Systems analysis of molecular networks based on biochemical computational simulations

点击翻译按钮获取中文摘要

Many biological phenomena are regulated by their intrinsic biological networks. These networks contain both linear and non-linear relationship, which make difficult to understand the whole networks at the systems level. In this study, to overcome these difficulties we tried to utilize two interdependent approaches, prediction by in silico computational simulation and validation by in vivo experiments. We focused on two phenomena, 1) molecular network in myosin light chain (MLC) phosphorylation and 2) quantitative analysis of cerebellar long-term depression (LTD).1)Dynamics of throbin-induced MLC phosphorylation in endothelium is composed of initial and sustained phases. By the literature-based simulation model we succeeded in the reproduction of only the initial but not the sustained phase. The model pinpointed us the missing pathway may underlie downstream of Rho-kinase. We experimentally validated the sustained activation of Rho-kinase by the phosphorylation of MYPT1 (myosin phosphatase target subunit 1), a substrate of Rho-kinase, which exhibited hysteresis.2)We developed a computational model of postsynaptic Ca^<2+> increase in cerebellar LTD. The model revealed that Ca^<2+> increase is induced by a regenerative cycle by Ca^<2+>-induced Ca^<2+> release (CICR) via IP3 receptors, and that the latency of IP3 generation by PF stimulation produces timing window of spike-timing dependent LTD in the cerebellum. Moreover, we developed a unified model of the Ca^<2+> increase and downstream signaling cascades. The unified model raised the possibility that postsynaptic spines of the cerebellum tune their own synaptic strength for adequate Ca^<2+> increase for robust LTD.