METTL3-mediated m6A RNA methylation promotes the anti-tumour immunity of natural killer cells

Abstract

Natural killer (NK) cells exert critical roles in anti-tumor immunity but how their functions are regulated by epitranscriptional modification (e.g., N⁶-methyladenosine (m⁶A) methylation) is unclear. Here we report decreased expression of the m⁶A "writer" METTL3 in tumor-infiltrating NK cells, and a positive correlation between protein expression levels of METTL3 and effector molecules in NK cells. Deletion of Mettl3 in NK cells alters the homeostasis of NK cells and inhibits NK cell infiltration and function in the tumor microenvironment, leading to accelerated tumor development and shortened survival in mice. The gene encoding SHP-2 is m⁶A modified, and its protein expression is decreased in METTL3-deficient NK cells. Reduced SHP-2 activity renders NK cells hyporesponsive to IL-15, which is associated with suppressed activation of the AKT and MAPK signaling pathway in METTL3-deficient NK cells. These findings show that m⁶A methylation safeguards the homeostasis and tumor immunosurveillance function of NK cells.

Fig. 1. METTL3 is positively correlated with effector functions of NK cells.

a Heatmap showing the NK cell effector function-related genes in tumor-infiltrating NK cells compared with those in normal liver NK cells from a published dataset (GSE120123). b Correlation between METTL3 TPM (transcripts per kilobase of exon model per million mapped reads) and module TPM of NK effector function (NCR1, EOMES, TBX21, CD69, DNAM1, KLRK1, PERF, IFNG, and GZMB) in lung squamous cell carcinoma (LUSC), lung adenocarcinoma (LUAD), thyroid carcinoma (THCA), uterine carcinosarcoma (UCS), and prostate adenocarcinoma (PRAD) tissues. Spearman’s correlation coefficient (R) and P values are shown. c A representative histogram showing METTL3 expression in peripheral blood NK cells from a healthy human. d–h WT C57BL/6 mice were harvested for analyses of lung NK cells 3–4 weeks after intravenous injection of 5 × 10⁵ B16/F10 cells. d Representative plots showing expression of NK1.1 and CD3 in lungs. e Percentages of NK cells from control (n = 5) or B16/F10-bearing (n = 3) mice. f A representative histogram (left) and MFI (right) of METTL3 expression in NK cells from control (n = 5) or B16/F10-bearing (n = 3) mice. g Representative histograms of indicated-molecule expression in NK cells from control or B16/F10-bearing mice. For detection of IFN-γ, lung mononuclear cells were isolated and stimulated with phorbol 12-myristate 13-acetate (PMA) and ionomycin for 4 h for flow-cytometric analyses. h Correlation between the MFI of METTL3 expression and MFI of the indicated molecule expression in NK cells, or NK cell numbers (n = 6 for the control group; n = 5 (IFN-γ) or 17 (other molecules) for B16/F10 group). i–m WT C57BL/6 mice were harvested for analyses of liver NK cells 3–4 weeks after intrasplenic injection of 2 × 10⁵ MC38 cells. i Representative plots showing expression of NK1.1 and CD3 in livers. j Percentages of liver NK cells from control (n = 5) or MC38-bearing (n = 3) mice. k A representative histogram (left) and MFI (right) of METTL3 expression in liver NK cells from control (n = 5) or MC38-bearing (n = 3) mice. l Representative histograms of indicated-molecule expression in liver NK cells from control or MC38-bearing mice. For detection of IFN-γ, liver mononuclear cells were isolated and stimulated with PMA and ionomycin for 4 h for flow-cytometric analyses. m Correlation between the MFI of METTL3 expression and MFI of the indicated molecule expression in NK cells, or NK cell numbers (n = 12 for the control group; n = 5 for MC38 group). Each symbol represents an individual mouse (e, f, h, j, k, m). Data are the mean ± SEM (unpaired two-tailed t test (e, f, j, k), two-tailed Pearson correlation-test (h, m)). Source data are provided as a Source Data file. Data represent at least two independent experiments (c, d–m).

Fig. 2. Mettl3 deletion in NK cells fails to control tumor development.

a Imaging of lungs (left) and tumor nodules in lungs (right) from WT (n = 6) or cKO (n = 4) mice 16 days after intravenous injection of 5 × 10⁵ B16/F10 cells. b Imaging of livers (left), and tumor nodules in livers (right) from WT (n = 6) or cKO (n = 4) mice 16 days after intravenous injection of 5 × 10⁵ B16/F10 cells. c Representative hematoxylin and eosin (H&E) staining images of the liver (top) and lung samples (bottom) from WT mice or cKO mice after intravenous injection of 2 × 10⁵ B16/F10 cells. Scale bars represent 1 mm. d Survival of WT mice (n = 7) and cKO mice (n = 5) after intrasplenic injection of 2 × 10⁵ MC38 cells. e Survival of WT mice (n = 8) and cKO mice (n = 12) after intrasplenic injection of 2 × 10⁵ B16/F10 cells. f Survival of WT mice (n = 9) and cKO mice (n = 7) after intravenous injection of 5 × 10⁵ MC38 cells. g Survival of WT mice (n = 5) and cKO mice (n = 5) after intravenous injection of 2.5 × 10⁵ B16/F10 cells. h WT mice or cKO mice were co-transferred with 5 × 10⁶ CFSE-labeled RMA cells and 1.5 × 10⁷ CTV labeled RMA-S cells, and splenocytes were harvested 16 h later. i Representative plots (left) showing expression of CTV and CFSE in the spleen, and a statistical graph showing the ratio of RMA-S cells: RMA cells (n = 7) or cKO (n = 5) mice in (i). Each symbol represents an individual mouse (a, b, d, e, f, g, i). Data are the mean ± SEM (unpaired two-tailed t test (a, b, i) or two-tailed log-rank (Mantel–Cox) test (d–g)). Source data are provided as a Source Data file. Data represent at least two independent experiments (a–d, g-i) or are pooled from two independent experiments (e, f).

Fig. 3. METTL3 deficiency disturbs NK cell homeostasis at a steady state.

a Representative plots showing expression of NK1.1 and CD3/CD19 in mononuclear cells from the indicated organs of WT and cKO mice. b Percentages and absolute numbers of NK cells in the indicated organs from WT and cKO mice (n = 5/group). c Percentages of NK cells at different stages in the indicated organs of WT and cKO mice (n = 5/group). (Stage 1: CD27⁻ CD11b⁻; stage 2: CD27⁺ CD11b⁻; stage 3: CD27⁺ CD11b⁺; stage 4: CD27⁻ CD11b⁺). d, e Representative histograms (d) and MFI (e) of KLRG1 expression in NK cells from the indicated organs of WT (n = 6) and cKO (n = 4) mice. f, g Representative histograms (f) and MFI (g) of expression of the indicated molecules in splenic NK cells from WT (n = 5 for 2B4 and Ly49D, n = 6 for Ly49H, and n = 4 for other molecules) and cKO mice (n = 3 for Ly49H, n = 4 for 2B4 and Ly49D, and n = 5 for other molecules). Each symbol represents an individual mouse (b, c, e, g). Data are the mean ± SEM (ns, not significant; unpaired two-tailed t-test). Source data are provided as a Source Data file. Data represent two (d, e) or at least three independent experiments (a–c, f, g).

Fig. 4. METTL3 supports NK cells accumulation and effector function in the TME.

a–d WT mice and cKO mice were harvested for flow-cytometric analyses 16 days after intravenous injection with 5 × 10⁵ B16/F10 cells or PBS. a, c Representative plots showing expression of NK1.1 and CD3 in the liver (a) or lung (c) mononuclear cells from WT and cKO mice. b, d Percentages or absolute numbers of the liver (b) or lung (d) NK cells from WT mice (n = 6/group) and cKO mice (n = 6 for the control group; n = 4 for B16/F10 group). e Immunofluorescence staining of lungs with anti-NKp46 (red) and DAPI (blue) from WT or cKO mice 25 days after intravenous injection with 2 × 10⁵ B16/F10 cells. Scale bar, 50 μm. The arrows indicate NK cells. f NKp46-positive cell numbers per field based on (e) (n = 6 for WT group; n = 5 for cKO group). g–l WT mice and cKO mice were harvested for flow-cytometric analyses 23 days after intravenous injection with 2 × 10⁵ B16/F10 cells or PBS. Liver mononuclear cells were isolated and stimulated with PMA and ionomycin for 4 h for flow-cytometric analyses. Representative plots and percentages for the expression of CD107a (g, h), GzmB (i, j), and IFN-γ (k, l) expression among liver NK cells from WT (n = 5 for the control group; n = 6 for B16/F10 group) and cKO mice (n = 5 for the control group; n = 4 for B16/F10 group). Each symbol represents an individual mouse (b, d, h, j, l) or individual field of the microscope (f). Data are the mean ± SEM (ns, not significant; unpaired two-tailed t test (f) or one-way ANOVA (b, d, h, j, l)). Source data are provided as a Source Data file. Data represent at least two independent experiments (a–l).

Fig. 5. METTL3-deficient NK cells are hyporesponsive to IL-15 stimulation.

a Fold changes of cytokine expression in tumor-bearing liver tissue (n = 4) from WT mice injected with 2 × 10⁵ MC38 cells into the spleen 22 days previously, compared with normal liver tissue (n = 4), were determined by RT-qPCR, normalized to β-actin expression. b–d WT (n = 5/group) or cKO mice (n = 6 for pLIVE group; n = 4 for the pLIVE-IL-15 group) that received a hydrodynamic injection of pLIVE-IL-15 or pLIVE control plasmid 1 month previously were harvested for flow-cytometric analyses. b Representative plots showing expression of NK1.1 and CD3 in liver mononuclear cells from WT and cKO mice. c Percentages or absolute numbers of liver NK cells from WT and cKO mice. d Percentages of CD107a-, GzmB-, IFN-γ-, and Perforin-positive NK cells among total NK cells in splenocytes 4 h after stimulated with PMA and ionomycin. e–h Splenocytes from WT or cKO mice were cultured with the presence or absence of rmIL-15 (100 ng/mL) for the indicated times, and then analyzed by flow cytometry. Percentages (e) and absolute numbers (f) of NK cells among splenocytes are shown (n = 4/group). g Percentages of Ki67-positive cells among NK cells (n = 3/group). h Percentages of Annexin V-positive cells among NK cells (n = 4/group). i, j Splenocytes from WT mice or cKO mice were stimulated with rmIL-15 (10 ng/mL) for 7 days, and then collected for stimulation with PMA and ionomycin, followed by flow cytometric analyses. Representative plots (i) and percentages (j) of expression of GzmB, IFN-γ, and TNF in NK cells from WT and cKO mice (n = 4/group). k Protein concentration of GzmB and IFN-γ in the supernatant of cultured NK cells in the condition of rmIL-15 (50 ng/mL) for 9 days from WT (n = 4) or cKO (n = 6) mice. Each symbol represents an individual mouse (c, d) or an individual well of a cell culture plate (e–h, j, k). Data are the mean ± SEM (ns, not significant; unpaired two-tailed t test (a, j, k) or one-way ANOVA (c, d) or two-way ANOVA (e–h)). Source data are provided as a Source Data file. Data represent at least two independent experiments (a–k).

Fig. 6. The IL-15-AKT pathway is impaired in METTL3-deficient NK cells.

a, b Representative histograms (a) and MFI (b) of CD122 and CD132 expression in splenic NK cells from WT (n = 6 for CD122, n = 4 for CD132) and cKO (n = 3 for CD122, n = 4 for CD132) mice. c, e, f Immunoblotting of indicated molecules in splenic NK cells upon stimulation with rmIL-15 (50 ng/mL) for 1 h. The NK cells were sorted from WT and cKO mice that received hydrodynamic injection of the pLIVE-IL-15 plasmid or pLIVE control plasmid 1–2 months previously. d Representative histogram (left) and MFI (right) of p-AKT (S473) expression in splenic NK cells upon stimulation with rmIL-15 (50 ng/mL) for 1 h (n = 4 for WT group; n = 6 for cKO group). g, h Splenic NK cells from WT or cKO mice were cultured with rmIL-15 overnight (10 ng/mL), and then the indicated compounds were added into each well at the indicated times. g The oxygen consumption rate was measured for NK cells. h Statistical graphs showing the maximal respiration and spare respiratory capacity of NK cells (n = 7 for WT group; n = 5 for cKO group). Each symbol represents an individual mouse (b, d) or an individual well of a cell culture plate (h). Data are the mean ± SEM (ns, not significant; unpaired two-tailed t test (b, d, h)). Source data are provided as a Source Data file. Data represent at least two independent experiments (a–h).

Fig. 7. METTL3-mediated m⁶A modification promotes SHP-2 protein expression.

a Venn diagram showing the genes or proteins modified by m⁶A, bound by METTL3, and affected by METTL3 deficiency. b Cellular locations of the 23 proteins in (a) using information obtained from Uniprot. c Gene track shows Ptpn11 locus as a histogram of normalized reads (y-axis) plotted by genome position (x-axis). d Immunoblotting of SHP-2 and METTL3 in purified splenic NK cells from WT and cKO mice. e Immunoblotting of indicated proteins in purified splenic NK cells after stimulated with rmIL15 (50 ng/mL) in the presence or absence of SHP099 (4 μM) for 1 h. f   The splenocytes were cultured with rmIL-15 (50 ng/mL) in vitro for 3 days, followed by SHP099 (4 μM) treatment for 1 h, and then collected for flow-cytometric analyses of the indicated molecule expression in NK cells (n = 5/group). g, h Splenocytes from WT mice were stimulated with rmIL-15 (50 ng/mL) in the absence or presence of SHP099 (4 μM) for 3 days (n = 6/group). g Representative plots (left) showing the expression of CD3 and NK1.1 in cultured splenocytes or Ki67 expression in splenic NK cells; statistical graphs (right) showing the percentages of NK cells or Ki67⁺ NK cells. (h) Representative plots (left) and percentages (right) for the expression of GzmB, IFN-γ, TNF in splenic NK cells 4 hours after being stimulated with PMA and ionomycin. Each symbol represents an individual well of a cell culture plate (f–h). Data are the mean ± SEM (ns, not significant; unpaired two-tailed t test). Source data are provided as a Source Data file. Data represent one (f) or at least two independent experiments (d, e, g, h).