Abstract
The RNA modification N6-methyladenosine (m6A) plays a key role in RNA processing. It is catalysed by the RNA methyltransferase complex (MTC) which includes METTL3, METTL14 and CBLL1. Recently, a METTL3 inhibitor demonstrated promising preclinical results in several cancer types, yet the therapeutic potential of targeting m6A in breast cancer (BCa) remains poorly understood. Utilising a large BCa cohort, we identified that increased METTL14 and CBLL1 expression was associated with a more favourable prognosis, whereas increased METTL3 expression was associated with poorer patient outcomes in Triple Negative BCa (TNBC). Using siRNA depletion, we identified distinct METTL3, METTL14 and CBLL1 regulated gene networks in BCa cell lines. METTL3 inhibition reduced proliferation and invasion of BCa cell lines and induced an immune activation transcriptional signature. These results provide insight into the clinical functions of METTL3, METTL14 and CBLL1 in BCa and support the therapeutic potential of targeting METTL3 in BCa, particularly in TNBC.
Similar content being viewed by others
Introduction
Breast cancer (BCa) is the most commonly diagnosed malignancy and the leading cause of cancer-related deaths in women globally1,2. BCa tumours are sub-grouped based on their hormone receptor status, human epidermal growth factor receptor 2 (HER2) status and intrinsic molecular subtypes3. Hormone receptor (estrogen receptor (ER) and progesterone receptor (PR))-positive luminal tumours, accounting for 70-80% of all BCa, are treated with hormone therapy but still a proportion recurs4,5. Triple negative (ER, PR and HER2 negative) BCa (TNBC), which is diagnosed in ~10% of BCa patients, is associated with a poor prognosis due to an aggressive phenotype and a lack of specific therapeutic targets3,6,7,8,9. Therefore, there is an urgent need for novel therapeutic approaches for BCa.
The methylation of the N6 position of adenosine (m6A) is the most prevalent RNA modification in eukaryotic cells10. Enzymatic writers and erasers of m6A catalyse the dynamic and reversible addition and removal of the methyl group, whilst readers bind to influence RNA processing, including mRNA splicing, translation, localisation and stability11,12. The m6A methyltransferase complex (MTC) (writer complex) comprises the catalytic subunit METTL3, bound in a stable complex with proteins including METTL14, CBLL1 (HAKAI), WTAP, VIRMA (KIAA1429), ZC3H13, RBM15, and its paralogue RBM15B13,14,15,16,17.
The MTC components have been reported to have complex roles in BCa and the relevance of the RNA MTC and m6A as therapeutic targets in BCa remains poorly understood. Previous studies examining the functions of METTL3 and METTL14 in BCa have reported both tumour suppressor and oncogenic roles and have implicated METTL3 and METTL14 in several mechanisms18,19,20,21,22,23,24. CBLL1 is an adaptor protein that binds to the METTL3-METTL14 heterodimer and is believed to be required for full m6A methylation activity17,25. Conflicting findings have also been reported in studies investigating CBLL1 in BCa. CBLL1 has been positively associated with increased invasive potential and metastasis of BCa cells26,27, whilst other authors reported an association between higher CBLL1 expression and better relapse-free and overall survival (OS) in BCa patients28.
Recently, a first-in-class METTL3 inhibitor has been reported29. This small molecule inhibitor, STM2457, is highly selective for the S-adenosyl methionine (SAM) binding site of METTL3 and was found to reduce m6A levels and have promising effects in several cancer contexts30,31,32,33. STC-15, a potent derivative of STM2457, has entered Phase I clinical trial for solid tumours (NCT05584111). METTL3 inhibition has been shown to induce an anti-cancer immune response and it has been suggested, therefore, that METTL3 inhibitors may complement immunomodulatory therapies, including anti-PD1 therapy34.
In this study, we evaluated the expression and clinical relevance of METTL3, METTL14 and CBLL1 in large publicly available BCa datasets (RNA level) and in a well-characterised operable BCa patient cohort (~2500 patients) by immunohistochemical (IHC) staining. To provide insights into the functional effects of RNA methylation in BCa, we examined the effects of siRNA knockdown of METTL3, METTL14 and CBLL1 on the transcriptome of BCa cell lines by RNA-sequencing (RNA-seq). Finally, the effects of the STM2457 METTL3 inhibitor on BCa cell phenotype and transcriptome were assessed.
Results
METTL3, METTL14 and CBLL1 expression in breast cancer
Within the TCGA Breast Invasive Carcinoma PanCancer Atlas dataset (n = 996) mutations of METTL3, METTL14 and CBLL1 were uncommon, with amplifications being the most common genetic alteration (Supplementary Fig. 1A). Comparison of tumour and normal mRNA expression using TCGA GDC data (n = 1284) revealed no significant difference for METTL3 or CBLL1 expression, whereas METTL14 expression was lower in tumour tissue compared with normal breast tissue (p < 0.0001) (Supplementary Fig. 1B-D). KM plots revealed increased METTL14 and CBLL1 expression correlated with increased OS whereas METTL3 showed no significant correlation (Supplementary Fig. 1 E-G). Regarding protein expression, varying intensities of METTL3, METTL14 and CBLL1 staining was observed in the nuclei of tumour epithelial cells by IHC (Fig. 1). H-scores obtained for METTL3 (n = 2132), METTL14 (n = 2321) and CBLL1 (n = 1524) were correlated with clinicopathological data.
Representative photomicrographs of (A) low and (D) high level METTL3 expression; (B) low and (E) high level METTL14 expression; and (C) low and (F) high level CBLL1 expression. H-scores assigned to intact cores with tumour coverage >10%. Distribution of H-scores with red line denoting median: (G) METTL3, median H-score 105 (n = 2132); (H) METTL14, median H-score 120 (n = 2321), and (I) CBLL1, median H-score 70 (n = 1524).
Clinicopathological correlations associated with METTL3, METTL14 and CBLL1 expression
High METTL3 expression was significantly associated with lower grade, lower pleomorphic score, low mitotic score, ER and PR positivity and non-TNBC status (all p < 0.05; Table 1). Correlations with molecular class revealed high nuclear METTL3 expression to be associated with luminal A and luminal B, whilst lower METTL3 expression was associated with TNBC and HER2 enriched molecular subtypes (Table 1).
Higher nuclear METTL14 expression was associated with smaller tumour size (<2cm), older age (>50 years), lower grade, with lower pleomorphism, high tubule formation and lower mitosis scores and a more favourable Nottingham Prognostic Index (NPI) group (p < 0.0001; Table 1). Higher expression was associated with non-TNBC status, with ER and PR positivity and HER2 negativity, presence of ductal carcinoma in situ (DCIS) and lobular carcinoma in situ (LCIS), lack of lymphovascular invasion (LVI) and lower Ki67 proliferation index. Lower METTL14 expression was associated with TNBC and HER2 enriched molecular subtypes (p < 0.0001; Table 1).
Nuclear CBLL1 expression was significantly higher in patients with a lower grade at diagnosis, a favourable NPI prognostic group, ER positive status, and with non-TNBC status (p < 0.0001; Table 1). Similarly, higher nuclear CBLL1 was significantly associated with a smaller tumour size, lower Ki67 labelling index, PR positivity and HER2 negativity, as well as histological tumour types of a more favourable prognosis (p < 0.0001; Table 1).
METTL3 expression negatively associated and CBLL1 and METTL14 expression positively associated with increased patient survival
To determine associations between METTL3, METTL14 and CBLL1 expression and survival outcomes, Kaplan-Meier estimates were performed. Considering the cohort as a whole, there was no significant association between METTL3 and outcomes for breast cancer-specific survival (BCSS), time to distant metastasis (TTDM) and disease-free survival (DFS) (Supplementary Fig. 2). However, in TNBC patients, high METTL3 expression was associated with decreased 15-year BCSS, TTDM and DFS (Fig. 2A–C). Higher nuclear METTL14 and higher nuclear CBLL1 expression were both significantly associated with longer BCSS, DFS and TTDM across the entire patient cohort (p < 0.001; Fig. 2D–I). This was consistent with transcriptomic data which showed a significant association between higher CBLL1 and METTL14 expression and increased OS (n = 2976) (Supplementary Fig. 1).
IHC staining of BCa patient samples was assessed by H-score, divided by median into high and low expression groups and correlated with breast cancer-specific survival (BCSS), time to distant metastasis (TTDM) and disease-free survival (DFS). METTL3 staining in TNBC patients only, low: ≤105, high: ≥106. METTL14 staining in all BCa patients, low: ≤120, high: ≥121. CBLL1 staining in all BCa patients, low: ≤70, high: ≥71. Correlations: (A) METTL3 and BCSS in TNBC (n = 329, events at 15 years, low: 56/215, high: 44/114), (B) METTL3 and TTDM in TNBC (n = 329, events at 15 years, low: 58/215, high: 51/114), (C) METTL3 and DFS in TNBC (n = 329, events at 15 years, low: 73/215, high: 57/114), (D) METTL14 and BCSS (n = 2297, events at 15 years, low: 324/1212, high: 223/1085), (E) METTL14 and TTDM (n = 2297, events at 15 years, low: 379/1212, high: 269/1085), (F) METTL14 and DFS (n = 2297, events at 15 years, low: 494/1212, high: 366/1085). (G) CBLL1 and BCSS (n = 1513, events at 15 years, low: 263/851, high: 127/662), (H) CBLL1 and TTDM (n = 1513, events at 15 years, low: 294/851, high: 164/662), (I) CBLL1 and DFS (n = 1513, events at 15 years low: 374/851, high: 235/662). Correlated using Kaplan Meier estimate and analysed by log-rank test. Combined multivariate Cox regression analysis of METTL3 nuclear expression, tumour stage, tumour size, tumour grade, patient age, ER status and LVI with 15-year (J) BCSS, (K) TTDM and (L) DFS (n = 2106). Combined multivariate Cox regression analysis of METTL14 nuclear expression, tumour stage, tumour size, patient age, ER status and LVI with 15-year (M) BCSS, (N) TTDM and (O) DFS (n = 2296). Combined multivariate Cox regression analysis of CBLL1 nuclear expression, tumour stage, tumour size, tumour grade, patient age, ER status and LVI with 15-year (P) BCSS, (Q) TTDM and (R) DFS (n = 1507). * = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, ns = not significant.
Multivariate Cox regression analyses were conducted to assess the independent prognostic value of METTL3, METTL14 and CBLL1 expression whilst accounting for the potential confounding variables of nodal stage, tumour size, patient age, ER status and LVI (Fig. 2). High nuclear METTL3 expression was significantly associated with shorter BCSS, TTDM and DFS independent of patient age, ER status, LVI, nodal stage, grade and size in both the whole cohort (P < 0.05, Fig. 2J–L) and in the TNBC subgroup of patients (P < 0.05, Supplementary Fig. 2). In contrast, nuclear CBLL1 expression was an independent positive prognostic factor for BCSS, TTDM and DFS (P < 0.05, Fig. 2P–R). Nuclear METTL14 expression was not an independent prognostic factor (Fig. 2M–O).
Correlations between METTL3, METTL14 and CBLL1 expression and other functionally relevant proteins
Expression of METTL3, METTL14 and CBLL1 was subsequently correlated with expression of a number of other functionally related proteins previously analysed by IHC in the Nottingham BCa TMA. A positive correlation was observed between METTL3, METTL14 and CBLL1. Nuclear METTL3, METTL14 and CBLL1 were all significantly positively correlated with ERα and ER-regulated genes including AR, and FOXA1, E-cadherin, C-MYC and BRCA1 expression (p < 0.0001; Supplementary Table 1).
METTL3, METTL14 and CBLL1 regulate distinct transcriptional networks
To assess the role of METTL3, METTL14 and CBLL1 in BCa, siRNA-mediated knockdowns were performed in cell lines MCF7 and MDA-MB-231, confirmed by western blot and differential gene expression determined by RNA-seq (Fig. 3, Supplementary Figs. 3–14). It was also noted that METTL3, METTL14 and CBLL1 protein expression was upregulated by estrogen signalling in MCF7 (Fig. 3A, F and K). Comparison of differentially expressed genes revealed that METTL3, METTL14 and CBLL1 regulated distinct gene networks in MDA-MB-231 and MCF7 under estrogen depletion and with estrogen treatment (Fig. 4A–B, Supplementary Fig. 15).
siRNA-mediated knockdown of METTL3, METTL14 and CBLL1 in MCF7 and MDA-MB-231 cells (with subsequent treatment for 72 hours with vehicle or E2 in MCF7 cells) was confirmed by western blot with β-Actin loading control. Differential gene expression was analysed following RNA-seq. A METTL3 expression in MCF7 cells following METTL3 siRNA-mediated knockdown -/+ E2 (n = 3). B Differential gene expression compared between vehicle-treated siSCR control and METTL3-depleted cells (n = 3) and (C) between E2-treated siSCR control and METTL3-depleted cells (n = 3). D METTL3 expression in MDA-MB-231 cells following METTL3 siRNA-mediated knockdown (n = 3). E Differential gene expression compared between siSCR control and METTL3-depleted cells (n = 3). F METTL14 expression in MCF7 cells following METTL14 siRNA-mediated knockdown -/+ E2 (n = 3). G Differential gene expression compared between vehicle-treated siSCR control and METTL14-depleted cells (n = 3) and (H) between E2-treated siSCR control and METTL14-depleted cells (n = 3). I METTL14 expression in MDA-MB-231 cells following METTL14 siRNA-mediated knockdown (n = 3). J Differential gene expression compared between siSCR control and METTL14-depleted cells (n = 3). K CBLL1 expression in MCF7 cells following CBLL1 siRNA-mediated knockdown −/+ E2 (n = 3). L Differential gene expression compared between vehicle-treated siSCR control and CBLL1-depleted cells (n = 3) and (M) between E2-treated siSCR control and CBLL1-depleted cells (n = 3). N CBLL1 expression in MDA-MB-231 cells following CBLL1 siRNA-mediated knockdown (n = 3). O Differential gene expression compared between siSCR control and CBLL1-depleted cells (n = 3). Significant differential gene expression= FC ± 1.5 and adjusted p-value < 0.05.
Comparison of all DEGs in METTL3, METTL14 and CBLL1-depleted (A) MCF7, and (B) MDA-MB-231. Significantly enriched KEGG pathways associated with genes with: (C) downregulated expression following METTL3 depletion and vehicle treatment in MCF7; (D) upregulated expression following METTL3 depletion in E2 treated MCF7; (E) downregulated and (F) upregulated expression following METTL3 depletion in MDA-MB-231; (G) downregulated expression following METTL14 depletion and vehicle treatment in MCF7, and (H) downregulated expression following METTL14 depletion in E2 treated MCF7; (I) downregulated expression following METTL14 depletion in MDA-MB-231; (J) downregulated and (K) upregulated expression following CBLL1 depletion in MDA-MB-231. Significant gene expression= FC ± 1.5 and adjusted p-value < 0.05. All n = 3. Significant pathways FDR < 0.05.
METTL3 knockdown affected both the basal and estrogen-regulated transcriptome in the ERα-expressing cell line MCF7 and basal transcriptome in the TNBC cell line MDA-MB-231, including several transcription factors and genes involved in BCa carcinogenesis and progression (Fig. 3A–E). In the absence of estrogen, 1172 genes were differentially expressed in MCF7 and with E2 (10 nM) treatment 2314 genes were differentially expressed. METTL3 regulated a large subset of E2-regulated genes. Of the genes upregulated following METTL3 depletion in MCF7, 31% were also upregulated by E2 treatment including prototypical E2-upregulated genes such as TFF1 (Supplementary Fig. 16). METTL3 depletion in vehicle-treated cells resulted in ten KEGG pathways enriched in genes higher in siSCR compared with siMETTL3 including Pathways in cancer (hsa05200) and Signalling pathways regulating pluripotency of stem cells (hsa04550) (Fig. 4C). Two significantly enriched pathways were identified within the genes higher in estrogen-treated and METTL3 depleted MCF7 compared with estrogen-treated siSCR control (Fig. 4D). These two pathways comprised Protein digestion and absorption (hsa04974) and Extracellular matrix receptor interaction (hsa04512). No pathways were enriched in the gene set upregulated by siMETTL3 in vehicle treated MCF7 or in the gene set downregulated by siMETTL3 in E2 treated MCF7. BCa and estrogen regulation associated genes were found to be altered by METTL3 knockdown, including EGF, CDK6, PIK3R1, CDKN2B, CDKN2A, GPER1, NQO1, DNMT3B, PGR, TERT and SNAI1. METTL3 depletion in MDA-MB-231 resulted in 848 significantly differentially expressed genes (DEG), with 513 decreased and 335 increased following METTL3 knockdown (Fig. 3E). Ten KEGG pathways were significantly enriched in the genes downregulated by METTL3 depletion (Fig. 4E). These included Metabolic pathways (hsa01100), PI3K-Akt signalling pathway (hsa04151) and several pathways involved in innate viral response. One pathway, the TGF-beta signalling pathway (hsa04350), was significantly enriched in genes upregulated by METTL3 depletion (FDR < 0.05; Fig. 4F).
METTL14 knockdown had modest effects on the MCF7 transcriptome. In the absence of estrogen, only five genes were significantly differentially expressed when METTL14 was depleted, whereas with estrogen 51 genes were significantly differentially expressed. KEGG pathway analysis revealed the Metabolism of xenobiotics by cytochrome P450 pathway (hsa00980) was enriched in the genes decreased when METTL14 was depleted in both vehicle and estrogen treated cells (Fig. 4G–H). A larger number of genes were significantly differentially expressed upon METTL14 depletion in MDA-MB-231 compared with MCF7 (Fig. 3F–J). Depletion of METTL14 resulted in 56 genes with significantly decreased expression and 17 genes with increased expression. In the gene set that decreased with METTL14 depletion, the significantly over-represented pathways were Metabolic pathways (hsa01100) and Fatty acid metabolism (hsa01212) (Fig. 4I). No pathways were enriched in the gene set increased upon METTL14 depletion however several of these genes were also involved in metabolism (CYP1B1, CYP1A1, ANGPTL4, PTGES).
Depletion of CBLL1 in vehicle-treated MCF7 resulted in the significant differential expression of 495 genes, of which 215 were induced by CBLL1 knockdown and 280 were downregulated by CBLL1 knockdown (including PPL, TFEB and NTN4) (Fig. 3K–O). In the presence of E2, depletion of CBLL1 resulted in 423 genes being differentially expressed, with 154 upregulated, and 269 genes (including PPL, TFEB and NTN4) downregulated (Fig. 3M). CBLL1 depletion subtly increased the expression of TFF1, a prototypical estrogen-induced gene (p < 0.0001). No DEG pathways were significantly enriched following CBLL1 depletion in vehicle-treated or in E2-treated MCF7. In MDA-MB-231, 980 genes were upregulated by CBLL1 knockdown and 459 were downregulated. The pathways enriched by the genes decreased by CBLL1 depletion were Ribosome (hsa03010) and Ribosome biogenesis in eukaryotes (hsa03008) (Fig. 4J). In the genes upregulated by CBLL1 depletion the following pathways were enriched: Steroid biosynthesis (hsa00100), Complement and coagulation cascades (hsa04610), TNF signalling pathway (hsa04668), Cell adhesion molecules (CAMs) (hsa04514), NOD-like receptor signalling pathway (hsa04621), Lysosome (hsa04142), Transcriptional misregulation in cancer (hsa05202), Cytokine-cytokine receptor interaction (hsa04060), Human papillomavirus infection (hsa05165), and PI3K-Akt signalling pathway (hsa04151) (Fig. 4K).
In both MCF7 and MDA-MB-231 cell lines, METTL3 inhibition and the depletion of METTL3, METTL14 or CBLL1 dramatically altered splicing, with skipped exons being the most common differentially spliced event (Supplementary Tables 2 and 3).
Regulation of METTL3, METTL14 and CBLL1 expression by m6A
METTL3 knockdown additionally reduced METTL14 protein expression, and conversely METTL14 knockdown also reduced METTL3 protein expression (Fig. 5E, F, H, I, Supplementary Figs. 10, 11 and 12), whereas mRNA expression was unchanged (Supplementary Fig. 17). This suggests auto-regulation of MTC component protein expression and stoichiometry. To determine whether the ubiquitin-ligase MTC component CBLL1 plays a role in this mutual regulation of METTL3 and METTL14 expression, the effect of combinatorial siRNA knockdown of METTL3 or METTL14 and CBLL1 on METTL14 and METTL3 expression respectively was assessed in MDA-MB-231. The addition of CBLL1 knockdown in MDA-MB-231 did not alter the effect of METTL3 or METTL14 knockdown on METTL14 and METTL3 expression (Fig. 5D-I, Supplementary Figs. 10, 11 and 12). The knockdown of CBLL1 also did not alter the expression of METTL3 or METTL14 in MDA-MB-231 or MCF7 (Fig. 5A, B and C, Supplementary Figs. 9, 13 and 14). We next assessed whether the METTL3, METTL14 and CBLL1 transcripts harboured m6A peaks in publicly available MeRIPseq data. Interestingly, m6A peaks in METTL14 and CBLL1 were enriched in the last exon whereas m6A peaks in METTL3 were enriched in the third exon in MCF7 and MDA-MB-231 (Fig. 5J–L, Supplementary Fig. 18).
Effect of siRNA-mediated knockdown of METTL3, METTL14 and CBLL1, both individually and in combination, on expression of METTL3, METTL14 and CBLL1 in MDA-MB-231. Quantification of expression of CBLL1, METTL3 and METTL14 determined by western blot with β-Actin loading control (n = 6) (A-I). Protein expression quantification of: (A) CBLL1 following siRNA-mediated knockdown of CBLL1; (B) METTL3 following CBLL1 siRNA-mediated knockdown; (C) METTL14 following CBLL1 siRNA-mediated knockdown. D CBLL1, (E) METTL3 and (F) METTL14 expression following siRNA-mediated knockdown of METTL3 and knockdown of METTL3 and CBLL1 in combination. G CBLL1, (H) METTL3 and (I) METTL14 expression following siRNA-mediated knockdown of METTL14 and knockdown of METTL14 and CBLL1 in combination. Evidence of the regulation of (J) METTL3, (K) CBLL1 and (L) METTL14 by m6A methylation in MDA-MB-231, MeRIP-seq data obtained from GEO series GSE185494. Peaks in second panel (m6A enriched) compared to first panel (no enrichment control) show area of m6A enrichment within gene. Red box highlights last exon of each gene and yellow box highlights the third exon of METTL3. * = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001 by t-test or ANOVA for multiple comparisons bar charts show mean ± SEM.
METTL3 inhibition reduces invasion and proliferation and induces an immune gene signature in BCa cell lines
The effects of pharmacological METTL3 inhibition on the phenotype, and differentially expressed and spliced transcriptome of BCa cell lines were investigated. The STM2457 small molecule METTL3 inhibitor reduced BCa cell line proliferation and invasion in both ER positive and TNBC cell lines (Fig. 6A–M). We also observed that STM2457 reduced proliferation of HMECs (Fig. 6A).
Effect of increasing STM2457 treatment (STM) concentrations on proliferation in: (A) non-malignant HMEC (n = 6), (B) malignant MCF7 + /- E2 (n = 9), (C) malignant T-47D + /- E2 (n = 9), (D) malignant MDA-MB-436 (n = 9), and (E) malignant MDA-MB-231 (n = 9). F Effect of STM2457 (STM) treatment and E2 on the invasion of MCF7 (n = 8). Representative photomicrographs of MCF7 treated with (G) vehicle (veh), (H) E2, (I) STM2457 (STM), and (J) STM2457 (STM) with E2. K Effect of STM2457 (STM) treatment on the invasion of MDA-MB-231 (n = 8). Representative images of MDA-MB-231 treated with (L) vehicle (veh) and (M) STM2457 (STM). Significantly enriched KEGG pathways associated with genes with: (N) upregulated, and (O) downregulated expression following STM2457 treatment in MCF7; (P) upregulated, and (Q) downregulated expression following STM2457 and E2 treatment in MCF7; (R) upregulated, and (S) downregulated expression following STM2457 treatment in MDA-MB-231. * = p < 0.05, ** = p < 0.005, *** = p < 0.001, **** = p < 0.0001, ns = not significant determined by ANOVA for multiple comparisons or t-test, bar charts show mean ± SEM. Significant KEGG pathways FDR < 0.05.
RNA-seq performed on cell line MCF7 revealed that pathways Cytokine-cytokine receptor interaction (hsa04060) and Influenza A (hsa05164) were significantly enriched in the gene set increased with STM2457 treatment (Fig. 6N). In the gene set decreased with STM2457 treatment, the pathways Axon guidance (hsa04360), Ribosome (hsa03010), Fatty acid degradation (hsa00071), Proteoglycans in cancer (hsa05205), Fatty acid metabolism (hsa01212), Valine, leucine and isoleucine degradation (hsa00280), beta-Alanine metabolism (hsa00410) and Carbon metabolism (hsa01200) were enriched (Fig. 6O).
Pathway analysis was also performed on genes significantly upregulated and downregulated by STM2457-mediated METTL3 inhibition in the presence of estrogen. Several of the pathways enriched in the gene set significantly increased with STM2457 treatment were related to cancer, viral response, and cell death (Fig. 6P). Notable pathways include TNF signalling pathway (hsa04668), Transcriptional misregulation in cancer (hsa05202), Pathways in cancer (hsa05200) and Cytokine-cytokine receptor interaction (hsa04060). Several immune related genes were common between MCF7 vehicle and E2 treated, including genes CXCL10, IFIT1 and DDX58 (RIG1). In the gene set decreased with STM2457 treatment, pathways related to cell division and metabolism were significantly enriched (Fig. 6Q) and included Metabolic pathways (hsa01100) and Cell cycle (hsa04110). Many of these metabolic pathways were also significantly enriched with METTL3 inhibition without estrogen treatment (Fig. 6O).
Fewer genes were significantly differentially expressed by STM2457 treatment of MDA-MB-231 compared to MCF7 (Supplementary Fig. 19). Several cancer-associated pathways were significantly enriched in the gene set induced by STM2457, including MAPK signalling pathway (hsa04010), TGF-beta signalling pathway (hsa04350) and Pathways in cancer (hsa05200) (Fig. 6R). There were also several pathways associated with immune and stress response: NOD-like receptor signalling pathway (hsa04621), TNF signalling pathway (hsa04668), Human papillomavirus infection (hsa05165), Complement and coagulation cascades (hsa04610) and Cytokine-cytokine receptor interaction (hsa04060). In the genes decreased by STM2457 in MDA-MB-231, over-represented genes were associated with the following KEGG pathways: Ribosome (hsa03010), Metabolic pathways (hsa01100), DNA replication (hsa03030), Terpenoid backbone biosynthesis (hsa00900), Carbon metabolism (hsa01200) and Proteasome (hsa03050) were enriched (Fig. 6S). Within the DNA replication pathway is the gene Proliferating Cell Nuclear Antigen (PCNA) a well-defined marker of cell division35. In the ribosome pathway, 58 of the 153 ribosome-associated genes decreased with STM2457 treatment.
To give further insight into the expression of METTL3 in BCa tissue samples, the TCGA RNA-seq data was utilised (n = 1078). The IL-17 signalling pathway (hsa04657) was significantly enriched in the gene set that was increased when METTL3 expression was low (Supplementary Fig. 20).
Discussion
In this large BCa patient cohort study, we found that higher nuclear CBLL1 and METTL14 expression was significantly associated with several favourable prognostic factors, most notably lower grade, better NPI prognostic group, ER and PR positivity, and non-TNBC. Kaplan-Meier survival analysis revealed that higher nuclear METTL14 and CBLL1 expression were both significantly associated with longer BCSS, TTDM and DFS. However, only CBLL1 was also found to be an independent prognostic indicator of BCSS, TTDM and DFS when considering the potential confounding variables of nodal stage, tumour size, patient age, ER status and LVI.
These findings were consistent with previous studies that reported a tumour suppressor role for CBLL1 in BCa, including association between higher CBLL1 expression and reduced proliferation and migration in BCa cells36, and better relapse-free survival and OS in BCa patients (n = 134)28. In agreement with our findings that increased METTL14 expression correlated with more favourable clinical prognostic factors and outcomes, TCGA GDC transcriptomic data showed decreased METTL14 expression in tumour tissue compared with normal breast tissue. Additionally, two studies have reported reduced METTL14 protein expression in BCa compared to normal breast tissue23,37.
Further supporting the role of m6A in BCa, we have previously shown that low mRNA and protein expression of m6A demethylase ALKBH5 was associated with poorer clinical outcomes and prognosis38. Additionally, the other known components of the m6A MTC have also been implicated in BCa. Downregulated ZC3H13 and upregulated VIRMA have been shown to correlate with poor prognosis in BCa37,39. RBM15 has been found to be elevated in basal-like BCa and TNBC and associated with poorer outcomes, likely through regulation of serine and glycine metabolism40. WTAP has been shown to be predominantly associated with a cancer-promoting role in BCa41,42.
Despite higher METTL3 expression being significantly associated with lower tumour grade, ER and PR positivity, and non-TNBC status, METTL3 expression did not correlate with survival outcomes in our patient cohort, when considering the cohort as a whole. METTL3 expression also did not correlate with survival in TCGA data. However, higher nuclear METTL3 expression was significantly associated with decreased BCSS, TTDM and DFS in TNBC patients. High nuclear METTL3 expression was also an independent prognostic indicator of poorer BCSS, TTDM and DFS independent of patient age, ER status, LVI, nodal stage, tumour grade and size in both the whole cohort and the TNBC sub-cohort. Therefore, the associations between increased METTL3 expression with favourable clinicopathological variables are likely due to the association of METTL3 expression with estrogen signalling. Similarly, METTL3 expression, together with METTL14 and CBLL1, was found to be positively correlated with the expression of proteins associated with better BCa prognosis such as BRCA1, ERα and FOXA143,44,45.
The individual siRNA depletion of METTL3 and METTL14 also reduced expression of METTL14 and METTL3 protein, respectively, without altering mRNA expression. Similarly, Chelmicki and colleagues reported that METTL3 knockout resulted in loss of METTL14 protein, and METTL14 knockout reduced METTL3 protein expression in mouse embryonic stem cells46. Similar results were also obtained in Arabidopsis, demonstrating that MTA and MTB (METTL3 and METTL14 homologues, respectively) have interdependent effects on protein accumulation and localisation47. The exact mechanism of this evolutionarily conserved reciprocal regulation of MTC components remained unknown. For this reason, we assessed whether the CBLL1 ubiquitin-ligase component of the MTC may play a role in sensing and regulating the relative stoichiometry of the METTL3 and METTL14 MTC components. To do this we conducted individual and combinatorial knockdown of METTL3 or METTL14 with CBLL1. Knockdown of CBLL1 did not affect the reciprocal mutual regulation of METTL3 and METTL14, suggesting an alternative mechanism, potentially involving m6A regulation of METTL3 and METTL14 expression. To address this we next assessed publicly available MeRIPseq data from BCa cell lines to evaluate whether the METTL3, METTL14 and CBLL1 transcripts harboured m6A. METTL3, METTL14 and CBLL1 transcripts were found to have m6A peaks. Thus, knockdown of either METTL3 or METTL14 would impact MTC function, reducing global m6A, including of METTL3 and METTL14 transcripts, thereby affecting their protein expression. It is also possible that when one component of the catalytic METTL3:METTL14 heterodimer was depleted, then the heterodimer was unable to form, and the other component became more susceptible to protein degradation. Further studies are warranted to determine the exact mechanism of MTC regulation and the functional effect of the m6A peak within the third exon of METTL3 transcripts.
RNA-seq analyses following the knockdown of METTL3, METTL14 and CBLL1 revealed that each regulates distinct gene networks and pathways in the BCa cell lines examined here. Several KEGG pathways including Pathways in cancer (hsa05200) were enriched in genes higher in siSCR Veh compared to siMETTL3 Veh. METTL3 depletion had distinct functions and more pathway-targeted effects in the absence of E2 signalling in MCF7. Multiple KEGG pathways were significantly enriched in the genes downregulated by METTL3 depletion in MDA-MB-231 suggesting selective transcriptional effects, not transcriptome-wide, non-specific effects. These included Metabolic pathways (hsa01100), PI3K-Akt signalling (hsa04151) and several pathways involved in viral response. This is consistent with MeRIP-seq analysis reported by Li and colleagues that showed m6A modification peaks were enriched in the Metabolic, TNF signalling, and PI3K-Akt–related transcripts in glioma stem cells48. The reduction in the pro-survival and pro-proliferative PI3K-Akt signalling pathway under METTL3 depletion may underlie the poor prognosis associated with high METTL3 expression in the TNBC patient cohort in this study. Indeed, the PI3K-Akt signalling pathway is often activated in BCa through PIK3CA, AKT1 or MTOR activation and/or loss of PTEN and INPP4B, and therapeutics targeting this pathway are being developed49,50. Interestingly, several immune pathways were enriched in the gene set downregulated by METTL3 depletion in contrast to the inhibition of METTL3 by STM2457 in which immune pathways were increased. Despite the immune-related pathways being enriched, viral RNA response genes such as OASL, IFIT1, IFIT2 and IFIT3 were upregulated following METTL3 depletion in MDA-MB-231. Inconsistencies between METTL3 inhibition and depletion may be the result of METTL3 depletion resulting in a residual level of functional METTL3, disruption of the MTC formation and function or due to non-m6A methyltransferase roles of METTL351,52.
Of the 51 genes commonly significantly downregulated by METTL3 depletion in both MCF7 and MDA-MB-231, several have been implicated in BCa including ZNF652, GPER1, and STC1. ZNF652 encodes a transcription factor reported to promote expression of genes associated with cancer development and progression53. GPER1 (G protein-coupled estrogen receptor) binds E2 and rapidly induces non-genomic E2 effects and its expression has recently been found to be associated with poor outcomes in TNBC54,55. GPER1 has been reported to interact with and possibly induce ERα36, a truncated and transcriptionally inactive isoform of ERα56,57. STC1 transcription has already been reported to be upregulated indirectly by m6A21. STC1 is reported to promote BCa growth and metastasis and may also promote homologous recombination-mediated DNA damage repair by recruiting BRCA158,59.
In contrast to METTL3 and CBLL1 depletion, METTL14 depletion affected the expression of few genes reinforcing the importance of METTL3 as the catalytic component of the MTC. Indeed, the reduction of METTL3 observed following METTL14 depletion did not induce the transcriptional changes observed following direct METTL3 depletion. One explanation is that the downregulation of METTL3 following METTL14 depletion may have been insufficient to induce the transcriptional changes or may have resulted in the reduction of METTL3 protein that, due to sub-cellular distribution and function, did not function to influence the transcriptome. METTL14-regulated DEGs included both oncogenic and tumour suppressor genes and demonstrate a role of METTL14 in regulating metabolism in breast cancer. These transcriptomic effects may be due to both m6A changes and the role of METTL14 in chromatin regulation60.
RNA-seq analysis of CBLL1 depleted MCF7 cells revealed a number of DEGs associated with BCa, including PPL, TFEB and NTN4, that offer a mechanistic basis for the association of CBLL1 with better outcomes. PPL, TFEB and NTN4 expression levels were reduced by CBLL1 depletion. Downregulated PPL (periplakin) expression has been associated with cancer progression in a number of cancer types61,62,63, and has been implicated in brain metastasis in TNBC64. The TFEB gene encodes the transcription factor EB, a master regulator of several essential cellular processes, including cellular differentiation, autophagy, cellular energy metabolism, lysosome regulation, and immune response65,66,67,68,69. In early stage BCa, high TFEB expression has been associated with a poorer prognosis in patients70, although another study found that increased TFEB expression reduced breast tumour development through modulation of tumour-associated macrophages, and concluded that TFEB activation represented a promising therapeutic approach71. NTN4 was found to be downregulated following the siRNA-mediated knockdown of CBLL1, consistent with our findings that low CBLL1 correlates with poorer outcomes. It has been shown previously that low NTN4 expression was associated with poorer survival in BCa and associated with the infiltration of numerous immune cell types72.
In MDA-MB-231, CBLL1 knockdown resulted in enriched pathways consistent with CBLL1 being a positive prognostic indicator. In the genes upregulated by CBLL1 depletion, the Steroid biosynthesis, Complement and coagulation cascades, TNF signalling pathway, Cell adhesion molecules (CAMs), NOD-like receptor signalling pathway, Lysosome, Transcriptional misregulation in cancer, Cytokine-cytokine receptor interaction, Human papillomavirus infection and PI3K-Akt signalling pathways were enriched. The Ribosome and Ribosome biogenesis in eukaryotes pathways were enriched by genes downregulated following CBLL1 depletion, indicating a potential reduction in protein translation. As with METTL3 inhibition by STM2457, we observed an increase in immune signalling following CBLL1 depletion, further suggesting m6A mediation.
Our data suggests a role for these m6A regulators in modulating E2-regulated gene expression. METTL3, METTL14 and CBLL1 protein expression was upregulated by estrogen in MCF7. Both siRNA-depletion and pharmacological inhibition of METTL3 also resulted in the upregulation of prototypical estrogen-regulated genes, such as TFF1. This finding is supported by a recent study by Wan and colleagues that determined a role for METTL3-dependent m6A methylation in balancing estrogen and progesterone signalling73. They determined that conditional knockout of METTL3 causes hyperactivation of the estrogen response by stabilising mRNAs in murine uterine tissue73. This suggests that METTL3 functions as a transcriptional “dimmer switch” of basal expression of estrogen-regulated genes. Thus, high METTL3 expression could downregulate E2-driven proliferation and may explain why increased METTL3 expression is not significantly associated with a poorer prognosis in hormone receptor positive BCa. Similarly, CBLL1 depletion subtly increased the expression of TFF1, indicating this may be an m6A-mediated mechanism. However, another study found that through directly binding ERα and competing with coactivators, CBLL1 acts as a co-repressor of ERα36.
METTL3, METTL14 and CBLL1 depletion had an extensive effect on splicing in both MCF7 and MDA-MB-231 including many genes implicated in BCa. This result is consistent with another study that found METTL3 promoted alternative splicing events associated with the malignant phenotype of BCa, both directly through m6A and by the regulation of transcription factors such as MYC which led to decreased SRSF11 splicing factor protein levels74. We observed no significant change to the SRSF11 transcript level following METTL3 depletion. METTL3 inhibition by STM2457 has also been found to alter splicing in prostate cancer32, and may have a role in supressing tumorigenesis.
Treatment with the STM2457 small molecule inhibitor reduced cellular proliferation and invasion in both ER positive and TNBC cell lines, but also reduced proliferation of primary mammary epithelial cells (HMECs). This finding is in line with other studies that reported reduced proliferation and invasion of BCa cell lines with the knockdown of METTL3 and increased invasion and migration upon METTL14 overexpression19,22,74,75. Another study supporting our findings demonstrated that STM2457 treatment reduced growth and had a synergistic effect with carboplatin and Olaparib treatment in both MDA-MB-231 and a TNBC patient-derived organoid76. STM2457 treatment significantly altered the transcriptome and mRNA splicing, and several mechanisms that could reduce cancer cell growth were identified. Another potential mechanism is that the reduction or absence of m6A is hindering effective processing of RNA and protein translation, as demonstrated by the dramatic changes in transcripts and splicing alterations, which is impeding cellular processes and preventing proliferation and migration. We also demonstrated that STM2457 induces immune and inflammatory pathways, particularly those in response to viral infection. Although immune and inflammatory signalling activation is often associated with BCa progression and poor patient outcomes, the decrease in proliferation and invasion observed in phenotypic studies suggests that these pathways are not acting in a pro-tumourigenic manner, although the tumour microenvironment is critical for immune-related features.
The role of m6A in the innate immune system has been previously explored in several studies77,78,79,80,81. It was shown in foetal murine hematopoietic stem cells that loss of METTL3 and m6A resulted in the formation of endogenous double-stranded RNAs (dsRNAs) which activated an aberrant innate immune response82. It is therefore likely that m6A depletion is resulting in the activation of the antiviral response in BCa. Indeed, recently, inhibition of METTL3 with STM4257 and another METTL3 inhibitor (STM3006) was shown to stimulate an interferon response, augment antigen-dependent killing of cancer cells by T-cells in vitro and exhibit far greater in vivo pre-clinical activity when in combination with anti-PD1 therapy34. The analysis of TCGA RNA-seq data revealed evidence of an association between METTL3 and immune response in BCa tumour specimens, as low METTL3 expression correlated with an increased expression of genes involved in immune-related pathways. Further studies into the potential functional interactions between m6A and immune-targeted therapies in BCa are therefore warranted.
Further work is required to understand the distinct cell line and context specific transcriptional networks regulated by these MTC components. The established E3 ubiquitin ligase function of CBLL1 and the emerging m6A-independent functions of METTL3 and METTL14 provide potential explanations for the distinct differential gene expression observed between the different siRNA knockdowns25,51,52,60. However, the distinct gene expression profiles observed between MDA-MB-231 and MCF7 in estrogen depleted and estrogen treated conditions, indicate BCa subtype and endocrine conditions have a dramatic influence on the molecular consequences of METTL3, METTL14 and CBLL1 activity. Critically, the examination of transcript m6A levels by methods such as MeRIPseq or Nanopore Direct RNA sequencing (DRS), would enable the effects observed here due to knockdown and METTL3 inhibition, such as the modulation of E2 signalling, to be attributed to either the absence of m6A on specific transcripts, indirect m6A effects or to the other non-m6A roles of these proteins. Further work is now required to determine the effects of METTL3 inhibition in vivo. STC-15, the METTL3 inhibitor currently undergoing clinical trial, like STM2457, also binds the SAM pocket of METTL334. However, STC-15 has improved potency, oral bioavailability and metabolic stability and future in vivo studies may benefit from utilizing STC-15 over STM2457.
In conclusion, our findings suggest important clinical and mechanistic roles for the m6A methyltransferase complex proteins, METTL3, METTL14 and CBLL1, in the regulation of gene expression and alternative splicing in BCa and estrogen signalling. IHC analysis of ~2500 BCa tumour specimens revealed an association of increased METTL14 and CBLL1 with favourable clinical outcomes in BCa and increased METTL3 expression with poor clinical outcomes in hormone receptor negative BCa. Transcriptomic analysis of METTL3, METTL14 and CBLL1 depletion in BCa cell lines revealed the regulation of distinct transcriptional networks and demonstrated that METTL3 modulates estrogen signalling and has BCa-subtype specific roles. The inhibition of the methyltransferase activity of METTL3 reduced the invasive and proliferative phenotype of BCa cell lines and increased the expression of immune-related pathways. These results highlight the promising potential of METTL3 inhibitors as an effective therapeutic option in BCa, particularly TNBC. Crucially, the safety profile of STC-15, a derivative of STM2457, from the ongoing phase 1 trial will determine the potential of therapeutically targeting m6A to benefit patients with BCa.
Methods
Bioinformatic analysis of METTL3, METTL14 and CBLL1 in clinical specimens
METTL3, METTL14 and CBLL1 expression and gene alterations were analysed in The Cancer Genome Atlas (TCGA) breast invasive carcinoma cohort (PanCancer Atlas, n = 996)83 using the cBioPortal for Cancer Genomics (retrieved January 2024)84,85. Analysis of METTL3, METTL14 and CBLL1 expression in the GDC TCGA Breast Cancer patient cohort83 was carried out using the UCSC Xena Functional Genomics Browser (n = 1284)86. The Kaplan-Meier Plotter online tool87 was used to investigate the association of METTL3, METTL14 and CBLL1 expression with patient survival. RNA-seq data utilising best performing cut-off for dichotomisation for OS was used (n = 2,976). TCGA RNA-seq data was utilised to perform differential gene expression analysis based on METTL3 expression in primary female BCa samples (n = 1078). Count data, generated by HTSeq, was obtained from the Genomics Data Commons and was normalised within the Deseq2 package and quartiles assigned for METTL3 expression. DESeq2 was utilized to determine the differentially expressed genes between Quartile 1 (breast cancer with low METTL3 expression) and Quartile 4 (breast cancer with high METTL3 expression)88. Methylated RNA Immunoprecipitation Sequencing (MeRIP) data was obtained from Gene Expression Omnibus (GEO) for MCF7 (ER positive BCa cell line) Series GSE14344189 and MDA-MB-231 (TNBC cell line) Series GSE18549490 and viewed using Integrative Genomics Viewer (IGV)91.
Immunohistochemistry of BCa tissue specimens
A breast tumour tissue microarray (TMA) was constructed from a well-characterised Nottingham University Hospitals (NUH) NHS Trust patient cohort. The cohort comprises surgically removed tumour tissue from over 2800 BCa patients diagnosed between 1990 and 2006. All patients were neoadjuvant treatment-naïve at the time of surgery and all patient clinical data were maintained prospectively.
IHC was carried out using the Novolink Max Polymer Detection System (Leica Biosystems, UK) as previously described92. Sections (4 µm) were cut from the TMA blocks and subsequently stained with primary antibodies: anti-METTL3 (ab195352 PUR; Abcam, UK, 1:500), anti-METTL14 (ab220030; Abcam, UK, 1:3000), anti-CBLL1 (NBP1-83589; Novus Biologicals, USA, 1:100). All primary antibodies were incubated at room temperature for one hour. Following staining, all slides were scanned at high resolution with the Nanozoomer (Hamamatsu Photonics, UK). The siRNA-mediated knockdowns performed in this study support the specificity of these antibodies.
Nuclear staining was assessed using the semi-quantitative histo-score (H-score) system93. Staining was performed on intact cores with tumour coverage of >10% of the core. A second scorer independently assessed a minimum of 10% of all specimens in order to validate all H-scores, confirmed with concordance of >0.8 via intraclass correlation.
Protein expression scores were dichotomised into high and low expression groups by median H-score and correlated with clinicopathological parameters and patient outcomes. METTL3 staining was divided into H-score groups of low ≤105 and high ≥106, METTL14 into H-score groups of low ≤120 and high ≥121, and CBLL1 into H-score groups of low ≤70 and high ≥71. Protein expression was also correlated with markers previously studied94,95,96,97.
Ethics approval for the construction of the TMA for samples collected from 1990-1998 was granted by the Nottingham Research Ethics Committee 2 (Reference title: Development of a molecular genetic classification of breast cancer), with ethics approval for samples collected from 1998-2006 granted by the North West Greater Manchester Central Research Ethics Committee (Reference title: Nottingham Health Science Biobank; reference number: 15/NW/0685). The Human Tissue Act and Helsinki Declaration of Human Rights were strictly observed, and the General Data Protection Regulation (GDPR) was applied.
Cell lines and culture conditions
Human Mammary Epithelial Cells (HMECs), ER-expressing malignant cell lines MCF7 and T-47D, and ER-negative malignant cell lines MDA-MB-231 and MDA-MB-436 were utilised in this study. Cell line identity was confirmed by genotyping or STR profiling. All cells were cultured at 37 °C and 5% CO2.
MCF7, T-47D and MDA-MB-231 were cultured in RPMI-1640 medium with L-glutamine supplemented with 100 µg/mL penicillin, 100 µg/mL streptomycin, 1 mM sodium pyruvate (all from Gibco® by Life Technologies, USA), and 10% heat-inactivated foetal bovine serum (FBS) (Sigma-Aldrich, USA). For estrogen treatment experiments, cells were grown in phenol-red free RPMI-1640 media supplemented with 100 µg/mL penicillin, 100 µg/mL streptomycin, 1 mM sodium pyruvate, 2 mM L-glutamine (all from Gibco by Life Technologies, USA), and 10% charcoal-stripped FBS with β-estradiol (E2) (10 nM) (Tocris Biosciences, UK) or DMSO as a vehicle control. HMEC cells were cultured in HuMEC Basal Serum-Free Medium with addition of the HuMEC Supplement Kit (Gibco by Life Technologies, USA). MDA-MB-436 cells were cultured in DMEM F12 media supplemented with HEPES, 100 µg/mL penicillin, 100 µg/mL streptomycin and 1 mM sodium pyruvate (Gibco by Life Technologies, USA).
siRNA-mediated functional depletion of METTL3, METTL14 and CBLL1 and pharmaco-inhibition of METTL3
siRNA-mediated depletion of METTL3, METTL14 and CBLL1 was performed using DharmaFECT 1 Transfection Reagent (GE Dharmacon, USA) and siRNA targeting METTL3 (SMARTpool: ON-TARGETplus METTL3 siRNA L-005170-02-0010; GE Dharmacon, USA), METTL14 (SMARTpool: ON-TARGETplus METTL14 siRNA L-014169-02-0010; GE Dharmacon, USA) and CBLL1 (SMARTpool: ON-TARGETplus CBLL1 siRNA L-007069-00-0005; GE Dharmacon, USA), respectively, following the manufacturer’s instructions at a final concentration of 20 nM siRNA. The ON-TARGETplus non-targeting control siRNA (D-001810-10-20, GE Dharmacon, USA) was employed as a negative control (scrambled control, siSCR). Following the first transfection, the media was replaced and a second transfection performed 48 hours later. Cellular RNA or protein was harvested after a further 72 hours. MCF7 was additionally treated with E2 (10 nM) (Tocris Biosciences, UK) or DMSO as a vehicle control at the second transfection.
Functional inhibition of METTL3 was conducted in the MCF7 and MDA-MB-231 cell lines using the METTL3 inhibitor STM2457 (MedChemExpress, USA). The cells were treated with either STM2457 (10 μM) or DMSO (vehicle) for 6 days. MCF7 was additionally treated with E2 (10 nM) (Tocris Biosciences, UK) or DMSO as a vehicle control.
Gene expression analysis and western blotting
For mRNA and protein expression analysis, BCa cells were treated as described above. RNA isolation was carried out as per the manufacturer’s protocol using the GenElute™ Mammalian Total RNA Miniprep Kit with on-column DNase treatment (Sigma-Aldrich, USA). cDNA was synthesised using the qScript™ cDNA Synthesis Kit (Quanta Biosciences, USA). mRNA expression analysis was performed using real-time quantitative polymerase chain reaction (RT-qPCR) with TaqMan™ Fast Advanced Master Mix and Taqman™ Real-Time PCR probes (Thermo Fisher Scientific™, USA) and carried out in a CFX Connect Real-Time PCR Detection System (Bio-Rad Laboratories, USA). The Taqman™ probes used in this study were β-Actin (Hs01060665_g1), METTL3 (Hs00219820_m1) and METTL14 (Hs00383340_m1) from Thermo Fisher Scientific (Waltham, USA). Relative mRNA expression was calculated using the Pfaffl method98 and analysed relative to β-Actin.
Total protein analysis by western blotting was carried out as previously described32,99. The following antibodies were used in this study: anti-METTL3 (ab195352 PUR; Abcam, UK; 1:10,000), anti-METTL14 (ab220030; Abcam, UK; 1:1,000), anti-CBLL1 (NBP1-83589; Novus Biologicals, USA; 1:1,000), and anti-β-Actin (sc-130657; Santa Cruz, USA, 1:50,000 or MA5-15739; Invitrogen, USA, 1:10,000). Goat anti-mouse (ab97023; Abcam, UK, 1:10,000–50,000 or Sc-2005; Santa Cruz, USA, 1:10,000–50,000) and goat anti-rabbit (ab6721; Abcam, UK, 1:10,000-50,000 or Sc-2004; Santa Cruz, USA, 1:10,000-50,000) secondary antibodies were used. Western blots from all experimental repeats and full, uncropped and annotated western blot images are available in Supplementary Figs. 3–14.
RNA-seq and splicing analysis of siRNA depletion or pharmaco-inhibition of METTL3
RNA-seq analysis (Novogene, UK) was performed on MCF7 and MDA-MB-231 BCa cells transfected with siSCR control, siMETTL3, siMETTL14 or siCBLL1. Additionally, RNA-seq was performed on MCF7 and MDA-MB-231 cells treated either with vehicle (DMSO) or 10 μM STM2457. MCF7 experiments were also performed in vehicle (DMSO) or estrogen (10 nM E2) conditions. Analysis of the obtained Fastq files was carried out as previously described99. Genes were determined significant by fold change (FC) ± 1.5 and false discovery rate (FDR) < 0.05 and are listed in Supplementary File 1. Differential alternative splicing was also assessed using replicate multivariate analysis of transcript splicing (rMATS, version 3.2.5)100. Biologically and statistically significant differential alternative splicing was determined by difference in percentage spliced in (dPSI) ≥ 5% and false discovery rate (FDR) < 0.05. The Web-Based Gene Set AnaLysis Toolkit (WebGestalt) was utilised to identify significantly enriched Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways and weighted set cover filter applied101. Venny 2.1 (https://bioinfogp.cnb.csic.es/tools/venny/) was utilised to produce Venn diagrams to compare gene sets102 and the VolcaNoseR ShinyApp (https://huygens.science.uva.nl/VolcaNoseR/) was used to prepare volcano plots of significantly differentially expressed genes. Significantly differentially expressed gene sets were compared using Venny 2.1.0 (csic.es). All RNA-sequencing data reported here is available from NCBI-GEO (www.ncbi.nlm.nih.gov/geo) under the following accession numbers: GSE195532 (siMETTL3 in MCF7), GSE195487 (siMETTL14 in MCF7), GSE234571 (siMETTL3 and siMETTL14 in MDA-MB-231), GSE271326 (siCBLL1 in MCF7), GSE271330 (siCBLL1 in MDA-MB-231) and GSE234322 (STM2457 treatment of MCF7 and MDA-MB-231).
Phenotypic analysis
The effect of METTL3 inhibition on the proliferation of HMEC, MCF7, T-47D, MDA-MB-436 and MDA-MB-231 over six days was determined. Following plating, cells were treated 24 hours later with varying concentrations of STM2457 (0 µM, 10 µM, 20 µM, 30 µM and 40 µM) in the cell-appropriate medium, with media and treatment replaced after 72 hours. E2 (10 nM) or DMSO vehicle control conditions were also performed for MCF7 and T-47D. After a further 72 hours, cellular proliferation was measured using the CyQUANT™ Direct Cell Proliferation Assay (Invitrogen, USA) as per the manufacturer’s instructions, and the relative DNA content calculated.
Changes in cellular invasion following STM2457 treatment were assessed in cell lines MCF7 and MDA-MB-231. Cells were pre-treated for five days prior to the assay with either STM2457 (10 µM) or DMSO vehicle control. In MCF7, +/- E2 (10 nM) conditions were also applied. Changes in cellular invasion following STM2457 treatment were assessed using a Matrigel (Corning, USA) invasion assay as previously described99. 5 ×105 MCF7 cells and 2.5 ×105 MDA-MB-231 cells were plated prior to fixing at 24 hours. A total of 12 fields of view were systematically imaged per well using an inverted microscope (Leica, Germany), with the resultant images subsequently manually assessed to quantify the total number of cells that had invaded, with relative invasion calculated compared to vehicle control.
Statistical analysis
Statistical analysis was carried out using GraphPad Prism (GraphPad Software Inc., USA) and SPSS v26.0 (IBM, USA) statistical software. T-tests were carried out for the comparison of two means. For the comparison of multiple means a one-way analysis of variance (ANOVA) was performed. The median H-score was utilised to dichotomise protein expression into low and high expression groups, with 15-year breast cancer-specific survival (BCSS), disease-free survival (DFS) or time to distant metastasis (TTDM) used as an endpoint. Correlations with clinicopathological parameters were analysed by χ² test (asymptomatic significance, 2-sided) and the statistical significance of Kaplan Meier estimates calculated using the log-rank (Mantel-Cox) test. Combined multivariate Cox regression analysis (or proportional hazards model) was utilised to assess the prognostic independence of a marker from established prognostic indicators. Correlation between METTL3, METTL14 and CBLL1 with other previously assessed markers was determined by Spearman’s rank correlation coefficient. Graphs were generated with GraphPad Prism 9 (GraphPad Software, USA) and data were expressed as mean ± standard error of the mean (SEM). A p value of <0.05 was considered statistically significant.
Ethics approval and consent to participate
This study was reviewed and approved by the Nottingham Research Ethics Committee, (approval # REC202313), and the research ethics committee of the University of Nottingham School of Veterinary Medicine and Science (approval #3483 211102). The General Data Protection Regulation (GDPR) was applied, and informed consent obtained. The Helsinki Declaration of Human Rights was strictly observed.
Data availability
Data are available upon reasonable request. All RNA-sequencing data reported here is available from NCBI-GEO (www.ncbi.nlm.nih.gov/geo) under the following accession numbers: GSE195532 (siMETTL3 in MCF7), GSE195487 (siMETTL14 in MCF7), GSE234571 (siMETTL3 and siMETTL14 in MDA-MB-231), GSE271326 (siCBLL1 in MCF7), GSE271330 (siCBLL1 in MDA-MB-231) and GSE234322 (STM2457 treatment of MCF7 and MDA-MB-231).
Code availability
Example code used to analyse the data (which is available in GEO accession numbers: GSE195532 (siMETTL3 in MCF7), GSE195487 (siMETTL14 in MCF7), GSE234571 (siMETTL3 and siMETTL14 in MDA-MB-231), GSE271326 (siCBLL1 in MCF7), GSE271330 (siCBLL1 in MDA-MB-231) and GSE234322 (STM2457 treatment of MCF7 and MDA-MB-231). Trimgalore: trim_galore -q 30 --fastqc --paired sample.R1.fastq.gz sample.R2.fastq.gz. STAR alignment: STAR --genomeDir /starindexGrch38.83_star273a/ --sjdbGTFfile /starindexGrch38.83_star273a/Homo_sapiens.GRCh38.83.gtf --runThreadN 6 --readFilesIn sample_1_val_1.fq.gz sample_2_val_2.fq.gz --readFilesCommand zcat --outSAMtype BAM SortedByCoordinate --chimSegmentMin 20 --chimOutType WithinBAM --outFileNamePrefix sample_star_output; To trim reads for rMATs. python trimFastq.py sample_1_val_1.fq.gz 145_sample_1_val_1.fq.gz 145. rMATs: python2.7 /home/MATS.3.2.5/RNASeq-MATS.py -s1 -s2 -gtf Homo_sapiens.GRCh38.83.gtf -bi /home/star_v271a/ -novelSS 1 -len 145 -t paired -o sample_output.
References
Nolan, E., Lindeman, G. J. & Visvader, J. E. Deciphering breast cancer: from biology to the clinic. Cell 186, 1708–1728 (2023).
Sung, H. et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin. 71, 209–249 (2021).
Perou, C. M. et al. Molecular portraits of human breast tumours. Nature 406, 747–752 (2000).
Eroles, P., Bosch, A., Alejandro Pérez-Fidalgo, J. & Lluch, A. Molecular biology in breast cancer: intrinsic subtypes and signaling pathways. Cancer Treat. Rev. 38, 698–707 (2012).
McGuire, A., Brown, J. A. L., Malone, C., McLaughlin, R. & Kerin, M. J. Effects of age on the detection and management of breast cancer. Cancers 7, 908–929 (2015).
Dai, X. et al. Breast cancer intrinsic subtype classification, clinical use and future trends. Am. J. Cancer Res. 5, 2929–2943 (2015).
Sweeney, C. et al. Intrinsic subtypes from PAM50 gene expression assay in a population-based breast cancer cohort: differences by age, race, and tumor characteristics. Cancer Epidemiol. Biomark. Prev. 23, 714–724 (2014).
Onitilo, A. A., Engel, J. M., Greenlee, R. T. & Mukesh, B. N. Breast cancer subtypes based on ER/PR and Her2 expression: comparison of clinicopathologic features and survival. Clin. Med. Res. 7, 4–13 (2009).
Perou, C. M. Molecular stratification of triple-negative breast cancers. Oncologist 16, 61–70 (2011).
Dominissini, D. et al. Topology of the human and mouse m6A RNA methylomes revealed by m6A-seq. Nature 485, 201 (2012).
Zaccara, S., Ries, R. J. & Jaffrey, S. R. Reading, writing and erasing mRNA methylation. Nat. Rev. Mol. Cell Biol. 20, 608 (2019).
Wei, G. et al. Acute depletion of METTL3 implicates N6-methyladenosine in alternative intron/exon inclusion in the nascent transcriptome. Genome Res. 31, 1395–1408 (2021).
Knuckles, P. et al. Zc3h13/Flacc is required for adenosine methylation by bridging the mRNA-binding factor RbM15/spenito to the m 6 a machinery component Wtap/Fl(2)d. Genes Dev. 32, 415–429 (2018).
Ping, X.-L. et al. Mammalian WTAP is a regulatory subunit of the RNA N6-methyladenosine methyltransferase. Cell Res. 24, 177–189 (2014).
Liu, J. et al. A METTL3-METTL14 complex mediates mammalian nuclear RNA N6-adenosine methylation. Nat. Chem. Biol. 10, 93–95 (2014).
Su, S. et al. Cryo-EM structures of human m6A writer complexes. Cell Res. 32, 982–994 (2022).
Yue, Y. et al. VIRMA mediates preferential mA mRNA methylation in 3′UTR and near stop codon and associates with alternative polyadenylation. Cell Discov. 4, 10–10 (2018).
Cai, X. et al. HBXIP-elevated methyltransferase METTL3 promotes the progression of breast cancer via inhibiting tumor suppressor let-7g. Cancer Lett. 415, 11–19 (2018).
Wang, H., Xu, B. & Shi, J. N6-methyladenosine METTL3 promotes the breast cancer progression via targeting Bcl-2. Gene 722 https://doi.org/10.1016/j.gene.2019.144076 (2020).
Xie, J. et al. The m6A methyltransferase METTL3 promotes the stemness and malignant progression of breast cancer by mediating m6A modification on SOX2. J. BUON 26, 444–449 (2021).
Peng, F. et al. Oncogenic AURKA-enhanced N6-methyladenosine modification increases DROSHA mRNA stability to transactivate STC1 in breast cancer stem-like cells. Cell Res. 31, 345–361 (2021).
Yi, D. et al. METTL14 promotes the migration and invasion of breast cancer cells by modulating N6‑methyladenosine and hsa‑miR‑146a‑5p expression. Oncol. Rep. 43, 1375–1386 (2020).
Wu, L., Wu, D., Ning, J., Liu, W. & Zhang, D. Changes of N6-methyladenosine modulators promote breast cancer progression. BMC Cancer 19, 326 (2019).
Sun, T. et al. LNC942 promoting METTL14-mediated m6A methylation in breast cancer cell proliferation and progression. Oncogene 39, 5358–5372 (2020).
Ruzicka, K. et al. Identification of factors required for m(6) A mRNA methylation in Arabidopsis reveals a role for the conserved E3 ubiquitin ligase HAKAI. N. Phytol. 215, 157–172 (2017).
Liu, Z. et al. Celastrol inhibits the migration and invasion and enhances the anti-cancer effects of docetaxel in human triple-negative breast cancer cells. Med. Oncol. 39, 189 (2022).
Ma, L. et al. Ethoxysanguinarine induces apoptosis, inhibits metastasis and sensitizes cells to docetaxel in breast cancer cells through inhibition of Hakai. Chem. Biodivers. 20, e202200284 (2023).
Zheng, F. et al. Expression and clinical prognostic value of m6A RNA methylation modification in breast cancer. Biomark. Res 9, 28 (2021).
Yankova, E. et al. Small-molecule inhibition of METTL3 as a strategy against myeloid leukaemia. Nature 593, 597–601 (2021).
Arumugam, P. & Jayaseelan, V. P. Selective inhibition of METTL3 enhanced anticancer activity via targeting the c-Myc pathway. J. Clin. Oncol. 40, e15089–e15089 (2022).
Xu, Q.-C. et al. METTL3 promotes intrahepatic cholangiocarcinoma progression by regulating IFIT2 expression in an m6A-YTHDF2-dependent manner. Oncogene 41, 1622–1633 (2022).
Haigh, D. B. et al. The METTL3 RNA Methyltransferase Regulates Transcriptional Networks in Prostate Cancer. Cancers (Basel) 14, https://doi.org/10.3390/cancers14205148 (2022).
Sun, Y. et al. METTL3 promotes chemoresistance in small cell lung cancer by inducing mitophagy. J. Exp. Clin. Cancer Res. 42, 65 (2023).
Guirguis, A. A. et al. Inhibition of METTL3 results in a cell-intrinsic interferon response that enhances anti-tumour immunity. Cancer Discov. CD-23-0007 https://doi.org/10.1158/2159-8290.CD-23-0007 (2023).
Kurki, P., Ogata, K. & Tan, E. M. Monoclonal antibodies to proliferating cell nuclear antigen (PCNA)/cyclin as probes for proliferating cells by immunofluorescence microscopy and flow cytometry. J. Immunol. Methods 109, 49–59 (1988).
Gong, E. Y., Park, E. & Lee, K. Hakai acts as a coregulator of estrogen receptor alpha in breast cancer cells. Cancer Sci. 101, 2019–2025 (2010).
Gong, P.-J. et al. Analysis of N6-methyladenosine methyltransferase reveals METTL14 and ZC3H13 as tumor suppressor genes in breast cancer. Front. Oncol. 10, 578963 (2020).
Woodcock, C. L. et al. The role of the ALKBH5 RNA demethylase in invasive breast cancer. Discov. Oncol. 15, 343 (2024).
Zhu, W., Wang, J.-Z., Wei, J.-F. & Lu, C. Role of m6A methyltransferase component VIRMA in multiple human cancers (Review). Cancer Cell Int. 21, 172 (2021).
Park, S. H. et al. The m6A writer RBM15 drives the growth of triple-negative breast cancer cells through the stimulation of serine and glycine metabolism. Exp. Mol. Med. 56, 1373–1387 (2024).
Dragonetti, M. et al. The lncRNAMALAT1-WTAP axis: a novel layer of EMT regulation in hypoxic triple-negative breast cancer. Cell Death Discov. 10, 276 (2024).
Wang, C. Q. et al. Upregulated WTAP expression appears to both promote breast cancer growth and inhibit lymph node metastasis. Sci. Rep. 12, 1023 (2022).
Sørlie, T. et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc. Natl. Acad. Sci. USA 98, 10869–10874 (2001).
Thorat, M. A. et al. Forkhead box A1 expression in breast cancer is associated with luminal subtype and good prognosis. J. Clin. Pathol. 61, 327–332 (2008).
Kuchenbaecker, K. B. et al. Risks of breast, ovarian, and contralateral breast cancer for BRCA1 and BRCA2 mutation carriers. Jama 317, 2402–2416 (2017).
Chelmicki, T. et al. m6A RNA methylation regulates the fate of endogenous retroviruses. Nature 591, 312–316 (2021).
Shen, L. Functional interdependence of N6-methyladenosine methyltransferase complex subunits in Arabidopsis. Plant Cell 35, 1901–1916 (2023).
Li, F. et al. N6-methyladenosine modulates nonsense-mediated mRNA decay in human glioblastoma. Cancer Res. 79, 5785–5798 (2019).
Pascual, J. & Turner, N. C. Targeting the PI3-kinase pathway in triple-negative breast cancer. Ann. Oncol. 30, 1051–1060 (2019).
Nunnery, S. E. & Mayer, I. A. Targeting the PI3K/AKT/mTOR pathway in hormone-positive breast cancer. Drugs 80, 1685–1697 (2020).
Lin, S., Choe, J., Du, P., Triboulet, R. & Gregory, R. I. The m6A methyltransferase METTL3 promotes translation in human cancer cells. Mol. Cell 62, 335–345 (2016).
Choe, J. et al. mRNA circularization by METTL3-eIF3h enhances translation and promotes oncogenesis. Nature 561, 556 (2018).
Kumar, R. et al. Genome-wide mapping of ZNF652 promoter binding sites in breast cancer cells. J. Cell. Biochem. 112, 2742–2747 (2011).
Xu, T. et al. High GPER expression in triple-negative breast cancer is linked to pro-metastatic pathways and predicts poor patient outcomes. npj Breast Cancer 8, 100 (2022).
Hsu, L. H., Chu, N. M., Lin, Y. F. & Kao, S. H. G-protein coupled estrogen receptor in breast cancer. Int. J. Mol. Sci. 20, https://doi.org/10.3390/ijms20020306 (2019).
Wallacides, A. et al. Estrogens promote proliferation of the seminoma-like TCam-2 cell line through a GPER-dependent ERα36 induction. Mol. Cell Endocrinol. 350, 61–71 (2012).
Notas, G. et al. ERα36-GPER1 collaboration inhibits TLR4/NFκB-induced pro-inflammatory activity in breast cancer cells. Int. J. Mol. Sci. 22, https://doi.org/10.3390/ijms22147603 (2021).
Liu, A. et al. Stanniocalcin 1 promotes lung metastasis of breast cancer by enhancing EGFR–ERK–S100A4 signaling. Cell Death Dis. 14, 395 (2023).
Hou, J. et al. Molecular and clinical significance of stanniocalcin-1 expression in breast cancer through promotion of homologous recombination-mediated DNA damage repair. Front. Cell Dev. Biol. 9, 731086 (2021).
Dou, X. et al. METTL14 is a chromatin regulator independent of its RNA N6-methyladenosine methyltransferase activity. Protein Cell 14, 683–697 (2023).
Li, X. et al. Loss of periplakin expression is associated with the tumorigenesis of colorectal carcinoma. Biomed. Pharmacother. 87, 366–374 (2017).
Matsumoto, K. et al. Loss of periplakin expression is associated with pathological stage and cancer-specific survival in patients with urothelial carcinoma of the urinary bladder. Biomed. Res 35, 201–206 (2014).
Otsubo, T. et al. Aberrant DNA hypermethylation reduces the expression of the desmosome-related molecule periplakin in esophageal squamous cell carcinoma. Cancer Med 4, 415–425 (2015).
Choi, Y. K. et al. Brain-metastatic triple-negative breast cancer cells regain growth ability by altering gene expression patterns. Cancer Genom. Proteom. 10, 265–275 (2013).
Napolitano, G. et al. mTOR-dependent phosphorylation controls TFEB nuclear export. Nat. Commun. 9, 3312 (2018).
Settembre, C. et al. TFEB links autophagy to lysosomal biogenesis. Science 332, 1429–1433 (2011).
Settembre, C. et al. A lysosome-to-nucleus signalling mechanism senses and regulates the lysosome via mTOR and TFEB. EMBO J. 31, 1095–1108 (2012).
Zhang, Z. et al. Itaconate is a lysosomal inducer that promotes antibacterial innate immunity. Mol. Cell 82, 2844–2857.e2810 (2022).
Orfali, N. et al. All-trans retinoic acid (ATRA)-induced TFEB expression is required for myeloid differentiation in acute promyelocytic leukemia (APL). Eur. J. Haematol. 104, 236–250 (2020).
Giatromanolaki, A., Sivridis, E., Kalamida, D. & Koukourakis, M. I. Transcription factor eb expression in early breast cancer relates to lysosomal/autophagosomal markers and prognosis. Clin. Breast Cancer 17, e119–e125 (2017).
Li, Y. et al. TFEB is a master regulator of tumor-associated macrophages in breast cancer. J. Immunother. Cancer 8 https://doi.org/10.1136/jitc-2020-000543 (2020).
Yi, L. et al. NTN4 as a prognostic marker and a hallmark for immune infiltration in breast cancer. Sci. Rep. 12, 10567 (2022).
Wan, S. et al. METTL3-dependent m6A methylation facilitates uterine receptivity and female fertility via balancing estrogen and progesterone signaling. Cell Death Dis. 14, 349 (2023).
Achour, C., Bhattarai, D. P., Groza, P., Román, Á-C. & Aguilo, F. METTL3 regulates breast cancer-associated alternative splicing switches. Oncogene 42, 911–925 (2023).
Zhao, C., Ling, X., Xia, Y., Yan, B. & Guan, Q. The m6A methyltransferase METTL3 controls epithelial-mesenchymal transition, migration and invasion of breast cancer through the MALAT1/miR-26b/HMGA2 axis. Cancer Cell Int. 21, 441 (2021).
Cesaro, B. et al. Enhancing sensitivity of triple-negative breast cancer to DNA-damaging therapy through chemical inhibition of the m6A methyltransferase METTL3. Cancer Commun. 44, 282–286 (2024).
Lu, M. et al. Nonsegmented negative-sense RNA viruses utilize N6-methyladenosine (m6A) as a common strategy to evade host innate immunity. J. Virol. 95, e01939–01920 (2021).
Kim, G. W., Imam, H., Khan, M. & Siddiqui, A. N(6)-methyladenosine modification of hepatitis B and C viral RNAs attenuates host innate immunity via RIG-I signaling. J. Biol. Chem. 295, 13123–13133 (2020).
Lu, M. et al. N(6)-methyladenosine modification enables viral RNA to escape recognition by RNA sensor RIG-I. Nat. Microbiol 5, 584–598 (2020).
Qiu, W. et al. N(6)-methyladenosine RNA modification suppresses antiviral innate sensing pathways via reshaping double-stranded RNA. Nat. Commun. 12, 1582 (2021).
Durbin, A. F., Wang, C., Marcotrigiano, J. & Gehrke, L. RNAs containing modified nucleotides fail to trigger RIG-I conformational changes for innate immune signaling. mBio 7 https://doi.org/10.1128/mBio.00833-16 (2016).
Gao, Y. et al. m(6)A modification prevents formation of endogenous double-stranded RNAs and deleterious innate immune responses during hematopoietic development. Immunity 52, 1007–1021.e1008 (2020).
Liu, J. et al. An Integrated TCGA pan-cancer clinical data resource to drive high-quality survival outcome analytics. Cell 173, 400–416.e411 (2018).
Cerami, E. et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2, 401–404 (2012).
Gao, J. et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci. Signal 6, pl1 (2013).
Goldman, M. J. et al. Visualizing and interpreting cancer genomics data via the Xena platform. Nat. Biotechnol. 38, 675–678 (2020).
Győrffy, B. Survival analysis across the entire transcriptome identifies biomarkers with the highest prognostic power in breast cancer. Comput. Struct. Biotechnol. J. 19, 4101–4109 (2021).
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, https://doi.org/10.1186/s13059-014-0550-8 (2014).
Lee, J. H. et al. Enhancer RNA m6A methylation facilitates transcriptional condensate formation and gene activation. Mol. Cell 81, 3368–3385.e3369 (2021).
Li, N. et al. KIAA1429/VIRMA promotes breast cancer progression by m(6) A-dependent cytosolic HAS2 stabilization. EMBO Rep. 24, e55506 (2023).
Robinson, J. T. et al. Integrative genomics viewer. Nat. Biotechnol. 29, 24–26 (2011).
Kariri, Y. A. et al. The prognostic significance of interferon-stimulated gene 15 (ISG15) in invasive breast cancer. Breast Cancer Res. Treat. 185, 293–305 (2021).
McCarty, K. S. Jr, Miller, L. S., Cox, E. B., Konrath, J. & McCarty, K. S. Sr Estrogen receptor analyses. Correlation of biochemical and immunohistochemical methods using monoclonal antireceptor antibodies. Arch. Pathol. Lab Med 109, 716–721 (1985).
Habashy, H. O. et al. Forkhead-box A1 (FOXA1) expression in breast cancer and its prognostic significance. Eur. J. Cancer 44, 1541–1551 (2008).
Abd El-Rehim, D. M. et al. High-throughput protein expression analysis using tissue microarray technology of a large well-characterised series identifies biologically distinct classes of breast cancer confirming recent cDNA expression analyses. Int. J. Cancer 116, 340–350 (2005).
Albarakati, N. et al. Targeting BRCA1-BER deficient breast cancer by ATM or DNA-PKcs blockade either alone or in combination with cisplatin for personalized therapy. Mol. Oncol. 9, 204–217 (2015).
Aleskandarany, M. A. et al. Prognostic significance of androgen receptor expression in invasive breast cancer: transcriptomic and protein expression analysis. Breast Cancer Res Treat. 159, 215–227 (2016).
Pfaffl, M. W. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 29, e45–e45 (2001).
Lothion-Roy, J. et al. Clinical and molecular significance of the RNA m6A methyltransferase complex in prostate cancer. Front. Genet. 13, 1096071 (2023).
Shen, S. et al. rMATS: Robust and flexible detection of differential alternative splicing from replicate RNA-Seq data. Proc. Natl. Acad. Sci. USA 111, E5593–E5601 (2014).
Liao, Y., Wang, J., Jaehnig, E. J., Shi, Z. & Zhang, B. WebGestalt 2019: gene set analysis toolkit with revamped UIs and APIs. Nucleic Acids Res. 47, W199–W205 (2019).
Oliveros, J. C. Venny. An interactive tool for comparing lists with Venn’s diagrams. 2007–2015, https://bioinfogp.cnb.csic.es/tools/venny/index.html (2016).
Acknowledgements
We thank the Nottingham Health Science Biobank and Breast Cancer Now Tissue Bank for providing the tissue samples. We gratefully acknowledge the funding from the School of Veterinary Medicine and Science, University of Nottingham, philanthropic donation [36565532] and the BBSRC Doctoral Training Programme [BB/T008369/1; BB/I024291/1].
Author information
Authors and Affiliations
Contributions
A.H., J.L.R., R.T., M.H., C.W., M.A., A.D., Y.K.: performed experimental work; A.H., J.L.R., R.T., M.H., C.W., M.A., A.D., Y.K., M.T., L.G., A.G., C.A., M.D., S.I., K.H.P., S.M., N.M., C.R., R.F., E.R., J.J.: undertook data analysis and interpretation; A.H., J.L.R., R.T. wrote the main manuscript text and prepared figures; N.M., C.R., R.F., E.R.: supervision of projects; N.M. acquisition of funding. All authors reviewed and approved the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Harris, A.E., Lothion-Roy, J., Thompson, R.L. et al. Functional and clinical significance of the RNA m6A methyltransferase complex in breast cancer. npj Breast Cancer 11, 147 (2025). https://doi.org/10.1038/s41523-025-00861-5
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41523-025-00861-5
