Tim-3 expression on PD-1+ HCV-specific human CTLs is associated with viral persistence, and its blockade restores hepatocyte-directed in vitro cytotoxicity

Abstract

Having successfully developed mechanisms to evade immune clearance, hepatitis C virus (HCV) establishes persistent infection in approximately 75%-80% of patients. In these individuals, the function of HCV-specific CD8+ T cells is impaired by ligation of inhibitory receptors, the repertoire of which has expanded considerably in the past few years. We hypothesized that the coexpression of the negative regulatory receptors T cell immunoglobulin and mucin domain-containing molecule 3 (Tim-3) and programmed death 1 (PD-1) in HCV infection would identify patients at risk of developing viral persistence during and after acute HCV infection. The frequency of PD-1-Tim-3- HCV-specific CTLs greatly outnumbered PD-1+Tim-3+ CTLs in patients with acute resolving infection. Moreover, the population of PD-1+Tim-3+ T cells was enriched for within the central memory T cell subset and within the liver. Blockade of either PD-1 or Tim-3 enhanced in vitro proliferation of HCV-specific CTLs to a similar extent, whereas cytotoxicity against a hepatocyte cell line that expressed cognate HCV epitopes was increased exclusively by Tim-3 blockade. These results indicate that the coexpression of these inhibitory molecules tracks with defective T cell responses and that anatomical differences might account for lack of immune control of persistent pathogens, which suggests their manipulation may represent a rational target for novel immunotherapeutic approaches.

Figure 1. Differential Tim-3 expression on total CD4⁺ and CD8⁺ T cells in acute, chronic, and resolved HCV infection.

T cell populations from acutely infected patients (n = 16) early (<3 months) and late (>6 months) after infection as well as long-term chronic (n = 27), remotely resolved (n = 23), and normal control (n = 10) subjects were identified by staining PBMCs with antibodies against CD3, CD4, and CD8. (A and B) Percent CD4⁺ and CD8⁺ T cells expressing Tim-3. Horizontal bars denote means. (A) CD4⁺ T cells from acute→chronic patients had significantly more Tim-3⁺ T cells at both early and late time points after infection compared with acute→resolved patients. (B) CD8⁺ T cells from acutely infected patients had significantly higher Tim-3 levels regardless of viral outcome, but Tim-3 levels returned to normal in remote resolved patients. There was no statistical difference between acute→chronic and acute→resolved patients at comparable phases of infection.

Figure 2. Tim-3 expression on HCV-specific CTLs is higher in acute→chronic patients.

HCV-specific T cell populations were identified by staining with antibodies against CD3, CD8, and a panel of HCV-specific MHC class I pentamers (HLA-A2₁₄₀₆, HLA-A2₂₅₉₄, and HLA-A1₁₄₃₆) along with HLA-A2–rescricted CMV (pp65) pentamer. (A) Representative plots of Tim-3 expression on HCV-specific T cells from 1 acute→chronic and 1 acute→resolved patient. (B and C) Percent pentamer-positive cells expressing Tim-3 (B) and MFI of Tim-3 expression (C) from acute patients (34 responses from 14 patients at 2 time points), long-term chronic patients (13 responses from 11 patients), and remote resolved patients (11 responses from 11 patients). Horizontal bars denote means. Antigen-specific T cells from acute→chronic patients expressed significantly more Tim-3 than did those of acute→resolved patients. A similar expression pattern was seen when comparing long-term chronic with remote resolved patients. Expression of Tim-3 was higher, by both percentage and MFI, on HCV-specific T cells than on CMV-specific T cells in both acute and long-term chronic patients. There was no significant difference in the number of CMV-specific T cells expressing Tim-3 according to HCV outcome.

Figure 3. Dual expression of Tim-3 and PD-1 on virus-specific CTLs predicts acute development of persistence.

HCV-specific T cell populations were identified by staining with antibodies against CD3, CD8, and a panel of MHC class I pentamers (HLA-A2₁₄₀₆, HLA-A2₂₅₉₄, HLA-A2₁₀₇₃, and HLA-A1₁₄₃₆). (A) Representative plots showing dual expression of PD-1 and Tim-3 on total CD8⁺ T cells and HCV-specific T cells from an acute→chronic and an acute→resolved patient. (B) Percentage PD-1⁺Tim-3⁺ or PD-1^–Tim-3^– pentamer-positive cells from early and late acute→chronic and acute→resolved, long-term chronic, and remote resolved patients. Antigen-specific T cells from acute→resolved and remote resolved patients had significantly more PD-1⁺Tim-3⁺ than PD-1^–Tim-3^– cells (*P = 0.039, **P = 0.0062, ***P = 0.0009). Conversely, the frequencies of PD-1⁺Tim-3⁺ and PD-1^–Tim-3^– CTLs were equivalent in acute→chronic patients. (C) Boolean graphs from 2 representative acutely infected patients showing the percentage of HCV- and CMV-specific T cells that expressed Tim-3 and PD-1 at early and late time points after infection. (D) Percent intrahepatic HCV-specific CTLs expressing Tim-3, PD-1, or both. The predominant phenotype of liver-resident CTLs was PD-1⁺Tim-3⁺, and the least frequent was PD-1^–Tim-3⁺. Horizontal bars in B and D denote means.

Figure 4. PD-1^hiTim-3^hi CTLs are associated with diminished antiviral cytokine production and chronic infection.

(A) PD-1^hiTim-3^lo, PD-1^loTim-3^hi, PD-1^loTim-3^lo, and PD-1^hiTim-3^hi pentamer-positive T cells from acute→chronic, acute→resolved, long-term chronic, and remote resolved patients as well as intrahepatic T cells from long-term chronic (liver) patients. The differences between long-term chronic and remote resolved patients were highly statistically significant (see Results). (B and C) PBMCs from acute→chronic patients were stimulated with the HLA-A1–restricted NS3₁₄₃₆ epitope or PMA and ionomycin for 5 hours in the presence of the CD107a antibody. Cell surface staining for CD3, CD8, HLA-A1₁₄₃₆ pentamer, Tim-3, and PD-1 was carried out, followed by intracellular cytokine staining for IFN-γ and TNF-α. Flow cytometric analysis was used to determine the proportions of the PD-1^hiTim-3^hi, PD-1^loTim-3^lo, and PD-1^–Tim-3^– cells producing IFN-γ, TNF-α, and CD107a after stimulation. (B) Histograms from a representative patient with high pentamer frequency (2.2%). (C) Percent cells producing IFN-γ, TNF-α, and CD107a according to PD-1 and Tim-3 expression following stimulation with the NS3₁₄₃₆ peptide or PMA and ionomycin. Pentamer frequencies for patients acute 4 and acute 6 are shown in Supplemental Figure 3.

Figure 5. T cells coexpressing Tim-3 and PD-1 display a T_CM phenotype.

(A and B) CD4⁺ and CD8⁺ T cells from acutely infected patients (n = 9) were stained with antibodies against CD45RA, CCR7, and CD27 to determine differentiation phenotype (see Results). (A) Percent PD-1⁺Tim-3⁺ cells within each population. Cells expressing both PD-1 and Tim-3 were predominantly of the T_CM phenotype. (B) Percent (± SEM) PD-1^–Tim-3^–, PD-1⁺Tim-3^–, PD-1^–Tim-3⁺, and PD-1⁺Tim-3⁺ cells within each population. There were no significant differences in the phenotype of PD-1⁺Tim-3⁺ T cells from normal subjects and HCV patients (not shown). (C) CD4⁺ and CD8⁺ T cells were stained with antibodies against PD-1, Tim-3, CD69, HLA-DR, and CD45RO, and the PD-1⁺Tim-3⁺ and PD-1^–Tim-3^– phenotypes were compared. Horizontal bars denote means. A significant portion of PD-1⁺Tim-3⁺ T cells expressed all 3 activation/memory markers.

Figure 6. Blocking Tim-3 and PD-1 enhances T cell proliferation.

PBMCs from 5 acute→chronic patients (acute 2–acute 6) and 1 acute→resolved patient (acute 1; ref. 8) were stained with CFSE and stimulated for 7 days with HLA-A2–restricted NS3₁₄₀₆ or HLA-A1–restricted NS3₁₄₃₆ peptide, IL-2 (0.5 ng/ml), anti–Tim-3 antibody (10 μg/ml), anti–PD-L1/PD-L2 antibodies (10 μg/ml), or both anti–Tim-3 and anti–PD-L1/PD-L2. HCV-specific T cells were identified by staining with anti-CD8 and HLA-A2₁₄₀₆ or HLA-A1₁₄₃₆ pentamers and analyzed by flow cytometry for proliferation, as measured by CFSE^lo cells. (A) Representative histograms from 1 patient showing an increase in CFSE^lo T cells after dual blocking of Tim-3 and PD-1. Percent HCV-specific T cells that proliferated are shown. (B) Percent CFSE^lo cells from 6 patients following treatment with the indicated blocking antibodies. Blocking with either PD-1 or Tim-3 alone improved proliferation in 4 of the 6 patients, and dual blocking further enhanced proliferation.

Figure 7. Dual blocking of PD-1 and Tim-3 enhances CD107a expression by HCV-specific T cells.

PBMCs from patients acute 1, acute 4, acute 5, and acute 7 or intrahepatic lymphocytes from long-term chronic (liver) patients (IH 1–IH 3) were stimulated for 7 days with NS3₁₄₀₆ or NS3₁₄₃₆ peptide (10 μg/ml), IL-2 (0.5 ng/ml), and control IgG, anti–Tim-3, anti–PD-L1/PD-L2, or both anti–Tim-3 and anti–PD-L1/PD-L2 antibodies. Cells were restimulated with peptide, and CD107a antibody and brefeldin A were added to the cultures 5 hours prior to staining with anti-CD8 and HLA-A2₁₄₀₆ or HLA-A1₁₄₃₆ pentamers. CD107a expression on HCV-specific CTLs was determined by flow cytometric analysis. (A) Representative histograms from 1 acute patient showing an increase in CD107a⁺ CTLs after dual blocking of Tim-3 and PD-1. (B and C) Percent HCV-specific T cells expressing CD107a from (B) acute→chronic patients or (C) intrahepatic lymphocytes from long-term chronic (liver) patients. Blocking Tim-3 alone or Tim-3 and PD-1 increased CTL cytotoxicity in both PBMCs and intrahepatic lymphocytes. Dotted lines represent CD107a expression after stimulation with peptide alone.

Figure 8. Tim-3 blockade enhances antigen-specific CTL killing of hepatocytes.

(A) Experimental design to assess how Tim-3 and/or PD-1 blockade affects cytotoxicity of HCV-specific CTLs against a hepatocyte cell line expressing an NS3 epitope. Cytotoxicity of NS3₁₄₀₆-specific T cells against HepG2 cells transduced with a NS3_1406–1415 minigene was determined using a variation of the VITAL assay. (B) Intrahepatic lymphocytes were stimulated for 5 days with NS3₁₄₀₆ peptide and the indicated blocking antibodies, followed by CD8⁺ T cell bead isolation. HCV-specific T cells were then cocultured with HepG2 cells expressing the NS3₁₄₀₆ minigene at a 0.5:1 effector/target ratio. Percent specific lysis was calculated (see Methods). Values denote mean ± SEM of 6 patients; P values were determined using paired 2-tailed t test. (C) AST levels in the supernatants from B demonstrated an increase in cultures containing the Tim-3–blocking antibody. Values denote mean ± SEM of triplicate samples.

Figure 9. Gradient of function and exhaustion in HCV-specific T cells according to expression of Tim-3 and PD-1.

PD-1^–Tim-3^– T cells (i) demonstrate a highly polyfunctional profile, with production of antiviral cytokines, cytotoxicity, and phenotype consistent with T_EM. Higher frequencies of these cells in acute HCV infection correspond to spontaneous viral eradication. At the other end of the spectrum, PD-1⁺Tim-3⁺ T cells (ii) accumulate in early HCV infection prior to development of persistence, demonstrate impaired CD4 help (decreased IL-2 production), impaired ability to secrete antiviral cytokines, and a phenotypic profile consistent with T_CM. Initially, TNF-α production is impaired, even though CD8⁺ T cells maintain the ability to produce IFN-γ (ref. and Figure 4B). The liver contains high expression of the ligands for PD-1 and Tim-3 (PD-L1/PD-L2 and galectin-9, respectively). The intrahepatic T cells express the highest level of PD-1⁺Tim-3⁺ T cells, consistent with the most exhausted phenotype, but dual blockade (iii) can restore function in many of these cells, including improved CD107a expression and attendant cytotoxicity.