Structure-function analysis of proteins through the design, synthesis, and study of peptide models

Phosphorylation of calcineurin: effect on calmodulin binding

The effect of phosphorylation of calcineurin on calmodulin (CaM) binding was examined using a synthetic peptide which contains the CaM-binding domain and the serine phosphorylation site. The peptide, corresponding to residues 391-414 of brain calcineurin A subunit, was rapidly phosphorylated by protein kinase C and Ca2+/CaM-dependent protein kinase II but not by cAMP-dependent protein kinase. Phosphorylation of peptide 391-414 did not significantly alter the binding of CaM when compared to the non-phosphorylated peptide.

Peptide models of dynorphin A(1-17) incorporating minimally homologous substitutes for the potential amphiphilic beta strand in residues 7-15

Two peptide models of dynorphin A(1-17) have been synthesized. These peptides incorporate a minimally homologous substitute sequence for residues 6-17, including alternating lysine and valine residues substituting for the potential amphiphilic beta-strand structure in positions 7-15. Model 1 retains Pro10 from the native sequence, but model 2 does not. Compression isotherms of peptide monolayers at the air-water interface and CD spectra of peptide films adsorbed from aqueous solution onto siliconized quartz slides were evaluated by comparison to those of idealized amphiphilic alpha-helical, beta-sheet, and disordered peptides. Dynorphin A(1-17) was mostly disordered, whereas beta-endorphin was alpha helical. Dynorphin model 1 had properties similar to those of dynorphin A(1-17) at these interfaces, but model 2 formed strongly amphiphilic beta sheets. In binding assays to mu-, delta-, and kappa-opioid receptors in guinea pig brain membranes, model 1 reproduced the high potency and selectivity of dynorphin A(1-17) for kappa receptors, and model 2 was only 3 times less potent and less selective for these receptors. Both peptide models retained the high, kappa-selective agonist activity of dynorphin A(1-17) in guinea pig ileum assays, and like dynorphin A(1-17), model 1 had little activity in the rat vas deferens assay. In view of the minimal homology of the modeled dynorphin structures, these studies support current models of membrane-catalyzed opioid ligand-receptor interactions and suggest a role for the amphiphilic alpha-helical and beta-strand structures in beta-endorphin and dynorphin A(1-17), respectively, in this process.

Proton NMR and CD solution conformation determination and opioid receptor binding studies of a dynorphin A(1-17) model peptide

The opioid peptide dynorphin A(1-17) contains a peptide segment in residues 7-15 with the potential to form an amphiphilic beta-strand. This amphiphilic structure may, like the amphiphilic alpha-helices found in many other peptide hormones, be an important determinant of its interactions with membranes and receptors. In order to investigate and characterize these interactions, we have synthesized a 17-residue dynorphin analogue (YGGFLKKVKPKVKVKSS) that incorporates a peptide model of this amphiphilic secondary structure with minimized homology (25%) relative to the native sequence. This peptide exhibits the full biological potency of dynorphin in assays of kappa-opioid receptor binding, and is more selective for this type of opioid receptor than the natural peptide. The conformation of the model peptide in aqueous solution has been investigated in detail by NMR spectroscopy. The values of the NH-CH alpha coupling constants together with rotating frame NOEs indicate the presence of an amphiphilic structure together with some beta-strand structure in residues 7-15, and demonstrate that a peptide model that stabilizes this structure in aqueous solution and enhances kappa-opioid receptor selectivity can be successfully designed using using alternating lysine and valine residues.

The structure of the mouse lymphocyte pore-forming protein perforin

Purified murine lymphocyte pore-forming protein (PFP or perforin) was partially sequenced. Oligonucleotides synthesized on the basis of this sequence information were used to screen a murine cytotoxic T lymphocyte (CTL) cDNA library. Seven clones were obtained, two of which were sequenced, providing full-length sequence information on PFP. Murine PFP (534 a.a.) is 68% identical to human PFP. Hydropathic analysis revealed a predominantly hydrophilic protein with some hydrophobic domains, including a region (a.a. 191-251) that could contain putative membrane-spanning domains. PFP is approx. 20% identical to human C7, C8 and C9 within a region encompassing 270 a.a., confirming previous immunological cross-reactivity studies. Northern blot analysis showed that expression of PFP but not of a serine esterase transcript is enhanced in a CTL line by antigen receptor-stimulation. Southern blot analysis of mouse genomic DNA indicated that PFP is encoded as a single-copy gene with the coding region contained within 10 kilobases of genomic DNA.