Receptor tyrosine kinases are crucial for cell signaling, survival, metabolism, migration, and cell-cycle control. In cancer, dysregulation and activation occur through various mechanisms, including gene fusions that supplant the regulation that is afforded through ligand binding to the extracellular domain to drive constitutive activity through oligomerization. While oligomerization is a common mechanism for activating kinases, not all oligomers activate kinases and not all gene fusions are pathogenic. We hypothesize that specific rules governing structural orientation, conformation, configuration, recombination, and oligomerization affect kinase activation. To uncover these rules, we systematically examine fusion protein assemblies using Met Kinase as a model system. We invented a novel high-throughput synthetic biology platform to generate a targeted domain-domain recombination Met Fusion library, fusing naturally occurring oligomeric domain genes systematically at all possible junction points of Met Kinase in a pooled fashion and employ deep sequencing-coupled assays to assess fusion protein activity and folding. We used IL3-independent growth in Ba/F3 cells to rank fusion variants based on their oncogenic potential. We also used the complementation assay based on split-GFP system to measure the expression levels and folding potential of each fusion variant. We will further integrate the deep sequencing data with machine learning models to predict important sequence-based and structure-based features responsible for causing fusion protein oncogenesis, thereby, exposing rules for how the extent of oligomerization and particular orientations of fusion proteins cause cancer. Comprehensive understanding of fusion proteins will enable structure-guided approach to design novel inhibitors that can be used to target these fusion domains, which could expand the landscape of cancer care.
Bajaj et al. (Sun,) studied this question.