m6A mRNA demethylase FTO regulates melanoma tumorigenicity and response to anti-PD-1 blockade

Abstract

Melanoma is one of the most deadly and therapy-resistant cancers. Here we show that N⁶-methyladenosine (m⁶A) mRNA demethylation by fat mass and obesity-associated protein (FTO) increases melanoma growth and decreases response to anti-PD-1 blockade immunotherapy. FTO level is increased in human melanoma and enhances melanoma tumorigenesis in mice. FTO is induced by metabolic starvation stress through the autophagy and NF-κB pathway. Knockdown of FTO increases m⁶A methylation in the critical protumorigenic melanoma cell-intrinsic genes including PD-1 (PDCD1), CXCR4, and SOX10, leading to increased RNA decay through the m⁶A reader YTHDF2. Knockdown of FTO sensitizes melanoma cells to interferon gamma (IFNγ) and sensitizes melanoma to anti-PD-1 treatment in mice, depending on adaptive immunity. Our findings demonstrate a crucial role of FTO as an m⁶A demethylase in promoting melanoma tumorigenesis and anti-PD-1 resistance, and suggest that the combination of FTO inhibition with anti-PD-1 blockade may reduce the resistance to immunotherapy in melanoma.

Fig. 1

FTO is upregulated in human melanoma. a Immunofluorescence staining of FTO and MART1 in normal human skin and melanoma tissues. Nuclei are counterstained with DAPI in blue. Scale bar, 50 μm. The arrows indicate representative melanocytes in normal human epidermis. b Percentage of tumors (in stacked column format) for each score of FTO. 0 (Negative), 1 (Weak), 2 (Medium), and 3 (Strong); *P < 0.05; **P < 0.01; ****P < 0.0001; ns, P > 0.05, not significant; Mann–Whitney U-test. c Immunoblot analysis of the protein levels of FTO, and β-actin (loading control) in normal melanocytes and melanoma cells

Fig. 2

Effect of FTO knockdown or forced FTO expression in melanoma cells. a Confirmation of knockdown or forced expression of FTO in Mel624, CHL-1, B16F10, and WM35 by immunoblot analysis. b Cell proliferation assay in Mel624, CHL-1, B16F10, and WM35 with shNC (negative control), shFTO (FTO knockdown), GFP (vector control), or GFP-FTO (FTO overexpression). c Cell migration assay with cells as in b. d Cell invasion assay with cells as in b. e Cell viability assay in cells as in b but placed in suspension. f Average tumor volume (mm³) of Mel624, CHL-1, B16F10, and WM cells as in b at different days after subcutaneous injection in nude mice or C57BL/6 mice (n = 3). g Final tumor weight from F (n = 3). Data are shown as mean ± S.D. (n ≥ 3). *P < 0.05; **P < 0.01; ***P < 0.001; Student’s t-test

Fig. 3

m⁶A enrichment in mRNA, and the function of m⁶A in melanoma cells. a, b m⁶A dot blot assays using total RNA of Mel624 and B16F10 cells with or without FTO knockdown. Methylene blue staining was used as a loading control. c, d m⁶A dot blot assays using poly(A) + mRNA of Mel624 and B16F10 cells with or without FTO knockdown. e Immunoblot analysis confirming forced overexpression of METTL3 and METTL14 in Mel624 cells. f m⁶A dot blot assays using poly(A) + mRNA in Mel624 with or without forced overexpression of METTL3 and METTL14. g Cell proliferation assay in Mel624 cells with or without forced overexpression of METTL3 and METTL14. h Cell migration assay in cells in G. i Cell invasion assay in cells in g. j–m Cell proliferation (j), migration (k), invasion (l), and cell viability (m) assays in suspension in shNC and shFTO of Mel624 with or without knockdown of METTL3 (M3) and METTL14 (M14). Data are shown as mean ± S.D. (n ≥ 3). *P < 0.05; **P < 0.01; ***P < 0.001; Student’s t-test

Fig. 4

Identification of potential targets of FTO in melanoma. a Schematic summary for analysis of FTO target genes and number of genes identified. b Relative expression of FTO, PD-1 (PDCD1), CD274, and CD47 in Mel624 cells with or without FTO knockdown. c Distribution of m⁶A peaks across the length of mRNA. Region of 5-untranslated region (5′UTR) was binned into 10 segments, coding region (CDS) was binned into 50 segments, and 3-untranslated region (3′UTR) was binned into 40 segments, and the percentage of m⁶A peaks that fall within each bin was determined. d The proportion of m⁶A peak distribution in the indicated regions across the entire set of mRNA transcripts (top) and the appearance of new m⁶A peaks (unique peaks in shFTO), or loss of existing m⁶A peaks (unique peaks in shNC) after FTO knockdown (bottom). e Number of m⁶A-modified genes identified in m⁶A-seq in shNC and shFTO Mel624 cells. Common m⁶A genes contain at least one common m⁶A peak, while unique m⁶A genes contain no common m⁶A peaks. f Top consensus m⁶A motif identified by HOMER with m⁶A peaks in Mel624 cells with or without FTO knockdown. g qPCR analysis of mRNA levels of genes in Mel624 cells with shNC or shFTO. h Gene-specific m⁶A qPCR analysis of gene-specific m⁶A enrichment in Mel624 cells with shNC or shFTO. Data are shown as mean ± S.E. (n ≥ 3). *P < 0.05; **P < 0.01; ***P < 0.001; Student’s t-test

Fig. 5

PD-1 (PDCD1), CXCR4, and SOX10 are critical target genes of FTO in melanoma cells. a, b Immunoblot analysis of PD-1, p-p70S6K, p70S6K, CXCR4, SOX10, FTO, and β-actin in Mel624 stable cells with or without knockdown (a) and forced overexpression (b) of FTO. c Immunoblot analysis of PD-1, p-p70S6K, p70S6K, FTO, and β-actin in Mel624 with or without FTO knockdown, and/or PD-1 (PDCD1) overexpression. d Cell proliferation assay in cells as in c. e Cell migration assay in cells as in c. f, g Cell proliferation (f) and migration analysis (g) in Mel624 with or without FTO knockdown, and/or CXCR4 overexpression. h, i Cell proliferation (h) and migration analysis (i) in Mel624 with or without FTO knockdown, and/or SOX10 overexpression. j–n qPCR analysis of the expression of PD-1 (PDCD1) (j), CXCR4 (k), SOX10 (l), CTSV (m), and NOP16 (n) in Mel624 cells with or without FTO knockdown, in combination with siRNA knockdown of both METTL3 and METTL14. o Immunoblot analysis of PD-1, CXCR4, SOX10, FTO, METTL3, METTL14, and GAPDH in cells as in j–n. Data are shown as mean ± S.D. (d–i), or mean ± S.E. (j–n) (n ≥ 3). *P < 0.05; **P < 0.01; ***P < 0.001; Student’s t-test

Fig. 6

FTO regulates its target gene expression through suppressing m⁶A/YTHDF2-mediated mRNA decay. a qPCR analysis of mRNA levels in Mel624 stable cells of shNC or shFTO. b–d qPCR analysis of the mRNA levels of PD-1 (PDCD1) (b), CXCR4 (c), and SOX10 (d) in Mel624 cells with or without FTO knockdown, in combination with siRNA knockdown of the m⁶A readers YTHDF1-3. e–g qPCR analysis of the mRNA stability of PD-1 (PDCD1) (e), CXCR4 (f), and SOX10 (g) in Mel624 cells with or without FTO knockdown. h-j qPCR analysis of the mRNA stability of PD-1 (PDCD1) (h), CXCR4 (i), and SOX10 (j) in Mel624 cells with or without FTO knockdown in combination with or without siRNA knockdown of YTHDF2. k, l Cell proliferation (k) and migration (l) analysis in Mel624 with or without knockdown or forced overexpression of YTHDF2. m Tumor growth of Mel624 cells with or without knockdown or forced overexpression of YTHDF2 after subcutaneous injection in nude mice (n = 3). Data are shown as mean ± S.E. (a–k), or mean ± S.D. (l and m) (n ≥ 3). *P < 0.05; **P < 0.01; Student’s t-test

Fig. 7

Metabolic stress induces FTO expression through NF-kB and autophagy. a qPCR analysis of FTO and PD-1 (PDCD1) mRNA levels in Mel624 cells cultured with control medium (10% FBS DMEM), 0.2% FBS DMEM, serum-free DMEM, Hanks’ balanced salt solution containing calcium and magnesium (HBSS), or a combination of DMEM and HBSS. b Immunoblot analysis of FTO, PD-1, p62, LC3-I/II (short and long exposure), and GAPDH in Mel624 cells treated as in a. c m⁶A dot blot assays using total RNA of Mel624 cultured with control medium or HBSS. Methylene blue staining was used as a loading control. d qPCR analysis of FTO and PD-1 (PDCD1) mRNA levels in Mel624 cells with or without knockdown of ATG5 or ATG7 and with or without metabolic stress. e Luciferase reporter analysis of NF-κB response element in Mel624 cells as in d. f qPCR analysis of FTO and PD-1 (PDCD1) mRNA levels in Mel624 cells with or without FTO knockdown treated with or without metabolic stress. g qPCR analysis of FTO and PD-1 (PDCD1) mRNA levels in Mel624 cells with or without siRNA knockdown of RELA treated as in e. Data are shown as mean ± S.E. (a, d, f, g), or mean ± S.D. (e) (n ≥ 3).*P < 0.05; **P < 0.01; ***P < 0.001; Student’s t-test

Fig. 8

Role of FTO in melanoma cell response to anti-PD-1 antibody in vivo or to interferon gamma (IFNγ) in vitro. a, b Tumor growth kinetics of control or FTO-knockdown B16F10 cells in C57BL/6 mice (a) and NSG mice (b) treated with anti-PD-1 or isotype control antibody (n = 4–6). *P < 0.05. c, d Flow cytometric analysis of the number of CD4+ and CD8+ TILs (c) and IFNγ-producing TILs (d) per gram of tumor tissue (n = 9). Mann–Whitney U-test. *P < 0.05. n.s., not significant. e Immunoblot analysis of FTO, ALKBH5, METTL3, METTL14, and GAPDH in Mel624 cells treated with or without IFNγ (100 ng/ml) for 24 h. f m⁶A dot blot assays using total RNA of Mel624 cells treated with or without IFNγ (100 ng/ml) for 24 h. g Apoptosis assay in Mel624 cells with or without FTO knockdown and treatment with or without IFNγ (50 ng/ml) for 48 h. h Apoptosis assay in Mel624 cells overexpressing vector, FTO WT, mutant 1, or mutant 2, and treated with or without IFNγ (50 ng/ml) for 48 h (n = 3). **P < 0.01, Student’s t-test. i–k Apoptosis assay in Mel624 cells with or without FTO knockdown overexpressing vector or PD-1 (i), CXCR4 (j), or SOX10 (k) treated with or without IFNγ (50 ng/ml) for 48 h (n = 3). *P < 0.05; **P < 0.01; ***P < 0.001; Student’s t-test. l Tumor growth of control or FTO-knockdown B16F10 cells in C57BL/6 mice treated with the anti-IFNγ antibody or isotype control IgG (n = 4–6). Data are shown as mean ± S.E. (a–e, and l), or mean ± S.D. (g–k) (n ≥ 3). *P < 0.05; **P < 0.01; ***P < 0.001; Student’s t-test. m Proposed model of the regulatory and functional role for FTO in melanoma pathogenesis and response to anti-PD-1 blockade