The origins of asymmetry in the folding transition states of protein L and protein G

Abstract

Topology has been shown to be an important determinant of many features of protein folding; however, the delineation of sequence effects on folding remains obscure. Furthermore, differentiation between the two influences proves difficult due to their intimate relationship. To investigate the effect of sequence in the absence of significant topological differences, we examined the folding mechanisms of segment B1 peptostreptococcal protein L and segment B1 of streptococcal protein G. These proteins share the same highly symmetrical topology. Despite this symmetry, neither protein folds through a symmetrical transition state. We analyzed the origins of this difference using theoretical models. We found that the strength of the interactions present in the N-terminal hairpin of protein L causes this hairpin to form ahead of the C-terminal hairpin. The difference in chain entropy associated with the formation of the hairpins of protein G proves sufficient to beget initiation of folding at the shorter C-terminal hairpin. Our findings suggest that the mechanism of folding may be understood by examination of the free energy associated with the formation of partially folded microstates.

Fig. 1.

A comparison of the structures of (a) protein L and (b) protein G. The PDB accession codes are 2PTL and 1PGB, respectively. The figures were generated using MOLSCRIPT (Kraulis 1991).

Fig. 2.

Thermodynamic functions used in the initial characterization of protein L (solid line) and protein G (dotted line) folding. (a) The temperature dependence of the heat capacity, C_v. (b) The temperature dependence of the fraction of native contacts formed, q. (c) The free energy, F, as a function of q at the transition temperature.

Fig. 3.

The free energy at the transition temperature as a function of several progress variables for protein L (a,c,e) and protein G (b,d,f). (a,b) The free energy as a function of the fraction of native contacts formed in the N-terminal hairpin (q_N-terminal) and the fraction of total native contacts formed (q). (c,d) The free energy as a function of the fraction of native contacts formed in the C-terminal hairpin (q_C-terminal) and the fraction of total native contacts formed (q). (e,f) The free energy as a function of the fraction of helical native contacts formed (q_helix) and the fraction of total native contacts formed (q). The free energy difference between adjacent contour lines corresponds to k_BT_f.

Fig. 4.

The free energy at the transition temperature as a function of the fraction of C-terminal hairpin native contacts formed (q_C-terminal) and the fraction of N-terminal hairpin native contacts formed (q_N-terminal), for (a) protein L and (b) protein G. The free energy difference between adjacent contour lines corresponds to k_BT_f.

Fig. 5.

The φ-values predicted by the FOLD-X algorithm. The C-terminal hairpin is found to fold ahead of the N-terminal hairpin for structures of both protein L (2ptl, 1HZ6) and protein G (1pgb), but the results are ambiguous for the protein G structure 1igd. The φ values were scaled by 100.