Abstract 58: Understanding target biology using protein interactomes

Proceedings: AACR 102nd Annual Meeting 2011‐‐ Apr 2‐6, 2011; Orlando, FL The use of systematic genome-wide physical protein-protein interaction detection methods has led to the generation of large networks of putatively interconnected human proteins including a number of therapeutically relevant disease-associated proteins. These networks have intrinsic value for understanding the pathophysiology of proteins, identification of tractable drug targets within a network, and hypothesis generation with regard to potential pharmacodynamic and patient selection biomarkers for a drug target of interest. Here, we present a novel approach to utilize a high-confidence human focal adhesion kinase (FAK) protein interaction network to better understand FAK target biology. FAK is a well-studied signaling protein known to directly control cellular motility via mechanosensing. As a large 125 kDa protein with both protein tyrosine kinase and scaffolding functions, FAK is particularly rich in its potential to interact with numerous proteins through SH2 and other interaction domains. By forming a diverse array of target interactions that regulate cytoskeletal structural changes, cellular adhesion, cell migration and cellular proliferation, FAK is believed to be a key regulator of several critical processes in tumorigenesis. It has also been established that human tumors often exhibit FAK gene amplification with resulting increased FAK expression which correlates with cancer progression and poor prognosis. Although the plausible roles of FAK have been well-elaborated, there are limited data available on the precise interplay of biological entities controlling the FAK signaling pathway. In this study, we built a prototypical model representing cross-talk between multiple proteins believed to be involved in the FAK signaling pathway. In doing so, we describe a step by step procedure to assemble a FAK interactome using integrative data mining procedures with publicly available interaction databases such as MINT, BioGRID, DIP and HPRD. Next, we filter this data set by incorporating confidence scores and lines of evidence available corresponding to each interaction to finally derive a high-confidence FAK interactome. We propose that this approach to generate high-confidence protein interactomes could also be applied to other proteins of interest to facilitate a systems biology approach to elucidate additional biological pathways influencing tumorigenesis, metastasis and other disease states amenable to rational drug discovery. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 58. doi:10.1158/1538-7445.AM2011-58