Achuthan, S. | Search bioRxiv

bioRxiv2020

Phenotypic switching and adaptive strategies of cancer cells in response to stress: insights from live cell imaging and mathematical modeling

A major challenge in cancer therapy is drug resistance, which is typically attributed to acquired mutations and tumor heterogeneity. However, emerging evidence suggests that dynamic cellular interactions and group behavior also contribute to drug resistance, although, the details of such mechanisms are poorly understood. Here, by combining real time cellular growth data with mathematical modeling, we showed that the cisplatin-sensitive and tolerant lung cancer cells when co-cultured in cisplatin-free and cisplatin-treated environments, exhibit drastically different group strategies in response to environmental changes. While tolerant cells exhibited a persister-like behaviour and were attenuated by sensitive cells, sensitive cells learned to evade chemotherapy from tolerant cells when co-cultured. Further, tolerant cells could switch phenotypes to become sensitive, although high cisplatin concentrations suppressed this switching. Finally, switching cisplatin administration from continuous to intermittent suppressed the emergence of tolerant cells, suggesting that intermittent rather than continuous chemotherapy may result in better outcomes in lung cancer.

cancer biology↗

bioRxiv2019

Machine Learning for Prioritization of Thermostabilizing Mutations for G-protein Coupled Receptors

Although the three-dimensional structures of G-protein-coupled receptors (GPCRs), the largest superfamily of drug targets, have enabled structure-based drug design, there are no structures available for 87% of GPCRs. This is due to the stiff challenge in purifying the inherently flexible GPCRs. Identifying thermostabilized mutant GPCRs via systematic alanine scanning mutations has been a successful strategy in stabilizing GPCRs, but it remains a daunting task for each GPCR. We developed a computational method that combines sequence, structure and dynamics based molecular properties of GPCRs that recapitulate GPCR stability, with four different machine learning methods to predict thermostable mutations ahead of experiments. This method has been trained on thermostability data for 1231 mutants, the largest publicly available dataset. A blind prediction for thermostable mutations of the Complement factor C5a Receptor retrieved 36% of the thermostable mutants in the top 50 prioritized mutants compared to 3% in the first 50 attempts using systematic alanine scanning.\n\nStatement Of SignifiganceG-protein-coupled receptors (GPCRs), the largest superfamily of membrane proteins play a vital role in cellular physiology and are targets to blockbuster drugs. Hence it is imperative to solve the three dimensional structures of GPCRs in various conformational states with different types of ligands bound. To reduce the experimental burden in identifying thermostable GPCR mutants, we report a computational framework using machine learning algorithms trained on thermostability data for 1231 mutants and features calculated from analysis of GPCR sequences, structure and dynamics to predict thermostable mutations ahead of experiments. This work represents a significant advancement in the development, validation and testing of a computational framework that can be extended to other class A GPCRs and helical membrane proteins.

biophysics↗