Shedding light on Gaussia Luciferase conformational space using multi-conformation enhanced sampling simulation

Abstract

Gaussia luciferase (GLuc) has broad biotechnological applications owing to its bioluminescence, yet its catalytic mechanism remains poorly understood and rational engineering efforts are hindered by the absence of a structural model. This gap stems from GLuc's highly dynamic nature, which allows it to populate a wide range of conformations rather than adopt a single, well-defined fold. Aiming to bridge this gap, in this work we employ structure predictors and conformational sampling techniques to prod GLuc's structure. Our results indicate that structure predictors such as AlphaFold, as well as ensemble predictors such as BioEmu, fail to adequately sample this conformational space, and even extended molecular dynamics simulations do not yield converged conformational ensembles. However, enhanced sampling via PT-WTE provides a more satisfactory description of GLu's conformational landscape, allowing us to cluster distinct conformational basins, identify loosely defined substrate-binding pockets, and characterize the conformational changes induced by substrate binding. These insights advance our understanding of GLuc's catalytic mechanism and inform future efforts to engineer improved variants.