Preferential loss of IL-2–secreting CD4+ T helper cells in chronic HCV infection


  • Potential conflict of interest: Nothing to report.


Hepatitis C virus (HCV) becomes persistent in the majority of infected individuals. In doing so, the virus evades host adaptive immune responses, although the mechanisms responsible in this evasion are not clear. Several groups have demonstrated weak or absent HCV-specific CD4+ T cell responses during chronic HCV infection using proliferation assays and, more recently, class II tetramers. However, the functional status of HCV-specific CD4+ T cells in resolved and persistent infection is poorly understood. Using interferon γ (IFN-γ) and interleukin 2 (IL-2) enzyme-linked immunospot assays, we analyzed cytokine secretion patterns in chronically infected patients and compared them with those with resolved infection. In the spontaneous resolver group, strong IL-2 secretion in relation to IFN-γ secretion was observed. However, in the persistently infected group, a consistent and significant loss of IL-2–secreting cells, compared with IFN-γ–secreting cells, was identified. In vitro addition of IL-2 had a substantial effect in restoring CD4+ T cell activity. In conclusion, failure of IL-2 secretion, as opposed to physical deletion or complete functional unresponsiveness, appears to be an important determinant of the status of CD4+ T cell populations in chronic HCV infection. Loss of IL-2 secretory capacity may lead to disruption of IFN-γ and proliferative function in vivo—a status that characterizes the cellular immune response in both CD4+ and CD8+ compartments in chronic disease. (HEPATOLOGY 2005;41:1019–1028.)

Hepatitis C virus (HCV) is a major cause of morbidity and mortality worldwide. Persistent infection is readily established and is associated with the evolution of liver fibrosis, cirrhosis, and hepatocellular carcinoma.1, 2 The mechanisms behind viral persistence are poorly understood.3, 4 HCV is thought to evade host innate immune responses through, for example, interactions with interferon-signaling molecules.5–7 Adaptive immune responses do, however, play a significant role in viral control.8–13 A fraction of individuals (≈20%) appears to be able to contain the virus after initial exposure and develop a stable polymerase chain reaction (PCR) state. In these individuals, there is functional and genetic evidence that CD8+ and CD4+ T cells play a role in the prevention of persistence.10, 14, 15

Once persistent infection is established, T cell responses in blood appear to be weak in most studies, compared with those individuals who spontaneously resolve infection.16 In particular, loss of CD4+ T cell proliferative responses appears to be a hallmark of persistent infection.13, 17, 18 Failure to establish proliferative T cell responses during acute infection may contribute to the establishment of persistence,11 although in cross-sectional studies it is not clear whether such changes are cause or effect. Failure to detect T cells in proliferation assays may derive from deletion of such cells, functional anergy, or evolution of effector cell populations that lack proliferative capacity.17 In HIV infection, recent data suggest that, in the presence of high viral loads, HIV-specific CD4+ T cell populations do exist but lack proliferative capacity19; this status is associated with the production of interferon γ (IFN-γ) upon antigen stimulation, but little or no interleukin 2 (IL-2). Once antigen loads are lowered through drug treatment, cell proliferation and IL-2 secretion can both be restored.

In HCV infection, even in those individuals for whom chronic disease has existed for decades, treatment with antiviral regimes such as IFN-α/ribavirin combination therapy can lead to restoration of CD4+ T cell proliferative capacity, and to a lesser extent, IFN-γ secretion.20–22 Therefore, IFN-γ and IL-2 secretion by HCV-specific CD4+ T cell responses upon antigen stimulation was compared in individuals who had spontaneously controlled HCV and those with chronic HCV.


HCV, hepatitis C virus; IFN-γ, interferon γ; IL-2, interleukin 2; PCR, polymerase chain reaction; ELISpot, enzyme-linked immunospot; PBMC, peripheral blood mononuclear cell; SFC, spot-forming cell.

Patients and Methods


Informed consent in writing was obtained from each patient, and the study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval from the ethics committee at the John Radcliffe Hospital, Oxford. For the analysis of HCV-specific T cell responses, blood was obtained from 23 untreated patients (Table 1; CH1-CH23) with persistent HCV infection (PCR+), defined as the detection of HCV RNA via PCR (detection limit of 300 HCV RNA copies per milliliter of plasma) (v2.0 Amplicor assay; Roche Diagnostics, Basel, Switzerland). Blood was also obtained from 11 patients (see Table 1; R1-R11) with spontaneously resolved HCV infection (PCR−), defined as the absence of detectable HCV RNA by PCR in the presence of HCV antibodies (Abbott HCV EIA 3.0; Abbott Laboratories, Abbott Park, IL) on at least two consecutive occasions 6 months apart. Thirteen healthy individuals were also recruited as negative controls (healthy controls) and were tested via IFN-γ enzyme-linked immunospot (ELISpot) assay. In 11 of these individuals, IL-2 ELISpots were also performed. Biopsies on PCR+ patients were scored using the method of Ishak to assess fibrosis (range 0–6) and inflammation (range 0–18).23

Table 1. Baseline Characteristics of the Patients
SubjectAgeSexStaging (0-6)*Grading (0-18)*ALT (normal <45 U/L)Genotype
  • Abbreviations: Inflamm., inflammation; ALT, alanine aminotransferase; n.a., result not available; —, no biopsy performed.

  • *

    Staging and grading was clinically scored according to Ishak's scoring method.23


Separation of Lymphocytes.

Peripheral blood mononuclear cells (PBMCs) were obtained from whole blood via density gradient centrifugation over Lymphoprep (Nycomed, Oslo, Norway).

ELISpot Assay.

PBMCs were tested in both IFN-γ (MABTECH, Stockholm, Sweden) and IL-2 (BD Biosciences, Oxford, UK) ELISpot assays as per each manufacturer's instructions. According to the manufacturer's reference, reproducibility of the IL-2 assay appears similar to that of the IFN-γ ELISpot assay, and the sensitivity of this assay lends itself to ex vivo measurement of even very low frequencies of cytokine-producing cells (potentially down to 1/300,000).

Briefly, whole or CD8-depleted PBMCs (0.2 million/well) were plated in 96-well ELISpot plates precoated with anti–IFN-γ or anti–IL-2. Antigens used for analyses were HCV core–derived peptides (20 mers overlapping by 10) spanning aa 1-191 arranged into three pools of five peptides and one pool of three peptides (10 μg/mL final concentration for each peptide; pool 1, 2, 3, 4). The recombinant proteins NS3, NS3/4, NS4, and NS5 (Chiron, Emeryville, CA) were used at a final concentration of 1 μg/mL as previously described.22 Medium alone was used as a negative control, and phytohemagglutinin was used as a positive control for all assays. Tetanus toxoid (2 μg/mL; Evans Vaccines, Liverpool, UK) and cytomegalovirus lysate (4 μg/mL; Virusys, North Berwick, ME) were also included in all assays. The plates were developed after 18 to 20 hours using AEC Chromogen color reagent for IL-2 (BD Biosciences) and AP color reagent A+B for IFN-γ (BioRad, Hercules, CA) analyzed for spot-forming cells (SFCs) using an ELISpot plate reader (AID Reader System; Strassberg, Germany) (ELISpot 3.1 SR program). Assays with a high background (average 10 SFC/well in negative control wells) or no phytohemagglutinin response were excluded. The frequency of IFN-γ+ or IL-2+ T cells specific for each antigen was calculated by subtracting the average number of SFCs in negative control duplicate wells from the average number of SFCs in stimulated duplicate wells and expressed as HCV-specific IFN-γ or IL-2 SFCs/106 PBMCs.

A positive response required wells with SFCs/well greater than 3 SD above mean negative control response or if the probability of a spot appearing in the stimulated well was significantly different (P < .05) from the probability of a spot appearing in the negative control well, assuming a Poisson distribution (Excel BINOMDIST statistics program; Microsoft).22 The mean background in the negative control wells was 5 spots/200,000 PBMCs for IFN-γ (SD 1.8) and 3 spots/200,000 PBMCs for IL-2 (SD 2.1).

Modulation of Function Through Addition of Cytokines.

For cytokine modulation, human recombinant cytokine IL-2 (Proleukin, Chiron) was added at 100 U/mL at the beginning of the ELISpot assay, while IL-15 (Peprotech, Rocky Hill, NJ) was added at 100 ng/mL to the PBMCs with or without antigen. The doses of cytokines were selected based on published dose–response experiments in which SFC responses were observed to peak at 100 U/mL for IL-2 and 100 ng/mL for IL-15. These doses for IL-2 and IL-15 have been reported as being optimal for preventing T cell apoptosis.24

IFN-γ/IL-2 Secretion Assay.

PBMCs (10 × 106) were stimulated at 37°C with 10 μg/mL HCV peptides or 1 μg/mL proteins as used in the ELISpot assay for 6 hours and 16 hours, respectively. IFN-γ– or IL-2–secreting cells were isolated using a human IFN-γ or IL-2 secretion assay (Miltenyi Biotec, Bergisch Gladbach, Germany) as per the manufacturer's instructions.

Cells were further stained with CD14-PerCP, CD19-PerCP, and ViaProbe (to exclude monocytes, B cells and dead cells, all of which may bind nonspecifically to the phycoerythrin/allophycocyanin magnetic beads) (all BD Pharmingen, San Diego, CA) and CD4-FITC (Miltenyi Biotec) and incubated on ice for 20 minutes.

Anti-phycoerythrin and anti-allophycocyanin microbeads were used to enrich cells stained with phycoerythrin (IL-2) and allophycocyanin (IFN-γ) by two rounds on magnetic separation columns (MS+ columns, Mini MACS, Miltenyi Biotec).

To calculate the percentage of cytokine+/CD4+ cells, where cytokine-positive cells were not detectable pre-enrichment, we used the following formula: absolute number of cytokine-positive cells in the postenrichment sample/(total number of CD4+ T cells in the preenrichment sample) × 9. The factor 9 is due to the acquisition of 10% of the stained sample as the pre-enrichment sample with the remainder (90%) of the cells being the source of the post-enrichment sample. This formula has been previously adopted for the calculation of the percentage of cytokine-secreting or tetramer-positive cells following magnetic enrichment (Miltenyi Biotec).25, 26

CD8 and CD25 Depletion.

CD8+ and CD25+ cell depletions were performed using magnetic bead separation (Dynal, Oslo, Norway) as per the manufacturer's instructions. CD8+ T cell depletion was efficient at removing 99% of CD8+ T cells as determined via flow cytometry.

Because CD8 depletion resulted in relative CD4 enrichment, calculation of HCV-specific CD4 T cell frequency in CD8-depleted PBMCs included the CD4 enrichment factor (%CD4 in CD8-depleted PBMCs/%CD4 in total PBMCs) as determined via flow cytometry. For example, with twofold CD4 enrichment following CD8 depletion (e.g., 15% CD4 in whole PBMCs, 30% CD4 in CD8-depleted PBMCs), IFN-γ+ or IL-2+ SFCs in 200,000 CD8-depleted PBMCs/well were divided by 2 to compensate for the twofold CD4 enrichment in each well (compared with 200,000 whole PBMCs/well). Thus, for 50 IFN-γ+ or IL-2+ SFCs/200,000 CD8-depleted PBMCs/well, the result was calculated as (50 IFN-γ+ or IL-2+ SFCs/0.2 million PBMCs per well)/2 = 125 IFN-γ+ or IL-2+ SFCs/million PBMCs.

Flow Cytometry.

Flow cytometric analysis was performed on a FACSCalibur, and analysis was performed using CellQuest software (BD Biosciences).

Statistical Analysis.

Pooled data are presented as the mean. The Mann-Whitney and Fisher's exact statistical tests and the paired t test were performed using Prism V3 software (Graphpad, San Diego, CA). A P value of .05 or less was considered statistically significant.


Ex Vivo Cytokine Release in Spontaneously Resolved (PCR−) and Chronically Infected (PCR+) Individuals.

Firstly, PBMCs isolated from individuals with spontaneously resolved HCV infection (PCR−) were stimulated with HCV antigens, and their ability to secrete IFN-γ and IL-2 was analyzed using a cytokine secretion assay. Strong CD4+ T cell responses were detected in 8/8 individuals tested (Fig. 1A). An example of CD4+ T cells secreting cytokine in response to HCV peptides visible on flow cytometry after magnetic bead enrichment from a PCR+ subject is shown in Fig. 1B.

Figure 1.

Analysis of cytokine secretion by antigen-specific CD4+ T cells. (A) Analysis of function of HCV-specific CD4+ T cells shown from three representative patients with spontaneously resolved HCV infection (R5, R6, R9). PBMCs were stimulated either for 16 hours with a pool of recombinant NS3, NS4, and NS5 antigens (R9) or for 6 hours with a pool of the core peptide pools 1-4 (R5 + R6), respectively, and isolated with an IL-2 and/or IFN-γ cell enrichment and detection assay. The percentage in the upper right quadrants of each FACS plot represents cytokine-secreting CD4+ T cells among enriched cells calculated from input cell number (see Patients and Methods); left-hand FACS plots, combined IL-2/IFN-γ double-positive CD4+ T cells; middle FACS plots, IL-2–producing CD4+ T cells; right-hand FACS plots, IFN-γ–producing CD4+ T cells. (B) PBMCs of a chronic HCV patient were stimulated for 6 hours with core peptide pools 1-4. The responding cells were stained and isolated according to secretion of IFN-γ using the IFN-γ secretion assay–cell enrichment and detection kit. Upper plots, negative control; lower plots, frequency after stimulation with core peptide pools 1-4. (C) CD4+ T cell–mediated IFN-γ and IL-2 responses in ELISpot assays. The upper two panels show an IFN-γ ELISpot against NS4 (R3), core peptide pool 2 (R9), and NS3 (R6) in three representative HCV-resolved patients before and after CD8 depletion; the lower two panels show an IL-2 ELISpot against NS4 (R3) and core peptide pool 3 (R5 + R6) before and after CD8 depletion (purple spots, IFN-γ; red spots, IL-2). Example for correcting for the number of CD4 T cells after CD8 depletion: number of IL-2 SFCs/106 PBMCs in patient R3 tested in NS4 = 45 SFCs/200,000 in CD8-depleted PBMCs (corrected = 113 SFCs/106 PBMCs compared with 55/106 in the ex vivo sample). The formula for correcting for the number of CD4 T cells after CD8 depletion is described in Patients and Methods. (D) IFN-γ ELISpot assay and IFN-γ cytokine secretion assay correlation, shown for eight resolved individuals tested against the nonstructural proteins NS3-NS5. In this assay, both responses were tested after 18 hours of exposure to antigen. Similar correlations were also seen for analysis of responses against all epitopes for IFN-γ (r2 = 0.3; P = .038) and for IL-2 (r2 = .88; P < .0001). IFN-γ, interferon γ; IL-2, interleukin 2; Neg., negative; PBMC, peripheral blood mononuclear cells; SFC, spot-forming cells.

To screen larger patient numbers and multiple antigens simultaneously and avoid the requirement for positive selection using magnetic beads, IL-2 and IFN-γ ELISpot assays were used. The magnitude of the T cell response against HCV in patients CH1-CH23 and R1-R11 was measured using an ELISpot assay, as previously described22 and discussed in Patients and Methods, to assess responses to core, NS3, NS4, and NS5. Depletion of CD8+ cells from PBMCs confirmed that the majority of responses to the HCV antigens for both IFN-γ and IL-2 was independent of CD8+ T cells (Fig. 1C). Similarly, a correlation between CD4+ T cells secreting cytokine (as measured via flow cytometric assay) and the ELISpot could be seen (Fig. 1D). These assays together suggest the majority of the responses observed were due to CD4+ T cells.

Having validated the assays, we next compared clinical groups (PCR+ vs. PCR−). Overall, significantly larger responses could be measured using the IFN-γ ELISpot in the PCR− group compared with the PCR+ group. Interestingly, the response difference across all antigens, although significant, was only about 2.2-fold greater in the PCR− group than in the PCR+ group (mean 232 spots per million vs. mean 107 spots per million; P = .0137) (data not shown). A positive IFN-γ response was detected in all (11/11) PCR− individuals, but only in 13/23 PCR+ individuals (P = .0135; Fisher's exact test). These data are consistent with previous analyses indicating weak CD4+ T cell responses in PCR+ individuals using proliferative assays, but the magnitude of the differences is smaller than might be predicted from such analyses.

IL-2 ELISpot assays were also performed in parallel. In the PCR− group, responses across all antigens were similar in overall magnitude to IFN-γ; however, in the PCR+ group, only weak responses were obtained (mean 138 spots per million vs. mean 10 spots per million; P = .0125) (data not shown). The majority of individuals within the PCR+ group had no detectable IL-2 ELISpot response at all (4/23 PCR+ vs. 7/11 PCR−; P = .016; Fisher exact test). No responses using either IFN-γ or IL-2 ELISpot assays were detectable in the healthy negative control group (data not shown).

The differences between the groups were maintained when we analyzed the responses of different specificities (Fig. 2). PCR− individuals maintained a broad response with detectable responses against all antigens tested for both IFN-γ and IL-2 (P value not significant). IFN-γ responses against core peptides were common in the PCR+ group, but were significantly weaker using the IL-2 assay (P = .006; paired t test). The IFN-γ responses in the PCR+ group against NS3-5 were less common (4/23), and an IL-2 response was detectable in only one individual.

Figure 2.

Magnitude of HCV-specific T cell responses as determined via ELISpot assay for both IFN-γ and IL-2. PBMCs were stimulated using HCV core peptide pools 1-4 and HCV nonstructural proteins NS3, NS3/4, NS4, and NS5 as described in Patients and Methods. Overall responses of IFN-γ– and IL-2–producing PBMCs comparing 23 chronic HCV patients (PCR+) with 11 HCV resolvers (PCR−). Responses represent the mean of duplicate wells after subtraction of the background and responses of core peptides and NS proteins are displayed separately. SFC, spot-forming cells; PBMC, peripheral blood mononuclear cells; IFN-γ, interferon γ; IL-2, interleukin 2; PCR, polymerase chain reaction.

Comparing individual antigens (for those antigens for which a positive response was obtained), the ratio of IL-2 producing cells to IFN-γ producing cells in the PCR− group was between 1:1 and 1:2, and approximately 1:10 for the PCR+ group (mean 0.5 vs. mean 0.08; P = .035) (Fig. 3A). Overall, the ratio of total level of IL-2 production (summed across antigens) compared with total IFN-γ production was approximately 0.6 in resolvers and 0.06 in chronically infected persons (data not shown). The relative deficit in IL-2 production in PCR+ individuals was similar in responses to core peptides and NS3-5 antigens (P value not significant) (Fig. 3B). The IL-2/IFN-γ ratios in responses to both core peptides and NS3-5 proteins remained lower in PCR+ individuals than in PCR− negative individuals when tested separately (P = .03 and P = .04, respectively).

Figure 3.

Ratio of IL-2– to IFN-γ–producing T cells in ELISpot assay. (A) Ratio of IL-2 to IFN-γ production shown for each individual antigen (HCV core peptide pools 1, 2, 3, and 4 and nonstructural proteins NS3, NS3/4, NS4, and NS5) in both PCR+ and PCR− groups (only those in which at least one cytokine was positive were included). (B) Ratio of IL-2 to IFN-γ production shown for each individual antigen by separation of the core peptide pools 1-4 and the nonstructural proteins NS3-NS5. IL-2, interleukin 2; IFN-γ, interferon γ; PCR, polymerase chain reaction; ns, not significant.

To determine whether the weak IL-2 response seen in patients with chronic HCV infection was a characteristic of HCV-specific cells only, we analyzed responses to cytomegalovirus, a persistent virus associated with expanded pools of effector T cell memory, and tetanus toxoid, a nonpersistent protein antigen associated with resting central T cell memory. The overall magnitude of IFN-γ and IL-2 responses to cytomegalovirus did not differ between seronegative healthy controls, PCR+, and PCR− groups (Fig. 4A). A strong IFN-γ response compared with the IL-2 response was seen in each group, with a ratio of approximately 1 IL-2–secreting cell to 5 IFN-γ–secreting cells in each case (P = .005, P = .01, and P = .005 for healthy controls, PCR−, and PCR+ groups, respectively). The tetanus toxoid responses were generally weaker and more comparable to HCV, but no significant differences were observed between the groups (Fig. 4B). The ratio of IL-2– to IFN-γ–secreting cells for the tetanus toxoid–specific response was higher than for cytomegalovirus, between 1:1 and 1:2 (P = .02, P value not significant, P value not significant for healthy controls, PCR−, and PCR+ groups, respectively).

Figure 4.

Overall magnitude of IFN-γ and IL-2 responses to cytomegalovirus and tetanus toxoid as determined via ELISpot assay. (A) Cytomegalovirus responses for IFN-γ (upper left panel) and IL-2 (lower right panel) tested in the PCR+, PCR− and healthy control groups. Differences remained nonsignificant even after very high responses (>1,000 IFN-γ SFCs/106 PBMCs) were excluded. (B) Tetanus toxoid responses for IFN-γ (upper left panel) and IL-2 (lower right panel) tested in the PCR+, PCR−, and healthy control groups. IFN-γ, interferon γ; SFC, spot-forming cells; PBMC, peripheral blood mononuclear cells; ns, not significant; PCR, polymerase chain reaction; IL-2, interleukin 2.

Thus responses against HCV were weaker in the PCR+ group, but the most striking finding overall was a lack of IL-2 secretion rather than a lack of IFN-γ secretion, especially against the core peptide pools.

Restoration of CD4+ T Cell Responsiveness by IL-2.

IL-2 alone is able to stimulate IFN-γ secretion in T cells through activation of intracellular Jak/Stat pathways.27 Therefore, the lack of IL-2 secretion seen in PCR+ individuals might account for loss of overall functional capacity in HCV-specific CD4+ T cell populations. To determine whether it was possible to restore IFN-γ production by HCV-specific CD4+ T cells in vitro, the effect of IL-2 addition was studied. In addition, CD25+ regulatory CD4+ T cells have been shown to play important roles in the control of a variety of responses against pathogens in both human and animal models,28–32 and a role for such activity in control of HCV-specific CD8+ T cell responses has been proposed.33 CD25+ T cells may compete for IL-2, thus effectively lowering the concentration available for antigen-specific T cells.34 Hence, the effect of CD25+ cell depletion on IFN-γ secretion was also examined.

Whole PBMCs, CD25+ cell–depleted PBMCs, and PBMCs in the presence of additional recombinant IL-2 were tested for HCV responsiveness in IFN-γ ELISpot assays. Figure 5A demonstrates the restoration of immune responsiveness in vitro through the addition of IL-2. Interestingly, IL-2 not only restored responses in terms of magnitude, but also revealed new responses—thus increasing the breadth of the responses detected in the ELISpot assay (Fig. 5B). Overall, the effect was highly significant when analysis of HCV-specific responses was performed, but not significant for the control tetanus toxoid response. Responses to tetanus toxoid for each tested individual with and without IL-2 treatment are shown in Fig. 5C. The addition of IL-15 in vitro led to major increases in background IFN-γ release, thus making analysis of specific T cell populations impossible (data not shown). Depletion of CD25+ cells, which includes CD4+CD25+ T regulatory cells as well as activated T cells, had a less marked effect on the magnitude and breadth of HCV-specific responses (see Fig. 5A–B). There was no significant effect on control responses in these individuals (see Fig. 5C). The increase in responsiveness to HCV antigens was observed both against NS proteins and core peptides, although most clearly in the former (Fig. 5D). The effect of IL-2 could be titrated out, and the addition of IL-2 had little effect at these concentrations on background staining (Fig. 5E).

Figure 5.

Immunomodulatory effect of IL-2 on IFN-γ HCV-specific responses via in vitro stimulation. (A) HCV-specific IFN-γ ELISpot responses in a group of chronically infected HCV patients were tested using whole PBMCs, CD25+ cell–depleted PBMCs, and PBMCs with addition of exogenous IL-2. The dots shown represent the overall magnitude of IFN-γ–producing T cells in each tested individual. The total responses were derived by antigens used for the stimulation with core peptide pools 1, 2, 3, and 4 and the nonstructural proteins NS3, NS3/4, NS4, and NS5. The mean and SD for background in negative controls did not change significantly after IL-2 stimulation. (B) Breadth of HCV-specific T cell responses represented as the number of positive antigens in the ELISpot (no. of Ag's) shown for the same group and same antigens with whole PBMCs, CD25+ cell–depleted PBMCs, and PBMCs with addition of exogenous IL-2. (C) IFN-γ ELISpot for the same chronically infected HCV group tested for tetanus toxoid–specific responses with whole PBMCs, CD25+ cell–depleted PBMCs, and PBMCs with addition of exogenous IL-2. *Paired t test for comparison of 10 individuals. The unpaired t test comparing 18 versus 10 individuals showed similar results. (D) HCV-specific IFN-γ ELISpot responses separated into nonstructural proteins NS3-NS5 (upper left panel) and core peptide pools 1-4 (lower middle panel) before and after exogenous IL-2 stimulation. **Single-tailed paired t test. (E) Immunomodulatory effect of IL-2 on IFN-γ HCV core peptide pool 1-4 specific responses in one representative PCR+ individual, shown by a dose titration curve with four different doses of IL-2 (12.5 IU/mL, 25 IU/mL, 50 IU/mL, and 100 IU/mL). Continuous line: IFN-γ responses represent the mean magnitude of duplicate wells via stimulation with core peptide pools 1-4 at different IL-2 concentrations. Dashed line: IFN-γ responses of the negative controls at different IL-2 concentrations, showing no significant increase in the background. IFN-γ, interferon γ; SFC, spot-forming cells; PBMC, peripheral blood mononuclear cell; ns, not significant; IL-2, interleukin 2.


Resolved HCV infection is characterized by relatively strong T cell responses that are detectable for years after initial infection, even after loss of detectable antibody.16, 35 In particular, the hallmark of such individuals is the maintenance of strong T cell proliferative responses to a range of antigens. Here we demonstrate that such responses are characterized by strong IL-2 secretion among CD4+ T cell populations. Data using cytokine capture assays with magnetic bead enrichment confirmed that such populations contained large populations of CD4+ T cells secreting IL-2 and many IFN-γ/IL-2 double-positive cells. The ELISpot assays confirmed similar proportions of IL-2– and IFN-γ–secreting cells, although these assays cannot verify whether such cells were double-positive for the cytokines tested. By analogy with HIV, these are similar to CD4+ T cell populations found under conditions when the viral load is low.19

Chronic HCV infection results in a range of clinical severities. The majority of the individuals tested in this study were found to have low IL-2 secretion. The study was, however, skewed toward those with lower levels of fibrosis. This was because we were interested in testing patients who had not been previously treated, because IFN-α and ribavirin can influence T cell function and also cytokine secretion profiles.22, 36 Additionally, we did not assess the relationship between IL-2 secretion and viral load. A prospective study aiming to look at the IL-2 secretion in more severely affected patients over a range of viral loads is underway, as well as assessing the potential impact of therapy.

Compared with studies that largely use proliferation assays, a significant fraction of persistently infected individuals did maintain IFN-γ–secreting cell populations, but very few were found to secrete IL-2. This is consistent with the low proliferative capacity typically associated with HCV-specific populations during chronic infection, even after therapy.37–40 Low levels of IL-2 secretion have been shown in independent studies in humans and mice to be accompanied by loss of proliferative capacity both in vitro and in vivo.19, 41–43 We did not assess proliferative responses in detail in this group. However, a previous study, using identical antigens in a similarly recruited cohort showed strong and multispecific proliferative responses in 5/15 spontaneously resolved individuals, but only in 1/21 with untreated persistent infection. The concentration of protein used in our assays (1 μg/mL) is relatively low, which may potentially reduce the frequency of strong responders in proliferation assays. However, the response rates in ELISpot assays are high as shown here, and a clear difference was seen between the clinical groups exposed to similar antigens and concentrations.

Lack of IL-2 secretion is most readily explained by overstimulation of CD4+ T cells, which is analogous to the state of T cells in untreated HIV infection.19, 44, 45 Because T cell populations proliferate in response to antigen, they lose IL-2 secretory capacity as they acquire effector characteristics. In cases where stimulation is extreme and/or prolonged, loss of cytokine secretion may occur successively, before deletion occurs. In this case, we expect that reduction in viral load through combination therapy may contribute to the well-documented restoration of CD4+ T cell responsiveness through a return to an IL-2 secreting state.22, 46 This hypothesis will be tested in future longitudinal prospective studies.

Interestingly, we were able to markedly influence IFN-γ responsiveness of HCV-specific CD4+ T cell populations through in vitro IL-2 treatment. The depletion of CD25+ cells had a modest influence on IFN-γ responsiveness in these assays.47 However, we only performed simple CD25+ T cell depletion, which may remove activated T cells in addition to other populations that are not CD4+CD25+ T regulatory cells. Additionally, differences in the bead preparation used may also influence the level of CD25 expression on T cells that are depleted. However, despite these limitations, we used this depletion as a simple method to remove cells expressing high levels of the IL-2 receptor, which might then lead to “release” of IL-2 for use by antigen-specific cells.34

The effect of adding IL-2 in vitro was obvious both in terms of response breadth and magnitude. The effect was clear-cut on HCV-specific T cell populations only, although it is known that IL-2 can restore T cell responsiveness in other disease settings, such as HIV infection.48, 49 The mechanism is likely to be a direct action on T cells, although in such assays an indirect effect through other T or non–T cell populations cannot be excluded.

These findings contrast in some ways with the limited comparison previously made using class II tetramers.26 Here, a set of three tetramers containing NS3/4-derived peptides and a single class II molecule (DRB1*0401) were used to study small groups of PCR+ and PCR− individuals. While tetramer-positive cells were detectable in those with resolved infection, no responses were detectable in four individuals with persistent viremia. In the present study, responses to nonstructural proteins were rarely detected in PCR+ individuals; however, responses are still detectable overall, particularly to epitopes in core. It is possible that some CD4+ T cell responses are indeed lost, through escape and exhaustion, while others are maintained—perhaps those of lower avidity. The fine mapping and restriction of IFN-γ+/IL-2 responses in PCR+ individuals will allow further detailed comparisons using novel class II tetramer reagents. One important feature of class II tetramers is the capacity to examine the surface and intracellular phenotype of antigen-specific CD4+ T cells ex vivo. In previous studies, we have observed that HCV-specific memory populations are typically CD62L- and CCR7-high (i.e., they have features of “central” memory T cells).26 While examination of phenotype is theoretically possible using flow cytometric–based cytokine assays, even short-term stimulation can markedly modify the surface expression of CD62L and CCR7 (data not shown), making them unsuitable for this purpose. Thus, we have little current data on the status of the antigen-specific T cells in PCR+ patients. A recent study of HIV-specific CD4+ T cells using class II tetramers, which are typically IL-2-low/IFN-γ-high in untreated infection,19 revealed these to be CCR7-low and CD62L-low, with CCR7 expression correlating inversely with viral load.50

The responses identified here are defined as CD4+ T cell responses through a combination of depletion experiments and confirmation via flow cytometric analysis. CD8+ T cell responses have been extensively examined by our group and others.17, 35 Interestingly, we observed very few responses to overlapping core peptides, using a comprehensive screening technique, especially in PCR+ patients (0/20).37 In future studies of the function of HCV-specific CD8+ T cell populations, it will be important to assess IL-2 secretion as well as that of IFN-γ.

In practical terms, the analysis of CD4+ T cells through IL-2 ELISpot assays is simple and robust and provides important information regarding the functional activity of specific T cell populations. As shown in Fig. 1A–B, it is also possible to identify these cells using flow cytometric techniques, but these require magnetic bead enrichment to gain accurate staining ex vivo.25, 26 The sensitivity for IL-2 secretion using this method seemed to be slightly higher than that of the ELISpot assay, although the ELISpot technique is directly quantitative without enrichment and much more suited to clinical studies using a range of antigens. In vitro IL-2 treatment reveals de novo responses in some chronic patients. These responses may potentially be restored through conventional therapy or immunomodulation. Finally, the fact that such responses are intact and can respond to IL-2 in vitro indicates that cytokine treatment in vivo may be beneficial.


The authors thank all of the patients for their participation; Katie Jeffery, Annie Lorton, Jane Phillips, Jane Collier, and Eleanor Barnes for assistance with samples and patients; Joan Warde for secretarial support; and Rodney Phillips for continued support in the Peter Medawar Building.