Cutting Edge: Antitumor Immunity by Pathogen-Specific CD8 T Cells in the Absence of Cognate Antigen Recognition

Abstract

Cancer prognosis often correlates with the number of tumor-infiltrating CD8 T cells, but many of these cells recognize pathogens that commonly infect humans. The contribution of pathogen-specific "bystander" CD8 T cells to antitumor immunity remains largely unknown. Inflammatory cytokines are sufficient for memory CD8 T cell activation and gain of effector functions, indicating tumor-derived inflammation could facilitate pathogen-specific CD8 T cells to participate in tumor control. In this study, we show in contrast to tumor-specific CD8 T cells that pathogen-specific primary memory CD8 T cells inside tumor were not able to exert their effector functions and influence tumor progression. However, infection-induced memory CD8 T cells with defined history of repeated Ag encounters (i.e., quaternary memory) showed increased sensitivity to tumor-derived inflammation that resulted in activation, gain of effector functions, and better control of tumor growth. Thus, memory CD8 T cells with heightened ability to recognize environmental inflammatory stimuli can contribute to antitumor immunity in the absence of cognate Ag recognition.

Figure 1.. Pathogen-specific 1°M CD8 TILs do not show measurable Ag-independent effector functions in the tumor.

(A) Experimental design. Mice received adoptive transfer of 10⁴ naïve OT-I (Thy1.1⁺/1.1⁺) and 10⁴ naïve P14 (Thy1.1⁺/1. 2⁺) cells followed by i.v. injection of 1x10⁷ PFU Listeria monocytogenes expressing ovalbumin and glycoprotein 33 (LM-OVA/GP33). After 30 d mice received s.c. injection of 2x10⁴ B16-OVA in the hind flank. After 28 d mice received i.v. injection of fluorescently-conjugated CD45.2 mAb 3 min prior to harvest. (B) OT-I and P14 CD8 T cells were distinguished in tumor samples based on Thy disparity. Cells were divided into vasculature-excluded (CD45.2⁻/i.v.⁻) and vasculature-derived (CD45.2⁺/i.v.⁺) populations that were phenotypically distinct. Representative histograms of direct ex vivo staining for CD69, IFN-γ and Granzyme B (in the absence of Brefeldin A or monensin (15, 16) on OT-I cells were shown. (C) The frequency of OT-I and P14 CD8 T cells in i.v.⁺ and i.v.⁻ fractions expressing CD69 (D) IFN-γ and (E) Granzyme B. Data are representative of two independent experiments with 3–5 mice per group per experiment. **=p<0.01; *** = p<0.001 as determined by student t-test, fold change between groups is shown. Error bars are ± SEM.

Figure 2.. Pathogen-specific 1°M CD8 T cells do not influence cancer progression.

(A) Experimental Design. Mice received adoptive transfer of 10⁴ naïve P14 cells followed by i.p. injection of 2x10⁵ CFU LCMV Armstrong 10 or 40 days before s.c. injection of 2x10⁴ B16-OVA in the hind flank. (B) Tumor volume (mm³) at indicated day after B16 injection. (C) Correlation of tumor volume (mm³) 23 days after B16 injection and the frequency of P14 cells in the blood at the time of B16 injection. Data are representative of two independent experiments with at least 5 mice per group. NS=not significant, determined by 2-way ANOVA. Error bars are ± SEM

Figure 3.. Pathogen-specific 4°M CD8 TILs perform Ag-independent effector functions in the tumor.

(A) Experimental design. Mice received adoptive transfer of 10⁴ naïve P14 cells followed by infection with cognate Ag. At a memory time point, 1-5x10⁵ memory P14 cells were adoptively transferred and mice infected to generate higher order memory. (B) The gene expression of IL-12 and IL-18 receptor components in the indicated order of memory CD8 T cells compared to naïve counterparts. (C) Experimental design. Mice received adoptive transfer of 10⁴ naïve P14 (Thy1.1⁺/1.2⁺) and 5x10⁵ 3°M P14 (Thy1.1⁺/1.1⁺) cells followed by i.p. injection of 2x10⁵ CFU LCMV Armstrong. After 22 d mice received s.c. injection of 2x10⁴ B16-OVA in the hind flank. After 9 d mice received i.v. injection of fluorescently-conjugated CD45.2 mAb 3 min prior to harvest. (D) 1°M and 4°M P14 cells (representing 18% and 0.9% of total CD8 T cells, respectively) were distinguished in tumor samples based on Thy disparity. (E) Frequency of KLRG-1 and (F) CD62L expressing P14 cells in the spleen. (G) Representative histograms and gating strategy for indicated marker in i.v.⁻ P14 cells in the tumor. (H) The frequency of 1°M and 4°M P14 cells in i.v.⁻ fraction of tumor samples that express CD69 (I) IFN-γ and (J) Granzyme B. Data are representative of two independent experiments with 4 mice per group. *=p<0.05, **=p<0.01; *** = p<0.001 as determined by student t-test, fold change between groups is shown. Error bars are ± SEM. GEO accession number GSE21360 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE21360).

Figure 4.. Pathogen-specific 4°M CD8 T cells influence growth of B16 tumors.

(A) Experimental Design. Mice received adoptive transfer of 10⁴ naïve or 5x10⁵ 3°M P14 cells followed by i.p. injection of 2x10⁵ CFU LCMV Armstrong 16 days before s.c. injection of 2x10⁴ B16-OVA in the hind flank. (B) Percent survival and (C) Tumor volume (mm³) at indicated day after B16 injection. (D) Correlation of tumor volume (mm³) 18 days after B16 injection and the frequency of 1°M P14 cells at the time of B16 injection. (E) Correlation of tumor volume (mm³) 18 days after B16 injection and the frequency of 4°M P14 cells at the time of B16 injection. Data are representative of two independent experiments with indicated number of mice per group shown. ****= p<0.0001 as determined by 2-way ANOVA, fold change between groups is shown. Error bars are ± SEM.