A novel computer simulation method for simulating the multiscale transduction dynamics of signal proteins

• Published in: The Journal of chemical physics 2012-03, Vol.136 (12), p.124112-124112-14
• Main Authors: Peter, Emanuel, Dick, Bernhard, Baeurle, Stephan A.
• Format: Article
• Language: English
• Subjects: Algorithms; Avena - chemistry; Computer Simulation; Humans; Models, Theoretical; Molecular Dynamics Simulation; Protein Conformation; Proteins - chemistry; Signal Transduction
• ISSN: 0021-9606; 1089-7690
• DOI: 10.1063/1.3697370

Signal proteins are able to adapt their response to a change in the environment, governing in this way a broad variety of important cellular processes in living systems. While conventional molecular-dynamics (MD) techniques can be used to explore the early signaling pathway of these protein systems at atomistic resolution, the high computational costs limit their usefulness for the elucidation of the multiscale transduction dynamics of most signaling processes, occurring on experimental timescales. To cope with the problem, we present in this paper a novel multiscale-modeling method, based on a combination of the kinetic Monte-Carlo- and MD-technique, and demonstrate its suitability for investigating the signaling behavior of the photoswitch light-oxygen-voltage-2-Jα domain from Avena Sativa (AsLOV2-Jα) and an AsLOV2-Jα-regulated photoactivable Rac1-GTPase (PA-Rac1), recently employed to control the motility of cancer cells through light stimulus. More specifically, we show that their signaling pathways begin with a residual re-arrangement and subsequent H-bond formation of amino acids near to the flavin-mononucleotide chromophore, causing a coupling between β-strands and subsequent detachment of a peripheral α-helix from the AsLOV2-domain. In the case of the PA-Rac1 system we find that this latter process induces the release of the AsLOV2-inhibitor from the switchII-activation site of the GTPase, enabling signal activation through effector-protein binding. These applications demonstrate that our approach reliably reproduces the signaling pathways of complex signal proteins, ranging from nanoseconds up to seconds at affordable computational costs.