METTL3 Inhibits Antitumor Immunity by Targeting m6A-BHLHE41-CXCL1/CXCR2 Axis to Promote Colorectal CancerHuarong Chen, Yasi Pan, Qiming Zhou et al.|Gastroenterology|2022 Background & AimsN6-Methyladenosine (m6A) is the most prevalent RNA modification and recognized as an important epitranscriptomic mechanism in colorectal cancer (CRC). We aimed to exploit whether and how tumor-intrinsic m6A modification driven by methyltransferase like 3 (METTL3) can dictate the immune landscape of CRC.MethodsMettl3 knockout mice, CD34+ humanized mice, and different syngeneic mice models were used. Immune cell composition and cytokine level were analyzed by flow cytometry and Cytokine 23-Plex immunoassay, respectively. M6A sequencing and RNA sequencing were performed to identify downstream targets and pathways of METTL3. Human CRC specimens (n = 176) were used to evaluate correlation between METTL3 expression and myeloid-derived suppressor cell (MDSC) infiltration.ResultsWe demonstrated that silencing of METTL3 in CRC cells reduced MDSC accumulation to sustain activation and proliferation of CD4+ and CD8+ T cells, and eventually suppressed CRC in ApcMin/+Mettl3+/− mice, CD34+ humanized mice, and syngeneic mice models. Mechanistically, METTL3 activated the m6A-BHLHE41-CXCL1 axis by analysis of m6A sequencing, RNA sequencing, and cytokine arrays. METTL3 promoted BHLHE41 expression in an m6A-dependent manner, which subsequently induced CXCL1 transcription to enhance MDSC migration in vitro. However, the effect was negligible on BHLHE41 depletion, CXCL1 protein or CXCR2 inhibitor SB265610 administration, inferring that METTL3 promotes MDSC migration via BHLHE41-CXCL1/CXCR2. Consistently, depletion of MDSCs by anti-Gr1 antibody or SB265610 blocked the tumor-promoting effect of METTL3 in vivo. Importantly, targeting METTL3 by METTL3-single guide RNA or specific inhibitor potentiated the effect of anti–programmed cell death protein 1 (anti-PD1) treatment.ConclusionsOur study identifies METTL3 as a potential therapeutic target for CRC immunotherapy whose inhibition reverses immune suppression through the m6A-BHLHE41-CXCL1 axis. METTL3 inhibition plus anti-PD1 treatment shows promising antitumor efficacy against CRC. N6-Methyladenosine (m6A) is the most prevalent RNA modification and recognized as an important epitranscriptomic mechanism in colorectal cancer (CRC). We aimed to exploit whether and how tumor-intrinsic m6A modification driven by methyltransferase like 3 (METTL3) can dictate the immune landscape of CRC. Mettl3 knockout mice, CD34+ humanized mice, and different syngeneic mice models were used. Immune cell composition and cytokine level were analyzed by flow cytometry and Cytokine 23-Plex immunoassay, respectively. M6A sequencing and RNA sequencing were performed to identify downstream targets and pathways of METTL3. Human CRC specimens (n = 176) were used to evaluate correlation between METTL3 expression and myeloid-derived suppressor cell (MDSC) infiltration. We demonstrated that silencing of METTL3 in CRC cells reduced MDSC accumulation to sustain activation and proliferation of CD4+ and CD8+ T cells, and eventually suppressed CRC in ApcMin/+Mettl3+/− mice, CD34+ humanized mice, and syngeneic mice models. Mechanistically, METTL3 activated the m6A-BHLHE41-CXCL1 axis by analysis of m6A sequencing, RNA sequencing, and cytokine arrays. METTL3 promoted BHLHE41 expression in an m6A-dependent manner, which subsequently induced CXCL1 transcription to enhance MDSC migration in vitro. However, the effect was negligible on BHLHE41 depletion, CXCL1 protein or CXCR2 inhibitor SB265610 administration, inferring that METTL3 promotes MDSC migration via BHLHE41-CXCL1/CXCR2. Consistently, depletion of MDSCs by anti-Gr1 antibody or SB265610 blocked the tumor-promoting effect of METTL3 in vivo. Importantly, targeting METTL3 by METTL3-single guide RNA or specific inhibitor potentiated the effect of anti–programmed cell death protein 1 (anti-PD1) treatment. Our study identifies METTL3 as a potential therapeutic target for CRC immunotherapy whose inhibition reverses immune suppression through the m6A-BHLHE41-CXCL1 axis. METTL3 inhibition plus anti-PD1 treatment shows promising antitumor efficacy against CRC.
In Colorectal Cancer Cells With Mutant KRAS, SLC25A22-Mediated Glutaminolysis Reduces DNA Demethylation to Increase WNT Signaling, Stemness, and Drug ResistanceBackground & AimsMutant KRAS promotes glutaminolysis, a process that uses steps from the tricarboxylic cycle to convert glutamine to α-ketoglutarate and other molecules via glutaminase and SLC25A22. This results in inhibition of demethylases and epigenetic alterations in cells that increase proliferation and stem cell features. We investigated whether mutant KRAS-mediated glutaminolysis affects the epigenomes and activities of colorectal cancer (CRC) cells.MethodsWe created ApcminKrasG12D mice with intestine-specific knockout of SLC25A22 (ApcminKrasG12DSLC25A22fl/fl mice). Intestine tissues were collected and analyzed by histology, immunohistochemistry, and DNA methylation assays; organoids were derived and studied for stem cell features, along with organoids derived from 2 human colorectal tumor specimens. Colon epithelial cells (1CT) and CRC cells (DLD1, DKS8, HKE3, and HCT116) that expressed mutant KRAS, with or without knockdown of SLC25A22 or other proteins, were deprived of glutamine or glucose and assayed for proliferation, colony formation, glucose or glutamine consumption, and apoptosis; gene expression patterns were analyzed by RNA sequencing, proteins by immunoblots, and metabolites by liquid chromatography–mass spectrometry, with [U-13C5]-glutamine as a tracer. Cells and organoids with knocked down, knocked out, or overexpressed proteins were analyzed for DNA methylation at CpG sites using arrays. We performed immunohistochemical analyses of colorectal tumor samples from 130 patients in Hong Kong (57 with KRAS mutations) and Kaplan-Meier analyses of survival. We analyzed gene expression levels of colorectal tumor samples in The Cancer Genome Atlas.ResultsCRC cells that express activated KRAS required glutamine for survival, and rapidly incorporated it into the tricarboxylic cycle (glutaminolysis); this process required SLC25A22. Cells incubated with succinate and non-essential amino acids could proliferate under glutamine-free conditions. Mutant KRAS cells maintained a low ratio of α-ketoglutarate to succinate, resulting in reduced 5-hydroxymethylcytosine—a marker of DNA demethylation, and hypermethylation at CpG sites. Many of the hypermethylated genes were in the WNT signaling pathway and at the protocadherin gene cluster on chromosome 5q31. CRC cells without mutant KRAS, or with mutant KRAS and knockout of SLC25A22, expressed protocadherin genes (PCDHAC2, PCDHB7, PCDHB15, PCDHGA1, and PCDHGA6)—DNA was not methylated at these loci. Expression of the protocadherin genes reduced WNT signaling to β-catenin and expression of the stem cell marker LGR5. ApcminKrasG12DSLC25A22fl/fl mice developed fewer colon tumors than ApcminKrasG12D mice (P < .01). Organoids from ApcminKrasG12DSLC25A22fl/fl mice had reduced expression of LGR5 and other markers of stemness compared with organoids derived from ApcminKrasG12D mice. Knockdown of SLC25A22 in human colorectal tumor organoids reduced clonogenicity. Knockdown of lysine demethylases, or succinate supplementation, restored expression of LGR5 to SLC25A22-knockout CRC cells. Knockout of SLC25A22 in CRC cells that express mutant KRAS increased their sensitivity to 5-fluorouacil. Level of SLC25A22 correlated with levels of LGR5, nuclear β-catenin, and a stem cell-associated gene expression pattern in human colorectal tumors with mutations in KRAS and reduced survival times of patients.ConclusionsIn CRC cells that express activated KRAS, SLC25A22 promotes accumulation of succinate, resulting in increased DNA methylation, activation of WNT signaling to β-catenin, increased expression of LGR5, proliferation, stem cell features, and resistance to 5-fluorouacil. Strategies to disrupt this pathway might be developed for treatment of CRC. Mutant KRAS promotes glutaminolysis, a process that uses steps from the tricarboxylic cycle to convert glutamine to α-ketoglutarate and other molecules via glutaminase and SLC25A22. This results in inhibition of demethylases and epigenetic alterations in cells that increase proliferation and stem cell features. We investigated whether mutant KRAS-mediated glutaminolysis affects the epigenomes and activities of colorectal cancer (CRC) cells. We created ApcminKrasG12D mice with intestine-specific knockout of SLC25A22 (ApcminKrasG12DSLC25A22fl/fl mice). Intestine tissues were collected and analyzed by histology, immunohistochemistry, and DNA methylation assays; organoids were derived and studied for stem cell features, along with organoids derived from 2 human colorectal tumor specimens. Colon epithelial cells (1CT) and CRC cells (DLD1, DKS8, HKE3, and HCT116) that expressed mutant KRAS, with or without knockdown of SLC25A22 or other proteins, were deprived of glutamine or glucose and assayed for proliferation, colony formation, glucose or glutamine consumption, and apoptosis; gene expression patterns were analyzed by RNA sequencing, proteins by immunoblots, and metabolites by liquid chromatography–mass spectrometry, with [U-13C5]-glutamine as a tracer. Cells and organoids with knocked down, knocked out, or overexpressed proteins were analyzed for DNA methylation at CpG sites using arrays. We performed immunohistochemical analyses of colorectal tumor samples from 130 patients in Hong Kong (57 with KRAS mutations) and Kaplan-Meier analyses of survival. We analyzed gene expression levels of colorectal tumor samples in The Cancer Genome Atlas. CRC cells that express activated KRAS required glutamine for survival, and rapidly incorporated it into the tricarboxylic cycle (glutaminolysis); this process required SLC25A22. Cells incubated with succinate and non-essential amino acids could proliferate under glutamine-free conditions. Mutant KRAS cells maintained a low ratio of α-ketoglutarate to succinate, resulting in reduced 5-hydroxymethylcytosine—a marker of DNA demethylation, and hypermethylation at CpG sites. Many of the hypermethylated genes were in the WNT signaling pathway and at the protocadherin gene cluster on chromosome 5q31. CRC cells without mutant KRAS, or with mutant KRAS and knockout of SLC25A22, expressed protocadherin genes (PCDHAC2, PCDHB7, PCDHB15, PCDHGA1, and PCDHGA6)—DNA was not methylated at these loci. Expression of the protocadherin genes reduced WNT signaling to β-catenin and expression of the stem cell marker LGR5. ApcminKrasG12DSLC25A22fl/fl mice developed fewer colon tumors than ApcminKrasG12D mice (P < .01). Organoids from ApcminKrasG12DSLC25A22fl/fl mice had reduced expression of LGR5 and other markers of stemness compared with organoids derived from ApcminKrasG12D mice. Knockdown of SLC25A22 in human colorectal tumor organoids reduced clonogenicity. Knockdown of lysine demethylases, or succinate supplementation, restored expression of LGR5 to SLC25A22-knockout CRC cells. Knockout of SLC25A22 in CRC cells that express mutant KRAS increased their sensitivity to 5-fluorouacil. Level of SLC25A22 correlated with levels of LGR5, nuclear β-catenin, and a stem cell-associated gene expression pattern in human colorectal tumors with mutations in KRAS and reduced survival times of patients. In CRC cells that express activated KRAS, SLC25A22 promotes accumulation of succinate, resulting in increased DNA methylation, activation of WNT signaling to β-catenin, increased expression of LGR5, proliferation, stem cell features, and resistance to 5-fluorouacil. Strategies to disrupt this pathway might be developed for treatment of CRC.