EMT Features in Claudin-Low versus Claudin-Non-Suppressed Breast Cancers and the Role of Epigenetic Modifications

Abstract

Background: Breast cancers are heterogeneous and are classified according to the expression of ER, PR and HER2 receptors to distinct groups with prognostic and therapeutic implications. Within the triple-negative group, with no expression of these three receptors, molecular heterogeneity exists but is currently not exploited in the clinic. The claudin-low phenotype is present in a subset of triple-negative breast cancers and constitutes together with basal-like cancers the most extensive groups within triple-negative breast cancers. Suppression of epithelial cell adhesion molecules in claudin-low cancers is also a hallmark of Epithelial Mesenchymal Transition (EMT).

Methods: The groups of claudin-low and claudin-non-suppressed breast cancers from the extensive publicly available genomic cohorts of the METABRIC study were examined to delineate and compare their molecular landscape. Genetic and epigenetic alterations of key factors involved in EMT and potentially associated with the pathogenesis of the claudin-low phenotype were analyzed in the two groups.

Results: Claudin-low cancers displayed up-regulation of several core transcription factors of EMT at the mRNA level, compared with claudin-non-suppressed breast cancers. Global promoter DNA methylation was increased in both groups of triple-negative cancers and in claudin-low ER-positive cancers compared with the rest of ER-positive cancers. Histone modifier enzymes, including methyltransferases, demethylases, acetyltransferases and deacetylases displayed amplifications more frequently in claudin-non-suppressed triple-negative cancers than in claudin-low counterparts and the expression of some of these enzymes differed significantly between the two groups.

Conclusion: Claudin-low and claudin-non-suppressed triple-negative breast cancers differ in their landscape of EMT core regulators and epigenetic regulators. These differences may be explored as targets for therapeutic interventions specific to the two groups of triple-negative breast cancers.

Keywords: claudins; epigenetic; epithelial-to-mesenchymal transition; methylation; plasticity.

Figure 1

mRNA expression calculated as z-scores relative to all samples (log RNA Seq RPKM) of claudins 3, 4 and 7, E cadherin and occludin in representative breast cancer cases. (A). ER-negative/PR-negative, basal-like cancers, (B). ER-negative/PR-negative, claudin-low cancers, (C). ER-positive/HER2-negative/proliferation-low, luminal A cancers, (D). ER-positive/HER2-negative/proliferation-low, claudin-low cancers. Data are from the METABRIC cohort. Red color denotes up-regulation and blue denotes suppression. Gene symbols CLDN3: Claudin 3, CLDN4: Claudin 4, CLDN7: Claudin 7, CDH1: E cadherin, OCLN: Occludin.

Figure 1

mRNA expression calculated as z-scores relative to all samples (log RNA Seq RPKM) of claudins 3, 4 and 7, E cadherin and occludin in representative breast cancer cases. (A). ER-negative/PR-negative, basal-like cancers, (B). ER-negative/PR-negative, claudin-low cancers, (C). ER-positive/HER2-negative/proliferation-low, luminal A cancers, (D). ER-positive/HER2-negative/proliferation-low, claudin-low cancers. Data are from the METABRIC cohort. Red color denotes up-regulation and blue denotes suppression. Gene symbols CLDN3: Claudin 3, CLDN4: Claudin 4, CLDN7: Claudin 7, CDH1: E cadherin, OCLN: Occludin.

Figure 2

mRNA expression calculated as z-scores relative to all samples (log RNA Seq RPKM) of transcription regulators ZEB1, ZEB2, SNAIL (Gene symbol: SNAI1), Slug (Gene symbol: SNAI2), FOXC2 and TWIST1 in representative breast cancer cases. (A). ER-negative/PR-negative, basal-like cancers, (B). ER-negative/PR-negative, claudin-low cancers, (C). ER-positive/HER2-negative/proliferation-low, luminal A cancers, (D). ER-positive/HER2-negative/proliferation-low, claudin-low cancers. Data are from the METABRIC cohort. Red color denotes up-regulation and blue denotes suppression.

Figure 2

mRNA expression calculated as z-scores relative to all samples (log RNA Seq RPKM) of transcription regulators ZEB1, ZEB2, SNAIL (Gene symbol: SNAI1), Slug (Gene symbol: SNAI2), FOXC2 and TWIST1 in representative breast cancer cases. (A). ER-negative/PR-negative, basal-like cancers, (B). ER-negative/PR-negative, claudin-low cancers, (C). ER-positive/HER2-negative/proliferation-low, luminal A cancers, (D). ER-positive/HER2-negative/proliferation-low, claudin-low cancers. Data are from the METABRIC cohort. Red color denotes up-regulation and blue denotes suppression.

Figure 3

Percentage of amplifications of transcription regulators ZEB1, ZEB2, SNAIL (Gene symbol: SNAI1), Slug (Gene symbol: SNAI2), FOXC2 and TWIST1 in ER-negative/PR-negative, basal-like breast cancers (grey bars) and ER-negative/PR-negative, claudin-low breast cancers (black bars). Data are from the METABRIC cohort. Fisher’s exact test p = 0.02 for the comparison of the presence of amplifications in any core EMT transcription factor in claudin-low versus basal-like breast cancers.

Figure 4

mRNA expression calculated as z-scores relative to all samples (log RNA Seq RPKM) of methylation-sensitive genes in representative breast cancer cases. (A). ER-negative/PR-negative, basal-like cancers, (B). ER-negative/PR-negative, claudin-low cancers, (C). ER-positive/HER2-negative/proliferation-low, luminal A cancers, (D). ER-positive/HER2-negative/proliferation-low, claudin-low cancers. Data are from the METABRIC cohort. Red color denotes up-regulation and blue denotes suppression.

Figure 5

mRNA expression calculated as z-scores relative to all samples (log RNA Seq RPKM) of DNA methyltransferases DNMT1, DNMT3A and DNMT3B in representative breast cancer cases. (A). ER-negative/PR-negative, basal-like cancers, (B). ER-negative/PR-negative, claudin-low cancers, (C). ER-positive/HER2-negative/proliferation-low, luminal A cancers, (D). ER-positive/HER2-negative/proliferation-low, claudin-low cancers. Data are from the METABRIC cohort. Red color denotes up-regulation and blue denotes suppression.

Figure 6

Percentage of amplifications of genes encoding for histone methyltransferases in ER-negative/PR-negative, basal-like breast cancers (grey bars) and ER-negative/PR-negative, claudin-low breast cancers (black bars). Data are from the METABRIC cohort. Fisher’s exact test p < 0.0001 for the comparison of the presence of amplifications in any histone methyltransferase in claudin-low versus basal-like breast cancers.

Figure 7

Percentage of amplifications of genes encoding for histone demethylases in ER-negative/PR-negative, basal-like breast cancers (grey bars) and ER-negative/PR-negative, claudin-low breast cancers (black bars). Data are from the METABRIC cohort. Fisher’s exact test p < 0.0001 for the comparison of the presence of amplifications in any histone demethylase in claudin-low versus basal-like breast cancers.

Figure 8

mRNA expression calculated as z-scores relative to all samples (log RNA Seq RPKM) of genes encoding for histone methyltransferases in representative breast cancer cases. (A). ER-negative/PR-negative, basal-like cancers, (B). ER-negative/PR-negative, claudin-low cancers, Methyltransferases with significantly different expression between the 2 groups are shown. Red color denotes up-regulation and blue denotes suppression.

Figure 9

Percentage of amplifications of genes encoding for histone acetyltransferases in ER-negative/PR-negative, basal-like breast cancers (grey bars) and ER-negative/PR-negative, claudin-low breast cancers (black bars). Data are from the METABRIC cohort. Fisher’s exact test p = 0.02 for the comparison of the presence of amplifications in any histone acetyltransferase in claudin-low versus basal-like breast cancers.

Figure 10

Percentage of amplifications of genes encoding for histone deacetylases in ER-negative/PR-negative, basal-like breast cancers (grey bars) and ER-negative/PR-negative, claudin-low breast cancers (black bars). Data are from the METABRIC cohort. Fisher’s exact test p = 0.06 for the comparison of the presence of amplifications in any histone deacetylase in claudin-low versus basal-like breast cancers.