Chatzigoulas, A. | bioRxiv Science

bioRxiv2022

Evaluation of K-Ras4B dimer interfaces and the role of Raf effectors

K-Ras4B is one the most frequently mutated proteins in cancer, yet mechanistic details of its activation such as its homodimerization on the membrane remain elusive. The structural determinants of K-Ras4B homodimerization have been debated with different conformations being proposed in the literature. Here, we perform microsecond all-atom Molecular Dynamics (MD) simulations on the K-Ras4B monomer in solution, the K-Ras4B monomer on the membrane, and two experimentally-based K-Ras4B dimer models of the 4-5 interface to investigate the stability of these structures bound to GTP on a model cell membrane. We then evaluate the complexes for their propensity to form stable dimers on the plasma membrane in the presence and absence of Raf[RBD-CRD] effectors. We find that Raf[RBD-CRD] effectors enhance dimer stability, suggesting that the presence of effectors is necessary for K-Ras4B dimers stabilization on the cell membrane. Moreover, we observe, for the first time, a dynamic water channel at the K-Ras4B dimer interface, and identify putative allosteric connections in the K-Ras4B dimer interface. To discover novel K-Ras4B interfaces, we perform coarse-grained MD simulations in two dissociated K-Ras4B monomers on the membrane, which reveal that the dominant dimer interface is the 4-5 interface. Finally, a druggability analysis is performed in the different K-Ras4B structures in the monomeric states. Strikingly, all known binding pockets of K-Ras4B are identified only in the structure that is membrane-bound, but not in the solution structure. Based on these results, we propose that modulating the protein-membrane interactions can be an alternative strategy for inhibiting K-Ras4B signaling.

molecular biology↗

bioRxiv2021

Predicting protein-membrane interfaces of peripheral membrane proteins using ensemble machine learning

Abnormal protein-membrane attachment is involved in deregulated cellular pathways and in disease. Therefore, the possibility to modulate protein-membrane interactions represents a new promising therapeutic strategy for peripheral membrane proteins that have been considered so far undruggable. A major obstacle in this drug design strategy is that the membrane binding domains of peripheral membrane proteins are usually not known. The development of fast and efficient algorithms predicting the protein-membrane interface would shed light into the accessibility of membrane-protein interfaces by drug-like molecules. Herein, we describe an ensemble machine learning methodology and algorithm for predicting membrane-penetrating amino acids. We utilize available experimental data in the literature for training 21 machine learning classifiers and a voting classifier. Evaluation of the ensemble classifier accuracy produced a macro-averaged F1 score = 0.92 and an MCC = 0.84 for predicting correctly membrane-penetrating amino acids on unknown proteins of an independent test set. The python code for predicting protein-membrane interfaces of peripheral membrane proteins is available at https://github.com/zoecournia/DREAMM.

bioinformatics↗