Sosnick Lab

Experimental and computational studies of protein folding

General rules of protein folding: We are interested in understanding the general rules that governs how protein folds. Specifically, we have characterized the folding transition states (TS) of the B domain of protein A (BdpA) using ψ-analysis. We identified a TS that adopts 60-70% of the native state topology (as measured by the relative contact order (RCO) of Plaxco et al.). Taken with previous work on Ub and ctAcp, we propose that small proteins fold through TSs that adopt 70% of the native state topology. This "70% Rule" places strong limitations on possible TSEs and can therefore be used to construct initial TS models. These TS models can then be evaluated experimentally to see how well this "70% Rule" applies to other proteins that satisfy the ln k_f - RCO correlation.

Structural Origins of fractional ψ-values: ψ's of 1 and 0 unambiguously reflect native-like and unfolded-like structure in the TS. Fractional ψ may arise from TS heterogeneity (ctAcp) or from site distortion in the TS (BdpA). We sought to elucidate the structural origins of both phenomena using all-atom Langevin dynamics simulations of TS models for Ubiquitin. Of the six sites with fractional ψ^exp, we find that the sites close to the unity sites are constrained to adopt distorted site geometries while the more topologically distal sites sample unfolded- and native-like configurations. Regardless, the unity ψ-values alone are sufficient to generate a well-defined TSE with a highly native-like topology for Ub.

Implications for calculating φ:We have characterized a TS independent of mutational φ-analysis and are therefore in a position to investigate the ability of simulations to accurately predict experimental φ. One should not necessarily expect agreement between φ^exp and φ^sim because they are probing different phenomena. Experimental φ report the energetic effect of sidechain deletion while the simulated values count native contacts. It may be that φ^exp reflect the ability of the TS to accommodate the new side chain rather than the presence or absence of structure per se.

Current projects:

• Computational modeling of the unfolded state ensemble of Ubiquitin
• Predicting hydrogen exchange protection factors
• LD simulations of Aβ_16-22 monomers and dimers