NTU Undergraduates' research April 2014 - Biosciences | Seite 75

The Molecular Basis of Breast Cancer Subtypes

Janette K-Boadu
School of Science and Technology, Nottingham Trent University, Clifton Campus, Clifton Lane, NG11 8NS

Abstract
Breast cancer (BC), is the most common cancer amongst women. Although incidence rates are increasing, our
understanding of the identification of risk factors, is slowly improving. The use of gene expression profiling and
hierarchal clustering analysis, has revealed that BC can be classified into five intrinsic subtypes; Basal-like,
ERBB2, Luminal A, Luminal B, and Normal-like, solely based on their gene expression patterns. In this study the
molecular basis of BC subtypes was investigated by gene expression profiling, using the Artificial Neuronal
Network’s Stepwise analysis method to generate a set of top genes represented by each subtype. These genes were
analysed to identify similarities and differences in expression amongst the subtypes, and further studied to identify
their biological functions in cell pathways, tumour progression, disease, and association with clinical outcome
including prognosis. Results showed distinct patterns in gene expression. In particular, the level of gene expression
was significantly lower (p<0.001), in the basal subtype compared to the remaining four subtypes. Many genes
associated with high proliferation, high tumour grade and poor prognosis were closely linked to the basal subtype.
In contrast, genes associated with a better clinical outcome were linked to the Luminal A subtype. Genes that were
significantly over/under-expressed in the subtypes may be potential gene targets for the development of new
alternative therapies in BC, especially the less researched normal-like subtype. Gene expression has allowed
scientists and clinicians to improve understanding of how to diagnose and treat patients, expanding knowledge of
the clinical manifestations of this heterogeneous disease.
Key words: Breast cancer, Basal, ERBB2, Luminal A, Luminal B, Normal-like, Artificial Neural Networks,
Microarray Data, Gene expression profiling, Probe-set, Prognosis, Treatment