skip to primary navigationskip to content
 

Molecular principles by which gene fusions affect protein interaction networks in cancer

last modified Jun 18, 2015 12:17 PM
Natasha Latysheva, MRC Laboratory of Molecular Biology

Molecular principles by which gene fusions affect protein interaction networks in cancer

Natasha Latysheva, MRC Laboratory of Molecular Biology

Abstract

Gene fusions arising from structural mutations are common cancer causing entities. Although some individual instances have been well studied, the molecular properties and mechanisms of action of gene fusions are relatively poorly understood. In this work, we integrate and analyse numerous databases to characterise the protein interaction network properties and molecular features of thousands of potential fusion proteins. Our computational analysis demonstrates that gene fusions tend to incorporate highly-connected hub genes in interaction networks, and that their protein products have high numbers of interaction-mediating features, such as certain domains, complex-forming interfaces, interacting linear motifs, and specific PTM profiles. These results suggest that fusion proteins could rewire molecular interactions and change the topology of signaling and regulatory pathways by recombining interactionprone protein segments. We also find that higher network centrality of fusion genes is associated with worse patient outcomes, correlating with progression to metastasis. For the most common fusion genes, we find that highly non-random segments of the parent proteins are selected, leading to hypotheses for the oncogenicity of these fusions. Finally, we highlight fusion proteins in which oncogenes lose segments that can be ubiquitinated or tumour suppressor genes fuse to segments that can be ubiquitinated. These findings bolster the role of gene fusions in cancers and provide a holistic overview of the interactomic role of fusion proteins.