Computational analysis of sumoylation in protein kinases: a biomedical perspective

Abstract

Sumoylation is a general post-translational modification that re- gulates several cell processes such cell cycle control, transcription, DNA repair, and chromosome segregation. It is a covalent attach- ment of a small ubiquitin like modifier SUMO protein to target pro- tein lysines. This attachment affects biological functions of target proteins by blocking interactions sites, creating novel sites, or in- ducing conformational changes. Some kinases are regulated by this mechanism. For instance, AuroraB kinase, MEK, FAK and ERK5 are known targets of sumoylation. Give the importance of kinases as a main player in signal transduction processes, as well as being very relevant to disease and development, we aimed to address whether this mechanism is anecdotal or general within the kinase superfa- mily, and also if there are over-represented in developmental path- ways. To this purpose, we set to identify suitable features from these proteins that could be associated to infer and predict the impact of SUMOylation. Using bioinformatic/computational approaches we have 1) analysed potential SUMOylation in kinases subfamilies, 2) analysed them according to structural constraints, and 3) estima- ting its presence in cancer samples that exhibit different mutation rate, and 3)mapping SUMOylation distributions in kinases involved in developmental pathways. Our results demonstrate that protein kinases could be highly regulated at both family and superfamily level, where potential sumoylated lysines are significantly distribu- ted close to active sites, away from binding sites, and interestingly buried within the protein. We have not found any significant pre- sence of SUMOylable lysines in cancer, so according to our data, SUMOylation alteration may not be a general mechanism contri- buting to cancer development in the Kinase superfamily. Keywords: Computational Biology, Sumoylation, Protein kinases, Bioinformatics, Disease, Cancer, Python/R.