Molecular subtypes of breast cancer are associated with characteristic DNA methylation patterns

Abstract

Introduction: Five different molecular subtypes of breast cancer have been identified through gene expression profiling. Each subtype has a characteristic expression pattern suggested to partly depend on cellular origin. We aimed to investigate whether the molecular subtypes also display distinct methylation profiles.

Methods: We analysed methylation status of 807 cancer-related genes in 189 fresh frozen primary breast tumours and four normal breast tissue samples using an array-based methylation assay.

Results: Unsupervised analysis revealed three groups of breast cancer with characteristic methylation patterns. The three groups were associated with the luminal A, luminal B and basal-like molecular subtypes of breast cancer, respectively, whereas cancers of the HER2-enriched and normal-like subtypes were distributed among the three groups. The methylation frequencies were significantly different between subtypes, with luminal B and basal-like tumours being most and least frequently methylated, respectively. Moreover, targets of the polycomb repressor complex in breast cancer and embryonic stem cells were more methylated in luminal B tumours than in other tumours. BRCA2-mutated tumours had a particularly high degree of methylation. Finally, by utilizing gene expression data, we observed that a large fraction of genes reported as having subtype-specific expression patterns might be regulated through methylation.

Conclusions: We have found that breast cancers of the basal-like, luminal A and luminal B molecular subtypes harbour specific methylation profiles. Our results suggest that methylation may play an important role in the development of breast cancers.

Figure 1

Unsupervised clustering of 189 tumours based on the 332 most variably methylated CpGs. (a) Hierarchical clustering. The heatmap shows relative methylation levels (red, more methylated; green, less methylated). Clustering results in three clusters associated with lumB, lumA and basal-like tumours, respectively. (b) Kaplan-Meier demonstrating longest survival in lumA-associated Cluster 2 and shortest in basal-like-associated Cluster 3. P-value was calculated using log-rank test. (c) Fraction of genome altered (FGA) highest in basal-like-associated Cluster 3 and lowest in lumA-associated Cluster 2. P-value was calculated using analysis of variance (ANOVA). (d) S-phase fraction highest in basal-like-associated Cluster 3 and lowest in lumA-associated Cluster 2. P-value was calculated using ANOVA. The number of tumours in each subtype is shown at top.

Figure 2

Boxplot stratified by subtype for methylation frequencies of the 196 subtype-associated CpGs. These CpGs are more frequently methylated in lumB tumours and less methylated in basal-like tumours. P-value was calculated using analysis of variance. The number of tumours in each subtype is shown at top.

Figure 3

Heatmaps with average relative methylation and expression levels stratified by subtype. Subtype expression markers according to Hu et al. [14] having subtype-specific methylation are displayed. The number of samples in each subtype (top) and the Pearson correlation between methylation (red, more methylated; green, less methylated) and expression levels (red, high; green, low) are shown. The expression levels correspond well with methylation status. Gene expression data were unavailable for CXCL9.

Figure 4

Relative gene expression levels and genomic gain of EZH2 in the different subtypes. (a) Relative expression levels of EZH2 across subtypes. Basal-like tumours had the highest expression of EZH2. P-value was calculated using analysis of variance for all subtypes. (b) Fraction of samples with gain of EZH2. Gain of this gene is more frequent in basal-like tumours. P-value was calculated using Fisher's exact test between basal-like and the other subtypes. The number of tumours in each subtype is shown at the top.

Figure 5

Relative expression and methylation of PRC2 target genes derived from ES and MDA-MB-231 HOTAIR cells. PRC2 targets identified by Lee et al. in (a to c) ES cells [28] and Gupta et al. [29] by over-expressing HOTAIR in (d to f) MDA-MB-231 cells, and present in our gene expression data set or methylation panel, respectively, were used. (a and d) Average relative expression levels of PRC2 target genes. Basal-like and lumB tumours both have low expression of these genes compared with the other subtypes. P-values were calculated using analysis of variance. (b and e) Average relative methylation levels of PRC2 target genes. Low methylation levels are found in basal-like tumours while lumB tumours display high levels of methylation of these CpG sites. P-values were calculated using t-test between basal-like and lumB tumours. (c and f) Average relative methylation levels for PRC2 target genes compared with other genes for basal-like and lumB tumours. P-values were calculated using t-test. The number of tumours in each subtype is shown at the top.

Figure 6

Relative expression and methylation of SUZ12 and PRC2 target genes derived from MCF7 breast cancer cells. (a and c) SUZ12 targets identified by Squazzo et al. [30] and (b and d) PRC2 targets identified by Tan et al. [31], and present in our gene expression data set or methylation panel, respectively, were used. (a and b) Average relative expression of SUZ12 and PRC2 targets, respectively. LumA and especially LumB tumours, have low expression of these genes. P-values were calculated using analysis of variance (ANOVA). (c and d) Average relative methylation of SUZ12 and PRC2 targets, respectively. Higher methylation levels are found for lumB than lumA tumours. P-values were calculated using ANOVA. The number of tumours in each subtype is shown at top.

Figure 7

Potential model for the relations between luminal differentiation and breast cancer subtypes. PRC2-mediated gene silencing through trimethylation of H3K27 is common in stem/progenitor cells and would be characterised by high EZH2 expression and PRC2 targets having both low expression and unmethylated CpG sites. These characteristics match our findings for basal-like tumours. PRC2 is then displaced (upper path) and PRC2 targets are preferentially activated to promote differentiation. Such a committed cell state would be characterised by low EZH2 expression and PRC2 targets with both high expression and unmethylated promoters. These characteristics match our findings for lumA tumours. In cancer cells, an alternative route for differentiation (lower path), would be to more stably silence PRC2 target genes by promoter methylation. PRC2 associates with DNA methyltransferases (DNMTs) leading to hypermethylation of PRC2 targets. Such a committed cell state would be characterised by low EZH2 expression and PRC2 targets with both low expression and hypermethylated CpG sites. These characteristics match our findings for lumB tumours.