Modeling, Design and Engineering of Allosteric Regulatory Mechanisms Mediated by Molecular Chaperones in Signal Transduction Networks: An Integrative Platform of Computational Systems Biology, Network Science and Machine Learning Approaches
The allosteric interactions and regulation of molecular chaperones and protein kinases allow for molecular communication and event coupling in signal transduction networks. The overarching goal of understanding molecular principles underlying differentiation of protein kinase clients and chaperone-based modulation of kinase activity is fundamental to understanding activity of many tumor-inducing signaling proteins. The synergistic roles of the Hsp90-Cdc37 chaperone machinery and protein kinases in biology and disease have stimulated extensive structural and functional studies of regulatory mechanisms underlying the Hsp90-kinase interactions. Allosteric interactions of the Hsp90 with cochaperones and protein kinase clients can determine regulatory mechanisms and cellular functions of many signaling proteins and cascades. We report the results of integrative systems biology studies of the Hsp90 chaperone and protein kinases with an atomic level analysis of the communication pathways regulating conformational equilibrium of theses protein systems in signaling networks. Biophysical modeling of allosteric regulation in the protein kinases has offered additional insights into organizing principles of kinase activation by molecular chaperones that may be orchestrated by a cross-talk between key regulatory regions.
The results of biophysical and computational systems biology analyses combined with proteomics experiments have been integrated into a graph-based network model of allosteric regulation. Among our primary findings is the emerging evidence that a small number of functional motifs may be utilized by the chaperone and protein kinases to act collectively as central regulators of the intermolecular communications, ATP hydrolysis, and protein client binding in signaling networks. Integration of computational systems biology and machine learning analysis of the Hsp90 interactions with oncogenic kinase mutants is then used to construct models of allosteric regulation of oncogenic proteins by molecular chaperones in signaling cascades. Network modelling and percolation analysis approaches were used to emulate thermal unfolding and characterize conformational landscapes of a wide range of protein kinases, revealing that chaperone dependency of protein kinase clients may be linked with the elevated conformational mobility of their inactive states induced by dynamic and energetic polarization of kinase lobes. By showcasing a family of cyclin-dependent (CDK) kinases that display a broad repertoire of chaperone dependencies, we discovered that unique functional dynamics signatures and chaperone addiction of CDK4 and CDK7 client proteins can explain divergences in their regulatory mechanisms that require a confluence of events, including formation of the inhibitory ternary complex, substrate recruitment and activation loop phosphorylation.
Based on these findings, we developed a computational synthetic biology framework for design and re-engineering signal transduction networks and pathways that involve cross-talk between molecular chaperones and protein kinase clients. Network-defined modularity of interacting components and cross-talk in signal transduction networks are quantified through molecular mechanism of allosteric regulation. The increasingly growing fraction of cancer driver mutations emerging from sequencing studies of protein kinase genes appeared to be inactivating or kinase-dead leading to the loss of function. Our studies have revealed how constitutively activating and kinase-dead mutations could play context-dependent opposing roles in cancer and may be simultaneously present in a variety of oncogenic kinases that are regulated by their interactions with the Hsp90 chaperone. We have also analyzed mechanisms by which kinase drugs and allosteric Hsp90 modulators that may act synergistically and exert their pharmacological effect by depriving the client kinase of access to the molecular chaperone. Our study offers a systems-based perspective on drug design and re-engineering of signaling networks by unravelling relationships between protein kinase networks with molecular chaperones and binding specificity of targeted kinase drugs.