Evidence for a transcriptional signature of breast cancer

Cancer arises from a step-wise accumulation of genetic and epigenetic changes in oncogenes and tumor suppressor genes, followed by changes in transcription and protein profiles. To identify the intrinsic transcriptional features of breast cancer and to explore in more detail the molecular basis of breast carcinogenesis, genes differentially expressed between cancers and their paired normal breast samples in nine breast cancer patients were screened using microarray. Nine normal breast tissues and 49 breast cancer tissue samples were then clustered based on the set of differentially expressed genes. A transcriptional signature of breast cancer consisting of 188 differentially expressed genes was identified. This signature allowed the normal breast tissues to be distinguished from all of the breast cancer samples, and primary breast cancers could be classified into two phenotype-associated subgroups with different ER status and clinical outcome. Furthermore, the classification accuracy of the set of differentially expressed genes was validated in publically available breast microarray data. Moreover, the differentially expressed genes could be grouped into five subclusters involved in different biological processes of carcinogenesis. Most genes in a given subcluster interacted within an independent subnetwork, and subnetworks could cross-talk through a set of signal molecules. Thus, the transcriptional signature identified here may be an intrinsic feature of breast cancer, and it may constitute to the molecular basis of breast carcinogenesis and different phenotypes of breast cancer.