Search

Understanding how a linear sequence of amino acids guides acquisition of the native state of a protein — the protein folding problem — is key to interpreting information present in the human genome and to developing therapeutic approaches to diseases caused by protein misfolding. While considerable progress has been made in elucidating the interplay of local and long-range forces in the folding of α-helical proteins, such understanding has lagged for primarily β-sheet proteins [1]. We have undertaken studies aimed at determining the roles of local and global sequence information in the folding of the predominantly β-sheet protein, cellular retinoic acid binding protein I (CRABP I). CRABP I is a 136 residue β-clamshell protein (Figure 1) whose physiological role is the sequestration and transport of the hydrophobic ligand retinoic acid. CRABP I is a member of the large family of intracellular lipid binding proteins (iLBPs) [2], which contains 52 members with sequence identity greater than 30% with respect to CRABP I. Past work in our laboratory using stopped-flow (SF) mixing to follow the fluorescence of CRABP I has allowed us to define a series of kinetic phases from the urea-denatured conformational ensemble to the native fold [3]. Within the ≈10 ms dead time of the SF instrument, the unfolded ensemble forms a hydro-phobically collapsed state with considerable secondary structure. The presence of significant secondary structure along with hydrophobic collapse suggests that both global and local forces are acting in the earliest folding events.

Human calcitonin (hCt) is a peptide hormone of 32 amino acid residues that plays a central role in calcium metabolism [1,2]. The receptor-bound conformation of hCt has not yet been determined. In an effort to understand the bioactive conformation of hCt we have recently designed a lactam-bridged hCt analogue with the structure cyclo17,21-[Asp17,Orn21]hCt (1) and found it to be 400-times more potent than hCt and 4-times more potent than salmon Ct (sCt) in the in vivo hypocalcemic assay in mice, making this the most potent known hCt analogue [3]. By contrast, a linear control peptide [Asp17, Orn21]hCt (2) showed similar in vivo potency to hCt [3]. These results suggested that the conformation of 1 may correspond to a bioactive conformation of hCt [3]. Here, we present studies on the solution conformation of hCt and 1 based on NMR in combination with molecular modeling and CD spectroscopy. We also present data on the hCt receptor-binding affinity of 1 as compared to 2, hCt and sCt.