Targeting SRSF1 improves cancer immunotherapy by dually acting on CD8+T and tumor cells

Abstract

Serine arginine-rich splicing factor 1 (SRSF1) is a key oncogenic splicing factor in various cancers, promoting abnormal gene expression through post-translational regulation. Although the protumoral function of SRSF1 is well-established, the effects of inhibiting tumor-intrinsic SRSF1 on the tumor microenvironment and its impact on CD8⁺ T cell-mediated antitumor immunity remain unclear. Our findings indicate that depleting SRSF1 in CD8⁺ T cells improve antitumor immune function, glycolytic metabolism, and the efficacy of adoptive T cell therapy. The inactivation of SRSF1 in tumor cells reduces transcription factors, including c-Jun, c-myc, and JunB, facilitating glycolytic metabolism reprogramming, which restores CD8⁺ T cell function and inhibits tumor growth. The small-molecule inhibitor TN2008 targets SRSF1, boosting antitumor immune responses and improving immunotherapy effectiveness in mouse models. We therefore introduce a paradigm targeting SRSF1 that simultaneously disrupts tumor cell metabolism and enhances the antitumor immunity of CD8⁺ T cells.

Fig. 1

SRSF1 levels are associated with exhausted phenotypes in CD8⁺T cells. a Diagram illustrating the workflow of single-cell RNA sequencing (scRNA-seq) analysis. The figure was obtained by using Biorender (https://app.biorender.com/). b UMAP visualizations depicting various clusters within human CD8⁺ T cells. Clusters 0, 1, and 6 represent naive CD8⁺ T cells; Cluster 2 corresponds to effector CD8⁺ T cells; Clusters 3, 4, and 5 consist of exhausted CD8 + T cells. c Dotplots illustrate the expression levels of SRSF1 and T cell marker genes across various cell clusters. d Scatter plots illustrate differentially expressed genes (DEGs) between cytotoxic and exhausted CD8⁺ T cells. e UMAP plots for the cell clusters in murine CD8 + T cells. Cluster 0 consists of exhausted CD8⁺ T cells; Clusters 1, 2, 3, 4, and 6 comprise naive CD8⁺ T cells; Cluster 5 contains effector CD8 + T cells. f Dotplots showing marker genes for distinct cell clusters. g Expression of SRSF1 in exhausted CD8⁺ T cells from both tumor and normal tissues. h Western blot showing SRSF1 expressed more in exhausted CD8⁺T cells than activated CD8⁺T cells. i Schematic diagram of scRNA-seq analysis flow for HCC patients undergoing neoadjuvant anti-PD-1 therapy. The figure was obtained by using Biorender (https://app.biorender.com/). j UMAP visualizations for various clusters in human CD8⁺ T cells. k UMAP plots illustrate cluster variations between anti-PD-1 non-responsive and responsive groups. l Barplots showing different CD8⁺T cell clusters in different anti-PD-1 nonresponsive and responsive groups. m Comparison of SRSF1 expression in exhausted CD8⁺T cells between nonresponsive and responsive groups. Data presented as mean ± S.E.M. Statistical significance was determined by Wilcoxon signed rank test and two-tailed unpaired t test. The p value is 0.05, the p value is 0.0001

Fig. 2

SRSF1 depletion in CD8 + T cells enhance its cytotoxicity. a The schematic diagram of establishment of Srsf1 conditional knockout in CD8⁺T cell and Srsf1 expression between two groups from CD8 + T cells. b, c Comparison of HCC progression between Srsf1^fl/flCd4-cre and Srsf1^+/+Cd4-cre mice, with n = 6 for each group. d The study examined the tumor count (left panel) and the liver-to-body weight ratio (right panel) in Srsf1^fl/flCd4-cre and Srsf1^+/+Cd4-cre mouse groups, with each group consisting of six mice. e Kaplan-Meier survival analysis for Srsf1^fl/flCd4-cre and Srsf1^+/+Cd4-cre mice (n = 6 per group). f,g The UMAP of T cells from murine HCC tissues. NC: Srsf1^+/+Cd4-cre mice, KO: Srsf1^fl/flCd4-cre. h The T cell subset proportions in Srsf1^fl/flCd4-cre and Srsf1^+/+Cd4-cre mice (n = 3 per group) were analyzed. i Analysis of scRNA-seq data reveals CD38 expression and distribution differences between Srsf1^fl/flCd4-cre and Srsf1^+/+Cd4-cre mice. j GZMK expression in UMAP plot from scRNA-seq analysis. k The schematic diagram for HCC or B16F10 tumor model combined with PD-1 therapy (Hep1-6 or B16F10 cells: 2 × 10^5) and SRSF1 knockout in CD8⁺ T cell synergized with PD-1 therapy in tumor models. l Flow cytometry assays revealed a higher increase of CD38 + CD8 + T cells in Srsf1^fl/flCd4-cre mice compared to Srsf1^+/+Cd4-cre mice. m Flow cytometry assays indicated a higher increase of CD44⁺CD62L^lowCD8 + T cells in Srsf1^fl/flCd4-cre mice compared to Srsf1^+/+Cd4-cre mice. Data presented as mean ± S.E.M. Statistical significance was determined by log rank test, two-way ANOVA, and two-tailed unpaired t test. **p < 0.01, ***p < 0.001, ****p < 0.0001, ns indicates no significance

Fig. 3

SRSF1 depletion in CD8 + T cells increased CD38 expression by regulating FOXO1. a Heatmap showed upregulated or downregulated DEGs in Srsf1^+/+Cd8+T and Srsf1^–/–Cd8+T groups. b KEGG analysis of CD8 + T cell (Srsf1^+/+Cd8 T vs Srsf1^–/–Cd8 T). c–e Gene cluster analysis of CD8 + T cell (Srsf1^+/+Cd8 T vs Srsf1^–/–Cd8 T). f qPCR analysis of cytotoxic genes between Srsf1^+/+Cd8+T and Srsf1^–/–Cd8+T. g Western blot showed PTEN and S6 expression in Srsf1^–/–Cd8+T and Srsf1^+/+Cd8+T. h The proportion of effector CD8 + T cell in the Srsf1^–/–Cd8+T, Srsf1^+/+Cd8+T and Srsf1^–/–Cd8+T with rapamycin groups. i The Seahorse assays showed Cd8+T cells depleting Srsf1 could increase glycolytic metabolism and can be rescued by mTOR inhibitor. j Luciferase analysis showed SRSF1-RRM1 was binding to 3ʹUTR of PTEN mRNA. k The intersect result between upregulated transcription factors (TF) in SRSF1-KO CD8 + T cells and mTOR pathway related TFs. KO: Knockout. l Foxo1 expression in UMAP plot from scRNA-seq analysis. m Western blot assays showed CD38 and Foxo1 protein levels in Srsf1^+/+ and Srsf1^–/– CD8 + T cells when Foxo1 was silenced. n ChIP assays showed Foxo1 transcriptionally upregulated Cd38. o Western Blot assay showed Srsf1^–/–Cd8+T cells could decrease the expression of Nlk. p Luciferase analysis showed SRSF1-RRM1 was binding to 3ʹUTR of Nlk mRNA. q The study compares the mRNA expression levels of Nlk, Foxo1, and Pten in control versus Srsf1^+/– Cd8+ T cells. Data presented as mean ± S.E.M. A two-tailed unpaired t test was used to assess statistical significance. Significance levels are denoted as follows: * for p < 0.05, ** for p < 0.01, *** for p < 0.001, **** for p < 0.0001, and ‘ns’ indicates no significance

Fig. 4

Reducing tumor-intrinsic SRSF1 suppresses glycolysis by downregulating several bZIP transcription factors and c-myc. a KEGG of SRSF1-downregulated genes. b KEGG of SRSF1-regulated AS events. c, d Western Blot and qPCR assays showed SRSF1 regulated glycolytic genes. e, f SRSF1-RIP-seq peaks showed enrichment in the 5ʹ UTR, CDS, and intronic regions. RIP-seq peaks were classified based on their distribution across various genomic elements and analyzed against the genomic background. g The RIP assays showed SRSF1 was binding to c-myc, c-Jun and JunB. PCR (h) and Western Blot (i) assays showed SRSF1 regulated c-myc, c-Jun and JunB transcription factors. j The schematic diagram of SRSF1 or mutants of SRSF1 proteins. k Western blot analysis of transcription factor expression in SRSF1-sh cells transfected with both endogenous and exogenous SRSF1 using anti-HA. SRSF1 mutants include ΔRRM1 (RRM1 deletion), ΔRRM2 (RRM2 deletion), and ΔRS (RS deletion). All mutants were tagged with HA. l–n SRSF1 could protect c-myc, c-Jun and JunB transcription factors by binding to 3ʹUTR of mRNAs. Data are presented as mean ± S.E.M. and analyzed using one-way ANOVA with a multiple comparison test, where significance is indicated as follows: ***p < 0.001, **p < 0.01, *p < 0.05

Fig. 5

Characterization of the SRSF1 inhibitor TN2008. a The schematic diagram of SRSF1 inhibitor screening. b In vitro proliferative assays screened top SRSF1 inhibitor candidates. c The schematic diagram of SRSF1 inhibitor acting on CD8⁺T cells. d Heatmap showed SRSF1 inhibitors increased cytotoxic or memory genes of CD8⁺T cells. e Construction of a co-culture system for murine organoids and CD8⁺T cells. f Bar plots showed SRSF1 inhibitor TN2008 provided the highest inhibition rate when co-cultured HCC organoids and CD8⁺T cells. g, h SRSF1 inhibitor TN2008 inhibited the tumor growth in PDX mice model. i Overview of structural complex of SRSF1 bound with TN2008. j In vitro assessment of IC50 to evaluate TN2008’s inhibitory effect on SRSF1-mediated tumor growth inhibition (n = 3). k SPR assays showed SRSF1 recombination protein specifically bound with TN2008. l The homology model of the SRSF1 protein is in complexity with TN2008 by molecular dynamic simulation. m Key residues and their energy contributions were determined through MD simulation and MM/PBSA analysis. n Surface plasmon resonance (SPR) analysis demonstrated the interaction between TN2008 and the mutated SRSF1 protein. o qPCR assays showed SRSF1-mutated residues were responsible for glycolytic genes expression in tumor cells. p Western blot analysis demonstrated that TN2008 reduced glycolytic gene expression in tumor cells. q Western blot analysis demonstrated that TN2008 elevated Foxo1 expression in murine CD8⁺ T cells. r Colony formation assays showed TN2008 significantly inhibited tumor cell growth in vitro. Data presented as mean ± S.E.M. Statistical significance was assessed using a two-tailed unpaired t test and two-way ANOVA. Significance levels are denoted as follows: * for p < 0.05, ** for p < 0.01, *** for p < 0.001, **** for p < 0.0001, and ‘ns’ indicates no significance

Fig. 6

The SRSF1 inhibitor TN2008 enhances T cell infiltration and functionality and works synergistically with anti-PD-1 therapy. a The inhibitory effect of TN2008 on subcutaneous tumors in ll2rg-/-NOD-Scid mice. b, c TN2008 synthesizes PD-1 blockade in subcutaneous tumor model. d TN2008 could alleviate the function of murine liver from HCC tumor mice model. e Percentage of reduced tumor volume in two different mice groups. f TN2008 synthesizes PD-1 blockade in murine HCC spontaneous model. g liver tumor burden in four different groups. h Kaplan-Meier survival analysis between four different groups. i Liver-to-body weight ratio across various groups of mice. j Percentage of effector CD8 + T cell in four different groups. k Other cell percentage in different four groups. l The ratio of regulatory T cells (Tregs) to effector CD8 + T cells across four groups. m CyTOF analysis was conducted to examine T cells within the tumor microenvironment across three distinct groups. n The percentage of CD38⁺CD8⁺T cells among different groups. o The percentage of CD69⁺CD8⁺T cells among different groups. p The expression of SRSF1 in PD-1 responsive or non-responsive HCC cases. q The expression of SRSF1 in melanoma tissues among respond or non-respond cases from public database. r The expression of SRSF1 in basal cell carcinoma among respond or non-respond cases from public database. Data presented as mean ± S.E.M. Statistical significance was assessed using a two-tailed unpaired t test, two-way ANOVA, log-rank test, and Wilcoxon signed-rank test. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001