Folding Intermediates, Heterogeneous Native Ensembles and Protein Function.

Single domain proteins fold via diverse mechanisms emphasizing the intricate relationship between energetics and structure, which is a direct consequence of functional constraints and demands imposed at the level of sequence. On the other hand, elucidating the interplay between folding mechanisms and function is challenging in large proteins, given the inherent shortcomings in identifying metastable states experimentally and the sampling limitations associated with computational methods. Here, we show that free energy profiles and surfaces of large systems (>150 residues), as predicted by a statistical mechanical model, display a wide array of folding mechanisms with ubiquitous folding intermediates and heterogeneous native ensembles. Importantly, residues around the ligand binding or enzyme active site display a larger tendency to partially unfold and this manifests as intermediates or excited states along the folding coordinate in ligand binding domains, transcription repressors, and representative enzymes from all the six classes, including the SARS-CoV-2 receptor binding domain (RBD) of the spike protein and the protease Mpro. It thus appears that it is relatively easier to distill the imprints of function on the folding landscape of larger proteins as opposed to smaller systems. We discuss how an understanding of energetic-entropic features in ordered proteins can pinpoint specific avenues through which folding mechanisms, populations of partially structured states and function can be engineered.

Structural dynamics of single SARS-CoV-2 pseudoknot molecules reveal topologically distinct conformers.

The RNA pseudoknot that stimulates programmed ribosomal frameshifting in SARS-CoV-2 is a possible drug target. To understand how it responds to mechanical tension applied by ribosomes, thought to play a key role during frameshifting, we probe its structural dynamics using optical tweezers. We find that it forms multiple structures: two pseudoknotted conformers with different stability and barriers, and alternative stem-loop structures. The pseudoknotted conformers have distinct topologies, one threading the 5' end through a 3-helix junction to create a knot-like fold, the other with unthreaded 5' end, consistent with structures observed via cryo-EM and simulations. Refolding of the pseudoknotted conformers starts with stem 1, followed by stem 3 and lastly stem 2; Mg2+ ions are not required, but increase pseudoknot mechanical rigidity and favor formation of the knot-like conformer. These results resolve the SARS-CoV-2 frameshift signal folding mechanism and highlight its conformational heterogeneity, with important implications for structure-based drug-discovery efforts.