the 2 acute HCV groups is also presented as a function of the estimated time after infection (figure 8D); notably, 4 of the 8 patients who developed chronic infection demonstrated increasing suppression of at least 20%, whereas none of the patients whose infections resolved did so.
To further define whether subtle changes in Treg cell function were not discerned in our suppression model because the effector (responder) population was also potentially changing over time, co-mixing experiments using effector cells from a constant time point (baseline) and Treg cells from different time points were performed for a subset of patients. As shown in figure 8E, the suppressive activity of 6-monthTreg cells on responder cells was greater than that of the contemporaneous (baseline) Treg cells in patients who developed chronic infections (e.g., TN104 and TN108) but not those with recovery where there was a relative loss of regulatory function over time (HS112). Taken together, these results suggest that, although preexisting differences in Treg cell function do not determine virologic outcome (i.e., persistence versus recovery), the development of chronicity may lead to expansion of Treg cells that regulate effector immune responses.
Cytokine production and virologic outcome. The precise suppressive mechanisms of Treg cells are not clear, but there is evidence for both cell-cell contact and cytokine-mediated inhibition  (for more information, see appendix A). We found no qualitative or quantitative differences in the secretion of these cytokines in acute chronic versus acute resolving infection (figure 9).
The most striking clinical feature of infection with HCV is its high propensity to develop persistent viremia. Infections that follow a chronic course are usually marked by low frequencies of viralspecific CD4+ and CD8+ T cells , suggesting that the appropriate immune response required for spontaneous recovery either fails to develop or is suppressed.Alogical hypothesis is that the ability of Treg cells to inhibit the responder cell population (including HCVspecific T cells) correlates with the risk of developing persistence, and that, in contrast, patients who show relatively abrogated suppressive Treg cell function will recover from acute HCV infection. Our results provide the first ex vivo description of the population dynamics and functions of Treg cells during the early stages of HCV infection and provide new insights into their potential causative role in establishing chronicity. Indeed, to our knowledge, this is the first study to longitudinally characterizeTreg cells in the acute setting of any human infection.
We found that the proportional frequency (as a percentage of total PBMCs and CD4+ T cells) and phenotype of FoxP3+ Treg cells did not differ during the earliest stages of infection between those who subsequently developed chronic infections versus those whose infections spontaneously resolved. Our data thus provide insight into the long-standing observation that HCV-specific CTLs are poorly functional during acute HCV infection irrespective of the final outcome of disease . Both acute HCV groups had significantly higher frequencies of circulating FoxP3< Treg cells than did healthy control subjects, suggesting that expansion of Treg cells occurs during the earliest phases of acuteHCVinfection. Our data are in accordance with the findings of a recent preliminary study in chimpanzees providing indirect evidence that Treg cells can be induced in vivo duringHCVinfection ; FoxP3mRNAlevels were increased in the livers of 6 chimpanzees that were studied during acute HCV infection and were not predictive of subsequent virologic outcome (i.e., HCV clearance or persistence; B. Rehermann [NIDDK, NIH], personal communication).
With regard to the phenotype of Treg cells, CTLA-4 and CD62L were both relatively down-regulated in Treg cells derived patients with acute HCV infection, compared with that in healthy subjects. CTLA-4 expression is tightly controlled; in a resting cell, the majority of CTLA-4 molecules are sequestered in a subcellular Golgi compartment . However, with T cell receptor triggering, CTLA-4 molecules move out of this compartment onto the T cell surface. CD62L (L-selectin) is also downregulated upon T cell receptor activation. Thus, lower intracellular CTLA-4 and CD62L may be a reflection of T cell activation resulting from antigenic stimulation during acute infection.
Treg cells are heterogeneous and include subsets that develop naturally in the thymus (“natural” or “central”) or are acquired peripherally (“inducible” or “adaptive”) ; these may have complementary and overlapping functions in the control of immune responses . Markers distinguishing these Treg cell subpopulations currently do not exist, and both natural and adaptive Treg cells [4, 24] express FoxP3. Treg cells can be induced by polyclonal or antigen-specific stimulation, particularly in the absence of CD28 costimulation . In this regard, HCV infection has been associated with down-regulation of CD80 and CD86 on antigen-presenting cells such as dendritic cells . Congruent with the hypothesis that repeated stimulation in the context of defective CD28 costimulation could promote the generation of CD4+CD25high Treg cells in the periphery, we found enhanced Treg cell suppressive activity at 6 months in patients who developed chronic infection even though the Treg cell frequency was comparable in the 2 acute HCV groups. There is a precedent for functional divergence between CD25high Treg cell frequency and suppression: Viglietta et al.  found the same frequency of circulating Treg cells in healthy control subjects and patients with multiple sclerosis, but Treg cells in the latter group had decreased suppressive function.
Our results are in keeping with and extend considerably the findings of a recent study by Boettler et al.  that demonstrate greater suppressive activity of Treg cells derived from patients with long-term chronic infection . Importantly, prior studies did not detect suppression by CD4+CD25+ Treg cells in most patients who had recovered fromHCVinfection [9, 10]; however, assessment of HCV-specific T cell responses after depletion of Treg cells does not provide information about the relative suppressive function because the effector population is significantly different in resolved and chronicHCVinfection . One of the strengths of our current analysis is that each individual could serve as their own control, allowing, for the first time, temporal comparison of Treg cell suppressive activity. As shown in figure 8E, the use of effectors from a constant time point cocultured with Treg cells from various time points revealed that Treg cell function changed over time. An illustrative example is patient HS112, whose CD25high Treg cells derived at the first time point (when serum HCV RNA was detectable) showed greater suppressive activity against autologous effectors than did Treg cells derived 6 months later (when serum HCV RNA was no longer detectable). In contrast, Treg cells derived from TN104 and TN108 conferred greater suppressive function at later time points than at baseline. Once chronic infection is established, Treg cells may actively inhibit the eradication of HCV and simultaneously control excessive collateral damage and immunopathology [2, 21]. Whether expansion of Treg cells is dependent on specific viral gene products is not known, but HCV core protein has been shown to mediate the suppression of proliferation by inhibition of IL-2 and interferon (IFN)-γ production  and suppress the in vivo antiviral CTL response after infection with murine vaccinia virus . In this regard, a recent report indicates that HCV proteins can induce FoxP3+ Treg cells , and thus, our findings may be explained by the induction of HCV-specific Treg cells.
Taken together, these results indicate that the development of persistence or recovery after acute HCV infection does not reflect preexisting developmental, phenotypic, or functional differences in FoxP3+ Treg cells but instead that chronic HCV infection may lead to maintenance of Treg cells exerting suppressive activity that may contribute to the development of persistence in some patients. In contrast, spontaneous recovery is associated with temporal loss of suppressive Treg cell function.
We thank the patients for their time and willingness to participate in this study as well as Nicole Castelblanco for technical assistance.