A novel methodology for the identification of drugable enzymes as targets in cancer via meta analysis of gene expression microarrays

Abstract

9602 Background: Many studies have used expression microarrays to identify cancer signatures for diagnostic and prognostic purposes. We hypothesized that identifying enzymes with known therapeutic inhibitors, which are highly over-expressed in cancer as compared to benign tissue, may uncover drugable targets shared by multiple cancers and those specific to a cancer type. To test the methodology, we analyzed arrays from untreated prostate and non-small cell lung cancer (NSCLC) patients for their variation in tumor biology and large patient samples. Methods: Utilizing ONCOMINE (www.oncomine.com), we meta-analyzed eight prostate & six lung cancer arrays published in literature. Prior well characterized drug targets such as DNA topoisomerase 1 & 2 were excluded. Differential expression analysis was done utilizing t-statistics and corrected for multiple comparisons using adjusted Qvalues (Q = p* n/I; where n = total number of genes and I=sorted rank of p value). Results: We identified 10 significant enzymes (Q ≤ 1.0e-2): 4 shared, 3 unique to prostate and 4 to NSCLC. Shared targets include two epigenetic enzymes, histone deacetylase1 (HDAC1) and poly (ADP-ribose)polymerase1 (PARP1), guanine monophosphate synthetase (GMPS), an enzyme in nucleoside catabolism involved in DNA synthesis and cellular metabolism, and cyclin-dependent kinase 4 (CDK4), a cell-cycle regulatory protein. Targets unique to prostate cancer are endothelin converting enzyme1 (ECE1) which forms endothelin-1 from its precursor, and two critical enzymes in fatty acid systhesis pathways implicated in prostate cancer pathogenesis: fatty acid synthase (FASN) and acetyl-Coenzyme A carboxylase alpha (ACACA). NSCLC enzymes are a protein that regulates p53 function, NADPH dehydrogenase (NQO1), macrophage migration inhibitory factor (MIF), minichromosome maintenance deficient 6 (MCM6) which is a DNA helicase and casein kinase 2 (CSNK2A1), a putative target of the active anti-oxidant of red wine, resveratrol. Conclusions: Meta-analysis of expression microarrays is a novel methodology which can quickly uncover targets over expressed in multiple cancers and help identify potential therapeutic inhibitors. No significant financial relationships to disclose.