Analysis of IDH1 and IDH2 mutations as causes of the hypermethylator phenotype in colorectal cancer

Abstract

The CpG island methylator phenotype (CIMP) occurs in many colorectal cancers (CRCs). CIMP is closely associated with global hypermethylation and tends to occur in proximal tumours with microsatellite instability (MSI), but its origins have been obscure. A few CRCs carry oncogenic (gain-of-function) mutations in isocitrate dehydrogenase IDH1. Whilst IDH1 is an established CRC driver gene, the low frequency of IDH1-mutant CRCs (about 0.5%) has meant that the effects and molecular covariates of those mutations have not been established. We first showed computationally that IDH2 is also a CRC driver. Using multiple public and in-house CRC datasets, we then identified IDH mutations at the hotspots (IDH1 codons 132 and IDH2 codons 140 and 172) frequently mutated in other tumour types. Somatic IDH mutations were associated with BRAF mutations and expression of mucinous/goblet cell markers, but not with KRAS mutations or MSI. All IDH-mutant CRCs were CIMP-positive, mostly at a high level. Cell and mouse models showed that IDH mutation was plausibly causal for DNA hypermethylation. Whilst the aetiology of hypermethylation generally remains unexplained, IDH-mutant tumours did not form a discrete methylation subcluster, suggesting that different underlying mechanisms can converge on similar final methylation phenotypes. Although further analysis is required, IDH mutations may be the first cause of hypermethylation to be identified in a common cancer type, providing evidence that CIMP and DNA methylation represent more than aging-related epiphenomena. Cautious exploration of mutant IDH inhibitors and DNA demethylating agents is suggested in managing IDH-mutant CRCs. © 2025 The Author(s). The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Keywords: DNA methylation; cancer driver genes; cancer genomics; colorectal cancer; epigenetics; isocitrate dehydrogenase.

Figure 1

CpG methylation of IDH‐mutant CRCs in (A) TCGA‐COADREAD and (B) S:CORT. (i) The mean DNA methylation probe β value per cancer of IDH‐wildtype (WT, green) and IDH‐mutant (purple) CRCs. (ii) Association between global methylation levels (mean probe β values) and the four CIMP clusters, calculated from the CpG probes used for RPMM‐based cluster analysis, for each CRC within the four RPMM clusters. CRCs harbouring IDH1 ^R132C (red), IDH1 ^R132G (light blue), IDH2 ^R140W (purple), or IDH2 ^R172S (orange) mutations are highlighted, with IDH‐wildtype (WT) cancers in grey. ND = not defined. (iii) DNA methylation β value distributions of the probes used for RPMM clustering by IDH mutation status within North Shore (2 kb upstream of CpG island), CpG Island, and South Shore regions (2 kb downstream of CpG island).

Figure 2

RPMM‐based classification of CIMP status in TCGA‐COADREAD CRCs. β values from 2,062 CpG probes (y‐axis, where point colour indicates probe β value) were subjected to unsupervised recursively partitioned mixture model clustering. The annotation above the heatmap depicts the methylation cluster of each cancer, IDH mutation, and other chosen molecular and clinicopathological features. WT, wildtype; ND, not defined; MSI, microsatellite instability; MSS, microsatellite stable.

Figure 3

RPMM‐based classification of CIMP status in S:CORT CRCs. Clustering used 1,475 CpG probes. Other features are shown as per Figure 2. WT, wildtype; ND, not defined; MSI, microsatellite instability; MSS, microsatellite stable.

Figure 4

Gene expression characteristics of IDH‐mutant CRCs. (A) Volcano plot of IDH‐mutant (n = 4) versus IDH‐wildtype (n = 411) CRCs from the TCGA‐COAD study. Differentially expressed genes (DEGs) with p _FDR  < 0.05 reflect lower expression (log₂(fold‐change) < −1, blue) and higher expression (log₂(fold‐change) > 1, red) in IDH‐mutant versus IDH‐wildtype CRCs, respectively. The top 20 significant DEGs, potential components of an IDH‐mutant gene expression signature, are labelled and coloured green. (B) GSEA normalised enrichment scores (NES) for significant hallmark or cell type gene‐sets from a comparison of IDH‐mutant and IDH‐wildtype CRCs. Proportion of DEGs in gene‐set values indicate the fraction of the gene‐set that is significantly differentially expressed. (C) GSEA for goblet cells showing strong enrichment. (D) Expression levels in IDH‐mutant (purple) and IDH‐wildtype (green) CRCs of selected genes marking secretory progenitors (SPC), Goblet and Paneth cell‐specific transcription factors (GC/PC TFs), Goblet cells (GC markers) and intestinal stem cells (SC niche). Boxes show data between first and third quartiles, with the median as a horizontal line. Whiskers show 1.5× interquartile range. *p < 0.05, **p < 0.01, ***p < 0.001. (E) Goblet cell scores calculated for IDH‐mutant CRCs, colorectal adenocarcinomas with/without mucinous phenotype, and normal colonic tissue samples from TCGA. Tumours reported to have mucinous histology are shown in red, other tumours in green, and normal tissues in blue.

Figure 5

Analysis of Idh1‐mutant mice. (A) Genotyping of Vil1‐Cre (upper) and the recombined Idh1 ^fl(R132H)/+ allele (lower) in three Vil1‐Cre;Idh1 ^+/+ and three Vil1‐Cre;Idh1 ^fl(R132H)/+ animals. Primer sequences and PCR conditions are provided in the supplementary material, Table S1. (B) Idh1 RNA sequence derived from intestines of three Vil1‐Cre;Idh1 ^+/+ mice (upper) and three Vil1‐Cre;Idh1 ^fl(R132H)/+ mice (lower), showing expression of the mutant allele specifically in the latter. (C) Global methylation (mean β value of all probes) in the intestines of test and control Idh1‐mutant mice. Individual data points for each mouse are shown as dots. The boxes show the data between the first and third quartiles, with the median as a horizontal line. Whiskers show a maximum of 1.5× interquartile range. A linear mixed effect model, incorporating Idh1 status, sex, and age as fixed‐effect variables and BeadChip as a random effect was constructed, where mean DNA methylation β value (n = 265,710 probes) was the outcome variable. Test animals (n = 12; purple) were Vil1‐Cre;Idh1 ^fl(R132H)/+ and control animals were Vil1‐Cre;Idh1 ^+/+ (n = 5; green). The magnitude of the difference between Idh1‐mutant and wildtype mice is small in absolute terms, but its value is likely to be reduced by several factors, including a preponderance of methylation‐invariate sites, a polyclonal cell population and the relatively narrow time window for the Idh1 mutation to act compared with the life history of a human cancer.

Figure 6

DNA methylation changes in CRISPR‐edited IDH1‐mutant Caco‐2 cells. (A) D‐2HG levels for two IDH1‐wildtype (WT, grey) replicates, four IDH1 ^R132C replicates (comprised of two technical replicates of two independent clones, green), and two replicates of a single IDH1 ^R132G clone (purple). *p < 0.05 from two‐tailed Student's t‐test for pairwise difference. (B) The mean DNA methylation β values of the 728,496 EPIC array probes in Caco‐2 cells, comprising IDH1‐wildtype (WT, grey, n = 3), IDH1 ^R132C (green, n = 4), or IDH1 ^R132G (purple, n = 2). *p < 0.05, ***p < 0.001 from two‐tailed Student's t‐test for pairwise difference. Every mutant clone (n = 6) had higher methylation than any of the three wildtype clones (p = 0.020, Student's t‐test). (C, D) Volcano plots and bar charts showing differentially methylated probes in IDH1 ^R132C (C) and IDH1 ^R132G (D). Plotted in the volcano plots are the log₂ fold‐change (log₂FC) in methylation and −log₁₀ Benjamini−Hochberg false discovery rate‐corrected p value (−log₁₀ P) for each probe. Hypermethylated probes in IDH1‐wildtype cells are shown in turquoise and hypermethylated probes in IDH1‐mutant cells in red. Probes in grey show no significant difference between groups (p _FDR > 0.05). Also shown are the distribution of the β values of these significantly differentiated probes located within North Shore (2 kb upstream of CpG island), CpG Island, and South Shore regions (2 kb downstream of CpG island) in IDH1‐wildtype (WT, grey), IDH1 ^R132C (green), and IDH1 ^R132G (purple) Caco‐2 cells.