Identification of novel targets in non-muscle invasive bladder cancer: A systems biology approach.

Abstract

396 Background: Most bladder cancers are diagnosed as non-muscle invasive bladder cancer (NMIBC). Because of the high rate of recurrence and limited treatment options for recurrent disease, there is an urgent need to identify new treatment options. In this study, a network-based drug discovery approach was applied to identify novel targets in NMIBC. Methods: A computational systems-biology framework integrating genomic, transcriptomic, protein-protein, and protein-drug interactions was applied to identify and prioritize potential NMIBC targets and pathways. First, a NMIBC disease network module was constructed from genetic associations augmented with known protein-protein interactions. Independent NMIBC case/control expression data sets were leveraged to develop a phenotypic network module comprising genes differentially expressed in NMIBC. Proteins and pathways linking the disease model with the phenotypic module were identified, ranked, and benchmarked against known NMIBC pathways and drug targets from the literature. Results: NMIBC seed genes cluster significantly within the human interactome, reflecting an underlying disease module. The expanded disease module comprises 103 genes that show functional and literature-based enrichment. Genes differentially expressed in NMIBC were significantly connected to disease module genes compared with random expectation (z = 23.3), and are enriched in several biologically relevant pathways. Based on similarity to disease biology and network topology, 11 NMIBC pathways were identified as bridging genetic predisposition for and phenotypic expression of NMIBC. Within these overlapping pathways, known and novel drug targets were identified. Of these in silico targets, 19 are currently under development for treatment of bladder cancer, validating the approach. To further validate the approach, we observed overlap between in silico targets and hits from a chemogenomic library phenotypic screen. Conclusions: A network-driven systems biology approach identified both known NMIBC targets and novel proteins/pathways that are potential targets for the treatment of NMIBC. Further target validation is required.