tRNA m1A modification regulates cholesterol biosynthesis to promote antitumor immunity of CD8+ T cells

Abstract

Activation of CD8+ T cells necessitates rapid metabolic reprogramming to fulfill the substantial biosynthetic demands of effector functions. However, the posttranscriptional mechanisms underpinning this process remain obscure. The transfer RNA (tRNA) N1-methyladenine (m1A) modification, essential for tRNA stability and protein translation, has an undefined physiological function in CD8+ T cells, particularly in antitumor responses. Here, we demonstrate that the tRNA m1A "writer" gene Trmt61a enhances the tumor-killing capacity of CD8+ T cells by regulating cholesterol biosynthesis. Deletion of Trmt61a in CD8+ T cells leads to a compromised tumor-killing function in both in vivo and in vitro assays. Mechanistically, tRNA m1A promotes antitumor immunity in CD8+ T cells by enhancing the translation of ATP citrate lyase, a key enzyme for cholesterol biosynthesis. Cholesterol supplementation rescues the impaired tumor-killing function and proliferation of TRMT61A-deficient CD8+ T cells. Our findings highlight tRNA m1A modification as a regulatory checkpoint in cholesterol metabolism in CD8+ T cells, suggesting potential novel strategies for cancer immunotherapy.

Graphical abstract

Figure 1.

TRMT61A expression correlates with T cell tumor-killing function in the TME. (A) Box plots depict the expression levels of the m¹A modification writer gene TRMT61A in paired normal (blue) and tumor (red) tissues. The boxes indicate the median ± 1 quartile range, with whiskers extending to the smallest or largest values within 1.5× IQR from the box boundaries. (B) Kaplan–Meier survival curves for patients with high (red) or low (blue) TRMT61A expression levels in tumor tissues from the COAD cohort. P < 0.05 in the two-sided log-rank test is considered statistically significant. (C) RT-qPCR analysis of Trmt61a mRNA levels in CD8⁺ T cells sorted from the spleen of untreated WT mice (SPL) and from TIL and LN compartments of MC38 tumor-bearing WT mice (n = 6). (D) RT-qPCR analysis of Trmt61a mRNA levels in CD8⁺ T cells from untreated, tumor culture medium (TCM) treated for 6 and 20 h groups, and tumor co-culture groups (n = 3). Naïve CD8⁺ T cells were sorted from WT mice and activated with anti-CD3/CD28 antibodies. (E) Violin plots displaying the signature scores of T cell–mediated cytotoxicity in relation to TRMT61A expression (red) or silence (blue) in tumor-infiltrating CD8⁺ T cells from CRC patient scRNA-seq data. (F) Western blot quantification of TRMT61A protein levels in naïve CD8⁺ T cells from WT mice stimulated with anti-CD3/CD28 antibodies for 0, 6, and 24 h. BLCA, bladder urothelial carcinoma; BRCA, breast invasive carcinoma; CHOL, cholangiocarcinoma; COAD, colon adenocarcinoma; ESCA, esophageal carcinoma; HNSC, head and neck squamous cell carcinoma; KICH, kidney chromophobe; KIRP, kidney renal papillary cell carcinoma; LIHC, liver HCC; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; PRAD, prostate adenocarcinoma; READ, rectal adenocarcinoma; STAD, stomach adenocarcinoma. Data are representative of three (C, D, and F) independent experiments. Error bars represent mean ± SEM; *P < 0.05, ***P < 0.001, ****P < 0.0001; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (C and D). Source data are available for this figure: SourceData F1.

Figure S1.

Generation and characterization of Trmt61a-cKO mice. (A) Schematic representation of the strategy for generating Trmt61a-cKO mice, utilizing the Cre-loxP recombination system. (B) RT-qPCR quantification of Trmt61a mRNA levels in WT and Trmt61a-cKO CD8⁺ T cells, demonstrating efficient gene ablation (n = 6). (C) Western blot of TRMT61A and TRMT6 protein levels in WT and Trmt61a-cKO CD8⁺ T cells, confirming the specificity of the knockout. (D) Appearance of immune organs in Trmt61a^flox/flox, Trmt61a^flox/floxCd4^Cre, and true WT mice (different genders are shown separately). (E–G) Flow cytometric analysis of CD4⁺ T and CD8⁺ T cell composition in thymus from WT and Trmt61a-cKO mice under steady-state conditions (n = 7–9). (H–L) Flow cytometric analysis of CD4⁺ T and CD8⁺ T cell composition in the spleen, mLN, iLN, and peripheral blood mononuclear cell (PBMC) from WT and Trmt61a-cKO mice under steady-state conditions (n = 7–9). Data are representative of three (B, C, and E–L) independent experiments. Error bars represent mean ± SEM; *P < 0.05, ****P < 0.0001; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (F, G, and I–L) and unpaired t test (B). Source data are available for this figure: SourceData FS1.

Figure 2.

Deletion of Trmt61a in T cells promotes tumor growth in vivo. (A) Quantification of the m¹A/A ratio in total tRNA purified from activated WT and Trmt61a-cKO CD8⁺ T cells (48 h) as determined by liquid chromatography with mass spectrometry (MS) (n = 3). (B–F) Flow cytometric analysis of CD8⁺ T cell activation in the spleen, mLN, iLN, and PBMC from WT and Trmt61a-cKO mice under steady-state conditions (n = 7–9). (G–I) Tumor growth (G and H) and tumor weight (I) in WT and Trmt61a^flox/floxCd4^Cre mice injected subcutaneously with 5 × 10⁵ MC38 colon cancer cells (n = 4–8). (J–L) Flow cytometric analysis of the proportion (K) and cell number (per gram tumor) (L) of tumor-infiltrating CD8⁺ T cells from WT and Trmt61a^flox/floxCd4^Cre MC38 tumor-bearing mice on day 14 (n = 4–8). (M and N) Flow cytometric analysis of Ki-67 expression in tumor-infiltrating CD8⁺ T cells from WT and Trmt61a^flox/floxCd4^Cre MC38 tumor-bearing mice (n = 4–8). (O and P) Flow cytometric analysis of granzyme B production in tumor-infiltrating CD8⁺ T cells from WT and Trmt61a^flox/floxCd4^Cre MC38 tumor-bearing mice (n = 4–8). Data are representative of three (A–P) independent experiments. Error bars represent mean ± SEM; **P < 0.01, ***P < 0.001, ****P < 0.0001; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (A and C–F) and unpaired t test (H, I, K, L, N, and P).

Figure S2.

The CD4 ⁺ T cell functional populations of Trmt61a-cKO mice. (A–E) Flow cytometric analysis of CD4⁺ T cell activation in the spleen, mLN, iLN, and PBMC from WT and Trmt61a-cKO mice under steady-state conditions (n = 7–9). (F–H) Flow cytometric analysis of FOXP3⁺ Treg cells in the spleen, mLN, iLN, and PBMC from WT and Trmt61a-cKO mice under steady-state conditions (n = 7–9). Data are representative of three (A–H) independent experiments. Error bars represent mean ± SEM; ****P < 0.0001; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (B–E, G, and H).

Figure S3.

Trmt61a deficiency does not affect total CD4 ⁺ T cell infiltration or function. (A and B) Flow cytometric analysis of IFN-γ production in tumor-infiltrating CD8⁺ T cells from WT and Trmt61a^flox/floxCd4^Cre MC38 tumor-bearing mice (n = 6). (C and D) Flow cytometric analysis of the proportion (C) and cell number (D) of tumor-infiltrating CD4⁺ T cells in WT and Trmt61a-cKO MC38 tumor-bearing mice on day 14. (E–H) Flow cytometric analysis of the expression of IFN-γ (E and F) and granzyme B (G and H) in tumor-infiltrating CD4⁺ T cells from WT and Trmt61a-cKO MC38 tumor-bearing mice. (I and J) Flow cytometric analysis of the proportion of FOXP3⁺ Treg cells in tumor-infiltrating CD4⁺ T cells from WT and Trmt61a-cKO MC38 tumor-bearing mice. Data are representative of three (A–J) independent experiments. Error bars represent mean ± SEM; **P < 0.01; NS, nonsignificant. Unpaired t test (B–D, F, H, and J).

Figure 3.

Trmt61a is indispensable for the tumor-killing function of CD8⁺T cells. (A) Schematic diagram of the transfer tumor model using pre-activated OT-I CD8⁺ T cells and MC38-OVA bearing Rag1^−/− mice. (B–E) Flow cytometric analysis of granzyme B (B and C) and IFN-γ (D and E) production in the in vitro–activated live WT and cKO OT-I CD8⁺ T cells before transfer (n = 5). (F–H) Tumor weight (F and G) and tumor growth (H) in MC38-OVA bearing Rag1^−/− mice that received WT and cKO OT-I CD8⁺ T cells (n = 5–8). (I–K) Flow cytometric analysis of the proportion (J) and cell number (per gram tumor) (K) of tumor-infiltrating OT-I CD8⁺ T cells from MC38-OVA bearing Rag1^−/− mice that received WT and cKO OT-I CD8⁺ T cells (n = 5–8). (L–O) Flow cytometric analysis of granzyme B (L and M) and IFN-γ (N and O) production in tumor-infiltrating OT-I CD8⁺ T cells from MC38-OVA bearing Rag1^−/− mice that received WT and cKO OT-I CD8⁺ T cells (n = 5–8). Data are representative of three (B–K) independent experiments. Error bars represent mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; NS, nonsignificant. Unpaired t test (C, E, G, H, J, K, M, and O).

Figure 4.

TRMT61A-mediated tRNA m ¹ A promotes CD8 ⁺ T cell–mediated tumor immunity in vitro. (A and B) Flow cytometric analysis of the CD8⁺ T cell killing assay for WT and Trmt61a^flox/floxCd4^Cre mice. Splenocytes from WT-OT-I and Trmt61a-cKO-OT-I mice were stimulated with OVA peptide (257–264) in vitro and co-cultured with 2 × 10⁴ MC38-OVA cells for 16 h. Killed tumor cells were identified using ANNEXIN-5 and 7-AAD (n = 8). (C–F) Flow cytometric analysis of granzyme B (C and D) and IFN-γ (E and F) expression in CD8⁺ T cells from WT and Trmt61a^flox/floxCd4^Cre mice. Naïve CD8⁺ T cells from Trmt61a-cKO and control mice were stimulated with anti-CD3/CD28 antibodies for 48 h (n = 6). (G and H) Flow cytometric analysis of splenic CD8⁺ T cell activation in WT and Trmt61a^flox/floxCd4^Cre mice upon in vitro stimulation with CD3/CD28 antibodies for 24 h (n = 5). (I and J) Assessment of WT and Trmt61a-deficient naïve CD8⁺ T cell proliferation using CellTrace dilution after in vitro stimulation with CD3/CD28 antibodies for 72 h (n = 8). (K) CD8⁺ T cell killing assay for WT and Trmt61a-cKO CD8⁺ T cells, after retroviral overexpression of GFP-control, TRMT61A-WT, and catalytic-dead mutant TRMT61A (TRMT61A-dead) plasmids (n = 6). (L) Assessment of WT and Trmt61a-deficient CD8⁺ T cell proliferation after retroviral overexpression of GFP-control, TRMT61A-WT, and TRMT61A-dead plasmids (n = 3). Data are representative of three (A–L) independent experiments. Error bars represent mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (K and L) and unpaired t test (B, D, F, H, and J).

Figure S4.

TRMT61A deficiency does not impact apoptosis or exhaustion of CD8 ⁺ T cells. (A–D) Flow cytometric analysis of exhaustion markers PD-1, TIM-3, and LAG-3 in polyclonal CD8⁺ T cells that chronically stimulated with constantly refreshed anti-CD3 antibody (n = 4–6). (E–H) Flow cytometric analysis of the granzyme B and IFN-γ production in polyclonal CD8⁺ T cells that chronically stimulated with constantly refreshed anti-CD3 antibody (n = 4–6). (I and J) Flow cytometric analysis of early and late apoptosis in splenic CD8⁺ T cells from WT and Trmt61a-cKO mice upon in vitro stimulation with CD3/CD28 antibodies for 24 h (n = 3). (K and L) m¹A dot blot of WT and Trmt61a-cKO CD8⁺ T cells after retroviral overexpression of GFP-control, TRMT61A-WT, and TRMT61A-dead plasmids. (M) Western blot of TRMT61A protein in WT and Trmt61a-cKO CD8⁺ T cells after retroviral overexpression of GFP-control, TRMT61A-WT, and TRMT61A-dead plasmids. (N) Schematic representation of the strategy for generating Trmt6-cKO mice, utilizing the Cre-loxP recombination system. (O) RT-qPCR quantification of Trmt6 mRNA levels in WT and Trmt6-cKO CD8⁺ T cells (n = 3). (P) Western blot of TRMT6 protein levels in WT and Trmt6-cKO CD8⁺ T cells. (Q and R) Assessment of naïve WT and Trmt6-cKO CD8⁺ T cell proliferation by CellTrace dilution following in vitro stimulation with CD3/CD28 antibodies for 72 h (n = 6). Data are representative of three (A–M and O–R) independent experiments. Error bars represent mean ± SEM; **P < 0.01, ***P < 0.001, ****P < 0.0001; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (J and L) and unpaired t test (B, D, F, H, O, and R). Source data are available for this figure: SourceData FS4.

Figure 5.

TRMT61A-mediated tRNA m ¹ A promotes CD8 ⁺ T cell cholesterol biosynthesis. (A and B) Volcano plots of RNA-seq data from Trmt61a-cKO naïve CD8⁺ T cells compared with WT naïve CD8⁺ T cells (A) and Trmt61a-cKO activated CD8⁺ T cells compared with WT activated CD8⁺ T cells (B), with cells activated by anti-CD3/CD28 antibodies for 48 h. Wald test (two-sided; adjustment method, Benjamini–Hochberg [BH]). (C) Summary of the number of DEGs from RNA-seq data. (D) KEGG enrichment analysis of downregulated transcripts in Trmt61a-cKO activated CD8⁺ T cells compared with WT activated CD8⁺ T cells. Hypergeometric test (one-sided; adjustment method, BH). (E) Heatmap showing the DEGs enriched in the steroid/cholesterol biosynthesis pathway between Trmt61a-cKO activated CD8⁺ T cells and WT activated CD8⁺ T cells. (F) Real-time PCR quantification of Nsdhl, Cyp51, Msmo1, Sqle, Dhcr24, Fdft1, Dhcr7, Hsd17b7, Tm7sf2, and Soat2 mRNA levels in WT and Trmt61a-cKO activated CD8⁺ T cells (n = 3). (G) KEGG enrichment analysis of downregulated proteins in Trmt61a-cKO activated CD8⁺ T cells compared with WT activated CD8⁺ T cells from proteomics data. Hypergeometric test (one-sided; adjustment method, BH) (n = 5). (H) Schematic diagram illustrating the downregulated transcripts of RNA-seq data (in both black and red) and downregulated proteins of proteomics data (in red) in the cholesterol biosynthesis pathway. (I) Violin plots displaying the signature scores of cholesterol metabolism in relation to TRMT61A expression (red) and TRMT61A repression (blue) in tumor-infiltrating CD8⁺ T cells from scRNA-seq data of CRC patients. Data are representative of three (F) independent experiments. Error bars represent mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (F).

Figure S5.

tRNA m ¹ A promotes antitumor immunity of CD8 ⁺ T cells by regulating the translation of ACLY rather than cholesterol esterification. (A and B) Detection of lipid droplets in naïve or activated WT and Trmt61a-cKO CD8⁺ T cells using Bodipy staining (n = 3). (C) Heatmap showing the content of different classes of lipids in WT and cKO CD8⁺ T cells (n = 5). (D) Histogram of absolute content of cholesterol esters (CE) in WT and cKO CD8⁺ T cells from quantitative lipidomics data. (E and F) Assessment of naïve WT and Trmt61a-cKO CD8⁺ T cell proliferation by CellTrace dilution, 72 h after retroviral overexpression of GFP-control, ACLY-WT, and ACLY-Mut plasmids (n = 3). (G and H) Flow cytometric analysis of granzyme B expression in GFP⁺CD8⁺ T cells from WT and Trmt61a^flox/floxCd4^Cre mice, after retroviral overexpression of GFP-control, ACLY-WT, and ACLY-Mut plasmids (n = 3). Data are representative of three (A, B, and E–H) independent experiments. Error bars represent mean ± SEM; *P < 0.05, **P < 0.01, ****P < 0.0001; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (B, F, and H).

Figure 6.

Cholesterol deficiency impairs the tumor-killing function of CD8 ⁺ T cells. (A–C) Detection of cholesterol levels in naïve and activated WT and Trmt61a-cKO CD8⁺ T cells using the Amplex Red assay (A, n = 6) and filipin III staining (B and C, n = 3). (D) CD8⁺ T cell killing assay in Trmt61a^flox/flox OT-I and Trmt61a^flox/floxCd4^Cre OT-I mice, with or without cholesterol supplementation in the culture medium. Effector:target = 1:1 (n = 4). (E and F) Flow cytometric analysis of granzyme B expression in CD8⁺ T cells from WT and Trmt61a^flox/floxCd4^Cre mice, with or without cholesterol supplementation during in vitro activation (n = 5). (G) Assessment of WT and Trmt61a-deficient naïve CD8⁺ T cell proliferation in vitro in the presence of anti-CD3/CD28 antibodies, with or without cholesterol supplementation (n = 3). (H and I) Detection of cholesterol levels in activated WT and Trmt61a-cKO CD8⁺ T cells using the filipin III staining, after retroviral overexpression of GFP-control, TRMT61A-WT, and TRMT61A-dead plasmids (n = 3). Data are representative of three (A–I) independent experiments. Error bars represent mean ± SEM; **P < 0.01; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (A, C, D, F, G, and I).

Figure 7.

tRNA m ¹ A modulates cholesterol biosynthesis through the translation of ACLY. (A) Venn diagram showing the intersecting genes that are unchanged at the transcriptional level and decreased at the protein level in Trmt61a-deficient CD8⁺ T cells. (B) KEGG enrichment analysis of the intersecting genes from A. Hypergeometric test (one-sided; adjustment method, BH). (C) Schematic representation of the mechanism by which ACLY generates acetyl-CoA from citrate for cholesterol biosynthesis, linking carbohydrate metabolism with cholesterol metabolism. (D and E) Western blot and RT-qPCR quantification of TRMT61A protein (D) and mRNA (E) levels in activated WT and Trmt61a-cKO CD8⁺ T cells (n = 3). (F) Schematic diagram of the Acly codon-switch assay. (G) The protein levels of ACLY in WT and Trmt61a-cKO CD8⁺ T cells were quantified by western blot, after retroviral overexpression of ACLY-WT and ACLY-codon-mutant (ACLY-Mut). (H and I) Detection of cholesterol levels in activated WT and Trmt61a-cKO CD8⁺ T cells using the filipin III staining, after retroviral overexpression of GFP-control, ACLY-WT, and ACLY-Mut plasmids (n = 3). Data are representative of three (D, E, and G–I) independent experiments. Error bars represent mean ± SEM; **P < 0.01; NS, nonsignificant. One-way ANOVA with Tukey’s multiple comparison test (I) and unpaired t test (E). Source data are available for this figure: SourceData F7.