Report

HCV Persistence and Immune Evasion in the Absence of Memory T Cell Help

See allHide authors and affiliations

Science  24 Oct 2003:
Vol. 302, Issue 5645, pp. 659-662
DOI: 10.1126/science.1088774

Abstract

Spontaneous resolution of hepatitis C virus (HCV) infection in humans usually affords long-term immunity to persistent viremia and associated liver diseases. Here, we report that memory CD4+ Tcells are essential for this protection. Antibody-mediated depletion of CD4+ Tcells before reinfection of two immune chimpanzees resulted in persistent, low-level viremia despite functional intra-hepatic memory CD8+ Tcell responses. Incomplete control of HCV replication by memory CD8+ Tcells in the absence of adequate CD4+ Tcell help was associated with emergence of viral escape mutations in class I major histocompatibility complex–restricted epitopes and failure to resolve HCV infection.

Acute hepatitis C virus (HCV) infection resolves spontaneously in about 30% of humans, with the majority becoming chronic carriers who are at risk for developing progressive liver disease (1). Resolution of primary HCV infection is temporally associated with a diverse and enduring HCV-specific CD8+ cytotoxic T lymphocyte response (25) that may contribute to rapid resolution of subsequent infections in chimpanzees (610) and possibly humans (11). Consistent with this view, we recently reported that HCV replication is prolonged in the absence of intrahepatic memory CD8+ T cells (12). Resolution of infection is also characterized by a robust CD4+ T helper cell response (35, 1316) that may be essential for effective antiviral CD8+ T cell immunity.

We investigated whether an inadequate host CD4+ T lymphocyte response is responsible for the failure to resolve infection in chimpanzees, the only animal model of HCV persistence. Two chimpanzees, 4X287 and CBO627, were infected with the genotype 1a virus HCV-1/910 (17). Chimpanzee 4X287 was HCV naïve when first infected in 2001. HCV RNA in plasma peaked on day 21 [1 × 105 genome equivalents (GE) per milliliter of plasma (GE/ml)] and was undetectable after day 56, with the exception of a transient rebound on day 126 (67 GE/ml) (Fig. 1A). Chimpanzee CBO627 was first infected with HCV-1/910 in 1994 and then rechallenged with the virus in 2001. Both infections resolved spontaneously, but peak viremia was 100-fold lower in the second compared with the first infection and terminated faster (42 compared with 98 days) (18) (Fig. 1B). This indicated a potentially important role for a memory immune response in controlling viremia. Indeed, 6 months after resolution of the primary (4X287) and secondary (CBO627) infections, high frequencies of HCV-reactive T cells targeting all HCV proteins (Core-NS5) were detected in blood (Fig. 1, A and B, 2001). Neither animal had a detectable HCV envelope glycoprotein–specific antibody response (18). Results from these animals were thus similar to those reported in other acute resolving HCV infections of humans (3, 4) and chimpanzees (2, 12, 16), with resolution correlating specifically with a strong sustained T cell response.

Fig. 1.

HCV replication and immune responses [blood in (A) and (B) and liver in (C)] after infection with HCV-1/910. Course of viremia during (A) primary (2001) infection and secondary (2002) reinfection after CD4 depletion of 4X287 and (B) secondary (2001) and tertiary (2002) infection after CD4 depletion of CBO627. The arrows indicate the day of infection. HCV-specific T cell responses in blood were measured in an IFN-γ elispot assay (U-Cytech, Utrecht, Netherlands) with the use of nine pools of overlapping peptides representing HCV structural and nonstructural proteins as previously described (12, 22). Reactivities against Core-NS2 (black), NS3 (red), NS4 (green), and NS5 (yellow) are shown. The vertical dashed lines represent the three doses of cM-T412 mAb (22). (C) Intrahepatic HCV-specific CD8+ T cell response after depletion of CD4+ T cells from chimpanzee CBO627. CD8+ T cells were expanded from liver and tested for reactivity to the HCV peptide pools with the use of an IFN-γ elispot assay (12, 22). The asterisks indicate time points for which liver biopsy samples for propagation of intrahepatic T cells were unavailable. (+) indicate time points for which samples were available but tested negative.

The link between an effective CD4+ T cell response and CD8+ T cell–mediated control of virus replication has been documented in several murine models of infection (19). However, these observations have not yet been extended to virus infections of humans. HCV infection of chimpanzees provided a highly relevant animal model to address this question. We therefore temporarily depleted 4X287 and CBO627 of CD4+ T cells immediately before rechallenging with HCV. A mouse-human chimeric monoclonal antibody (mAb), designated cM-T412, that recognizes the human and chimpanzee CD4 proteins [Supporting Online Material (SOM)] Text was administered intravenously 14, 10, and 7 days before reinfection with the same HCV dose and strain used in prior infections (2022). A 99% reduction in the absolute number of circulating CD4+CD3+ T cells was observed in both animals for up to 300 days (Fig. 2, A and B). The number of CD8+ T cells was relatively stable after cM-T412 treatment (Fig. 2, A and B). Importantly, intrahepatic CD4+CD3+ T cells were also efficiently depleted, because they were not recovered from the liver of either animal at any time after cM-T412 treatment (22).

Fig. 2.

Flow cytometric analysis of CD4+ T cell depletion after cM-T412 treatment of (A) 4X287 and (B) CBO627. Absolute number of CD4+CD3+ T cells (squares) and CD8+CD3+ T cells (triangles) per mm3 of blood are plotted.

As noted above, without cM-T412 pretreatment HCV infections resolved within 42 (CBO627) or 168 (4X287) days of virus challenge (Fig. 1, A and B, 2001). In contrast, reinfection of the animals after CD4+ T cell depletion resulted in persistence of HCV RNA for more than 300 days (Fig. 1, A and B, 2002) but at levels that were 5- to 10-fold lower than those observed after primary infection of 4X287 (Fig. 1A, 2001) and CBO627 (18). The magnitude and breadth of HCV-specific T cell responses in blood were also substantially reduced after cM-T412 treatment (Fig. 1, A and B). The low-level, oscillating pattern of viremia nonetheless suggested that immune control was partially effective in the absence of CD4+ T cells but inadequate for resolution of infection (Fig. 1, A and B, 2002).

We therefore postulated that memory CD8+ T cells were still present in the liver and exerted antiviral activity even in the absence of CD4+ T cells. A strong intrahepatic HCV-specific CD8+ T cell response was indeed detected in the liver of 4X287 (18) and CBO627 (Fig. 1C) after cM-T412 treatment and reinfection. These observations are consistent with very recent findings in murine models (2325). CD4+ T cells were required to initially prime CD8+ T cells in mice, but their absence during repriming still resulted in activation of memory CD8+ T cell responses (2325). Our findings in this nonhuman primate model nevertheless indicate that the capacity of CD8+ T cells alone to terminate infection is limited, at least for RNA viruses like HCV that have a propensity to persist.

In chimpanzees (26) and possibly humans (27), persistence of HCV is associated with mutations within multiple class I major histocompatibility complex (MHC)–restricted epitopes. We therefore postulated that incomplete CD8+ T cell–mediated control of virus replication after antibody-mediated depletion of CD4+ T cells was because of accumulation of escape mutations in the class I MHC-restricted epitopes. Intrahepatic CD8+ T cells from CBO627 targeted epitopes in structural (Core-NS2) and nonstructural (NS3 and NS5) proteins (Fig. 1C and table S1), including two, designated Core-130 and NS5A-2251, presented by the Patr class I molecule B0401 (table S1) (22). Sequencing of the HCV genomes revealed only wild-type HCV-1/910 sequences in the Core-130 epitope through the first 84 days of infection (Fig. 3A). However, at day 113 postinfection the viral titer increased to 13,000 GE/ml, and two nonsynonymous nucleotide substitutions were observed in this epitope that resulted in position 5 (P5) M-to-V (28) and P7 Y-to-F (28) mutations (Fig. 3A). The dominant epitope sequence shifted again at the next time point, when a P5 M-to-L (28) mutation became fixed in the viral population. This change facilitated immune evasion, because Core-130–reactive CD8+ T cells expanded from the liver of CBO627 produced interferon-γ (IFN-γ) in response to the wild-type HCV-1/910 sequence used for infection but not the P5 M-to-L–substituted epitope (Fig. 3B). Analysis of the NS5A-2251 epitope also revealed wild-type sequence during the early time points after infection. A mutation [L to F (28) at P7] was identified on day 126 (Fig. 3C) but was not fixed until day 224 postinfection, when all 14 molecular clones displayed this change. Again, a CD8+ cytotoxic T lymphocyte line that recognized the wild-type NS5A-2251 epitope failed to produce IFN-γ in response to the mutated peptide (Fig. 3D).

Fig. 3.

Evolution of class I MHC-restricted epitopes in CBO627. (A) HCV RNA extracted from the plasma was amplified by reverse transcription polymerase chain reaction (RT-PCR), and multiple clones were sequenced through the Core-130 epitope (22). The amino acid sequences (28) at each time point were compared with HCV-1/910 consensus (top row). The number of independent clones displaying variant sequences is shown on the right side. Dashed lines indicate identity to the consensus sequence of HCV-1/910 virus. PI, postinfection. (B) Flow cytometric analysis of intracellular IFN-γ production by Core-130 reactive CD8+ T cells from the liver of CB0627. T cells were stimulated with control (Ova) [SIINFEKL (28)], Core-130, or Core130-M (mutated peptide containing M-to-L change) peptides and stained as described in (22). Numbers represent the percentage of T cells staining for CD8 and IFN-γ after peptide stimulation. (C) The region of HCV NS5A containing the NS5A-2251 epitope was amplified from the plasma of CBO627 with the use of nested RT-PCR primers. (D) IFN-γ intracellular cytokine staining of a NS5A-2251–specific CD8+ T cell line stimulated with control (Ova), NS5A-2251, or NS5A-2251-M (L to Fat p7) peptides.

CBO627 expressed MHC class I molecules for presentation of three NS5B epitopes defined in earlier studies of HCV-infected chimpanzees, including NS5B-2541 (Patr-A0401–restricted), NS5B-2509 (Patr-A0301–restricted), and NS5B-2661 (Patr-B2401–restricted) (table S1 and Fig. 4). Sequencing of a contiguous 585-base segment of NS5B that encoded the three closely spaced epitopes revealed that two of them (NS5B-2509 and NS5B-2661) contained amino acid substitutions by day 84 postinfection (Fig. 4). All three were mutated at the next time point (day 113). These epitopes were intact in viruses from animals that were persistently infected for at least 10 years but lacked the Patr class I molecules required for presentation (18). We observed at least two oscillating virus populations in CBO627 that contained partially overlapping sets of mutations in the epitopes and in flanking regions (Fig. 4). This is illustrated by the appearance of a new virus population on day 126, when a 20-fold decrease in viral titer (from 12,881 to 541 GE/ml) was observed. This new HCV population (Fig. 4, gray shading) contained a nonsynonymous mutation only within the NS5B-2541 epitope but also had a series of silent mutations in the adjoining region that distinguished it from viruses present at earlier time points. These mutations suggested that it had evolved independently, perhaps at a reduced replication rate. It is noteworthy that the dominant viral population during the first peak of viremia on day 113 was the only one detected on day 259, when virus titers peaked again. These data suggest that independent HCV populations with a distinct pattern of mutations in class I MHC-restricted epitopes exist contemporaneously and oscillate depending on viral fitness for replication and immune selection pressure, a model that has also been proposed for human immunodeficiency virus–1 infection (29).

Fig. 4.

Evolution of three class I MHC-restricted epitopes in the HCV NS5B protein (28). A 585-nucleotide region of HCV NS5B containing the NS5B-2509, NS5B-2541, and NS5B-2661 epitopes was amplified with the use of nested RT-PCR primers from the plasma of CBO627 (22). The boxed regions indicate CD8+ T cell epitopes. Missense mutations resulting in amino acid changes within the epitopes are shown in red; silent mutations, in blue. The red diamonds reflect missense mutations in the flanking regions, and the blue diamonds indicate silent mutations. The number of clones with the indicated sequence and the total number of clones examined are shown on the right. In some cases, minor populations with an additional silent mutation were identified (days 21, 70, 126, 140, and 224) but are not shown for clarity. P.I., postinfection.

Our results demonstrate that emergence of viruses with mutations in class I MHC-restricted epitopes is the direct result of inadequate virus-specific CD4+ T cell help that is common to persistent HCV infections. Our data favor a model where transient HCV-specific CD4+ T cell responses observed in some humans and chimpanzees during acute infection (3, 4, 1316) facilitate priming of CD8+ T cells. However, without sustained CD4+ T cell help, CD8+ T cells may not be able to keep pace with the evolution of viruses that rapidly accumulate escape mutations.

These observations may also have implications for priming durable vaccine-mediated protection against HCV. There is evidence that CD8+ T cell memory is more stable over time than CD4+ T cell help (30). Despite recent studies in mice showing good recall responses of memory CD8+ T cell without any requirement for help (2325), our studies suggest that these cells alone will not afford protection for highly mutable viruses like HCV.

Supporting Online Material

www.sciencemag.org/cgi/content/full/302/5645/659/DC1

Materials and Methods

SOM Text

Tables S1 and S2

References and Notes

References and Notes

View Abstract

Navigate This Article