A High Throughput Method for Exploring the Sequence Space of Polypeptides That Exhibit Thermoresponsive Phase Behavior

Abstract

Intrinsically disordered protein polymers (IDPPs) are built up of repeating units and they can demonstrate two types of thermoresponsive phase behavior. Systems characterized by lower critical solution temperatures (LCST) undergo phase separation above the LCST whereas systems characterized by upper critical solution temperatures (UCST) undergo phase separation below the UCST. The amino acid composition and sequence of the repeat as well as the number of repeats will determine the thermoresponsive phase behavior of IDPPs. There is congruence between thermoresponsive coil-to-globule transitions and phase behavior. The theta temperature above or below which the IDPPs transition from coils to globules serve as useful proxies for the LCST/UCST values. This implies that one can design sequences with desired values for the theta temperature and either increasing or decreasing radii of gyration (Rg) as a function of increasing temperature (T). However, the vastness of sequence space makes it impossible to combine conformational sampling with sequence design algorithms. We recently discovered that the Monte Carlo simulations performed in the intrinsic solvation (IS) limit version of the temperature-dependent ABSINTH model, a limit that includes all terms excepting the descreened electrostatic interactions, affords highly efficient estimates of the temperature-dependence of the sequence-specific free energies of solvation. For sequences that demonstrate LCST behavior, the Rg vs. T plots have negative slopes whereas sequences with UCST behavior have Rg vs. T plots with near zero or positive slopes. Taking advantage of the efficiency of IS limit simulations and the robust classifications they afford, we have developed a high-throughput sequence design engine based on evolutionary algorithms to design novel sequences with LCST vs. UCST phase behavior. We are closing the loop with experiments to expand the libraries of known thermoresponsive IDPPs for biomedical applications.