Tim-3 negatively regulates cytotoxicity in exhausted CD8+ T cells in HIV infection

Abstract

Cytotoxic CD8(+) T cells (CTLs) contain virus infections through the release of granules containing both perforin and granzymes. T cell 'exhaustion' is a hallmark of chronic persistent viral infections including HIV. The inhibitory regulatory molecule, T cell Immunoglobulin and Mucin domain containing 3 (Tim-3) is induced on HIV-specific T cells in chronic progressive infection. These Tim-3 expressing T cells are dysfunctional in terms of their capacities to proliferate or to produce cytokines. In this study, we evaluated the effect of Tim-3 expression on the cytotoxic capabilities of CD8(+) T cells in the context of HIV infection. We investigated the cytotoxic capacity of Tim-3 expressing T cells by examining 1) the ability of Tim-3(+) CD8(+) T cells to make perforin and 2) the direct ability of Tim-3(+) CD8(+) T cells to kill autologous HIV infected CD4(+) target cells. Surprisingly, Tim-3(+) CD8(+) T cells maintain higher levels of perforin, which was mainly in a granule-associated (stored) conformation, as well as express high levels of T-bet. However, these cells were also defective in their ability to degranulate. Blocking the Tim-3 signalling pathway enhanced the cytotoxic capabilities of HIV specific CD8(+) T cells from chronic progressors by increasing; a) their degranulation capacity, b) their ability to release perforin, c) their ability to target activated granzyme B to HIV antigen expressing CD4(+) T cells and d) their ability to suppress HIV infection of CD4(+) T cells. In this latter effect, blocking the Tim-3 pathway enhances the cytotoxcity of CD8(+) T cells from chronic progressors to the level very close to that of T cells from viral controllers. Thus, the Tim-3 receptor, in addition to acting as a terminator for cytokine producing and proliferative functions of CTLs, can also down-regulate the CD8(+) T cell cytotoxic function through inhibition of degranulation and perforin and granzyme secretion.

Figure 1. Total and antigen specific Tim-3⁺ CD8⁺ T cells have higher frequencies of perforin expression.

Ex vivo PBMC from untreated chronically HIV infected subjects were stained for perforin (clone B-D48) and Tim-3. In a), shown are FMO staining for Tim-3, and B-D48 perforin antibodies of a representative sample, in b), representative gating strategy for Tim-3, and B-D48 perforin and a representative experiment showing the percentage of perforin and Tim-3 expressions on total CD8⁺ CD3⁺ T cells. Summary of data for 22 individuals showing perforin content of Tim-3⁺ and Tim-3⁻ CD8⁺ T cells as a % in c) and MFI in d). In e) a representative experiment where ex vivo CD8⁺ T cells are stained with a pool of HIV specific tetramers (see Methods) and perforin expression on gated Tim-3⁺ and Tim-3⁻ T cells are shown, with a summary shown in f) for 5 chronically infected individuals. Each symbol represents a single individual. Red histogram represents FMO. In g) Ex vivo PBMC from 7 chronic progressors were stained for HIV-specific tetramers and examined for Tim-3 and T-bet expression. The individual in e) was not examined in g) (p value based on two tailed paired t test).

Figure 2. Perforin content of Tim-3⁺ CD8⁺ T cells as determined by two antibody clones.

In a), a representative experiment showing Tim-3 expression on ex vivo CD8⁺ T cells from a treatment naïve chronically HIV infected subject showing perforin expression by two antibodies detecting different conformations of perforin on Tim-3⁺ or Tim-3⁻CD8⁺ T cells. The δG9 clone has been proposed to predominantly detect stored (granule associated) perforin and clone B-D48 detects stored perforin plus majority of other perforin conformations , . In b) a representative figure showing the relationship between two different conformations of perforin and Tim-3 expressions after gating out the naïve and terminally differentiated CD8⁺ T cells by gating in all CD45RA^- memory subsets. In c), summary of data for all 9 chronically HIV infected individuals stratifying perforin antibodies and Tim-3 expression on total CD8⁺ T cells. In d) summary of data for 5 chronically HIV infected individuals showing the perforin expression with two clones and Tim-3 expressions of memory subsets of CD8⁺ T cells (p value based on two tailed paired t test).

Figure 3. Tim-3⁺ CD8⁺ T cells cannot degranulate as effectively as Tim-3⁻ CD8⁺ T cells.

Ex vivo PBMC from HIV chronically infected subjects were stimulated for 6 hours with a pool of HIV Gag peptides or SEB or DMSO and stained for CD107a and Tim-3. In a), a representative experiment showing higher CD107a expression in Tim-3 negative component of CD8⁺ T cells after stimulation with either pool of HIV antigens or SEB. In b), summary of all data for 10 chronically HIV infected individuals showing higher ability for degranulation in Tim-3 negative subpopulation of CD8⁺ T cells after stimulation with Gag peptides. In c) summary of all data for 8 chronically HIV infected individuals showing higher ability for degranulation in Tim-3 negative subpopulation of CD8⁺ T cells after stimulation with SEB (p value based on wilcoxon signed rank test).

Figure 4. Tim-3 pathway blocking improves the perforin release in antigen specific CD8⁺ T cells only in chronically HIV infected individuals.

2×10⁵ negatively sorted CD8⁺ T cells from ex vivo PBMC from 5 untreated chronically infected (Fig a) and four untreated viral controllers (Fig b) were stimulated with a pool of HIV Gag peptides (final concentration of 2 µg/ml/peptide) for 6 h. Perforin released in supernatant was measured in an ELISA experiment in pg/ml. (αTim-3  =  2E2 Tim-3 blocking antibody, Cntr  =  IgG1 isotype control antibody) (p value based on two tailed paired t test).

Figure 5. Tim-3 pathway blocking increases the antigen specific CD8⁺ T cells cytotoxicity and HIV suppression in chronic HIV infection.

The effect of HIV Gag-specific CD8⁺ T cells after blocking the Tim-3 pathway with a Tim-3 blocking antibody (clone 2E2) on eliminating HIV-infected CD4⁺ T cells was tested in a virus suppression assay. Autologous CD4⁺ T cells (targets) are infected with a primary HIV virus isolate. Autologous CD8⁺ T cells (effectors) are added in 1∶1 ratio at the time of infection with 2E2 antagonistic anti-Tim-3 antibody or isotype at 10 µg/ml. The co-culture is incubated at 37°C for three days. The final readout is the percentage of HIV (p24⁺) positive target cells on day three examined by intracellular flow cytometry. In a) are autologous CD4⁺ T cells in the absence of autologous CD8⁺ T cells taken from an HIV infected individual after exogenous infection by HIV in the presence of Tim-3 antibody or isotype. Tim-3 blockade had no effect on total and infected CD4⁺ numbers. In b) shown are CD4⁺ T cells from an HIV uninfected normal volunteer infected with exogenous HIV and co-cultured with autologous CD8⁺ T cells (1∶1 ratio). Again we could not appreciate any difference in survival of CD4⁺ T cells. Tim-3 blockade had no effect on CD8⁺ mediated suppression of HIV Infection. In c), is a representative experiment showing the percentage of infected p24⁺ CD4⁺ T cells in the two different conditions in a chronic HIV infected individual and in d) is a representative experiment showing the percentage of infected p24⁺ CD4⁺ T cells in the two different conditions in a viral controller, Shown in e), are summary data for four chronically HIV infected individuals and in f) are summary data for three viral controllers. Each solid circle represents the average of three independent experiments from each individual showing the percentage of p24⁺ CD4⁺ T cells (p value based on two tailed paired t test).

Figure 6. Blocking the Tim-3 signalling pathway enhances the per cell cytotoxicity of CD8⁺ T cells.

CD4⁺ T cells from chronically HIV infected individuals were transfected with HIV gag mRNA. These target cells were labeled with TFL-4 after transfection and then co-incubated with autologous CD8⁺ T cells. Granzyme B substrate is added to cell culture. Cytotoxicity of target cells is determined by the presence of activated GrnB substrate found within the target cells. Experiments are performed either in presence of 10 µg/ml 2E2 Tim-3 blocking or isotype control antibodies. In a) is one representative experiment showing the percentage of killed target cells (TFL-4⁺ and cleaved GrnB substrate⁺) in the presence of 2E2 blocking antibody and isotype control antibody respectively in 1∶1 ratio of effectors to target cells. Negative controls were shown in top two plots for either CD8⁺ or CD4⁺ T cells. In b), are summary of results from 11 different subjects, each solid dot represents the average of two independent experiments from each individual. In c), shown are summary data for five different individuals in three different ratios of effector to target cells (3∶1,1∶1,1∶3) (p value based on two tailed paired t test).

Figure 7. Tim-3 pathway blocking increases the ability of CD8⁺ T cells to degranulate upon further in vitro re-stimulation.

Ex vivo PBMC from HIV infected subjects were stimulated for 6 hours with a pool of HIV Gag peptides in the presence of sTim-3 (final concentration of 2 µg/ml) or just medium alone, and CD107a expression was measured. In a), a representative experiment showing CD107a expression with or without sTim-3 added to the cell culture from the same individual, In b), are summary of data for all 11 subjects (p value based on two tailed paired t test).