Slc7a5 promotes T cell anti-tumor immunity through sustaining cytotoxic T lymphocyte effector function

Abstract

Tryptophan (Trp) metabolites have emerged as key regulators of host tumor immunity and cancer patient response to immunotherapy. However, the function of and mechanism underlying Trp in tumor-activated CTLs in the tumor microenvironment are incompletely understood. Using a defined co-culture system of tumor-specific CTLs and cognate antigen-expressing tumor cells, we performed a genome-wide metabolomics screening and observed that Trp level is elevated in the tumor cell-activated CTLs. Parallel genome-wide RNA-Sequencing and ATAC-Sequencing analysis determined that tumor-specific CTLs respond to tumor cells by transcriptionally activating Slc7a5 expression. Pharmacological inhibition of Slc7a5 decreased Trp uptake in tumor-activated CTLs and suppressed CTL lytic activity in killing tumor cells in vitro. Mice with Slc7a5 deficiency only in T cells exhibited diminished level of tumor-infiltrating T cells and increased tumor growth and metastasis. scRNA-sequencing analysis revealed that Slc7a5 deficiency resulted in decreased activation of the aryl hydrocarbon receptor (AhR) pathway and repressed FasL expression in tumor-infiltrating T cells. Chromatin immunoprecipitation determined that AhR binds to Faslg promoter in tumor-infiltrating T cells. FasL blockade therapy promotes tumor growth and metastasis in tumor-bearing mice. In human cancer patients, AhR expression correlates with FasL expression in tumor-infiltrating T cells. Furthermore, FasL expression is correlated with patient response to pembrolizumab and survival time. Our finding determines that the Slc7a5-Trp metabolic pathway activates AhR to up-regulate FasL expression in tumor-infiltrating T cells to sustain CTL anti-tumor immunity. Targeting CAR-T cells to up-regulate Slc7a5 to maintain T cell proliferation and function therefore could be a promising direction in cancer immunotherapy.

Fig. 1. Tumor cell and tumor-specific T cell interaction leads to T cell activation.

A The tumor-specific 2/20 CTL phenotype. B CTL responds to tumor cells by activating PD-1 expression. C Tumor cell and tumor-specific CTL co-culture results in rapid CTL proliferation in vitro. Day1=one day after tumor and CTL co-culture. Day2=two days after tumor and CTL co-culture. Day3=three days after tumor cell and CTL co-culture. D–G Cytokine and chemokine profiles of tumor-activated CTLs in vitro. 2/20 CTLs were cultured with 4T1 and AB1 tumor cells at 1:1 ratio, respectively. Culture supernatants from tumor cell alone, CTLs alone, and tumor cell-CTL co-cultures were collected at days 1 and 3 and measured for the indicated cytokines.

Fig. 2. Tumor-specific CTLs respond to tumor cells by increasing tryptophan accumulation.

A Experimental design of metabolomics, RNA-Seq, and ATAC-Seq analysis of tumor-specific CTLs. B Metabolite set enrichment analysis (MSEA) of metabolites increased (left) or decreased (right) in antigen (AH1 peptide)-activated 2/20 CTLs compared to resting 2/20 CTLs. C Relative levels of amino acids, glycolysis and TCA cycle intermediates, and PPP metabolites and nucleotides in antigen peptide-activated 2/20 CTLs compared to resting 2/20 CTLs. D Metabolite set enrichment analysis (MSEA) of metabolites increased (left) or decreased (right) in tumor cell-activated CTLs as compared to the resting CTLs. E Relative levels of amino acids, glycolysis-TCA cycle intermediates, and PPP metabolites in tumor cell-activated CTLs as compared to the resting CTLs.

Fig. 3. Slc7a5 is transcriptionally activated in tumor cell-activated CTLs.

A Heatmaps illustrating differential chromatin accessibility at 2 kb windows centered at the summit of the ATAC-seq peaks identified between AH1-activated 2/20 CTLs and resting 2/20 CTLs (first three columns) and gene expression log2 fold changes at 24 hours between these two samples (4^th column on the right). Line graphs on top of the heatmaps demonstrate the average accessibility profiles of clusters 1–4, respectively. The box plot on top of the expression heatmap represents the average log2 fold changes of clusters 1–4. The ATAC-seq peak signals represent the average signals of two biological replicates. For genes associated with multiple ATAC-seq peaks, only peaks with the highest log2 fold changes of differential peak signals were shown. B Heatmaps illustrating differential chromatin accessibility at 2 kb windows centered at the summit of the ATAC-seq peaks identified between tumor cell-activated 2/20 CTLs and resting 2/20 CTLs samples (first three columns) and gene expression log2 fold changes at 24 hours between these two samples (4^th column on the right). C GSEA analysis results demonstrate the enrichment of up-regulated transporter genes in AH1-activated and tumor cell-activated 2/20 CTLs as compared with resting 2/20 CTLs. 142 differentially expressed transporter genes were used as a gene set and compared with rank-sorted gene expression from matched samples. The results suggest that more transporter genes are up regulated. D Heatmap illustrating the differential expression patterns of the 142 transporter genes at various time points of 2/20 CTL activation by antigen AH1 and tumor cells. E Wiggle plot of ATAC-seq peaks in Slc7a5 gene locus. The colors of the tracks represent the sample groups. Green, resting 2/20 CTLs; red, AH1-activated 2/20 CTLs; and blue, tumor cell-activated 2/20 CTLs. The unique motif peaks in tumor activated CTLs are indicated with the blue arrows. Red arrows point to peaks in activated CTLs.

Fig. 4. Slc7a5 is essential for CTL tryptophan uptake in tumor cell-activated CTLs in vitro.

A–D 2/20 CTLs were co-cultured with 4T1 and AB1 tumor cells, respectively in the presence of JPH203 (10 μM) for 24 h, and analyzed for the indicated metabolites. E–H 2/20 CTLs were co-cultured with 4T1 and AB1 tumor cells, respectively, in the presence of JPH203 at the indicated doses for 6 h. The CTL and tumor co-culture mixtures were stained with CD8-specific mAb, Annexin V, and PI. CD8⁺ and CD8⁻ cells were gated for cell death of the CTLs (E, F) and tumor cells (G, H).

Fig. 5. Slc7a5 is essential for TIL expansion and tumor growth control in mice.

A AT3 cells were injected orthotopically in breast in WT (n = 9) and Slc7a5.TKO (n = 8) C57BL/6 mice. Shown is tumor image. B Tumor size and weight as in A were quantified and presented. C The tumor-bearing mice as in A were sacrificed after 14 days and the lungs were inflated by India ink. The tumor nodules were quantified and presented. D B16F10 cells were injected subcutaneously in WT (n = 4) and Slc7a5.TKO (n = 7) C57BL/6 mice. Shown is tumor image. E Tumor size and weight as in D were quantified and presented. F Spleens were collected from AT3 tumor-bearing WT and Slc7a5.TKO C57BL/6 mice as in A and analyzed by flow cytometry for CD8⁺, CD4⁺ and CD11b⁺ cells. Shown are representative dot plots. G Quantification of indicated immune cells as in (F). H Tumor tissues were collected from AT3 tumor-bearing WT and Slc7a5.TKO C57BL/6 mice as in A and analyzed by flow cytometry for CD8⁺, CD4⁺ and CD11b⁺ cells. Shown are representative dot plots. I Quantification of indicated immune cells as in (H). J Spleens were collected from B16F10 tumor-bearing WT and Slc7a5.TKO C57BL/6 mice as in D and analyzed by flow cytometry for CD8⁺ and CD4⁺ cells. Shown are representative dot plots. K Quantification of indicated immune cells as in (J). L Tumor tissues were collected from B16F10 tumor-bearing WT and Slc7a5.TKO C57BL/6 mice as in D and analyzed by flow cytometry for CD8⁺ and CD4⁺cells. Shown are representative dot plots. M Quantification of indicated immune cells as in (L).

Fig. 6. Slc7a5 regulates TCR signaling and the AhR metabolic pathway in TILs in mice.

A UMAP plot of major cell subpopulations isolated from orthotopic breast tumor-bearing mice, colored by identified cell clusters (right panel). B Dot plot of signaling pathway activation as identified by expression of gene signatures in the indicated cell subpopulations in WT and Slc7a5.TKO (KO) mice. C Violin plot of signaling pathway activation as identified by expression of gene signatures in the indicated cell subpopulations in WT and Slc7a5.TKO mice. Dot (D) and violin (E) plots of expression of the indicated genes in the indicated cell subpopulations in WT and Slc7a5.TKO mice.

Fig. 7. AhR activates FasL expression to suppress tumor growth and metastasis in vivo.

A Wiggle plot of ATAC-seq peaks in the Faslg gene locus. Peaks observed in the tumor-activated CTLs are highlighted with arrows. B The Falg gene promoter structure with the predicted AhR-binding consensus sequence element (left panel). C Tumor-infiltrating T cells were isolated from AT3 tumor-bearing WT and SLC7A5.TKO mice. Shown is purity of the isolated CD8⁺ T cells from the tumor tissues. The purified CD8⁺ T cells were then subjected to ChIP using anti-AhR antibody. The input DNA was used as positive control and the IgG was used as negative control. D, E Breast tumor AT3 cells were injected orthotopically into mice. Sixteen mice with similar tumor size were randomly divided into two groups and treated with IgG or anti-FasL antibody every 3 days for four times. Shown are tumor images (D) and tumor size and weight (E) at the end of experiment. F The tumor-bearing mice as shown in (C) were inflated with ink to inflate lungs and tumor nodules in the lungs were quantified. G, H Melanoma B16F10 cells were injected subcutaneously to mice. Twelve mice with similar size of tumors were randomly divided into two groups and treated with IgG or anti-FasL mAb, respectively, every 3 days for 4 times. Shown are tumor image (G) and tumor size and weight (H). I TUNEL analysis of melanoma B16F10 tumors treated with IgG or anti-FasL antibody as in (G). Shown are representative images. Red arrows point to apoptotic cells. The apoptotic cells were quantified and presented at the right.

Fig. 8. FasL expression is correlated with human cancer patient response to PD-1 blockade immunotherapy and patient survival.

A Published human breast cancer patient scRNA datasets were extracted downloaded from the database [28] and analyzed using R package. UMAP plot major cell subpopulations in responders and non-responders of breast cancer patients treated with pembrolizumab. B–E UMAP plot of the major cell populations as shown in B, showing PD-1 (C, D) and FasL (E, F) expression level in TILs. Red dots represent PD-1⁺ (C) and FasL⁺ (E) TILs. PD-1 (D) and FasL (F) expression level is shown as violin plots. F Correlation between FasL expression and Slc7a5 expression in tumor-infiltrating T cells in human breast cancer patients. G TCGA dataset analysis showing correlation between FasL expression level and patient survival in breast cancer. H TCGA dataset analysis showing correlation between FasL expression level and patient survival in melanoma. I TCGA dataset analysis showing FasL expression level and patient survival in pan-cancer (33 types of human cancers).