Mutagenic effects on protein folding and stability

Abstract

Knowing how sequence information dictates the formation of protein structure is critical for accurate prediction of structure, for de novo protein design, and for understanding protein folding and misfolding. Based on extensive studies in peptide and protein systems, the pattern of polar and nonpolar amino acids, complementary packing of side chains in the protein core, electrostatic interactions, turns and helix-capping motifs, and secondary-structure propensity have been shown to be important for folding and stability to varying degrees. The pattern of polar and nonpolar residues (binary pattern) in a sequence appears to be most critical for determining the gross three-dimensional fold, whereas other interactions are responsible for the details of protein structure and stability. Residues 9-14 of wild-type P22 Arc repressor form a two-stranded 5-sheet and have the binary pattern of an amphipathic -sheet. Switching two residues in this region (NL11, LN12) results in the formation of two right-handed 310-helices and changes the binary pattern to that of an amphipathic helix. Arc NL11 has an ambiguous binary pattern and is in dynamic equilibrium between the sheet and helical structures. Characterization of mutants in which position 11 is replaced with different hydrophobic residues shows that binary pattern is primarily responsible for determining the structure of this region of Arc. In the context of an ambiguous binary pattern, however, the detailed chemical properties of the position- 1 side chain dictate which of the two competing folds is preferred.

(cont.) Although surface residues are typically unimportant for protein structure and stability, there are important exceptions to this rule. In Arc, Ser32 is solvent exposed and serves as the Ncap residue of helix B. Characterization of a library of position-32 variants shows that this position is critical both for determining the stability and activity of Arc. Except for cysteine, the stability of these variants correlates with the frequency at which the corresponding residues are found at Ncap positions in the protein database. Because the position-32 side chain is close to the DNA backbone in the repressor-operator complex, only variants with small, uncharged residues at this position show significant activity in vivo or in vitro.