Phosphoproteome-based Modeling And Analysis Of Kinase-substrate Phosphorylation Networks

Protein phosphorylation is one of the most important protein post-translational modifications,and it is mediated by protein kinases through transferring the phosphate group from ATP to the specific amino acid in substrate.Protein phosphorylation is a reversible reaction,and the dephosphorylation is mediated by protein phosphatases.As an important mechanism in regulation of biological processes,the abnormal phosphorylation is always associated with the diseases or cancers.Recently,the explosive data were produced with the development of genomics and proteomics,and how to extract the the useful information is the key to study the phosphorylation and diseases.In this paper,we have performed the systematic analysis to study the association between phosphorylation and diseases in both enzyme and substrates level.Protein kinases are key regulators mediating the reversible phosphorylation reaction,and a better understanding of kinases is the fundamental for studying the phosphorylation.In this work,we performed a comprehensive study on the classification of kinases.Through the literature mining and database integration,we have collected 1,855 protein kinases.Based on the previous studies,we standardized and unified the classification for kinases and classified into 10 groups and 149 families.Furthermore,the Profile HMM model and ortholog search were performed to predict the potential kinases in 84 eukaryotes.Totally,we got 50,433 protein kinases.Finally,we built an online database EKPD,and provided the detail annotation for each protein.Using the same approach,the protein phosphatases were also identified in eukaryotes and further added to the database.So far,EKPD serves as the most comprehensive protein kinases and protein phosphatases database with most species and most detailed annotation.Based on the knowledge of classification and annotation of protein kinases,the function study on protein kinases is more important to elucide the mechanism of phosphorylation in disease.Recent studies demonstrated that dysregulation of autophagy may play a central role in the pathogenesis of neurodegenerative disorders.In this work,quantitative phosphoproteome profiling together with computational analysis were performed to delineate the distinct phosphorylation signaling networks regulated by two natural neuroprotective autophagy enhancers,Cory and Cory B.We developed a novel network-based algorithm of in silico Kinome Activity Profiling(iKAP)to computationally infer potential important regulators,namely protein kinases,from phosphorylation networks.Using this algorithm,we observed distinct kinases differentially regulated by two compounds.We predicted and validated RPS6KB1,to be down-regulated by Cory but not Cory B.We also discovered two kinases,MAP2K2 and PLK1,to be up-regulated by Cory,whereas the knock-down of MAP2K2 and PLK1 significantly inhibited Cory-induced autophagy.Furthermore,Cory promotes the clearance of ?-amyloid precursor protein and?-synuclein by enhancing autophagy,which was dramatically diminished by the inhibition of MAP2K2 and PLK1.Taken together,our study developed a powerful method for the identification of important kinases from phosphoproteomes,and also revealed the important role of MAP2K2 and PLK1 in neuronal autophagy.Since most complex diseases are due to the actions of multiple genes/proteins,the identification of potential important phospho-signatures within phosphoproteomics-based networks is more efficient than studying the single phosphorylated substrate or site.Genetic variations are always associated with diseases,however,the mechanisms how genetic variations "drive" the diseases were still unknown.In this work,we performed an analysis on nsSNPs that influence the phosphorylation(PhosSNPs).Using iGPS algorithm,we computationally detected 9,606 potential PhosSNPs and reconfigured the PhosSNPs influenced kinase-substrate phosphorylation network.Further analyses demonstrated that the proteins in the network are heavily associated in various signaling and cancer pathways,while cancer genes and drug targets are significantly enriched.We re-constructed four population-specific kinase-substrate networks and found that several inherited disease or cancer genes were differentially regulated by PhosSNPs.Moreover,we observed that a considerable number of phosphorylated proteins,especially those over-representing cancer genes,contain both PhosSNPs and phosCMs.Furthermore,it was observed that PhosSNPs were significantly enriched in amplification genes identified from breast cancers and tyrosine kinase circuits of lung cancers.Taken together,these results revealed the potential impact of PhosSNPs on phosphorylation.They may influence the susceptibility of diseases through reconfigure of phosphorylation network and drive the disease development together with other variations.With the increase of phosphoproteomics,the accurate prediction the kinase-substrate relations come up to be the essential to study the phosphorylation and disease.In this work,based on the previous study,we improved the kinase-specific phosphorylation sites prediction tool GPS 3.0.We collected 7,426 known phosphorylation sites,and imported the "motif length selection" and "weight mutation" to the model.The GPS 3.0 has been extended to be able to predict phosphorylation sites for more kinases,and included the species-specific and dual-specific kinase prediction.Through comparison with other tools,GPS 3.0 exhibited more excellent performance.Finally,we provided the local and online version of GPS 3.0.In this work,we have performed a systematic analysis on relationship between protein phosphorylation and diseases based on the kinase-substrate phosphorylation network modeling.We demonstrate that the genetic variation could influence diseases through phosphorylation network,and also revealed that the kinase activity analysis could serve as a useful approach for the study of drug target and clinical research.These results provided a new insight for further elucidation of the regulation mechanism of phosphorylation as well as the new leads for personalized medicine.