Frequent loss of endothelin-3 (EDN3) expression due to epigenetic inactivation in human breast cancer
- Equal contributors
1 Molecular Oncology Group, Institute of Pathology, University Hospital of the RWTH Aachen, Pauwelsstrasse 30, D-52074 Aachen, Germany
2 Cancer Epigenetics Group, Cancer Epigenetics and Biology Program (PEBC), Bellvitge Institute for Biomedical Research (ICO-IDIBELL), Av. Gran Via de L'Hospitalet 199-203, E-08907 Barcelona, Spain
3 Department of Internal Medicine IV (Haematology/Oncology), University Hospital of the RWTH Aachen, Pauwelsstrasse 30, 52074 D-Aachen, Germany
4 Department of Pathology, University of Erlangen, Krankenhausstrasse 12, D-91054 Erlangen, Germany
Breast Cancer Research 2009, 11:R34 doi:10.1186/bcr2319Published: 15 June 2009
Endothelin (EDN) signalling plays a crucial role in cell differentiation, proliferation and migration processes. There is compelling evidence that altered EDN signalling is involved in carcinogenesis by modulating cell survival and promoting invasiveness. To date, most reports have focused on the oncogenic potential of EDN1 and EDN2, both of which are overexpressed in various tumour entities. Here, we aimed at a first comprehensive analysis on EDN3 expression and its implication in human breast cancer.
EDN3 mRNA expression was assessed by Northern blotting in normal human tissues (n = 9) as well as in matched pairs of normal and tumourous tissues from breast specimens (n = 50). EDN3 mRNA expression in breast cancer was further validated by real-time polymerase chain reaction (PCR) (n = 77). A tissue microarray was used to study EDN3 protein expression in breast carcinoma (n = 150) and normal breast epithelium (n = 44). EDN3 promoter methylation was analysed by methylation-specific PCR in breast cell lines (n = 6) before and after demethylating treatment, normal breast tissues (n = 17) and primary breast carcinomas (n = 128). EDN3 expression and methylation data were statistically correlated with clinical patient characteristics and patient outcome.
Loss of EDN3 mRNA expression in breast cancer, as initially detected by array-based expression profiling, could be confirmed by Northern blot analysis (> 2-fold loss in 96%) and real-time PCR (> 2-fold loss in 78%). Attenuated EDN3 expression in breast carcinoma was also evident at the protein level (45%) in association with adverse patient outcome in univariate (P = 0.022) and multivariate (hazard ratio 2.0; P = 0.025) analyses. Hypermethylation of the EDN3 promoter could be identified as the predominant mechanism leading to gene silencing. Reversion of the epigenetic lock by 5-aza-2'-deoxycytidine and trichostatin A resulted in EDN3 mRNA re-expression in vitro. Furthermore, EDN3 promoter hypermethylation was detected in 70% of primary breast carcinomas with significant association to loss of EDN3 mRNA expression (P = 0.005), whilst normal matched breast tissues revealed no EDN3 promoter methylation.
EDN3 is a frequent target of epigenetic inactivation in human breast cancer, potentially contributing to imbalanced EDN signalling commonly found in this disease. The clinical implication supports the view that EDN3, in contrast to EDN1 and EDN2, may act as natural tumour suppressor in the human mammary gland.