Advanced MD Simulation Methods Uncover Mechanisms of SH3 Domain Functions in Small GTPase Signaling

Abstract

The protein complex comprising the SH3 domain and DLC1 proteins plays a vital role in various cellular processes and diseases, including cancer. Essential dynamics for the stability of this complex, which cannot be elucidated by static X-ray crystal structures, have significant implications for understanding cellular physiology and critical diseases. We thoroughly investigated this complex using advanced molecular dynamics, Adaptively Biased Force MD (ABF-MD), and conventional MD (cMD) simulation methods. Radial distribution function (RDF) calculations demonstrate that the interaction between the two proteins is highly specific, as all mutations exhibit a single peak, indicating no additional interacting sites. The probabilities of two key interactions, Glu298-Arg1114 and Lys292-Leu1239, were observed to increase in cancer-related mutations but not in other mutations known to disrupt complex formation. Using a Markov State Model (MSM), we identified a key intermediate in the wild type that was absent in other variants. Correlation analysis of deviations in distances among key interacting residues showed values greater than 0.95, indicating cooperativity among interacting residues. cMD simulations also revealed increased distance values between interacting residues in complex-disrupting mutations, but not in cancer-related mutations. Principal component analysis (PCA) further revealed significant conformational changes, indicating important distinct conformations potentially involved in complex formation. Specifically, the loop region between residues 1236-1261 exhibits distinct conformations upon mutations among variants. This distinct conformation, particularly in the L1267D mutation, leads to the displacement of the SH3 domain from the binding site, which may contribute to complex destabilization. Additionally, PCA analysis suggests that complex-disrupting mutations significantly increase the ability of the loop region to explore different conformations compared to the wild type. In contrast, the cancer-related mutation, V1227M, does not significantly affect protein flexibility or its capacity to stay in a stable conformation. The binding energy analysis reveals that the wild-type DLC1 complex has moderate stability (-8.87 +/- 1.31 kcal/mol), and the V1227M variant shows the most stable binding (-6.89 +/- 1.04 kcal/mol) among other mutants. In contrast, L1267D, R1114A, and R1114E variants exhibit weaker binding affinities (-5.89 +/- 1.01, -3.18 +/- 1.04, and - 0.58 +/- 1.11 kcal/mol, respectively), indicating reduced complex stability.