Recent technological developments that enable simultaneous measurement of many parameters (“omic” approaches) hold the promise of identifying new molecular prognostic and predictive markers. Accordingly the field of functional genomics has progressed rapidly due in large part to the development of microarray technology. We applied and implemented an algorithm for extracting information from the published gene expression microarray datasets. The approach is based on the hypothesis that gene products involved in the same biological process would be coupled at transcriptional level and was applied to the identification of correlated gene expression of ligands and receptors, as possible activators of signaling in ovarian cancer. By using, as resources, the Database of Ligand-Receptor Patterns (DLRP), the Human Protein Reference Database and PubMed, and selecting proteins whose interaction is at the cell surface we compiled a list of 535 ligand-receptor (L-R) pairs. For each pair, gene expression profiles were extracted from eight microarray ovarian cancer datasets and the Pearson correlation coefficient was calculated. Eighty-four L-R pairs showed a statistically significant correlation, after Bonferroni correction, in at least one dataset, and 12 were significantly correlated in 3 or more datasets. L-R pairs, previously implicated in ovarian cancer signaling, such as the colony-stimulating factor 1 (CSF1) ligand and its receptor (CSF1R), the hepatocyte growth factor (HGF) ligand and its receptor (MET) and the plasminogen activator urokinase (PLAU) ligand and its receptor (PLAUR), significantly correlated in 3/5, 6/6 and 5/7 datasets respectively, thus demonstrating the validity of this approach. Beside that, new L-R pairs were identified as potentially implicated in signaling in EOC. Subsequently, we calculated the correlation coefficients in patient samples subdivided according to the survival length. This analysis extracted four L-R pairs with significantly correlated gene expression patterns in two out of three datasets. One of L-R pairs, associated with short-term survival, is involved in motility and angiogenesis. The coordinated expression of this L-R was analyzed by western blot and immunohistochemistry using in vitro cell lines and surgical specimens from ovarian cancer. In conclusion, this study highlights the feasibility of analyzing large scale gene expression data to identify L-R signaling loops in ovarian cancer and, once supported and validated by biological assays, opens the way to the design of specifically targeted new diagnostic and therapeutic tools.
[Proc Amer Assoc Cancer Res, Volume 47, 2006]