Abstract
N6-methyladenosine (m6A), the most abundant internal RNA modification in humans, regulates most aspects of RNA processing. Prostate cancer is characterized by widespread transcriptomic dysregulation; therefore, we characterized the m6A landscape of 162 localized prostate tumors with matched DNA, RNA and protein profiling. m6A abundance varied dramatically across tumors, with global patterns emerging via complex germline–somatic cooperative regulation. Individual germline polymorphisms regulated m6A abundance, cooperating with somatic mutation of cancer driver genes and m6A regulators. The resulting complex patterns were associated with prognostic clinical features and established the biomarker potential of global and locus-specific m6A patterns. Tumor hypoxia dysregulates m6A profiles, bridging prior genomic and proteomic observations. Specific m6A sites, such as those in VCAN, drive disease aggression, associating with poor outcomes, tumor growth and metastasis. m6A dysregulation is thus associated with key events in the natural history of prostate cancer: germline risk, microenvironmental dysregulation, somatic mutation and metastasis.
Similar content being viewed by others
Main
Cancer is a disease of the genome, driven by somatic mutations1. The cancer genome is actuated by the cellular machinery of the central dogma, leading to transcriptomic and proteomic dysregulation and eventual clinical presentation. The proteomic and clinical features of cancer are not fully explained by the genome and the transcriptome, suggestive of widespread post-transcriptional and post-translational regulation2,3. Chemical modification of RNA is a leading candidate to explain these discrepancies4.
m6A is the most abundant internal chemical modification of RNA in humans. This epitranscriptomic mark regulates almost every facet of the RNA life cycle, including secondary structure formation, splicing, nuclear export, translation and degradation5. At least 28 enzymes read, write or erase m6A, facilitating dynamic transcript regulation6. Three lines of evidence suggest that m6A drives tumor initiation and progression4. First, global m6A abundance is altered relative to that of normal tissue in a cancer type-specific manner7,8. Second, mutations in m6A enzymes are common in many cancer types, and their perturbation influences tumor phenotypes in multiple model systems9,10,11,12,13. Third, specific m6A sites in oncogenes and tumor suppressors influence protein abundance via transcript stability and translation rates11,14,15.
Here, we mapped m6A modifications transcriptome wide in 162 intermediate-risk localized prostate tumors with extensive prior molecular, microenvironmental and clinical characterization2,16,17,18,19,20,21. m6A methylation defined clinically distinct tumor subgroups and influenced specific signaling pathways, with cross-talk between somatic mutational and epitranscriptomic profiles. m6A sites interacted with clinically relevant germline genotypes, with 1,350 m6A quantitative trait loci (m6A-QTLs) identified. Similarly, both global and specific m6A information reflected underlying tumor hypoxia, which is a strong prognostic factor. Consistent with this, m6A peaks predicted multiple clinical phenotypes, including patient survival. Finally, we demonstrate that specific, targetable m6A modifications functionally drive tumor growth and progression.
Results
The m6A landscape of localized prostate cancer
We quantified the epitranscriptomes of 162 primary localized prostate tumors using methylated RNA immunoprecipitation sequencing (meRIP-seq)22,23,24. All patients were clinically managed for National Comprehensive Cancer Network intermediate-risk disease via radical prostatectomy with curative intent25. Clinical annotation included pretreatment prostate-specific antigen (PSA) abundance, tumor size and extent (T category), surgical International Society of Urological Pathology (ISUP) Grade Group, biochemical relapse, subhistologies and age at diagnosis (Fig. 1a and Supplementary Table 1). Tumors were treatment naive, and samples were taken from index lesions as assessed by two uropathologists. Median follow-up was 6.72 years. Samples were profiled with multiple additional assays including tumor–reference whole-genome sequencing16,17,18,21, DNA methylation and histone H3 lysine 27 (H3K27) acetylation (H3K27ac) profiling16,18,20, RNA-seq19 and proteomics2 (Fig. 1a and Extended Data Fig. 1a).
a, Top to bottom, bar plots. The number of MeTPeak peaks across samples ordered from greatest to smallest. The number of genomic rearrangements (GRs) identified in each sample. The number of single-nucleotide variants (SNVs) identified in each sample. The PGA40 as a proxy of the total CNAs for each sample. The Conti PRS50 calculated for each sample. Top heatmap shows the associated clinical covariates including ISUP Grade Group, anatomical T category, PSA, age, IDC/CA, biochemical relapse (BCR) and metastasis. Bottom heatmap, complementary molecular profiling data collected for each sample including germline polymorphisms in samples of European ancestry18, somatic CNAs and simple somatic mutations17, DNA methylome18, H3K27ac20, ultra-deep RNA-seq19 and proteomics2. SSM, simple somatic mutation; N/A, not available. b–e, Exemplar plots for AR (the inset magnifies the region highlighted in yellow; b), MYC (c), TP53 (d) and PTEN (e). In the top polygon plots, the median IP (green) and input (purple) coverage (reads per kilobase per million mapped reads-normalized bigWig files) is represented using lines, while background colors represent the range of IP and input coverage across samples. Exons identified in GENCODE version 34 are annotated below in dark blue. Bottom heatmaps represent the distribution of IP and input coverage (log1p transformed) across samples (y axis), and darker colors correspond to greater read coverage. Samples are clustered using Euclidean distance and Ward’s minimum variance method with squared distances. chr, chromosome. f, Top bar plot, the number of peaks uniquely found in a given number of samples. Bottom scatterplot, the median adjusted m6A abundance of a joint peak (y axis) versus the number of samples in which the peak is identified (x axis). Colors indicate the deciles of the adjusted m6A abundance. Multiple joint peaks can be identified in a single gene, but the most prevalent and abundant peaks for a given gene are labeled where applicable. g, Distribution of the normalized mutual information (MI) for each data type pair. For visualization, values < 10−4 are shown as 10−4. Top heatmap indicates data type pair. Subsequent heatmaps show the number of genes with data available for each data type pair, numbers of genes for which mutual information is significant (Q < 0.1) and percentage of significant genes.
For each sample, two libraries were created and sequenced: an immunoprecipitation (IP) library generated by RNA IP with an anti-m6A antibody and an input library generated from the total RNA pool. After alignment and quantitation (Extended Data Fig. 1b), stringent quality control led to the exclusion of 14 samples (Extended Data Fig. 1c). In the final 148-patient cohort, input library results were strongly positively correlated with prior deep transcriptome sequencing (median Spearman’s ρ = 0.87; Extended Data Fig. 1d)19. Germline variants identified from input libraries validated sample identity (Extended Data Fig. 1e)18, and metrics from the ENCODE Consortium and other prior m6A studies were within expected ranges (Extended Data Fig. 1f)26.
m6A peaks were identified with both a highly specific algorithm (MeTPeak) and a highly sensitive one (exomePeak; Extended Data Fig. 2a)27,28. The number of peaks detected was modestly correlated with IP library size (Extended Data Fig. 2b), but read downsampling showed that libraries reached or approached saturation (Extended Data Fig. 2c,d)29. The vast majority of identified peaks overlapped known peaks (Extended Data Fig. 2e, inner panel)26, and the canonical RRACH motif was enriched in peaks identified in every sample30. Downstream analyses used high-specificity MeTPeak results (Fig. 1a and Extended Data Fig. 2e, outer panel).
To rationalize peaks cohort wide, we created an algorithm to integrate sample-level peaks into cohort-level joint peaks called HistogramZoo31. Briefly, peaks for each sample were aggregated onto transcript backbones, forming a coverage histogram for each transcript. These histograms were then segmented to create a set of joint peaks32,33,34,35. m6A methylation was quantitated for each sample on each peak as the number of IP reads, normalized for library size across the cohort36 and then adjusted for transcript abundance using matched input libraries37. In the full 148-patient cohort, HistogramZoo (Methods) identified 32,051 high-confidence m6A peaks across 9,571 genes; these were specifically enriched near stop codons (Extended Data Fig. 2f). The median tumor harbored 7,611 ± 2,922 m6A peaks (s.d.). Several key prostate cancer oncogenes such as MYC, AR, MALAT1 and FOXA1 had frequently methylated m6A peaks (Fig. 1b,c and Extended Data Fig. 2g,h), while tumor suppressors like TP53, PTEN and RB1 were very infrequently methylated (Fig. 1d,e and Extended Data Fig. 2i). One notable exception to the trend is NKX3-1, which was frequently methylated across samples (Extended Data Fig. 2j). About 16% of peaks were identified in a single sample (5,146 peaks), and ~20% were identified in at least half (6,467 peaks), with only 0.5% (167 peaks) detected in every sample (Fig. 1f).
To ascertain how m6A influences other aspects of the central dogma, we used mutual information, which can capture complex associations without assumptions of linearity or monotonicity38. For every gene, mutual information was calculated between each pair of molecular characteristics (Extended Data Fig. 1a). The number of genes with significant mutual information varied widely between different types of data (Fig. 1g and Supplementary Table 2). m6A was associated with RNA abundance for ~20% of analyzed genes (839 genes), with less frequent associations with copy number alterations (CNAs) (26 genes), DNA methylation (102 genes) and H3K27ac (one gene) (Fig. 1g). Genes with very low RNA abundance displayed fewer samples with a peak (Extended Data Fig. 2k), likely due to limitations in RNA signature detection. Similarly, a significant negative correlation was detected between m6A and RNA abundance (Extended Data Fig. 2l). The extent to which RNA explained protein abundance was weakly associated with mean gene-level m6A abundance but not with the number of samples for which the m6A peak was detected (Extended Data Fig. 2m,n). This provides minor support for previously reported regulatory roles of m6A in modulating RNA translation into protein23,39.
In addition to these 148 prostate tumor samples, we profiled benign tissue from seven individuals to allow tumor–normal comparisons (Supplementary Fig. 1 and Supplementary Tables 3 and 4). Furthermore, m6A-selective allyl chemical labeling and sequencing was applied to eight samples, facilitating single-nucleotide-resolution m6A profiling (Supplementary Fig. 2 and Supplementary Table 5). The analysis of these data is presented in the Supplementary Note.
m6A subtypes of localized prostate cancer
To identify global trends in m6A variation across patients, consensus clustering was performed on peak-level m6A abundance. We identified five patient subtypes (Supplementary Fig. 3a; P1 through P5) and five m6A subtypes (Supplementary Fig. 3b,c; M1 through M5; Fig. 2a and Supplementary Tables 6–8). m6A-derived patient subtypes were associated with multiple clinico-molecular features (Fig. 2b), including tumor size and extent (pathologic T category (pT); Fig. 2c), presence of the aggressive intraductal carcinoma and cribriform architecture subhistologies (IDC/CA; Fig. 2d), genomic instability assessed as the proportion of the genome with a CNA (percent genome altered (PGA)40; Supplementary Fig. 3d) and relapse rate after surgery (Fig. 2e; unadjusted P values shown) but not with tumor purity (Fig. 2a and Supplementary Fig. 3e).
a, Clustering of the top quartile of interquartile range (IQR)-ranked m6A peaks (n = 5,203) identified five patient subtypes and five m6A subtypes. Clinical covariates are shown to the right of the heatmap. Heatmap coloring indicates z-score-scaled peak intensities. b, Association between m6A patient subtypes and clinical features. Q values are from Pearson’s χ2 test for categorical variables and a one-way ANOVA for continuous variables. c,d, Overlap of m6A patient clusters with pathologic tumor extent (pT) (c) and IDC/CA status (d). Patient clusters are indicated in rows, and clinical features are shown in columns. Row and column totals are depicted in the right and bottom heatmaps. Independence of clusters and clinical variables was assessed via Pearson’s χ2 test (P value shown). The relative intensity of blue shading indicates the size of the group. e, Biochemical relapse rate across the five m6A patient subtypes. A Cox proportional hazards model was fit with P1 as the baseline group. Hazard ratios (HR) and P values are shown with confidence intervals in parentheses. f–h, Association between m6A patient subtypes and mutations in m6A regulators. f, Bar plot shows Q values from Pearson’s χ2 test. ETS, E26 transformation-specific (ETS) transcription factor family members. g, Overlap of m6A patient clusters with presence of copy number gain of YTHDF1. h, Overlap of m6A patient clusters with presence of copy number gain of FANCA. Patient clusters are indicated in rows, and mutations are shown in columns. Row and column totals are depicted in the right and bottom heatmaps. Independence of clusters and mutations was assessed via Pearson’s χ2 test (P value shown). The relative intensity of blue shading indicates the size of the group. i, BORCS6 m6A peak abundance varies by pISUP Grade Group. P value from one-way ANOVA is shown (n = 148). Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5× the IQR. j, NSD1 m6A peak abundance is greater with presence of IDC/CA. P value from two-sided U-test is shown; samples where IDC status was missing have been removed (n = 133). Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5× the IQR.
m6A subtypes were related to known messenger RNA (mRNA) and CNA subtypes (Supplementary Fig. 3f,g), which themselves are associated with genomic instability and patient outcome16,19. The m6A subtypes were characterized by patterns of somatic CNAs in m6A writers, readers and erasers (Fig. 2f) including gain of YTHDF1 (Fig. 2g) and loss of FANCA (Fig. 2h). Thus, there are multiple m6A subtypes with distinct clinical presentations, which partially overlap previous DNA and RNA subtypes.
Clinical correlates of locus-specific m6A
Several clinical features of prostate cancer influence patient management, including age at diagnosis, pathological tumor extent (pT), pretreatment serum PSA abundance, pathological ISUP (pISUP) Grade Group and aggressive intraductal carcinoma subhistology (IDC/CA). For each clinical indicator, we identified individual m6A peaks associated with their status (Q < 0.1; Extended Data Fig. 3a). Six m6A peaks were associated with age (Spearman’s rank correlation, Q < 0.1; Extended Data Fig. 3b and Supplementary Table 9), eight peaks were associated with grade (one-way ANOVA, Q < 0.1; Fig. 2i), and 13 were associated with IDC/CA (two-sided Mann–Whitney U-test, Q < 0.1). These included NSD1, which enhances AR transactivation (Fig. 2j)41. The number of m6A peaks in a tumor was elevated in IDC/CA-positive disease (Extended Data Fig. 3c,d), and loss of the tumor suppressor genes FANCA and TP53 was more common in patients with more peaks (two-sided Mann–Whitney U-test, P < 0.05; Extended Data Fig. 3e). Thus, locus-specific m6A patterns reflect clinical features of prostate cancer.
Heritable tumor-specific m6A regulation
Given that prostate cancer is the most heritable solid cancer42, we next considered whether germline genetics influence m6A methylation. First, we examined the potential cis regulatory effects of germline polymorphisms on m6A. Across all patients, around six million unique SNPs were identified18. Of these, 4,755 were located at the specific residue of a previously annotated m6A site43,44. We identified 35 germline polymorphisms that (1) overlapped an m6A peak identified in this cohort, (2) occurred at an m6A annotated site and (3) had an A allele on the transcribed strand (Extended Data Fig. 4a). A alleles in active m6A sites were present at much higher frequencies than non-A alleles (medianA, 0.83; rangeA, 0.11–0.95; mediannon-A, 0.45; rangenon-A, 0.053–0.95) (Extended Data Fig. 4b). A lower A allele frequency was observed at inactive m6A sites: that is, annotated sites lacking an m6A peak in localized prostate cancer (Extended Data Fig. 4b). This imbalance suggests that non-A alleles are associated with reduced m6A abundance. A quantitative analysis identified many SNPs associated with allelic imbalance of IP and/or input reads at the m6A site (Fig. 3a; paired t-test, Q < 0.1). These were often correlated with clinical features. For example, a SNP in GPR107 was associated with biochemical recurrence (Cox proportional hazards model, hazard ratio = 13.8; confidence interval, 3.2–60.3; Q = 6.3 × 10−3; ExaLT P = 2.0 × 10−2; Extended Data Fig. 4c). rs2240912 is neither a hit in prior GWAS studies45,46,47,48 nor in linkage disequilibrium with a GWAS SNP from those studies, likely due to the low minor allele frequency. Validation of its molecular effects and clinical consequences will be needed to better estimate its associations with disease aggression.
a, Allelic imbalance of reads at SNPs in annotated m6A sites in heterozygous samples in either the IP or input library. Allelic imbalance is evaluated using a paired t-test at a statistical threshold of Q < 0.1. Effect size (log2 (fold change (FC))) is calculated with respect to the A allele. Effect size and direction are represented by the size and color of the disks, respectively. Statistical significance is represented by the background color. b, Distribution of P values in associations of m6A peak status with Conti PRS (t-test)50. c, m6A methylation of SRRM2 (peak 4) is associated with a lower Conti PRS (t-test). Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5× the IQR. n = 133. d, Significant prostate cancer risk SNP local m6A-QTLs identified using a linear additive model. Effect size (β) and direction are represented by the size and color of the disks, respectively. Statistical significance is represented by the background color. e,f, Two local risk SNP quantitative trait loci are depicted: the associations of genotype with m6A methylation for B3GAT1 (rs878987 | peak 13) (e) and RAB29 (rs708723 | peak 1) (f), respectively. Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5× the IQR. n = 133. g, Manhattan plot of genome-wide m6A-QTL analysis. Results from a linear additive model implemented in Matrix eQTL83. Points representing the tag SNPs of m6A-QTLs where the gene has been annotated in the Cancer Gene Census56 or by Armenia et al.57, Fraser et al.16 and Quigley et al.58 are indicated with black squares, while the SNPs of significant (Q < 0.1) m6A-QTLs where the SNP has been annotated in RMVar43 or by Cotter et al.44 are indicated with black diamonds. Labeled black disks selectively identify top hits that fall into neither of the former categories. SLC9A3R2 (NHERF2). h, A genome-wide significant m6A-QTL: SNP rs4951018 with SLC45A3 m6A abundance. Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5× the IQR. n = 133.
Next, to understand potential epitranscriptomic alterations characteristic of prostate cancer incidence, we evaluated the association of m6A peaks with two polygenic risk scores (PRSs) for incidence49,50 and one for hazard (PHS290)51 (Fig. 3b, Extended Data Fig. 4d–f and Supplementary Table 10). The number of m6A peaks across samples was weakly associated with polygenic risk (Fig. 1a and Extended Data Fig. 4g). The P-value distribution suggested that many subthreshold associations of m6A peaks with the incidence PRS remain to be elucidated in larger cohorts (Fig. 3b), and an m6A site in SRRM2 was associated with genetic risk (Fig. 3c and Supplementary Table 11).
We then related the 272 individual risk SNPs to m6A peaks within 10 kbp and to more distal m6A peaks identified using three-dimensional spatial genomic data52 (Extended Data Fig. 4h,i). We identified 13 local prostate risk SNP m6A tags, including six associated with peaks in B3GAT1 and one with a peak in RAB29 (Fig. 3d–f and Supplementary Table 12).
To generalize this observation, we identified m6A-QTLs genome wide, comparing each m6A peak to all SNPs within 10 kbp. This yielded 14,775 significant m6A-QTLs representing 1,350 unique peaks (Q < 0.1; Fig. 3g and Supplementary Table 13). Eleven percent (151 of 1,350) of the m6A-QTL SNPs overlap with the peaks they are associated with (Extended Data Fig. 4j), in concordance with previous studies showing that most of the m6A-QTLs are not located within m6A peaks53,54. Among these SNPs, 1% (14 of 1,350) overlap with methylated ‘A’ sites in the peaks (Supplementary Table 13). For SNPs that are located in regulatory regions, applying publicly available55 and in-house H3K27ac high-throughput chromosome conformation capture with chromatin immunoprecipitation (HiChIP) data, we found evidence of physical interactions between the m6A-QTL SNPs and the associated peaks in 4% (60 of 1,350) of SNP–peak pairs (Supplementary Table 13). These m6A-QTLs affected 1,017 genes, including 60 cancer or prostate cancer driver genes16,56,57,58,59. Of the 1,350 m6A-QTLs, 101 were associated with RNA abundance changes of the corresponding gene, nine with protein abundance changes and four with both (Q < 0.1; Extended Data Fig. 4k). For example, rs4951018 was associated with changes in m6A, RNA and protein abundances of SLC45A3, a component of the common prostate cancer fusion gene SLC45A3-ELK4 (Fig. 3h and Extended Data Fig. 4l–n). Three independent m6A-QTL SNPs (rs143089027, rs57557217 and rs76338659) were associated with pISUP Grade Group (Pearson’s χ2 test, Q < 0.1; Extended Data Fig. 4o). In sum, these data reveal broad germline regulation of m6A sites, including on cancer driver genes.
Broad m6A dysregulation by tumor hypoxia
Hypoxia is an adverse prognostic feature in localized prostate cancer associated with grade, IDC/CA and disease relapse17. We identified 2,280 hypoxia-associated m6A peaks (Spearman’s rank correlation, Q < 0.05; Fig. 4a and Supplementary Table 14). Hypoxia-associated peaks tended to be less abundant in hypoxic tumors and preferentially occurred on genes involved in gene expression regulation (Fig. 4a). The RNA abundance of m6A regulators was also correlated with hypoxia (Spearman’s rank correlation, P < 0.1; Fig. 4a). Further highlighting the widespread nature of the hypoxia–m6A relation, m6A patient subtypes (Fig. 2a) were hypoxia associated, with the mutationally quiet P5 cluster being the most normoxic (one-way ANOVA, P = 8.1 × 10−3; Fig. 4b). Eight peaks showed additional mutational or clinical associations, including a peak located on NSD1 (Figs. 2j and 4c,d and Supplementary Table 9).
a, Hypoxia correlates with m6A peak abundance. Top bar plot indicates sample hypoxia score. Left–middle heatmaps show the corresponding z-scored m6A peak abundance per sample for peaks with a hypoxia correlation (Spearman’s correlation, Q < 0.05, 1,184 peaks). Peaks are clustered by abundance values, and the two largest resulting clusters are displayed in separate heatmaps. Right–middle heatmaps display m6A peak annotations. ‘RNA–hypoxia correlation’ indicates whether RNA abundance for the corresponding genes is also correlated with hypoxia. Hypoxia-correlated peaks are enriched for several biological pathways, and subsequent columns indicate whether each corresponding gene is tagged with each of the enriched pathway terms. Bottom heatmap shows z-scored RNA abundance for m6A writers, readers and erasers. Right bar plot shows P values for Spearman’s correlation between RNA abundance and hypoxia. Pol, polymerase. b, Patient m6A subtypes are associated with varying hypoxia levels. P value from one-way ANOVA. Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5× the IQR. n = 146. c,d, Associations between m6A abundance for an example peak on NSD1. Hypoxia (n = 146) (c), PGA (n = 148) (d). Plots annotated with results from Spearman’s correlation. e, Overlap between hypoxia-specific peaks in V16A and PC-3 cells. f,g, Patient hypoxia-associated peaks significantly overlap with hypoxia-specific peaks in prostate cancer cell lines. Contingency tables showing results from two-sided Fisher’s exact test for association between patient hypoxia-associated peaks and V16A (f) and PC-3 (g) hypoxia-specific peaks. The relative intensity of blue shading indicates the size of the group. OR, odds ratio. h, Correlation between tumor m6A abundance and hypoxia for a peak region in HNRNPLL. This peak shows a corroboratory hypoxia association in both V16A and PC-3 cell lines. Plots annotated with results from Spearman’s correlation (n = 146).
To validate these hypoxia-associated m6A peaks, we exposed two prostate cancer cell lines (V16A and PC-3) to 0.2% O2 for 24 h and profiled m6A using the same approach. Both cell lines had many hypoxia-specific peaks (Fig. 4e and Supplementary Table 15), with these peaks significantly overlapping the hypoxia-responsive m6A peaks in primary patients (Fig. 4f,g). Nine peaks showed a consistent hypoxia correlation across both cell lines in addition to patient samples, including HNRNPLL (Fig. 4h). Thus hypoxia correlates with m6A profiles in patients and model systems.
m6A is a biomarker of prostate cancer patient outcome
We next sought to evaluate the biomarker potential of m6A. The enzymes that read, write or erase m6A are frequently altered by somatic mutations, with a majority of tumors having a mutation affecting an m6A enzyme (Fig. 5a and Supplementary Table 16). Copy number loss of these genes was accompanied by reduced RNA and protein abundance in several instances (Extended Data Fig. 5a), and these modulations were associated with pISUP Grade Group for loss of METTL16 and gain of VIRMA, HNRNPA2B1 and YTHDF3 (Extended Data Fig. 5a). A majority (seven of nine) of these associations were replicated in the Cancer Genome Atlas (TCGA) Prostate Adenocarcinoma (PRAD) cohort (N = 421) (Supplementary Table 17). Similarly, RNA abundance of HNRNPA2B1 and YTHDC2 was also associated with ISUP Grade Group (Extended Data Fig. 5b).
a, m6A regulatory genes are frequently mutated. m6A enzyme-encoding genes are shown on the y axis, and patient samples are on the x axis. Top bar plot represents the number of mutations per sample, while the right bar plot shows the number of mutations in each m6A regulatory gene across samples. Central heatmaps display the mutations in each enzyme for each sample. b–d, Meta-analysis of six independent patient cohorts identifies mutations in m6A regulatory genes as predictive of patient biochemical recurrence. b, Left heatmap depicts m6A regulatory gene, mutation type and category of m6A regulatory action. Forest plot shows hazard ratios and 95% confidence intervals of each mutation on biochemical recurrence, and left bar plot indicates corresponding P values from a Cox model. The far right bar plot displays average sample mutation rate across the patient cohorts (protein, n = 54, m6A, n = 148; RNA, n = 92; gain or loss, n = 146). c,d, Gain of VIRMA (c) and gain of YTHDF3 (d) predict decreased biochemical recurrence-free rate. VIRMA gain, n = 22; loss, 191. YTHDF3 gain, n = 23; loss, 186. P values displayed are from Cox proportional hazards models. e–i, m6A provides complementary prognostic information in profiling prostate cancer driver genes. e, Left heatmap shows driver genes and molecular data type analyzed. Forest plot shows hazard ratio and 95% confidence intervals for biochemical recurrence, and the right bar plot indicates corresponding P values from a Cox model. f,g, Influence of m6A and copy loss on biochemical recurrence-free rate for tumor suppressors TP53 (f) and NKX3-1 (g). Kaplan–Meier survival curves show the results from median dichotomizing m6A abundance to create four patient groups, while inset forest plots show results from treating m6A as a continuous variable. P values displayed are from Cox proportional hazards models. h,i, Relative m6A abundance by biochemical recurrence status for TP53 (h) and NKX3-1 (i). Plots show copy number, z-scored RNA, m6A and protein abundance, and biochemical recurrence status at censoring time. Patients are depicted in columns and are ordered by biochemical recurrence status followed by m6A abundance. j,k, m6A peak status predicts biochemical recurrence for specific peak sites on INHBA, VCAN and ZFHX4. j, Patients are represented in columns. Top heatmap displays biochemical recurrence status at censoring time. Middle heatmap shows patient peak status (peak sites depicted in row labels). Right bar plot shows P values from a Cox model. k, Presence of a VCAN m6A peak increases risk of biochemical recurrence. VCAN peak, n = 23; no peak, 125. P values displayed are from Cox proportional hazards models.
To further investigate the clinical importance of these m6A regulator mutational profiles, we analyzed six independent cohorts comprising 1,239 patients16,17,59,60,61,62. CNAs were significantly predictive of disease relapse (Fig. 5b), with the strongest effect sizes being observed for gain of the m6A regulators encoded by VIRMA and YTHDF3 (Fig. 5b–d and Supplementary Table 18). Furthermore, CNAs in prognostic m6A regulatory genes co-occurred with spatially distant prostate cancer driver events, providing evidence of positive selection (Extended Data Fig. 5c). Of these significantly co-occurring driver–m6A CNA events, 52% (29 of 56) replicated in the TCGA-PRAD cohort (Supplementary Table 19).
Mutations in m6A enzymes may determine patient outcomes through direct modification of site-specific m6A. We investigated whether these mutations, in addition to canonical prostate cancer driver events, were associated with differential methylation of m6A peaks. Of the 6,467 m6A peaks identified in at least 50% of samples, over half (3,432 peaks) were differentially methylated in relation to a prostate cancer driver or m6A regulator, with clustering by effect size resulting in five peak clusters and five mutation clusters (MC; Extended Data Fig. 5d and Supplementary Table 20). A subset of m6A clustering reflects known mutational co-occurrence, such as for YTHDF1 gain and FANCA loss (MC4, Extended Data Fig. 5c,d). In many cases, unrelated somatic mutations yielded similar m6A profiles, suggesting convergent epitranscriptomic dysregulation (for example, ZC3H13 and SPOP (MC5)). Co-regulated peaks were enriched for specific biological functions. MC1 preferentially dysregulates several pathways relating to cellular organization, while MC4 was associated with regulation of nucleic acid metabolism and RNA methylation (Extended Data Fig. 5e and Supplementary Table 21). Thus, in addition to predicting patient outcome, the somatic mutation landscape influences the epitranscriptome in a driver- and regulator-specific manner.
We next examined the clinical importance of locus-specific m6A on known prostate cancer drivers. A comprehensive survival analysis of canonical prostate cancer driver genes across available biomolecular data types was performed (Fig. 5e and Supplementary Table 22). Consistent with previous findings, CNAs were predictive of biochemical recurrence for a subset of driver events (MYC, PIK3R1 and PTEN16,63). For tumor suppressors TP53 and NKX3-1, gene-level m6A abundance provided significant prognostic information (Fig. 5e). Median dichotomization of m6A revealed a trend toward increasing risk of biochemical recurrence upon TP53 loss in the context of higher m6A abundance. Further investigation identified a significant interaction effect, with a hazard ratio of 3.35 per standard deviation increase in log2 (m6A) abundance in patients with TP53 loss (Fig. 5f,i). Finally, we expanded the biomarker evaluation transcriptome wide by considering the 20,334 m6A peaks identified in at least six patients. Strong associations with biochemical recurrence were detected for peaks on the INHBA, VCAN and ZFHX4 transcripts (Fig. 5j, Extended Data Fig. 5f and Supplementary Tables 9 and 23). Additional peaks present on these transcripts showed a similar trend for association with biochemical recurrence but with nonsignificant effect sizes (Fig. 5j). The presence of each peak was associated with worse patient outcome (Fig. 5k and Extended Data Fig. 5g,h). In summary, m6A provides prognostic value in patients with prostate cancer via both epitranscriptomic and somatic mutational information.
m6A modification of VCAN associates with prostate cancer progression
To gain deeper insights into the intricate mechanisms underlying the relationship between m6A peaks and tumorigenesis, we conducted a more detailed exploration. The RNA abundances of VCAN, INHBA and ZFHX4 were significantly correlated in two large patient cohorts (Extended Data Fig. 6a). VCAN m6A peaks were observed in ~15% of tumors, while INHBA and ZFHX4 peaks were observed in 5%. A second VCAN peak correlated well with tumor hypoxia (Supplementary Table 14), prompting further investigation.
VCAN encodes a large chondroitin sulfate proteoglycan called versican (VCAN). It is a key component of the extracellular matrix64; the secretion of VCAN by fibroblasts promotes prostate cancer invasion65. The mRNA abundance of VCAN is significantly higher in tumors with m6A peaks than in those without (Extended Data Fig. 6b). The abundance of VCAN protein is also higher in tumors with m6A peaks (Extended Data Fig. 6c). Among the five prostate cancer cell lines tested, PC-3 had the highest level of VCAN m6A (Extended Data Fig. 6d), comparable to patient samples. High abundance of VCAN mRNA was significantly associated with a worse outcome in five cohorts comprising 981 independent patients (Fig. 6a and Extended Data Fig. 6e).
a, Kaplan–Meier survival curves showing biochemical recurrence stratified by VCAN mRNA abundance across five cohorts (Gerhauser, Ross-Adams, Taylor, TCGA-PRAD and International Cancer Genome Consortium (ICGC)-PRAD-CA (CA, Canada)), grouped by median abundance of VCAN mRNA. A Cox proportional hazards model was used to compare the hazard of biochemical recurrence between the groups. b, Kaplan–Meier survival curves showing biochemical recurrence stratified by VCAN average optical density from the tissue microarrays (TMA), grouped by median VCAN average optical density. A Cox proportional hazards model was used to compare the hazard of biochemical recurrence between the groups. c, Quantification of a migration and invasion assay after shRNA-mediated knockdown of VCAN in PC-3 cells. Data are represented as mean ± s.d. of three biological replicates. P values were calculated using one-way ANOVA with Dunnett’s multiple-comparisons test. d, Tumor growth of xenografts derived from injected PC-3 cells infected with shRNAs targeting VCAN or green fluorescent protein (GFP) as the control. Data are shown as mean ± s.e.m. of five biological replicates. P values are from one-way ANOVA with Dunnett’s multiple-comparisons test. e, Writing efficiency was assessed by m6A meRIP-quantitative PCR (qPCR) in PC-3 cells. SETD7 was introduced here as a negative control. Data are represented as mean ± s.d. of three biological replicates. P values are calculated using one-way ANOVA with Dunnett’s multiple-comparisons test. g, guide RNA; gNT, non-targeting guide RNA. f, VCAN protein abundance after dCasRx-METTL3-based m6A writing in PC-3 cells was detected by western blot. Ponceau S stain serves as the loading control. Numbers on the right indicate the positions of molecular mass (kDa) standards. g,h, Migration and invasion of PC-3 cells after writing by the dCasRx-METTL3-based RNA-editing system guided by VCAN mRNA-targeting guide RNA mixes. Representative images (g) and quantification (h, mean ± s.d. of three biological replicates; one-way ANOVA with Dunnett’s multiple-comparisons test) are shown. Scale bar, 200 μm. i, RIP–qPCR showing the physical association of VCAN mRNA with IGF2BPs in PC-3 cells. Data are represented as mean ± s.d. of three biological replicates. P values were calculated using an unpaired two-sided Student’s t-test. IgG, immunoglobulin G. j, Western blot showing VCAN protein abundance change after knocking down each IGF2BP protein (siIGF2BP1, siIGF2BP2 and siIGF2BP3) or all three IGF2BPs (siIGF2BP123) in PC-3 cells. Ponceau S stain serves as the loading control. Numbers on the right indicate the positions of molecular mass (kDa) standards. Ctrl, control. k, Graphical representation of VCAN regulation by m6A.
VCAN protein abundance showed a trend but was not significantly associated with patient survival, possibly due to the relatively small sample size (Supplementary Fig. 4a). To address this limitation, we performed immunohistochemistry staining of VCAN in a tissue microarray with a larger sample size (n = 154). The staining results revealed the presence of VCAN in both tumor cells and stromal cells, albeit in modest amounts, with its primary localization observed in the extracellular matrix (Supplementary Fig. 4b). Importantly, patients with higher levels of VCAN had significantly poorer survival outcomes than those with lower levels (Fig. 6b and Supplementary Table 24). Furthermore, we observed a significant increase in the total amount of VCAN protein in tumors compared to normal tissues (Supplementary Fig. 4c). Additionally, VCAN abundance was notably elevated in high-risk groups compared to low- and intermediate-risk groups (Supplementary Fig. 4d). These findings highlight the prognostic value of VCAN in prostate cancer.
Knockdown of VCAN using short hairpin RNA (shRNA) or small interfering RNA (siRNA) significantly reduced cell proliferation, migration and invasion in PC-3 cells (Fig. 6c, Supplementary Fig. 4e and Extended Data Fig. 7a–g). VCAN knockdown with shRNA significantly reduced xenograft tumor growth in vivo (Fig. 6d and Extended Data Fig. 7h). Cell extravasation, which is a key step during cancer metastasis, was reduced after VCAN suppression in vivo in a chick embryo model (Extended Data Fig. 7i and Supplementary Videos 1–3). RNA-seq of PC-3 mouse xenografts with and without shRNA-mediated knockdown of VCAN identified 2,050 differentially expressed genes (Extended Data Fig. 7j), with downregulated genes enriched in cell adhesion, consistent with decelerating cell growth and motility (Extended Data Fig. 7k). In sum, in vitro, in vivo and clinical data suggest that VCAN drives aggressive prostate cancer.
VCAN-mediated cross-talk in prostate cancer progression
VCAN is recognized as a pivotal constituent in prostate stroma, secreted by fibroblasts, and exerts substantial influence on tumor initiation and progression65,66. To explore cross-talk between tumor and stromal cells in VCAN-mediated prostate cancer progression, we used the human prostate fibroblast cell line WPMY-1, which expresses an abundance of VCAN comparable to PC-3 cells (Extended Data Fig. 8a). Knocking down VCAN in WPMY-1 cells significantly reduced proliferation, underscoring its pivotal role in facilitating cellular proliferation (Extended Data Fig. 8b). Co-culturing PC-3 and WPMY-1 cells enhanced PC-3 motility in a dose-dependent manner with increasing WPMY-1 cell numbers (Extended Data Fig. 8c,d). Furthermore, co-culturing PC-3 cells with WPMY-1 cells with either VCAN or control knockdown (Extended Data Fig. 8e,f) showed that WPMY-1 could rescue the phenotype caused by VCAN knockdown (Extended Data Fig. 8g,h), suggesting that the phenotype was VCAN dependent. These findings highlight the complex interplay between tumor and stromal cells mediated by VCAN in prostate cancer progression.
m6A on VCAN drives prostate cancer aggression
Given that tumors with m6A peaks have higher VCAN mRNA abundance, we hypothesize that the m6A modification stabilizes VCAN mRNA. We initially observed that VCAN mRNA abundance decreased with METTL3 knockdown and increased with ALKBH5 and FTO knockdown (Extended Data Fig. 9a–c). To confirm that VCAN m6A modifications directly influence VCAN gene expression, we used catalytically dead CasRx (dCasRx)/METTL3-based programmable site-specific base editing on VCAN transcripts. meRIP–qPCR showed that, guided by the VCAN-targeting guide RNAs, the modification level on VCAN was significantly increased, while SETD7 as a negative control was unchanged (Fig. 6e). After m6A writing, VCAN mRNA and protein abundance was significantly increased (Fig. 6f and Extended Data Fig. 9d), consistent with perturbation of m6A regulators. In contrast to knockdown of VCAN, writing additional m6A modification enhanced proliferation and migration of PC-3 cells (Fig. 6g,h and Extended Data Fig. 9e). These data suggest that specific m6A modifications on VCAN stabilize its mRNA and promote subsequent translation to functional protein.
To elucidate the regulatory mechanisms of m6A modification on VCAN expression, we investigated the involvement of m6A readers in regulating VCAN mRNA or protein abundance. YTHDF1 was initially considered a potential regulator of VCAN expression. However, analysis of publicly available YTHDF1 RNA immunoprecipitation sequencing (RIP-seq) data, including in PC-3 cells, revealed that VCAN is not a target of YTHDF1 (refs. 67,68,69,70). Furthermore, knockdown of YTH domain readers (YTHDF1–YTHDF3) in PC-3 cells did not affect VCAN protein abundance (Extended Data Fig. 9f–j). However, despite the lack of involvement of YTH domain-containing readers in VCAN regulation, previous studies have reported VCAN as a high-confidence target shared by the m6A readers IGF2BP1–IGF2BP3, which stabilize target mRNA and promote translation71. We tested whether the IGF2BP proteins also bind VCAN mRNA in PC-3 cells by the RIP assay. Significant enrichment of VCAN was detected in all three pulldowns (Fig. 6i and Extended Data Fig. 9k). Knockdown of these three readers also reduced VCAN mRNA and protein abundance (Fig. 6j and Extended Data Fig. 9l,m). Consistent with VCAN knockdown, cell proliferation and invasion were also significantly reduced upon suppression of these three readers (Extended Data Fig. 10a–c). Ribosome profiling revealed increased VCAN abundance in the polysome-free fraction upon IGF2BP knockdown (Extended Data Fig. 10d,e). A rescue assay introducing m6A modification onto VCAN mRNA under the background of IGF2BP2 knockdown restored VCAN expression and cell phenotypes (Extended Data Fig. 10f). These data support a mechanism for m6A modifications to increase VCAN mRNA abundance by stabilizing and promoting its translation through IGF2BP proteins, enhancing prostate cancer cell aggression (Fig. 6k).
Discussion
Post-transcriptional modifications such as m6A facilitate tight regulation of gene expression and are essential for normal tissue physiology67,72,73,74. Germline and somatic genetics directly influence site-specific m6A levels on thousands of transcripts, while m6A levels influence downstream RNA and protein abundance. Our data provide an initial functional characterization of m6A peaks and highlight the complexity of these regulatory relationships and the need for additional functional analyses.
Hundreds of studies have sought to identify genomic or transcriptomic features of prostate cancer or its microenvironment that can predict variable aggressiveness of diseases2,16,17,18,19,21,57,58,59,60,63,75,76,77,78,79,80,81. Our data support the hypothesis that the epitranscriptome is clinically predictive: it is influenced by germline, somatic and microenvironmental features (for example, hypoxia) and shows correlation with clinical behavior. It appears to synergize with other biomarkers and may be particularly useful with long-read, single-molecular sequencing technologies that simultaneously quantify abundance, structure and the epitranscriptome82. m6A may not only predict clinical behavior but directly drive aggressive phenotypes, such as proliferation, migration and invasion.
Our study demonstrates the role of IGF2BPs in recognizing m6A modifications on VCAN. Nevertheless, we acknowledge the possibility of other m6A readers and their potential collaboration in regulating VCAN mRNA or protein abundance. Future investigations in different cell lines and with additional m6A readers are warranted to comprehensively elucidate the regulatory landscape of VCAN.
While our study represents the most extensive m6A profiling endeavor in prostate cancer to the best of our knowledge, the use of a single discovery cohort (n = 148) limits the generalizability of our findings. Future studies with larger cohorts incorporating diverse ancestries are warranted to establish the broader clinical importance of m6A dysregulation in prostate cancer progression.
Our results propose m6A modification as a central player in the life history of cancer, influencing its initiation by intermediating genetic factors, being dysregulated by somatic mutations and ultimately predicting and driving aggressive clinical and molecular cancer biology.
Methods
Ethics statement and patient consent
All samples obtained in this study complied with the relevant ethical regulations approved by the institutional ethics committee and Research Ethics Board at the University Health Network (UHN). All procedures were performed in accordance with the International Guidelines for the Use of Animals and approved by the Animal Care Committee at the UHN. Informed consent was obtained at clinical follow-up and was consistent with local Research Ethics Board and ICGC guidelines.
Cohort development
Patient selection, tumor sectioning and sample processing followed established methods16. Tumor tissues had been collected following UHN Research Ethics Board and CHU de Québec-Université Laval-approved study protocols (UHN 06-0822-CE, UHN 11-0024-CE, CHUQc-UL 2012-913:H12-03-192). Patients were treated via radical prostatectomy, with biochemical recurrence after radical prostatectomy defined as two consecutive measurements of PSA > 0.2 ng ml−1 or the administration of salvage radiotherapy. All patients were classified as intermediate risk by National Comprehensive Cancer Network standards. ISUP score and tumor cellularity were evaluated by two genitourinary pathologists (T.v.d.K. and B.T.) using scanned hematoxylin- and eosin-stained slides from surgical samples16. This categorization yielded a distribution of nine patients in G1, 87 in G2, 36 in G3, eight in G4 and four in G5, in which G1–G5 classification determined prostate cancer-specific mortality. Tissue areas with >70% tumor cellularity were selected for macrodissection and multiomic data generation16. Cellularity was validated in silico using qpure version 1.1 with OncoScan array data84.
m6A meRIP and RNA sequencing
The m6A meRIP assay was performed as described in the ‘Refined RIP-seq’ protocol22. Briefly, 4.5 ng of Escherichia coli K-12 total RNA was added to 1 μg of total RNA for each sample as a spike-in control. The RNA samples were fragmented into ~180-nucleotide-long pieces at 70 °C with RNA Fragmentation Buffer. The reaction was stopped with EDTA, followed by ethanol precipitation. Ten percent of the fragmented RNA was saved as input. Protein A magnetic beads (30 μl, Thermo Fisher, 10002D) and 30 μl protein G magnetic beads (Thermo Fisher, 10004D) were mixed and washed twice using 1 ml IP buffer (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% IGEPAL CA-630) and resuspended in 500 μl IP buffer. Next, 5 μg anti-m6A antibody (Millipore, ABE572) was added to the beads, and samples were incubated at 4 °C for at least 6 h with constant rotation. The beads were then washed twice with 1 ml IP buffer and resuspended in 500 μl IP mixture containing 100 μl of 5× IP buffer, fragmented RNA and 5 μl RNasin Plus RNase Inhibitor (Promega, N2611) and then incubated at 4 °C for another 2 h. The beads were then washed twice with cold IP buffer, twice with cold low-salt buffer (10 mM Tris-HCl, pH 7.5, 50 mM NaCl, 0.1% IGEPAL CA-630) and twice with cold high-salt IP buffer (10 mM Tris-HCl, pH 7.5, 500 mM NaCl, 0.1% IGEPAL CA-630) for 10 min each at 4 °C. After the washing steps, RNA enriched on the beads was eluted from the beads with 200 μl RLT buffer supplied in the RNeasy Mini Kit (Qiagen, 74106) and then cleaned up with an RNeasy MinElute spin column. The anti-m6A antibody-enriched RNA was eluted with 14 μl ultrapure water.
MeRIP-seq libraries were prepared by using the SMARTer Stranded Total RNA-Seq Kit version 2 (Pico Input Mammalian, Takara, 634413) according to the manufacturer’s protocol, which incorporates an rRNA depletion step. This involves the targeted cleavage of rRNA by ZapR in the presence of the mammalian-specific R-Probes. Libraries for siRNA-mediated knockdown of VCAN in PC-3 cells were prepared by using the TruSeq Stranded mRNA kit (Illumina, 20020594). The libraries were sequenced on Illumina HiSeq X Ten with paired-end 150-bp read length to a depth of at least 30 million reads per sample. Samples were mapped, and counts were quantified using STAR (version 2.7.2), and differentially expressed genes were identified using DESeq2 (version 1.26.0)36. Padjusted < 0.05 and |FC| > 1.5 were used to define differentially expressed genes. GO enrichment analysis for significantly upregulated or downregulated genes after knocking down VCAN was performed using the R package clusterProfiler (version 3.14.3)85. Benjamini–Hochberg multiple-testing adjustment was used (Q value < 0.1).
Processing meRIP-seq data and quality control
Reads were trimmed using cutadapt (version 2.8) using adaptor sequences86. Samples were mapped, and reads were quantitated at the gene level using STAR (version 2.7.2) using two-pass mapping and default parameters for hg38 using annotations from gencode.v34.chr_patch_hapl_scaff87,88. Using the PCAWG pipeline as a reference, different parameters were chosen for the parameters outFilterMultimapNmax, the threshold for multiple mapping reads, and outFilterMismatchNoveReadLmax, the threshold for an acceptable number of mismatches, to determine the robustness of mapped reads89. The quality of each sample was determined using FastQC (version 0.11.8)90, RSeQC (version 3.0.1)91, Picard (version 2.21.7) and the ENCODE QC guidelines as described in ref. 26 and the output metrics of STAR. Using 26 QC metrics, the sum of negative z scores was computed, and 14 samples failing a number of QC metrics were excluded (Extended Data Fig. 1b). Input reads were counted using RSEM (version 1.3.2)92. Concordant with past observation, a high percentage (median: IP, 84.32%; input, 47.86% and range: IP, (0.51, 0.97); input, (0.28, 0.94)) of duplicated reads was observed26, potentially exacerbated by the use of low-input starting biological material93, but, as RNA-seq is prone to natural read duplicates, the duplicated reads were retained in subsequent peak calling and m6A quantitation. Germline variants were called using the GATK pipeline94,95,96 and compared to variants called by Houlahan et al.18 using vcfeval (rtg-tools, version 3.10.1) to verify sample identity97,98. TPMs were compared between the 92 overlapping samples from this study and previous ultra-deep bulk RNA-seq19.
Peak calling, m6A quantitation and motif identification
Peak calling employed MeTPeak (version 1.1) and exomePeak (version 2.16.0) with default parameters27,28. As MeTPeak is specialized for single-end sequencing, forward reads were extracted and used as input. Subsampling was conducted with Picard DownsampleSam (version 2.21.7) and three seeds (314,159, 271,828, 1,618) at 50%, 25%, 10% and 1%. Motifs were called for each sample using HOMER (version 4.1.1)30.
Joint peaks were called using HistogramZoo (version 1.0.1). Briefly, the coverage of gene exons by sample-level peak calls using MeTPeak was computed on a gene-based coordinate system with the removal of introns to generate a cross-cohort pileup. The pileup was binned into 1-bp bins as a gene-level histogram. Using a fine-to-coarse segmentation algorithm, a histogram segmentation was computed (hyperparameter ε = 0.005) and regions of low entropy were filtered32,33,34,35. Histograms were fit with probability distributions by optimizing against goodness-of-fit metrics (mean squared error, Kolmogorov–Smirnov statistic, area of intersection, Jaccard index, χ2 statistic) using the DEoptim R package to characterize distribution shape and skew99. The best-fitting distribution was selected using majority voting; ties were broken by the best fit using the Jaccard index.
Uniquely mapped reads were quantitated on the joint peaks using featureCounts from subread (version 2.0.0) using the parameter –largestOverlap to assign reads to the peak with the largest overlap. Uniquely mapped reads were quantitated on genes using RSEM (version 1.3.2). The two datasets were used to estimate peak-level and gene-level m6A abundance, respectively, using the following post-processing steps based on the protocol described by Zhang et al.37. Specifically, for the 1% of peaks or genes with the greatest counts, the mean fold enrichment of peak-level or gene-level IP counts over the corresponding gene-level input counts was calculated. These enrichment values were then used in library size factor calculation (RLE) with the DESeq2 (version 1.26.0)36 ‘estimateSizeFactorsForMatrix’ function. After normalization, IP reads in peaks or genes were adjusted by the normalized gene-level read counts of the input libraries and rounded to yield whole numbers37,54.
For m6A peaks, the resulting peak-by-sample count matrix was either (1) used as direct input or (2) ‘soft-masked’. The identification of a peak in a sample is a function of applying a statistical threshold on observed IP and input read counts, with the intention of determining a binary methylation status for a sample. A peak may therefore have nonzero counts in a sample for which the peak-calling threshold was not met. Masking thus addressed the goal of filtering spurious reads observed in ‘peak-less’ samples. Soft-masking was defined as setting the peak counts to zero after input adjustment in samples where the peak was not called.
Soft-masked peak-level m6A data were used for differential driver methylation analysis, while the unmasked peak-by-count matrix was used in subtype discovery, clinical associations, nominating m6A-QTLs, hypoxia associations and biomarker discovery. Gene-level m6A data were used in biomarker discovery for prostate cancer drivers. The binary status of whether a particular peak was observed in a given sample was used for mutual information analysis.
Identifying m6A-QTLs
m6A-QTLs were identified using a linear additive model implemented in Matrix eQTL (version 2.3) and validated using QTLtools (version 1.2)83,100,101. The ‘soft-masked’ peaks were filtered to remove peaks with zero reads in greater than 20% of samples (30,303 of 32,051 peaks remaining). Normalized, adjusted counts were then transformed using rank-inverse normal transformation102. The first seven principal components of the genotype matrix were identified using PCA with PLINK at default pruning for linkage disequilibrium103, and, in combination with age, were used as covariates. SNPs within 10-kbp windows to the peaks were tested for local m6A-QTLs, while SNPs identified as linked to genes by Yuan et al.52 were tested for spatial m6A-QTLs, and the top peak–SNP associations in the set of significant associations (Q < 0.1) were nominated as candidate m6A-QTLs. eQTLs and pQTLs were identified using the same transformation and covariates for significant hits.
m6A sites were downloaded from the RMVar database (hg38, accessed 21 August 2021) and filtered for high-confidence single-base pair m6A characterization methods in human cell lines. This dataset represented 132,428 single-base pair sites across 70 studies, 335 tissues, cell lines and conditions and 1,056 samples and were captured using one of six experimental techniques (miCLIP-seq, DART-seq, m6ACE-seq, m6A-REF-seq, PA-m6A-seq and m6A-label-seq). Sites were filtered for those found in lncRNAs, protein-coding genes and pseudogenes. Specifically, sites were required to be localized to exons, UTRs, stop codons and coding sequences. Next, sites were filtered for those that overlapped the 5,647 regions found in at least two cancer cell lines using meRIP-seq. Finally, sites were filtered for being identified in at least two studies, identified by at least two experimental techniques and identified in at least two tissues, resulting in a dataset of 2,883 m6A sites.
HiChIP loop calls from Wei et al.55 were obtained and used to determine whether m6A-QTL SNPs and their associated peaks were located in anchored loops.
The GenomicRanges (version 1.54.1) R package was used to intersect the genomic coordinates and calculate the distance between m6A-QTL SNPs and their associated m6A peaks.
m6A hypoxia analysis
Hypoxia scores were calculated using microarray-derived RNA abundance values in a previous study17. To maximize sample overlap, the corresponding gene-level mRNA abundance values16 were used when calculating correlation between m6A regulatory gene RNA abundance and sample hypoxia. m6A peaks that were identified in every sample (n = 20,811) were assessed for correlation with patient hypoxia score via Spearman’s rank correlation. Pathway enrichment analysis of the resulting genes with significant peaks was performed using the gprofiler2 R package (version 0.2.0)104 against the Gene Ontology Biological Process domain and the Reactome database, with the analysis being limited to pathway terms containing a maximum of 1,500 genes.
Prostate cancer drivers and m6A enzymes
Somatic mutations were selected based on prior work17. These data were obtained from the ICGC Data Portal under project ‘PRAD-CA’ (https://dcc.icgc.org/projects/PRAD-CA). Specifically, CNA calls from the Affymetrix OncoScan array were used. CNAs were annotated with genes of interest, and, if conflicting gains or losses overlapped a given gene, the CNA type with the largest overlap was selected. Single-nucleotide variants used were those called via SomaticSniper version 1.0.5.0 and indels as called by the ‘PCAWG Consensus INDEL caller’. Only simple somatic mutation events with the following predicted consequence type were used: ‘missense_variant’, ‘stop_gained’, ‘stop_lost’, ‘splice_donor_variant’, ‘splice_acceptor_variant’. The set of m6A regulatory enzymes was curated from ref. 105. The set of prostate cancer driver CNAs was curated from Rheinbay et al.106; a representative Cancer Gene Census gene was selected from every peak identified in prostate cancer through the Genomic Identification of Significant Targets in Cancer project that was annotated with at least one Cancer Gene Census gene106. The set of prostate cancer driver simple somatic mutations was curated from the TCGA-PRAD107. Structural variants resulting in ERG or ETV fusion events were called via identification of genomic breakpoints inferred through whole-genome sequencing or OncoScan array16 or the presence of TMPRSS2-ERG fusion transcripts inferred through RNA-seq. For statistical analyses, only genes mutated in >5% of samples were included, yielding 18 drivers and ten m6A enzymes. Mutation co-occurrence was assessed using a hypergeometric test.
Mutation differential methylation analysis
For differential methylation analysis, initial filtering of peaks was performed using a threshold of >0 normalized, ‘soft-masked’, IP-adjusted read counts in ≥50% of samples (74 of 144 samples with both meRIP-seq and mutation status data), resulting in a total of 6,467 peaks. To reduce the influence of outliers, the top and bottom 2.5% of samples for each peak were trimmed. For each mutation, differences in normalized IP-adjusted read counts for each peak were assessed between mutation-positive and mutation-negative samples using a Poisson random effect model as implemented in the diffIP function of the RADAR R package (version 0.2.4)37. Significance was called at Q < 0.1. Pathway enrichment analysis of significant genes was performed using ActivePathways (version 1.1.0)108 against the Gene Ontology and Reactome databases (data obtained from g:profiler release/2022-02-14). Terms were limited to between 5 and 500 genes. The threshold for merged P values and for pathway enrichment was set to 0.1. P values were merged using Brown’s method, and multiple-testing correction was conducted using Holm’s method.
Biomarker discovery using univariate survival analysis
CNAs in nine m6A regulatory genes were tested for association with time until biochemical recurrence. For each gene, a Cox proportional hazards model was used to compare the hazard of biochemical recurrence between the more common alteration event (that is, copy number gain or loss) versus neutral copy number. The model was fit using the survival R package (version 3.3.1)109 with Wald test P values. When modeling a given gene, the Cox model was fit separately to six different cohorts: Taylor et al. (n = 154)60, Hieronymus et al. (n = 104)61, Ross-Adams et al. 1 (n = 105)62, Ross-Adams et al. 2 (n = 76)62, ICGC-PRAD-CA (n = 379)16,17 and TCGA-PRAD (n = 421)59. A random effect meta-analysis was used to combine the six cohort-specific log hazard ratios into an overall hazard ratio via the meta R package110 (version 6.5.0).
Taylor et al., Ross-Adams et al. 1 and Ross-Adams et al. 2 CNA data were obtained from https://github.com/crukci-bioinformatics/camcAPP/tree/master/data. Taylor et al. biochemical recurrence data were obtained from https://www.cbioportal.org/study/clinicalData?id=prad_mskcc. Hieronymus et al. data were obtained from https://www.cbioportal.org/study/summary?id=prad_mskcc_2014, TCGA-PRAD data were downloaded using the TCGAbiolinks R package (version 2.25.3)111, and ICGC-PRAD-CA data are a superset of the CNA data described in the current study and were obtained as described above.
Survival analysis of prostate cancer drivers was performed similarly. m6A was considered as a continuous variable using gene-level normalized, IP-adjusted counts before median dichotomization. Other continuous variables (RNA and protein abundance) were also median dichotomized, and CNAs were analyzed as previously detailed.
For peak-level survival analysis, peaks were initially filtered to remove those identified in less than six samples (n = 20,334). A two-stage approach was then used to identify survival associations. First, Cox proportional hazards models were fitted with the R package survival (version 3.2-13) for each peak using presence or absence of the peak as a discrete variable and the Wald test for nonzero coefficients109. The proportional hazards assumption was tested for each peak. Next, to correct for the comparison of potentially imbalanced populations, ExaLT was used to validate P values using an approximation factor of 10 and a P-value threshold of 0.01 (ref. 112).
For survival analysis of VCAN RNA abundance, the assumptions of the Cox model were tested using the cox.zph function in the R survival package (version 2.41-3) with 0.1 as a cutoff. The Cox proportional hazards model was fit using patient groups dichotomized by the median RNA abundance, with biochemical recurrence as the endpoint. The analysis used bulk RNA-seq data from several studies, including TCGA-PRAD59, Taylor et al.60, Gerhauser et al.78, Ross-Adams et al.62 and ICGC-PRAD-CA19. TCGA-PRAD RNA-seq data were downloaded using the R package TCGA2STAT (version 1.2)113.
m6A datasets
Reference m6A sites were downloaded from RMVar (29 August 2021)43, and additional m6A sites from the prostate epithelial cell line RWPE and the cancer cell line LNCaP were downloaded from Cotter et al.44. For the latter, m6A sites were identified as the nearest adenine within 3 bp downstream of the cross-linking mutation by strand.
Additional datasets
Data from previous publications were used in their original state2,16,17,18,20 with the exception of genotyping data as described above and RNA-seq data from ref. 19, which were remapped to hg38 using annotations from gencode.v34.chr_patch_hapl_scaff, using the RNA-seq pipeline described above. Quality control of the input library, mutual information analysis, eQTLs, enzyme associations with RNA abundance, comparison of m6A subtypes to RNA subtypes and survival analysis of target genes based on transcript abundance were conducted using RNA-seq data19 due to greater sequencing depth. Additional RNA-seq and/or meRIP-seq datasets were processed using identical pipelines.
Statistics and reproducibility
Data analysis was conducted using R (version 3.6.1 to version 4.1.0). Unless indicated otherwise, multiple-hypothesis testing was conducted for each statistical analysis using the method outlined by Benjamini and Hochberg114, commonly referred to as the false discovery rate or Q value (denoted here as Q). In vitro and in vivo data were expressed as mean ± s.d. and mean ± s.e.m., respectively. Two-tailed unpaired Student’s t-test and one-way ANOVA were performed to identify significant differences between groups in our experiments. Data were confirmed to meet the assumptions of normality and equal variance for the statistical tests used. Details of the test method used for statistical analysis are specified in the relevant figure legends and the Methods. All values were considered significantly different at P < 0.05. Western blot experiments were performed at least twice with similar results. Data visualization was conducted using BoutrosLab.plotting.general (version 6.0.3)115, ggseqlogo (version 0.1)116 and Guitar (version 2.4.0)117. No statistical method was used to predetermine sample size. The meRIP-seq data of 14 samples were excluded from the analyses due to stringent quality control. The experiments were not randomized. The investigators were not blinded to allocation during experiments and outcome assessment.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
m6A meRIP-seq raw sequencing data are available at the EGA under accession EGAS00001006925. Other previously published molecular data characterizing the ICGC-PRAD-CA cohort analyzed in this study are available at the EGA under accession EGAS00001000900. PC-3 shRNA knockdown VCAN RNA-seq data were deposited in the Gene Expression Omnibus under the accession number GSE220432. Previously published mRNA microarray data are available at the Gene Expression Omnibus under accession code GSE84043. Source data are provided with this paper.
Code availability
HistogramZoo is available at https://github.com/uclahs-cds/public-R-HistogramZoo ref. 31. Scripts for data analysis and plotting are available at https://github.com/uclahs-cds/prostate-cancer-m6a-paper ref. 118.
References
Hanahan, D. & Weinberg, R. A. The hallmarks of cancer. Cell 100, 57–70 (2000).
Sinha, A. et al. The proteogenomic landscape of curable prostate cancer. Cancer Cell 35, 414–427 (2019).
Liu, J. et al. An integrated TCGA pan-cancer clinical data resource to drive high-quality survival outcome analytics. Cell 173, 400–416 (2018).
Huang, H., Weng, H. & Chen, J. m6A modification in coding and non-coding RNAs: roles and therapeutic implications in cancer. Cancer Cell 37, 270–288 (2020).
Zhao, B. S., Roundtree, I. A. & He, C. Post-transcriptional gene regulation by mRNA modifications. Nat. Rev. Mol. Cell Biol. 18, 31–42 (2017).
Shi, H., Wei, J. & He, C. Where, when, and how: context-dependent functions of RNA methylation writers, readers, and erasers. Mol. Cell 74, 640–650 (2019).
Wang, Q. et al. METTL3-mediated m6A modification of HDGF mRNA promotes gastric cancer progression and has prognostic significance. Gut 69, 1193–1205 (2020).
Gu, C. et al. Mettl14 inhibits bladder TIC self-renewal and bladder tumorigenesis through N6-methyladenosine of Notch1. Mol. Cancer 18, 168 (2019).
Panneerdoss, S. et al. Cross-talk among writers, readers, and erasers of m6A regulates cancer growth and progression. Sci. Adv. 4, eaar8263 (2018).
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, 556–560 (2018).
Zhang, S. et al. m6A demethylase ALKBH5 maintains tumorigenicity of glioblastoma stem-like cells by sustaining FOXM1 expression and cell proliferation program. Cancer Cell 31, 591–606 (2017).
Vu, L. P. et al. The N6-methyladenosine (m6A)-forming enzyme METTL3 controls myeloid differentiation of normal hematopoietic and leukemia cells. Nat. Med. 23, 1369–1376 (2017).
Chang, G. et al. YTHDF3 induces the translation of m6A-enriched gene transcripts to promote breast cancer brain metastasis. Cancer Cell 38, 857–871 (2020).
Su, R. et al. R-2HG exhibits anti-tumor activity by targeting FTO/m6A/MYC/CEBPA signaling. Cell 172, 90–105 (2018).
Fraser, M. et al. Genomic hallmarks of localized, non-indolent prostate cancer. Nature 541, 359–364 (2017).
Bhandari, V. et al. Molecular landmarks of tumor hypoxia across cancer types. Nat. Genet. 51, 308–318 (2019).
Houlahan, K. E. et al. Genome-wide germline correlates of the epigenetic landscape of prostate cancer. Nat. Med. 25, 1615–1626 (2019).
Chen, S. et al. Widespread and functional RNA circularization in localized prostate cancer. Cell 176, 831–843 (2019).
Kron, K. J. et al. TMPRSS2–ERG fusion co-opts master transcription factors and activates NOTCH signaling in primary prostate cancer. Nat. Genet. 49, 1336–1345 (2017).
Espiritu, S. M. G. et al. The evolutionary landscape of localized prostate cancers drives clinical aggression. Cell 173, 1003–1013 (2018).
Zeng, Y. et al. Refined RIP-seq protocol for epitranscriptome analysis with low input materials. PLoS Biol. 16, e2006092 (2018).
Meyer, K. D. et al. Comprehensive analysis of mRNA methylation reveals enrichment in 3′ UTRs and near stop codons. Cell 149, 1635–1646 (2012).
Dominissini, D. et al. Topology of the human and mouse m6A RNA methylomes revealed by m6A-seq. Nature 485, 201–206 (2012).
Schaeffer, E. M. et al. Prostate Cancer, Version 4.2023, NCCN Clinical Practice Guidelines in Oncology. J. Natl. Compr. Canc. Netw. 21, 1067–1096 (2023).
Liu, S., Zhu, A., He, C. & Chen, M. REPIC: a database for exploring the N6-methyladenosine methylome. Genome Biol. 21, 100 (2020).
Meng, J., Cui, X., Rao, M. K., Chen, Y. & Huang, Y. Exome-based analysis for RNA epigenome sequencing data. Bioinformatics 29, 1565–1567 (2013).
Cui, X., Meng, J., Zhang, S., Chen, Y. & Huang, Y. A novel algorithm for calling mRNA m6A peaks by modeling biological variances in meRIP-seq data. Bioinformatics 32, i378–i385 (2016).
Robinson, D. G. & Storey, J. D. subSeq: determining appropriate sequencing depth through efficient read subsampling. Bioinformatics 30, 3424–3426 (2014).
Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589 (2010).
Zhu, H., Eng, S., Boutros, P. & He, H. H. HistogramZoo. Zenodo https://doi.org/10.5281/ZENODO.14713036 (2025).
Delon, J., Desolneux, A., Lisani, J.-L. & Petro, A. B. A nonparametric approach for histogram segmentation. IEEE Trans. Image Process. 16, 253–261 (2007).
Delon, J., Desolneux, A., Lisani, J. L. & Petro, A. B. Color image segmentation using acceptable histogram segmentation. In Pattern Recognition and Image Analysis 239–246 (Springer, 2005).
Balaguer, A. B. P. Analytical Methods for the Study of Color in Digital Images. PhD thesis, Universitat de les Illes Balears (2006).
Lisani, J.-L. & Petro, A. B. Automatic 1D histogram segmentation and application to the computation of color palettes. Image Process. Line 11, 76–104 (2021).
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Zhang, Z. et al. RADAR: differential analysis of MeRIP-seq data with a random effect model. Genome Biol. 20, 294 (2019).
Shannon, C. E. A mathematical theory of communication. Bell Syst. Tech. J. 27, 623–656 (1948).
Mao, Y. et al. m6A in mRNA coding regions promotes translation via the RNA helicase-containing YTHDC2. Nat. Commun. 10, 5332 (2019).
Lalonde, E. et al. Tumour genomic and microenvironmental heterogeneity for integrated prediction of 5-year biochemical recurrence of prostate cancer: a retrospective cohort study. Lancet Oncol. 15, 1521–1532 (2014).
Heemers, H. V. & Tindall, D. J. Androgen receptor (AR) coregulators: a diversity of functions converging on and regulating the AR transcriptional complex. Endocr. Rev. 28, 778–808 (2007).
Lichtenstein, P. et al. Environmental and heritable factors in the causation of cancer—analyses of cohorts of twins from Sweden, Denmark, and Finland. N. Engl. J. Med. 343, 78–85 (2000).
Luo, X. et al. RMVar: an updated database of functional variants involved in RNA modifications. Nucleic Acids Res. 49, D1405–D1412 (2021).
Cotter, K. A. et al. Mapping of m6A and its regulatory targets in prostate cancer reveals a METTL3-low induction of therapy resistance. Mol. Cancer Res. 19, 1398–1411 (2021).
Al Olama, A. A. et al. A meta-analysis of 87,040 individuals identifies 23 new susceptibility loci for prostate cancer. Nat. Genet. 46, 1103–1109 (2014).
Wang, A. et al. Characterizing prostate cancer risk through multi-ancestry genome-wide discovery of 187 novel risk variants. Nat. Genet. 55, 2065–2074 (2023).
Xu, J. et al. Inherited genetic variant predisposes to aggressive but not indolent prostate cancer. Proc. Natl Acad. Sci. USA 107, 2136–2140 (2010).
He, Y. et al. The prostate cancer susceptibility variant rs2735839 near KLK3 gene is associated with aggressive prostate cancer and can stratify Gleason score 7 patients. Clin. Cancer Res. 20, 5133–5139 (2014).
Schumacher, F. R. et al. Association analyses of more than 140,000 men identify 63 new prostate cancer susceptibility loci. Nat. Genet. 50, 928–936 (2018).
Conti, D. V. et al. Trans-ancestry genome-wide association meta-analysis of prostate cancer identifies new susceptibility loci and informs genetic risk prediction. Nat. Genet. 53, 65–75 (2021).
Huynh-Le, M.-P. et al. Prostate cancer risk stratification improvement across multiple ancestries with new polygenic hazard score. Prostate Cancer Prostatic Dis. 25, 755–761 (2022).
Yuan, J. et al. Prostate cancer transcriptomic regulation by the interplay of germline risk alleles, somatic mutations, and 3D genomic architecture. Cancer Discov. 12, 2838–2855 (2022).
Xiong, X. et al. Genetic drivers of m6A methylation in human brain, lung, heart and muscle. Nat. Genet. 53, 1156–1165 (2021).
Zhang, Z. et al. Genetic analyses support the contribution of mRNA N6-methyladenosine (m6A) modification to human disease heritability. Nat. Genet. 52, 939–949 (2020).
Wei, Z. et al. MYC reshapes CTCF-mediated chromatin architecture in prostate cancer. Nat. Commun. 14, 1787 (2023).
Futreal, P. A. et al. A census of human cancer genes. Nat. Rev. Cancer 4, 177–183 (2004).
Armenia, J. et al. The long tail of oncogenic drivers in prostate cancer. Nat. Genet. 50, 645–651 (2018).
Quigley, D. A. et al. Genomic hallmarks and structural variation in metastatic prostate cancer. Cell 174, 758–769 (2018).
Cancer Genome Atlas Research Network. The molecular taxonomy of primary prostate cancer. Cell 163, 1011–1025 (2015).
Taylor, B. S. et al. Integrative genomic profiling of human prostate cancer. Cancer Cell 18, 11–22 (2010).
Hieronymus, H. et al. Copy number alteration burden predicts prostate cancer relapse. Proc. Natl Acad. Sci. USA 111, 11139–11144 (2014).
Ross-Adams, H. et al. Integration of copy number and transcriptomics provides risk stratification in prostate cancer: a discovery and validation cohort study. eBioMedicine 2, 1133–1144 (2015).
Weischenfeldt, J. et al. Integrative genomic analyses reveal an androgen-driven somatic alteration landscape in early-onset prostate cancer. Cancer Cell 23, 159–170 (2013).
Schmalfeldt, M., Dours-Zimmermann, M. T., Winterhalter, K. H. & Zimmermann, D. R. Versican V2 is a major extracellular matrix component of the mature bovine brain. J. Biol. Chem. 273, 15758–15764 (1998).
Read, J. T. et al. Androgen receptor regulation of the versican gene through an androgen response element in the proximal promoter. J. Biol. Chem. 282, 31954–31963 (2007).
Wu, Y. J., La Pierre, D. P., Wu, J., Yee, A. J. & Yang, B. B. The interaction of versican with its binding partners. Cell Res. 15, 483–494 (2005).
Wang, X. et al. N6-methyladenosine modulates messenger RNA translation efficiency. Cell 161, 1388–1399 (2015).
Liu, T. et al. The m6A reader YTHDF1 promotes ovarian cancer progression via augmenting EIF3C translation. Nucleic Acids Res. 48, 3816–3831 (2020).
Wang, S. et al. N6-methyladenosine reader YTHDF1 promotes ARHGEF2 translation and RhoA signaling in colorectal cancer. Gastroenterology 162, 1183–1196 (2022).
Li, P. et al. ELK1-mediated YTHDF1 drives prostate cancer progression by facilitating the translation of Polo-like kinase 1 in an m6A dependent manner. Int. J. Biol. Sci. 18, 6145–6162 (2022).
Huang, H. et al. Recognition of RNA N-methyladenosine by IGF2BP proteins enhances mRNA stability and translation. Nat. Cell Biol. 20, 285–295 (2018).
Wang, X. et al. N6-methyladenosine-dependent regulation of messenger RNA stability. Nature 505, 117–120 (2014).
Batista, P. J. et al. m6A RNA modification controls cell fate transition in mammalian embryonic stem cells. Cell Stem Cell 15, 707–719 (2014).
Geula, S. et al. m6A mRNA methylation facilitates resolution of naive pluripotency toward differentiation. Science 347, 1002–1006 (2015).
Wedge, D. C. et al. Sequencing of prostate cancers identifies new cancer genes, routes of progression and drug targets. Nat. Genet. 50, 682–692 (2018).
Baca, S. C. et al. Punctuated evolution of prostate cancer genomes. Cell 153, 666–677 (2013).
Barbieri, C. E. et al. Exome sequencing identifies recurrent SPOP, FOXA1 and MED12 mutations in prostate cancer. Nat. Genet. 44, 685–689 (2012).
Gerhauser, C. et al. Molecular evolution of early-onset prostate cancer identifies molecular risk markers and clinical trajectories. Cancer Cell 34, 996–1011 (2018).
Houlahan, K. E. et al. A polygenic two-hit hypothesis for prostate cancer. J. Natl. Cancer Inst. 115, 468–472 (2023).
Erho, N. et al. Discovery and validation of a prostate cancer genomic classifier that predicts early metastasis following radical prostatectomy. PLoS ONE 8, e66855 (2013).
Jairath, N. K. et al. A systematic review of the evidence for the decipher genomic classifier in prostate cancer. Eur. Urol. 79, 374–383 (2021).
Liu, H. et al. Accurate detection of m6A RNA modifications in native RNA sequences. Nat. Commun. 10, 4079 (2019).
Shabalin, A. A. Matrix eQTL: ultra fast eQTL analysis via large matrix operations. Bioinformatics 28, 1353–1358 (2012).
Song, S. et al. qpure: a tool to estimate tumor cellularity from genome-wide single-nucleotide polymorphism profiles. PLoS ONE 7, e45835 (2012).
Yu, G., Wang, L.-G., Han, Y. & He, Q.-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics 16, 284–287 (2012).
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 17, 10–12 (2011).
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Harrow, J. et al. GENCODE: the reference human genome annotation for the ENCODE Project. Genome Res. 22, 1760–1774 (2012).
PCAWG Transcriptome Core Group et al. Genomic basis for RNA alterations in cancer. Nature 578, 129–136 (2020).
Andrews, S. et al. FastQC. GitHub https://github.com/s-andrews/FastQC (2023).
Wang, L., Wang, S. & Li, W. RSeQC: quality control of RNA-seq experiments. Bioinformatics 28, 2184–2185 (2012).
Li, B. & Dewey, C. N. RSEM: accurate transcript quantification from RNA-seq data with or without a reference genome. BMC Bioinformatics 12, 323 (2011).
Parekh, S., Ziegenhain, C., Vieth, B., Enard, W. & Hellmann, I. The impact of amplification on differential expression analyses by RNA-seq. Sci. Rep. 6, 25533 (2016).
Van der Auwera, G. A. & O’Connor, B. D. Genomics in the Cloud: Using Docker, GATK, and WDL in Terra (O’Reilly Media, 2020).
Poplin, R. et al. Scaling accurate genetic variant discovery to tens of thousands of samples. Preprint at bioRxiv https://doi.org/10.1101/201178 (2017).
DePristo, M. A. et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat. Genet. 43, 491–498 (2011).
Krusche, P. et al. Best practices for benchmarking germline small-variant calls in human genomes. Nat. Biotechnol. 37, 555–560 (2019).
Cleary, J. G. et al. Comparing variant call files for performance benchmarking of next-generation sequencing variant calling pipelines. Preprint at bioRxiv https://doi.org/10.1101/023754 (2015).
Mullen, K. M., Ardia, D., Gil, D. L., Windover, D. & Cline, J. DEoptim: an R package for global optimization by differential evolution. J. Stat. Softw. 40, 1–26 (2011).
Ongen, H., Buil, A., Brown, A. A., Dermitzakis, E. T. & Delaneau, O. Fast and efficient QTL mapper for thousands of molecular phenotypes. Bioinformatics 32, 1479–1485 (2016).
Delaneau, O. et al. A complete tool set for molecular QTL discovery and analysis. Nat. Commun. 8, 15452 (2017).
McCaw, Z. R., Lane, J. M., Saxena, R., Redline, S. & Lin, X. Operating characteristics of the rank-based inverse normal transformation for quantitative trait analysis in genome-wide association studies. Biometrics 76, 1262–1272 (2020).
Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet. 81, 559–575 (2007).
Kolberg, L., Raudvere, U., Kuzmin, I., Vilo, J. & Peterson, H. gprofiler2—an R package for gene list functional enrichment analysis and namespace conversion toolset g:Profiler. F1000Res. 9, ELIXIR-709 (2020).
Jiang, X. et al. The role of m6A modification in the biological functions and diseases. Signal Transduct. Target. Ther. 6, 74 (2021).
Rheinbay, E. et al. Analyses of non-coding somatic drivers in 2,658 cancer whole genomes. Nature 578, 102–111 (2020).
Bailey, M. H. et al. Comprehensive characterization of cancer driver genes and mutations. Cell 174, 1034–1035 (2018).
Paczkowska, M. et al. Integrative pathway enrichment analysis of multivariate omics data. Nat. Commun. 11, 735 (2020).
Therneau, T. M. & Mayo Foundation. A Package for Survival Analysis in S sites.google.com/site/statsr4us/docs/SurvivalanalysisinR_package.pdf (1999).
Schwarzer, G. et al. meta: an R package for meta-analysis. R News 7, 40–45 (2007).
Colaprico, A. et al. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res. 44, e71 (2016).
Vandin, F., Papoutsaki, A., Raphael, B. J. & Upfal, E. Accurate computation of survival statistics in genome-wide studies. PLoS Comput. Biol. 11, e1004071 (2015).
Wan, Y.-W., Allen, G. I. & Liu, Z. TCGA2STAT: simple TCGA data access for integrated statistical analysis in R. Bioinformatics 32, 952–954 (2016).
Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. 57, 289–300 (1995).
P’ng, C. et al. BPG: seamless, automated and interactive visualization of scientific data. BMC Bioinformatics 20, 42 (2019).
Wagih, O. ggseqlogo: a versatile R package for drawing sequence logos. Bioinformatics 33, 3645–3647 (2017).
Cui, X. et al. Guitar: an R/Bioconductor package for gene annotation guided transcriptomic analysis of RNA-related genomic features. BioMed Res. Int. 2016, 8367534 (2016).
Xu, X. et al. The landscape of N6-methyladenosine in localized primary prostate cancer. Zenodo https://doi.org/10.5281/ZENODO.14655534 (2025).
Acknowledgements
We thank all members of the Boutros and He laboratory for insightful suggestions, all patients who made this study possible and X. Xiong for sending supporting GTEx data. The results here are in part based on data generated by the TCGA Research Network: https://www.cancer.gov/tcga. This work was supported by the Princess Margaret Cancer Foundation (886012001223 to H.H.H.), the Canadian Cancer Society (TAG2018-2061), CIHR operating grants (142246, 152863, 152864 and 159567 to H.H.H.) and a Terry Fox New Frontiers Program Project Grant (PPG19-1090 to H.H.H.). H.H.H. holds the Joey and Toby Tanenbaum Brazilian Ball Chair in Prostate Cancer and the Tier 1 Canada Research Chair in RNA Medicine. The data presented in this paper were partially funded by the Terry Fox Research Institute’s Marathon of Hope Cancer Centres Network. Xin Xu was supported by the Prostate Cancer Foundation Young Investigator Award (21YOUN06). H.Z. was supported by a CIHR doctoral award. K.E.H. was supported by a CIHR Vanier Fellowship. N.Z. was supported by the NIH through awards T32HG002536 and F31CA281168. S.C. was supported by the Prostate Cancer Foundation Young Investigator Award (21YOUN06) and a CIHR fellowship (181755). P.C.B. was supported by NIH grants P30CA016042, U2CCA271894, U24CA248265, U01CA214194 and R01CA272678. This work was supported by a Prostate Cancer Foundation Special Challenge Award to P.C.B. (award ID 20CHAS01) made possible by the generosity of L. Ruvo. This study is also supported by the Medical Research Data Center at Shanghai Medical College of Fudan University. The funders had no role in the study design, data collection and analysis, decision to publish or manuscript preparation.
Author information
Authors and Affiliations
Contributions
Designed studies: Xin Xu, H.Z., R.H.-W., J.L., P.C.B. and H.H.H. Sample and clinical data acquisition: M.O., V.P., H.H., A. Bergeron, L.L., B.T., Y.F., R.G.B., N.E.F., A. Berlin and T.v.d.K. Performed experiments: Xin Xu, Y.W., K.P., W.L., Xi Xu, M.S., W.Y.G., M.W., S.W., W.C., X.C., K.X., M.L., M.K., C.H. and H.L. Data analysis: H.Z., R.H.-W., J.L., Xin Xu, S.E., N.Z., S.C., K.E.H., S.L., J.A., Y. Song, Y.Z., G.Z., T.G., W.X., G.-H.W., X.W. and Y. Shao. Manuscript first draft: Xin Xu, H.Z., R.H.-W., J.L., P.C.B. and H.H.H. Revised and approved manuscript: all authors.
Corresponding authors
Ethics declarations
Competing interests
For a portion of the time during the preparation of this paper, H.Z. was an employee at Deep Genomics. All contributions to the design, analysis and interpretation of results of this project were completed outside of the term of employment. P.C.B. sits on the scientific advisory boards of Sage Bionetworks, Intersect Diagnostics and BioSymetrics. Y. Shao and X.W. are shareholders and/or employees of Geneseeq Technology. The other authors declare no competing interests.
Peer review
Peer review information
Nature Genetics thanks Jianjun Chen and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 Data Analysis & Quality Control.
a) The study workflow: we investigate the role of the post-transcriptional RNA modification m6A in primary prostate tumors by examining global m6A patterns, germline origins, clinical associations, microenvironmental associations and functional impact. This builds on previous molecular profiling studies of this primary prostate tumor cohort including 133 genotypes18, 148 CNA and 143 somatic mutation profiles17, 146 DNA methylation18, 29 H3K27ac ChIP-seq20, 92 ultra-deep RNA-sequencing19 and 54 proteomic profiles2. b) Data processing pipeline for meRIP-sequencing. c) Quality control metrics derived from the output of STAR, RSeQC, cutadapt and correlation of the Input library with previous transcriptome profiling datasets16,19. Top barplot: the sum of negative Z-scores for each sample. Bottom heatmap: metric by sample matrix where colors denote the magnitude of negative Z-scores. d) Left panel: sample-level correlation of transcript abundance in 92 samples common between Input libraries of this study and ultra-deep bulk RNA-seq from Chen et al.19 (Spearman’s ⍴). Right panel: gene-level correlation of transcript abundance in the same datasets (Spearman’s ⍴)19. e) Validation of sample identity by comparing germline variants identified in the Input libraries and Houlahan et al.18. Both axes are ordered identically, first, by ancestry (covariate: orange represents European ancestry, green represents non-European ancestry) and then, by sample ID. Heatmap color represents precision (left) and sensitivity (right). Precision is calculated as the intersection of RNA-seq and WGS variants divided by all RNA-seq variants. Sensitivity is calculated as the intersection of RNA-seq and WGS variants divided by all WGS variants. f) Comparisons of the PCR Duplicate Proportion (PDP), the Non-Redundant Fraction (NRF) and the PCR Bottlenecking Coefficients 1 and 2 (PBC1/2) for each sample. The metric values of the IP library (y-axis) are compared to the metric values of the Input library (x-axis).
Extended Data Fig. 2 Peak Calling & Central Dogma Integration.
a) The distribution of peaks identified by exomePeak and MeTPeak. Each dot in the scatterplot represents a sample, histograms show the distribution of m6A methylated genes and peaks across samples. Bottom histogram shows the distribution of the number of peaks identified per gene. b) Spearman’s correlation of the number of peaks and the library size. The size of the disks corresponds to the magnitude of Spearman’s ⍴ and the background color indicate the P value (two-tailed test). c-d) The number of peaks called using three IP libraries when randomly subsampled to 50%, 25% 10% and 1% of the original reads for exomePeak (c) and MeTPeak (d). e) Comparing the coverage of MeTPeak and exomePeak peaks using the Simpson Index (intersection/smallest set, median = 94.80%) and the Jaccard Index (intersection/union, median = 53.01%) for each sample. Inner panel: Overlap by genomic coverage with previously identified peaks in the REPIC database26. f) Distribution of peak density across a metagene as called by MeTPeak. Line shows the density of the peaks identified by HistogramZoo while the blue background shows the variation of peak density across individual samples. Bottom plot shows the distribution of peaks annotated across transcriptomic elements. g-j) Exemplar plots for MALAT1 (g), FOXA1 (h), RB1 (i) and NKX3-1 (j). The median IP (green) and Input (purple) coverage are represented using lines while background colors represent the range of IP and Input coverage across samples. Exons are annotated below in dark blue. Bottom heatmaps represent the distribution of IP and Input coverage (log1p-transformed) across samples. k) Relationship between the number of samples with an m6A peak at each gene and the corresponding gene-level RNA abundance. Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5x the IQR. l) Distribution of spearman’s correlation between m6A level and input abundance. m-n) Relationship between gene-level RNA and protein abundance correlation and (m) gene-level m6A abundance and (n) the number of samples with an m6A peak at each gene. Spearman’s correlation and corresponding P value displayed (n = 3,233).
Extended Data Fig. 3 Clinical Correlates of m6A.
a) m6A peak abundance correlates with patient clinical features. b) TAOK2 m6A peak abundance is correlated with age. Spearman’s correlation and corresponding P value displayed. c) Presence of IDC/CA co-occurs with greater patient peak number. P value is from a Student’s t-test (n = 133). Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5x the IQR. d) Association between patient peak number and patient clinical features. Dot size represents the indicated effect size (where applicable) and background color indicates P value. Spearman’s rank correlation was employed for PGA, age, and PSA, Mann Whitney U test was used for IDC status and pT, one-way ANOVA was used for analysis of ISUP Grade Group, and a Cox Proportional Hazards model was applied to analyse association between median-dichotomized sample peak number and biochemical recurrence. e) Association between patient peak number and mutations in prostate cancer driver genes. Samples are ranked by the number of peaks identified in descending order. Barplot indicates the statistical significance of each association, assessed via U-test.
Extended Data Fig. 4 Germline Correlates of m6A.
a) The overlap of m6A sites identified in RMVar42 and Cotter et al. 202143 with germline SNPs genotyped as described in Houlahan et al.18. b) The distribution of allele frequencies of the A for SNPs/m6A sites in prostate cancer joint peaks (red) compared to the A for SNPs/m6A sites outside of joint peaks (white). c) The SNP rs2240912 is associated with biochemical recurrence (Q < 0.1; Cox Proportional Hazards model). d) The difference in Conti PRS relative to m6A peak status (PRSPeak - PRSNo Peak) across peaks assessed using Student’s T test (two-sided). e-f) Left: The distribution of P values in associations (two-sided T test) between m6A peak status and Schumacher PRS (e) and PHS290 (f). Right: Volcano plots showing log10 P values versus effect sizes (Δ PRS). g) The association between Conti PRS versus the number of MeTPeak peaks per sample using Spearman’s correlation analysis (two-sided). h-i) Left: distribution of P values in modelling associations between risk SNPs and local m6A peak methylation (h) and spatially-associated m6A peak methylation (i) Right: Volcano plots of P values versus effect sizes from additive linear models (β). Data from linear additive model. j) Histogram of the distance (bps) between the m6A-QTL and associated m6A peak. k) The identification of significant RNA, protein and clinical associations based on tag SNPs and gene pairs of m6A-QTLs. l) RNA and protein associations identified using m6A-QTL tag SNP-gene pairs. The disk sizes and color are indicative of the magnitude and direction of the β values of a linear additive model while the background color indicates the Q value. m-n) The associations of the SNP rs4951018 with SLC45A3 RNA (m) and protein abundance (n). Box plots represent the median (center line) and upper and lower quartiles (box limits), and whiskers extend to the minimum and maximum values within 1.5x the IQR. P values from linear additive model (n = 133). o) Top: tables showing associations between ISUP-associated alleles and ISUP Grade Group. Q values are derived from the global Χ2 test. Bottom: tables showing associations between pairs of alleles.
Extended Data Fig. 5 Somatic Mutations Modulate the m6A Landscape.
a-b) Association of (a) copy number aberrations or (b) RNA abundance with 1) RNA or protein abundance of the corresponding gene, or total sample peak number (Student’s t-test, two-sided), 2) clinical features: ISUP Grade Group, IDC status, pathological T category (Pearson’s Χ2 test, two-sided), PSA, and age (Student’s t-test, two-sided). Black background indicates Q value < 0.1, whilst dot size represents (a) effect size or (b) Spearman’s correlation. c) Association of mutations in m6A regulatory genes and prostate cancer driver events. P values from a hypergeometric test (one-sided for over-representation) are represented by the color of the background, whilst the size of the dot represents the difference between observed and expected co-occurring mutations. Covariate bars indicate mutation type and, if applicable, GISTIC peak the mutation is located within. d) Consensus clustering of 3,432 differentially methylated m6A peaks (k = 5) associated with driver mutations or mutations to m6A regulators (k = 5). For each mutation, colors in the main heatmap indicate the log2 fold change in m6A level, and right barplot the number of m6A peaks significantly differentially m6A methylated, between samples with and without the corresponding mutation. MC: mutation cluster; DMPC: differentially-methylated peak cluster. e) Pathway enrichment analysis performed on genes with differential m6A abundance associated with mutations in m6A enzymes or prostate cancer drivers. Each node represents an enriched pathway term (from GO:BP or REACTOME ontologies). Node colors represent mutations from (d), with some pathways only enriched through joint consideration of all genes associated with a whole mutation cluster (indicated through ‘Contribution’ key). f) Forest plot showing the hazard ratio and 95% confidence intervals for presence of an m6A peak associated with biochemical recurrence, and left barplot indicates corresponding P values from a Cox Proportional Hazards model. g-h) Presence of a m6A peak on g) INHBA or h) ZFHX4, increases risk of biochemical recurrence. INHBA peak samples n = 6, ZFHX4 peak samples n = 9, total samples = 148. P values from a Cox Proportional Hazards model.
Extended Data Fig. 6 VCAN as a Candidate m6A Driver Event.
a) Correlations in RNA abundance of INHBA, VCAN, and ZFHX4 in the TCGA-PRAD and ICGC PRAD-CA cohorts. Dot size is Spearman’s correlation; background shading is P value. b) Comparison of the VCAN mRNA abundance between samples with and without peaks by the Input RNA-seq data. Samples with m6A peaks (n = 23) and without m6A peaks (n = 124) were derived from independent biological samples. The P value was calculated with two-sided Mann-Whitney U test between samples with and without m6A peaks on VCAN mRNA. The top and bottom of the box represent the 75th and 25th percentiles, respectively, with the line inside the box indicating the median. Whiskers extend to 1.5 times the interquartile range from the 25th and 75th percentiles. The fold change of peak mean TPM relative to no peak mean TPM is shown. c) Comparison of the VCAN abundance between samples with and without peaks by protein mass spectrometry data. Samples with m6A peaks (n = 15) and without m6A peaks (n = 39) were derived from independent biological samples. A two-sided Mann-Whitney U test was used to compare the difference between the two groups. The top and bottom of the box represent the 75th and 25th percentiles, respectively, with the line inside the box indicating the median. Whiskers extend to 1.5 times the interquartile range from the 25th and 75th percentiles. The fold change of peak mean value relative to no peak mean value is shown. d) m6A peak signals of VCAN transcripts from meRIP-seq libraries of several prostate cancer cell lines and patient samples with (CPCG0196) or without (CPCG0255) VCAN m6A peaks identified. The vertical red line indicates the location of the significant peak. e) Kaplan-Meier survival curves showing biochemical recurrence stratified by VCAN mRNA abundance in the Gerhauser, Ross-Adams, Taylor, TCGA-PRAD and ICGC-PRAD CA cohorts. A Cox Proportional Hazards model was used to compare the hazard of biochemical recurrence between the groups.
Extended Data Fig. 7 VCAN Drives Prostate Cancer Progression.
a-c) Knockdown efficiency of VCAN mRNA by shRNAs (a), protein abundance (b) and by siRNAs (c) in PC-3 cells. All other Western blots for VCAN in this study are treated by chABC by default. Ponceau S as the loading control. Numbers indicate the positions of molecular mass (kDa) standards. Data are represented as mean ± SD of three biological replicates. P values are calculated using one-way ANOVA with Dunnett’s multiple comparisons test. d) The representative images of the migration and invasion assay of PC-3 cells after shRNA-mediated knockdown of VCAN. The experiment was conducted using three biological replicates for each condition. The scale bar represents 200 μm. e) The growth curve after siRNA-mediated knockdown of VCAN in PC-3. Data are represented as mean ± SD of three biological replicates. P values are calculated using one-way ANOVA with Dunnett’s multiple comparisons test. f-g) The migration and invasion assay of PC-3 cells after siRNA-mediated knockdown of VCAN. The representative images (f), and the quantification (g). Data are represented as mean ± SD of three biological replicates. The scale bar represents 200 μm. h-i) The tumor weight (left) and the tumor size (right) of xenografts derived from injected PC-3 cells (h), and the extravasation ability of PC-3 (i) infected with shRNAs targeting VCAN or GFP as the control. Xenografts are shown as mean ± SEM of five biological replicates, and the extravasation data are presented as mean ± SD of biological replicates (shGFP, n = 9; VCAN sh1, n = 7; VCAN sh2, n = 6). P values are calculated using one-way ANOVA with Dunnett’s multiple comparisons test. j) The volcano plot showed the significantly up/down-regulated genes (average FC > log2(1.5), adjusted P value < 0.05, two-sided) between shGFP and shKD VCAN groups in PC-3 mouse xenografts. Adjusted P values were calculated using the Benjamini-Hochberg method. Vertical dashed lines indicate -log2(1.5) and log2(1.5), respectively, and the horizontal one marks adjusted P = 0.05. k) Dotmap representing top 10 enriched pathways after shKD VCAN in PC-3 mouse xenograft tumors. Dot size indicates the odds ratio and background color the Q values.
Extended Data Fig. 8 VCAN Regulates Tumor-Stromal Crosstalk.
a) Knockdown efficiency of VCAN by shRNA in WPMY-1 cells was detected by Western blot. Ponceau S stain serves as the loading control, and molecular mass (kDa) standards are indicated on the right. Additionally, a sample from PC-3 cells is included for relative quantification, demonstrating comparable levels of VCAN extracted from the same volume of culture medium collected from an equal number of cells. The experiment was conducted using three biological replicates for each condition. b) The growth curve after shRNA-mediated knockdown of VCAN in WPMY-1. Data are represented as mean ± SD of three biological replicates. P value is calculated using an unpaired two-sided Student’s T test. c-f) Migration and invasion analysis of indirect coculture of PC-3 and WPMY-1 cells. PC-3 cells (2 × 104) were seeded into the upper chamber containing serum-free medium, while WPMY-1 cells were seeded in the lower chamber containing 10% fetal bovine serum at varying ratios (PC-3 to WPMY-1) ranging from 1:2 to 1:16 as indicated (c-d). PC-3 cells (5 × 104) were seeded into the upper chamber containing serum-free medium, while WPMY-1 cells were seeded in the lower chamber containing 10% fetal bovine serum at a fixed ratio of 1:8 (PC-3 to WPMY-1) (e-f). Representative images (c, e) and quantification (d, f, mean ± SD of three biological replicates; one-way ANOVA with Dunnett’s multiple comparisons test) are presented. The scale bar represents 200 μm. g-h) PC-3 cells with VCAN knockdown (shVCAN) or control (shGFP) were cultured either with or without WPMY-1 cells. In the coculture condition, WPMY-1 cells were seeded in the lower chamber at a ratio of 1:8 (PC-3 to WPMY-1) in medium containing 10% fetal bovine serum (FBS); in the monoculture condition, the lower chamber contained only medium with 10% FBS. Representative images (g) and quantification (h, mean ± SD of three biological replicates; two-sided Student’s t-test) are presented. The treatments (knockdown) for each group are indicated in the figure. The scale bar represents 200 μm.
Extended Data Fig. 9 m6A Stabilizes VCAN mRNA and Promotes its Translation.
a-b) Knockdown efficiencies of the m6A regulators by shRNAs in PC-3 at RNA (a, Data are represented as mean ± SD of three biological replicates. P values are calculated using unpaired two-sided T test.) and protein level (b, Numbers indicate the positions of molecular mass (kDa) standards). c) Relative abundance of VCAN mRNA after knockdown m6A regulators in PC-3 cells. Ctrl:shRNA targets GFP, M3:METTL3, A5:ALKBH5. Data are represented as mean ± SD of three biological replicates; P values, unpaired Student’s T test. d-e) The VCAN mRNA abundance (d) and growth curve (e) of PC-3 cells after writing by VCAN mRNA-targeting guide RNA mixes. Data are represented as mean ± SD of three biological replicates. P values, one-way ANOVA with Dunnett’s multiple comparisons test. f-g) Knockdown efficiencies of YTHDFs (f) and corresponding VCAN protein abundance change (g) in PC-3 cells infected with siRNAs against each YTHDF were assessed by western blot. Numbers indicate the positions of molecular mass (kDa) standards. h-i) Knockdown efficiencies of YTHDC1 (h) and YTHDC2 (i) in PC-3 cells infected with siRNAs against each YTHDC were assessed by western blot. Vinculin was used as a loading control. Numbers on the right indicate the positions of molecular mass (kDa) standards. j) Western blot showing VCAN protein abundance change after knocking down YTHDC1 and YTHDC2 in PC-3 cells. Ponceau S stain serves as the loading control. Numbers on the right indicate the positions of molecular mass (kDa) standards. k) RIP efficiencies of the IGF2BP proteins were evaluated by western blot. Vinculin was included as negative control of the IPs. Numbers on the right indicate the positions of molecular mass (kDa) standards. l) Knockdown efficiencies of IGF2BPs in PC-3 cells infected with siRNAs against each IGF2BP were assessed by western blot. Vinculin was used as a loading control. Numbers on the right indicate the positions of molecular mass (kDa) standards. m) Relative VCAN mRNA abundance after siRNA-mediated knockdown of IGF2BPs in PC-3 cells. Data are represented as mean ± SD of three biological replicates. P values are from unpaired T test.
Extended Data Fig. 10 m6A Regulates VCAN Expression via IGF2BPs.
a) Cell proliferation after siRNA-mediated knockdown of IGF2BPs in PC-3 cells. Data are represented as mean ± SD of three biological replicates. P values are calculated using one-way ANOVA with Dunnett’s multiple comparisons test. b-c) Migration and invasion of PC-3 cells after siRNA-mediated knockdown of IGF2BPs in PC-3 cells. P values are calculated using one-way ANOVA with Dunnett’s multiple comparisons test. The representative images (b), and the quantification (c). Data are represented as mean ± SD of three biological replicates. The scale bar represents 200 μm. d) Ribosome profiling analysis of sample fractions. The graph illustrates ribosome profiling, with the intensity of the signal (voltage) plotted on the y-axis and sample fractions collected on the x-axis. The fractions are grouped based on the presence of polysomes, with polysome and polysome-free fractions distinguished and labeled accordingly on the graph. e) Relative enrichment of VCAN, GAPDH, and U6 in the polysome-free and polysome-associated fractions of PC-3 cells were measured by qPCR. GAPDH and U6 were employed as controls. Data are represented as mean ± SD of three biological replicates. P values are calculated using an unpaired two-sided Student’s T test. f) VCAN protein abundance after dCasRx-METTL3-based m6A writing with or without knocking down of IGF2BP2 was detected by western blot. Ponceau S stain serves as the loading control. Numbers on the right indicate the positions of molecular mass (kDa) standards.
Supplementary information
Supplementary Information
Supplementary Note, Figs. 1–4, Discussion and Methods.
Supplementary Tables 1–25
A workbook contains 25 Supplementary Tables in multiple tabs.
Supplementary Video 1
Videos of PC-3-zsGreen knockdown control cells during cancer cell extravasation and invadopodium formation. At t = 3.8 s, a mature invadopod is visualized by the green protrusion extending to the endothelium (purple). Lens culinaris agglutinin DyLight 649 to label the luminal surface of endothelial cells (purple). Scale bars, 10 μm.
Supplementary Video 2
Videos of PC-3-zsGreen knockdown of VCAN by sh1 cells during cancer cell extravasation and invadopodium formation. Cells (in red outline) arrested within the capillary bed underwent cell shrinkage. Hoechst to label the nucleus (cyan), dextran–fluorescein to label the lumen of the capillary bed (green), lens culinaris agglutinin DyLight 649 to label the luminal surface of endothelial cells (purple). Scale bars, 10 μm.
Supplementary Video 3
Videos of PC-3-zsGreen knockdown of VCAN by sh2 cells during cancer cell extravasation and invadopodium formation. Cells (in red outline) arrested within the capillary bed underwent cell shrinkage. Hoechst to label the nucleus (cyan), dextran–fluorescein to label the lumen of the capillary bed (green), lens culinaris agglutinin DyLight 649 to label the luminal surface of endothelial cells (purple). Scale bars, 10 μm.
Source data
Source Data Fig. 6 and Extended Data Figs. 7–9
Unprocessed western blots.
Source Data Fig. 1
Source data for Fig. 1.
Source Data Fig. 6
Statistical source data.
Source Data Extended Data Fig. 7
Source data for Extended Data Fig. 7.
Source Data Extended Data Fig. 8
Source data for Extended Data Fig. 8.
Source Data Extended Data Fig. 9
Source data for Extended Data Fig. 9.
Source Data Extended Data Fig. 10
Source data for Extended Data Fig. 10.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.
About this article
Cite this article
Xu, X., Zhu, H., Hugh-White, R. et al. The landscape of N6-methyladenosine in localized primary prostate cancer. Nat Genet 57, 934–948 (2025). https://doi.org/10.1038/s41588-025-02128-y
Received:
Accepted:
Published:
Version of record:
Issue date:
DOI: https://doi.org/10.1038/s41588-025-02128-y
This article is cited by
-
Translating the m6A epitranscriptome for prostate cancer
Nature Reviews Urology (2025)
