Construction of a chromosome 17 transcriptome in ovarian cancer identifies differentially expressed genes

¹Paulina M Wojnarowicz, ²Ali Filali-Mouhim, ^2,3Diane M Provencher, ^2,4Anne-Marie Mes-Masson, ^1,5,6Patricia N Tonin
¹Department of Human Genetics, McGill University, Montreal, Quebec, H3A 1B1, Canada, ²Centre de Recherche du Centre Hospitalier de L'Universite de Montreal/Institut de Cancer de Montreal, Montreal, Quebec, H2L 4M1, Canada, ³Division de gynecologie oncologique, Universite de Montreal, Montreal, Quebec, H2L 4M1, Canada, ⁴Departement de medicine, Universite de Montreal, Montreal, Quebec, H3T 3J7, Canada, ⁵The Research Institute of the McGill University Health Centre, Montreal, Quebec, H3G 1A4, Canada, ⁶Department of Medicine, McGill University, Montreal, Quebec, H3G 1A4, Canada

Epithelial ovarian cancer (EOC) is the most lethal, and second most common gynaecological cancer. The disease is further subcategorized by histopathological subtype, the most common being the serous subtype, and accounting for nearly 50% of all cases of ovarian cancer. Chromosome 17 has been implicated as playing an important role in the pathogenesis of ovarian cancer. Partial deletion or entire loss of chromosome 17 is found in 50-70% of tumours, consistent with the hypothesis that this chromosome harbours tumour suppressor genes important in tumourigenesis. By characterizing candidate tumour suppressor genes (TSGs) it may be possible to gain information about the tumour biology that can then be used in the development of screening programs, treatments and cures. As the serous histopathological subtype of EOC accounts for the greatest proportion of ovarian cancer cases and chromosome 17 is often implicated in the disease, identification of genes involved in this intersection may prove significant in our understanding of the genes and genetic interplay active in ovarian cancer. In order to identify chromosome 17 genes that play a role in ovarian cancer, a transcriptome map was generated by comparing primary cultures derived from 15 normal ovarian surface epithelium (NOSE) samples, 4 well-characterized EOC cell lines and 28 malignant ovarian tumour (TOV) samples, of which 17 belong to the serous histopathological subtype. The expression microarray analysis was carried out using the Affymetrix U133A GeneChip array. The normalized MAS5.0 derived data was then filtered by removing probe sets that did not exhibit at least a 3-fold difference between the mean of the NOSE samples and any TOV sample across a probe set, an expression value of at least 100 across the panel of samples for each probe set, and a TOV expression value outside the range of the NOSE expression values. This triple-filtered data was then analyzed with a specific focus on the profiles of the serous TOV samples. There were 253 genes that met these criteria, 158 of which were underexpressed relative to the NOSE samples. These underexpressed genes are most likely TSG candidates. The results of this work could identify genes and pathways that can lead to the development of biomarkers for clinical use. In addition, the results will enhance our understanding of the role of chromosome 17, and the genes harboured therein, in ovarian cancer tumourigenicity.