Zheng-Yu Peng

Manipulation of protein solubility is important for many aspects of protein design and engineering. Previously, we designed a series of consensus ankyrin repeat proteins containing one, two, three and four identical repeats (1ANK, 2ANK, 3ANK and 4ANK). These proteins, particularly 4ANK, are intended for use as a universal scaffold on which specific binding sites can be constructed. Despite being well folded and extremely stable, 4ANK is soluble only under acidic conditions. Designing interactions with naturally occurring proteins requires the designed protein to be soluble at physiological pH. Substitution of six leucines with arginine on exposed hydrophobic patches on the surface of 4ANK resulted in increased solubility over a large pH range. Study of the pH dependence of stability demonstrated that 4ANK is one of the most stable ankyrin repeat proteins known. In addition, analogous leucine to arginine substitutions on the surface of 2ANK allowed the partially folded protein to assume a fully folded conformation. Our studies indicate that replacement of surface‐exposed hydrophobic residues with positively charged residues can significantly improve protein solubility at physiological pH.

The ankyrin repeat is one of the most common, modular, protein–protein interaction motifs in nature. To understand the structural determinants of this family of proteins and extract the consensus information that defines the architecture of this motif, we have designed a series of idealized ankyrin repeat proteins containing one, two, three, or four repeats by using statistical analysis of ≈4,000 ankyrin repeat sequences from the PFAM database. Biophysical and x-ray crystallographic studies of the three and four repeat constructs (3ANK and 4ANK) to 1.26 and 1.5 Å resolution, respectively, demonstrate that these proteins are well-folded, monomeric, display high thermostability, and adopt a very regular, tightly packed ankyrin repeat fold. Mapping the degree of amino acid conservation at each position on the 4ANK structure shows that most nonconserved residues are clustered on the surface of the molecule that has been designated as the binding site in naturally occurring ankyrin repeat proteins. Thus, the consensus amino acid sequence contains all information required to define the ankyrin repeat fold. Our results suggest that statistical analysis and the consensus sequence approach can be used as an effective method to design proteins with complex topologies. These generic ankyrin repeat proteins can serve as prototypes for dissecting the rules of molecular recognition mediated by ankyrin repeats and for engineering proteins with novel biological functions.

Proteins in the molten globule state contain high levels of secondary structure, as well as a rudimentary, nativelike tertiary topology. Thus, the structural similarity between the molten globule and native proteins may have a significant bearing in understanding the protein‐folding problem. To explore the nature of side‐chain–side‐chain interactions in the α‐lactalbumin (α‐LA) molten globule, we determined the effective concentration for formation of the 28–111 disulfide bond in 14 double‐mutant proteins, each containing two hydrophobic core residues replaced by alanine. We compared our results with those of single‐alanine substitutions using the framework of double‐mutant cycle analysis and found that, in the majority of cases, the effects of two alanine substitutions are additive. Based on these results, we propose a model of side‐chain–side‐chain interactions in the α‐LA molten globule, which takes into consideration the dynamic nature of this partially folded species.