Klimov, D. K. ; Thirumalai, D. Stiffness of the distal loop restricts the structural heterogeneity of the transition state ensemble in SH3 domains. J Mol Biol 317, 721-37.
AbstractProtein engineering experiments and Phi(F)-value analysis of SH3 domains reveal that their transition state ensemble (TSE) is conformationally restricted, i.e. the fluctuations in the transition state (TS) structures are small. In the TS of src SH3 and alpha-spectrin SH3 the distal loop and the associated hairpin are fully structured, while the rest of the protein is relatively disordered. If native structure predominantly determines the folding mechanism, the findings for SH3 folds raise the question: What are the features of the native topology that determine the nature of the TSE? We propose that the presence of stiff loops in the native state that connect local structural elements (such as the distal hairpin in SH3 domains) conformationally restricts TSE. We validate this hypothesis using the simulations of a "control" system (16 residue beta-hairpin forming C-terminal fragment of the GBl protein) and its variants. In these fragments the role of bending rigidity in determining the nature of the TSE can be directly examined without complications arising from interactions with the rest of the protein. The TSE structures in the beta-hairpins are determined computationally using cluster analysis and limited Phi(F)-value analysis. Both techniques prove that the conformational heterogeneity decreases as the bending rigidity of the loop increases. To extend this finding to SH3 domains a measure of bending rigidity based on loop curvature, which utilizes native structures in the Protein Data Bank (PDB), is introduced. Using this measure we show that, with few exceptions, the ordering of stiffness of the distal, n-src, and RT loops in the 29 PDB structures of SH3 domains is conserved. Combining the simulation results for beta-hairpins and the analysis of PDB structures for SH3 domains, we propose that the stiff distal loop restricts the conformational fluctuations in the TSE. We also predict that constraining the distal loop to be preformed in the denatured ensemble should not alter the nature of TSE. On the other hand, if the amino and carboxy terminals are cross-linked to form a circular polypeptide chain, the pathways and TSs are altered. These contrasting scenarios are illustrated using simulations of cross-linked WT beta-hairpin fragments. Computations of bending rigidities for immunoglobulin-like domain proteins reveal no clear separation in the stiffness of their loops. In the beta-sandwich proteins, which have large fractions of non-local native contacts, the nature of the TSE cannot be apparently determined using purely local structural characteristics. Nevertheless, the measure of loop stiffness still provides qualitative predictions of the ordered regions in the TSE of Ig27 and TenFn3.
Klimov, D. K. ; Thirumalai, D. Is there a unique melting temperature for two-state proteins?.
J Comput Chem 23, 161-5.
AbstractThermal unfolding (or folding) in many proteins occurs in an apparent two-state manner, suggesting that only two states, unfolded and folded, are populated. At the melting temperature, Tm, the two states coexist. Using lattice models with side chains we show that individual residues become structured at temperatures that deviate from Tm, which implies that partially folded conformations make substantial contribution to thermodynamic properties of two-state proteins. We also find that the folding cooperativity for a given residue is linked to its accessible surface area. These results are consistent with the experiments on GCN4-like zipper peptide, which showed that local melting temperatures differ from Tm. Analysis of thermal unfolding of six proteins shows that deltaT/Tm approximately N(-1), where deltaT is the transition width and N is the number of residues. This scaling allows us to conclude that, when corrected for finite size effects, folding cooperativity can be captured using coarse grained models.
Dima, R. I. ; Thirumalai, D. Exploring the propensities of helices in PrP(C) to form beta sheet using NMR structures and sequence alignments. Biophys J 83, 1268-80.
AbstractNeurodegenerative diseases induced by transmissible spongiform encephalopathies are associated with prions. The most spectacular event in the formation of the infectious scrapie form, referred to as PrP(Sc), is the conformational change from the predominantly alpha-helical conformation of PrP(C) to the PrP(Sc) state that is rich in beta-sheet content. Using sequence alignments and structural analysis of the available nuclear magnetic resonance structures of PrP(C), we explore the propensities of helices in PrP(C) to be in a beta-strand conformation. Comparison of a number of structural characteristics (such as solvent accessible area, distribution of (Phi, Psi) angles, mismatches in hydrogen bonds, nature of residues in local and nonlocal contacts, distribution of regular densities of amino acids, clustering of hydrophobic and hydrophilic residues in helices) between PrP(C) structures and a databank of "normal" proteins shows that the most unusual features are found in helix 2 (H2) (residues 172-194) followed by helix 1 (H1) (residues 144-153). In particular, the C-terminal residues in H2 are frustrated in their helical state. The databank of normal proteins consists of 58 helical proteins, 36 alpha+beta proteins, and 31 beta-sheet proteins. Our conclusions are also substantiated by gapless threading calculations that show that the normalized Z-scores of prion proteins are similar to those of other alpha+beta proteins with low helical content. Application of the recently introduced notion of discordance, namely, incompatibility of the predicted and observed secondary structures, also points to the frustration of H2 not only in the wild type but also in mutants of human PrP(C). This suggests that the instability of PrP(C) proteins may play a role in their being susceptible to the profound conformational change. Our analysis shows that, in addition to the previously proposed role for the segment (90-120) and possibly H1, the C-terminus of H2 and possibly N-terminus may play a role in the alpha-->beta transition. An implication of our results is that the ease of polymerization depends on the unfolding rate of the monomer. Sequence alignments show that helices in avian prion proteins (chicken, duck, crane) are better accommodated in a helical state, which might explain the absence of PrP(Sc) formation over finite time scales in these species. From this analysis, we predict that correlated mutations that reduce the frustration in the second half of helix 2 in mammalian prion proteins could inhibit the formation of PrP(Sc).
exploring-the-propensities-of-helices-in-prpc-to-form-beta-sheet-using-nmr-structures-and-sequence-alignments.pdf