Poster Presentation 9th General Meeting of the International Proteolysis Society 2015

Characterization and prediction in a computational representation of the protease web (#119)

Nikolaus Fortelny 1 , Reinhild Kappelhoff 1 , Philipp F Lange 1 , Paul Pavlidis 1 , Chris M Overall 1
  1. UBC Vancouver, Vancouver, BC, Canada

Proteases generate shortened proteins with altered function and thus mold the proteome. In vivo studies of proteases are commonly based on comparing protease knock-out samples to wild-type. However, in vivo protease characterization, especially protease substrate assignment, is intricate because of the complexity of the biological system with indirect effects resulting from the activation of secondary proteases. An in silico representation of the protease web [1] allowed us to map complex protease interactions, demonstrate the pervasive connectivity between proteases, and to predict proteolytic pathways [2]. We further characterized the interplay of the protease web with other biological networks, thereby predicting additional protease interactions.

Building on concepts published last year, we modeled large-scale proteolytic data as a network and created the web-based tool PathFINDer that predicts proteolytic pathways. Furthermore, we assembled protein-protein interaction, co-expression, phylogenetic and kinase network data to predict additional protease interactions.

Our network model demonstrated the pervasiveness of the human protease web with interactions between various protease classes and cascades with protease inhibitors often forming network hubs. Proteases can thus potentially influence the cleavage of many more proteins than their direct substrates through protease activation pathways or inactivation of protease inhibitors. We create a web-based software termed PathFINDer (http://clipserve.clip.ubc.ca/topfind/pathfinder) that enables the query of the protease web to return proteolytic pathways and networks from the query protease to a set of substrates, each path representing a biochemical mechanism that can be validated. Recently, we contrasted the protease web model with other biological networks, thus predicting and biochemically validating additional protease interactions and demonstrating the imbedding of the protease web in the biological system.

We present systematically derived evidence for the existence of the protease web, a software tool PathFINDer that predicts proteolytic paths, and show the relationship of the protease web with other biological networks.

  1. Fortelny N, Cox JH, Kappelhoff R, Starr AE, Lange PF, Pavlidis P, Overall CM (2014), Network Analyses Reveal Pervasive Functional Regulation Between Proteases in the Human Protease Web, PLoS Biol, 12(5): e1001869
  2. Fortelny N, Yang S, Pavlidis P, Lange PF* and Overall CM* (2015) Proteome TopFIND 3.0 with TopFINDer and PathFINDer: database and analysis tools for the association of protein termini to pre- and post-translational events, Nucl Acids Res, 43 (D1): D290-D297