Shp-1 regulates the activity of low-affinity T cells specific to endogenous self-antigen during melanoma tumor growth and drives resistance to immune checkpoint inhibition

Abstract

Background: The presence of activated CD8 T cells in the tumor microenvironment is correlated with an effective immune response to immune checkpoint inhibitor (ICI) therapy. However, ICI predominantly targets high-affinity T cells, which may be less abundant in tumors with few neoantigens. Targeting the intracellular phosphatase Src homology region 2 domain-containing phosphatase-1 (Shp-1) in combination with ICI lowers the T cell activation threshold and enhances the ability of low-affinity T cells to mount a productive antitumor response.

Methods: In this study, we sought to determine whether temporal inhibition of Shp-1 during active tumor growth could rescue the activity of low-affinity T cells specific for endogenous self-antigens. To address this question, we implanted Yale University Mouse Melanoma (YUMM) tumor cell lines into WT mice and, on tumor establishment, administered an inhibitor of Shp-1 (TPI-1) with or without ICI treatment. We analyzed treatment-dependent changes in the immune infiltrate in the tumor via flow cytometry, major histocompatibility complex (MHC) tetramer-mediated detection of tyrosinase-related protein 2 (TRP-2)_180-188-specific T cells and a micropipette-based two-dimensional affinity assay to measure the T cell receptor (TCR) affinity.

Results: Administration of ICI and a Shp-1 inhibitor to mice with established YUMM tumors, but neither agent alone, resulted in a significant delay in tumor growth and an increased frequency of CD8 tumor-infiltrating T cells with enhanced effector and reduced exhaustion characteristics. In particular, combined treatment increased the frequency of CD8 T cells specific for the MHC Class I-restricted tumor self-antigen TRP-2_180-188. We found that the increase in effector T cells was almost entirely due to an increase in T cells with very low TCR affinity.

Conclusions: We conclude that approaches for altering TCR signaling threshold are effective in enhancing the antitumor response of low-affinity T cells specific for endogenous self-antigens in settings of ICI resistance and/or where neoantigens are not available to drive antitumor responses.

Keywords: Immune Checkpoint Inhibitor; Immunotherapy; Melanoma; T cell; T cell Receptor - TCR.

Figure 1. Inhibition of Shp-1 combines with ICI to inhibit YUMM tumor growth. (A) YUMM 1.1 or 5.2 cells were injected subcutaneously (s.c.) into the right rear flank of wildtype C57BL/6 mice. Tumor length and width were measured at various stages of tumor growth. Growth curves indicate tumor volume at each time point. (B) Bar graphs indicate the frequency of CD8 T cells, CD4 T cells and Tregs of total CD45+T cells isolated from the tumor, as determined by flow cytometric analysis, in both YUMM 1.1 (black) and YUMM 5.2 (blue) at day 21 of tumor growth. An additional bar graph indicates the ratio in the frequency of CD8 T cells to Tregs (n=10 mice per group, *p<0.05, **p<0.01, ***p<0.001). (C) Mice bearing YUMM 1.1 or YUMM 5.2 tumors were treated intraperitoneally (i.p.) with TPI-1 in combination with ICI (anti-PD-1 and anti-CTLA), or singly with either TPI-1 or ICI, every 3 days beginning at the time of tumor visibility (day 7–10). Graphs show tumor volume over time. Statistical significance was determined using a one-way ANOVA of the area under the curve (AUC). Results are representative of three independent experiments. (D) Graphs show differences in tumor volume at day 23 (YUMM 1.1) or day 20 (YUMM 5.2). Statistical significance was determined using a one-way ANOVA of the area under the curve (AUC) of tumor volume measurements over time (for C) or of the mean tumor volume of an individual time point (for D). (n=7–10 mice per group, **p<0.01, ***p<0.001, ****p<0.0001). ANOVA, analysis of variance; CTLA, cytotoxic T-lymphocytes-associated protein; ICI, immune checkpoint inhibitor; ns, not significant; PD-1, programmed cell death protein-1; Shp-1, Src homology region 2 domain-containing phosphatase-1; Treg, regulatory T cells; YUMM, Yale University Mouse Melanoma.

Figure 2. ICI+Shp-1i induces the presence of effector CD8 T cells in YUMM tumors. YUMM 1.1 and YUMM 5.2 tumors were established in C57BL/6 mice as in figure 1. Representative flow plots indicate the frequency of (A) CD8+CD3+ (gated on live CD45+cells), (B) CD8+PD-1+ cells (gated on live CD45+CD8+ cells), and (C) T-bet+Granzyme B (GzB)+cells (gated on live CD8+CD3+) isolated from YUMM 5.2 tumors at day 21 of tumor growth. Mice were treated with ICI alone, Shp-1i alone, ICI+Shp-1i, or were untreated (UT), as in figure 1. (D) Bar graphs indicate the frequency of CD8+CD3+ (of live CD45+cells), CD8+PD-1+ cells (of live CD45+cells), T-bet+GzB+ cells (of CD8+cells) and T-bet+GzB+ cells (“Effector CD8”), of live CD45+isolated from YUMM 1.1 and 5.2 tumors at day 21 of tumor growth in the indicated treatment groups. Statistical analysis was performed using one-way ANOVA (n=5–10 mice per group, **p<0.01, ****p<0.0001). Results are representative of two independent experiments. ANOVA, analysis of variance; ICI, immune checkpoint inhibitor; ns, not significant; PD-1, programmed cell death protein-1; Shp-1, Src homology region 2 domain-containing phosphatase-1; UT, untreated;YUMM, Yale University Mouse Melanoma.

Figure 3. ICI+Shp-1i does not increase the frequency of CD4 T cells or FoxP3-expressing Tregs in YUMM tumors. Flow cytometry analysis of infiltrating CD45+cells was performed for YUMM 1.1 and 5.2 tumors harvested at day 21. (A) Bar graphs indicate the frequency of CD4+CD3+ (of live CD45+cells) and CD4+FoxP3+ (Treg) cells (of live CD45+cells), MFI of Foxp3 among CD4 TIL, as well as the ratio of CD8+CD3+ T cell to Treg frequency for YUMM 1.1 and 5.2 tumors at day 21 of tumor growth in the indicated treatment groups. (B) Representative flow plots indicate the frequency of FoxP3-expressing CD4 T cells (gated on CD4+CD3+CD45+) in YUMM 1.1 tumors. (C–D) Bar graphs indicate the frequency of M-MDSC (CD11b+LY6C^hiLy6G^lo, gate on CD45+), PMN-MDSC (CD11b+LY6C^loLy6G^hi, gate on CD45+), and M1-Macrophages (MHCII+CD11b+F480+) in the indicated treatment groups in (C) YUMM 1.1 and (D) YUMM 5.2 tumors. (E–F) Bar graphs indicate the frequency of cDC1 (CD11c+CD11b- CD103+) and cDC2 (CD11c+CD11b+MHCII+ F4/80-) in (E) YUMM 1.1 and (F) YUMM 5.2 tumors. Statistical analysis was performed using one-way ANOVA (n=7–10 mice per group, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001). Results are representative of two independent experiments. ANOVA, analysis of variance; ICI, immune checkpoint inhibitor; MDSC, myeloid-derived suppressor cell; MFI, mean fluoresence intensity; PMN, polymorphonuclear;MHC, major histocompatibility complex; ns, not significant; Shp-1, Src homology region 2 domain-containing phosphatase-1; TIL, tumor-infiltrating lymphocyte; Treg, regulatory T cells; YUMM, Yale University Mouse Melanoma.

Figure 4. ICI+Shp-1i alters the frequency of activated and exhausted phenotype CD8 T cells in the tumor. Flow cytometry analysis of infiltrating CD45+cells was performed for YUMM 1.1 and 5.2 tumors harvested at day 21. (A) Representative flow plots indicate expression of 2B4 and PD-1 expression from both YUMM 1.1 (upper panel) and YUMM 5.2 tumors (lower panel), gated on live CD45+CD3+ CD8+ cells. (B) Bar graphs show frequency of PD-1+2B4 and PD-1+2B4+ in YUMM 1.1 tumors as a percentage of CD8+CD3+ T cells or total live CD45+cells. (C) Bar graphs show frequency of PD-1+2B4 and PD-1+2B4+ in YUMM 5.2 tumors as a percentage of CD8+CD3+ T cells or total live CD45+cells. Statistical analysis was performed using one-way ANOVA (n=7–10 mice per group, *p<0.05, **p<0.01, ****p<0.0001). Results are representative of two independent experiments. ANOVA, analysis of variance; ICI, immune checkpoint inhibitor; ns, not significant; PD-1, programmed cell death protein-1; Shp-1, Src homology region 2 domain-containing phosphatase-1; UT, untreated; YUMM, Yale University Mouse Melanoma.

Figure 5. ICI+Shp-1i increases the frequency of TRP-2-specific CD8 TIL. Flow cytometry analysis of infiltrating CD45+cells was performed for YUMM 1.1 and 5.2 tumors harvested at day 21. (A) Representative flow plots indicate the frequency of CD8 TIL in YUMM tumors showing dual binding to H-2K^b/TRP-2_180–188 tetramers (TRP-2+) conjugated to PE or APC (right plot), as compared with unstained control cells (left plot). (B) Representative flow plots show tetramer binding by CD8+CD3+ TIL in YUMM 1.1 tumors at day 21 following the indicated treatments. (C) Bar graphs indicate the frequency of total TRP-2+, TRP-2+PD-1+ and TRP-2+PD-1− CD8 TIL of CD45+cells in both YUMM 1.1 and YUMM 5.2 under the indicated treatment conditions. (D) Representative flow plots indicate expression of PD-1 and binding to tetramer by CD8 TIL in both YUMM 1.1 and YUMM 5.2 under the indicated treatment conditions. (E) Representative histograms indicate expression of 2B4 by PD-1+TRP-2+ CD8 TIL in both YUMM 1.1 and YUMM 5.2 under the indicated treatment conditions. Statistical analysis was performed using one-way ANOVA (n=7–10 mice per group, *p<0.05, ****p<0.0001). Results are representative of two independent experiments. ANOVA, analysis of variance; APC, allophycocyanin; ICI, immune checkpoint inhibitor; PD-1, programmed cell death protein-1; PE, phycoerythrin; Shp-1, Src homology region 2 domain-containing phosphatase-1; TIL, tumor-infiltrating lymphocyte; TRP-2, tyrosinase-related protein 2; UT, untreated; YUMM, Yale University Mouse Melanoma.

Figure 6. Effector phenotype TRP-2+cells are induced following ICI+Shp-1i treatment. Flow cytometry analysis of infiltrating CD45+cells was performed for YUMM 1.1 and 5.2 tumors harvested at day 21. Representative flow plots indicated expression of TCF-1 and Tbet by (A) TRP-2+PD-1– CD8+cells and (B) TRP-2+PD-1– CD8+cells in YUMM 1.1 and YUMM 5.2 tumors in the indicated treatment groups. (C) Bar graphs indicate the frequency of T-bet^hi TCF-1^lo cells among TRP-2+PD-1+ CD8 T cells (left) and TRP-2+PD-1– CD8 T cells (right) for both YUMM 1.1. Statistical analysis was performed using one-way ANOVA (n=7–10 mice per group, *p<0.05, ***p<0.001, ****p<0.0001). Results are representative of two independent experiments. ANOVA, analysis of variance; ICI, immune checkpoint inhibitor; PD-1, programmed cell death protein-1; Shp-1, Src homology region 2 domain-containing phosphatase-1; TRP-2, tyrosinase-related protein 2; UT, untreated; YUMM, Yale University Mouse Melanoma.

Figure 7. ICI+Shp-1i induces increased low-affinity CD8 TIL specific for TRP-2_180–188. CD8 TILs were isolated from YUMM 1.1 or YUMM 5.2 tumors in untreated or ICI+Shp-1i treated mice on day 21 of tumor growth. The frequency and affinity of TRP-2_180–188-specific T cells was assessed by a micropipette-based two-dimensional affinity assay. (A) Plot shows the adhesion frequency of individual T cells specific for K^b/TRP2_180–188, as compared with a negative control (K^b/OVA_257–264), in untreated and ICI+Shp-1i-treated YUMM 1.1 and YUMM 5.2 tumors. (B) Plots show the frequency of T cells binding to K^b/TRP-2_180–188 as a percentage of CD8 T cells or total live CD45+cells in the tumor. (C) Plot indicates the relative two-dimensional affinity of TRP-2-specific T cells isolated from YUMM 1.1 or YUMM 5.2 tumors under the indicated treatment conditions. Each circle represents an individual T cell, and results are pooled from four mice per group. The two-dimensional affinity of a known high-affinity TCR transgenic T cell line, OT-1 T cells specific for K^b/OVA_257–264, is shown for reference. (D) Plots show the frequency of low-affinity (<1×10⁻⁵ A_cK_amM⁶) and high-affinity (>1×10⁻⁵ A_cK_amM⁶) TRP-2-specific T cells in YUMM 1.1 and YUMM 5.2 tumors under the indicated treatment conditions. Statistical analysis was performed using one-way ANOVA (*p<0.05, ***p<0.01, ***p<0.001, ****p<0.0001). ANOVA, analysis of variance; ICI, immune checkpoint inhibitor; ns, not significant; OVA, Ovalbumin; Shp-1, Src homology region 2 domain-containing phosphatase-1; TCR, T cell receptor; TIL, tumor-infiltrating lymphocyte; TRP-2, tyrosinase-related protein 2; UT, untreated; YUMM, Yale University Mouse Melanoma.