Reconstructing the transport cycle in the sugar porter superfamily using coevolution-powered machine learning

Abstract

Sugar porters represent the largest group of secondary-active transporters. Some members, such as the glucose (GLUT) transporters, are well-known for their role in maintaining blood glucose homeostasis in mammals, with their expression upregulated in many types of cancers. Because only a few sugar porter structures have been determined, mechanistic models have been constructed by piecing together structural states of distantly-related proteins. Current GLUT transport models are predominantly descriptive, and oversimplified. Here, we have combined coevolution analysis and comparative modeling, to predict structures of the entire sugar porter superfamily in each state of the transport cycle. We have analysed the state-specific contacts inferred from coevolving residue pairs, and shown how this information can be used to rapidly generate free-energy landscapes consistent with experimental estimates, as illustrated here for the mammalian fructose transporter GLUT5. By comparing many different sugar porter models and scrutinizing their sequence, we have been able to define the molecular determinants of the transport cycle, which are conserved throughout the sugar porter superfamily. We have also been able to highlight differences leading to the emergence of proton-coupling, validating and extending the previously proposed latch mechanism. Our computational approach is transferable to any transporter, and to other protein families in general.

Determinants of sugar-induced influx in the mammalian fructose transporter GLUT5

“…S5). Indeed, independent co-evolution analysis corroborates that Y298 and H383 residues are forming an interaction in the occluded state of GLUT5 40 . Moreover, in yeast-based forward-evolution screens of hexose transporters, the equivalent residue to Y298 in TM7b was the only singe-residue-variant uncovered with a shifted sugar preference from D-glucose towards D-xylose 41,42 .…”

Substrate Recognition Properties from an Intermediate Structural State of the UreA Transporter

“…Thus, multiple states on systems that "flip" between conformations to perform their functions are precluded. Many efforts have been made to predict multistate structures that are not influenced by one specific state [33][34][35][36]41,42]. Del Alamo et al ( 2022) [21] attempted to expand the use of AI-based AF2 beyond the structural prediction of an already determined biological state for a structure, especially for membrane transporters and receptors, since some previous AF2 predictions were observed to be in intermediate state conformations [33,34].…”

Best Practices of Using AI-Based Models in Crystallography and Their Impact in Structural Biology

“…As mentioned above, AF2 and RF allow the prediction of contacts from sequence alignments. , The reliability of this prediction is estimated in the predicted aligned error (PAE) which can be used to predict domains, or to help protein–protein docking. , However, the prediction of contacts is further complicated by conformational changes that make the relationship between sequence evolution and contacts partially ambiguous. This issue triggered the development of strategies to explore the conformational landscape with AF2 and RF. − …”