Design, synthesis, and conformation of a model peptide of endothelin with cystine-stabilized alpha-helix motif

Cyclic and hairpin peptide complexes of heme

We have synthesized and characterized a new class of heme-peptide complexes using disulfide-linked hairpin-turn and cyclic peptides and compared these to their linear analogues. The binding affinities, helicities, and mechanism of binding of linear, hairpin, and cyclic peptides to [FeIII(coproporphyrin-I)]+ have been determined. In a minimalist approach, we utilize amphiphilic peptide sequences (15-mers), where a central histidine provides heme ligation, and the hydrophobic effect is used to optimize heme-peptide complex stability. We have incorporated disulfide bridges between amphiphilic peptides to make hairpin and even cyclic peptides that bind heme extremely well, roughly 5 x 106 times more strongly than histidine itself. CD studies show that the cyclic peptide heme complexes are completely alpha-helical. NMR spectra of paramagnetic complexes of the peptides show that the 15-mer peptides bind sequentially, with an observable monopeptide, high-spin intermediate. In contrast, the cyclic peptide complexes ligate both imidazoles cooperatively to the heme, producing only a low-spin complex. Electrochemical measurements of the E1/2 of the FeIII(coproporphyrin-I)+ complexes of these peptides are all at fairly low potentials, ranging from -215 to -252 mV versus NHE at pH 7.

The chimeric peptide [Lys(-2)-Arg(-1)]-sarafotoxin-S6b, composed of the endothelin pro-sequence and sarafotoxin, retains the salt-bridge staple between Arg(-1) and Asp8 previously observed in [Lys(-2)-Arg(-1)]-endothelin. Implications of this salt-bridge in the contractile activity and the oxidative folding reaction

The chimeric peptide [Lys(-2)-Arg(-1)]-sarafotoxin-S6b (KR-SRTb) designed from the Lys-2-Arg-1 dipeptide of the endothelin pro-sequence and the sarafotoxin-S6b sequence was synthesized. Its contractile activity was found to be decreased markedly when compared with that of the parent SRTb. In contrast, the extension by the Lys-Arg dipeptide was found to increase the formation of the native disulfide isomer (82/18 versus 96/4) when the reaction was carried out in the presence of redox reagents. The solution structure of KR-SRTb was determined by NMR as a function of pH. In the carboxylic acid state, the structure consists of the cystine-stabilized alpha-helical motif, with the alpha-helical part spanning residues 9-15, and of an unstructured C-terminal tail. In the carboxylate state, the structure is characterized by a salt-bridge between Arg(-1) and Asp8, which we identified previously in the [Lys(-2)-Arg(-1)]-endothelin-1 peptide (KR-ET-1). The fact that this salt-bridge is commonly observed in KR-SRTb and KR-ET-1, despite the 33% sequence difference between the corresponding parental peptides, highlights the remarkable adaptability of the Lys-Arg extension for the formation of a special salt-bridge. As a consequence, this salt-bridge, which does not depend on either the 4-7 sequence of the loop or the C-terminal sequence, appears to be particularly well suited to improve the stability of the cystine-stabilized alpha-helical motif. Therefore, because of its high yield in the native disulfide arrangement and its high permissiveness for sequence mutation in the 4-7 loop, such a stabilized cystine-stabilized alpha-helical motif could be a valuable scaffold for the presentation of a library of constrained short peptides.

Characterization of MB-1. A dimeric helical protein with a compact core

MB-1 is a de-novo protein designed to incorporate a large number of the nutritionally important amino acids methionine, lysine, leucine and threonine into a stable four-helix bundle protein. MB-1 has been expressed and purified from Escherichia coli, indicating it was resistant to intracellular proteases [Beauregard, M., Dupont, C., Teather, R.M. & Hefford, M.A. (1995) Bio/Technology 13, 974]. Here we report an analysis of the secondary, tertiary and quaternary structures in MB-1 using circular dichroism, fluorospectroscopy and size-exclusion chromatography. Our data indicate that the MB-1 structure is close to the target structure, an alpha-helical bundle, in many respects and is highly helical in solution. The single tyrosine incorporated into the designed protein as a spectrocopic probe of tertiary structure, is buried in a compact, folded core and becomes accessible on protein denaturation, as per design. Furthermore, MB-1 was found to be native-like in many respects: (a) protein denaturation induced by urea is cooperative and fully reversible; (b) its oligomeric state at moderate concentration is well defined; and (c) MB-1 has very low affinity for 8-anilino-1-naphthalenesulfonic acid (ANSA), leading to enhancement of ANSA fluorescence that resembles that of other native proteins. On the other hand, our analysis revealed two aspects that command further attention. The folding stability of MB-1 as assessed by urea and thermal denaturation is somewhat less than that found for natural globular proteins of similar size. Size-exclusion chromatography experiments and analysis of MB-1 denaturation indicate that MB-1 is dimeric, not monomeric as designed. In light of these results, the utility and the current limitations of our design approach are discussed.

Conformational analysis of biologically active truncated linear analogs of endothelin-1 using NMR and molecular modeling

Some linear truncated analogs of endothelin-1 display potent agonistic activity at the ET(B) receptor, especially when the side chain of Trp21 is N-formylated. Then, the three-dimensional arrangements of six structurally reduced linear analogs, three formylated and three nonformylated, have been investigated by high resolution NMR spectroscopy and molecular modeling, in order to pinpoint the conformational features related to the biological activity. Two-dimensional double-quantum-filtered correlation spectroscopy (DQFCOSY), total correlation spectroscopy (TOCSY) and nuclear Overhauser enhancement spectroscopy (NOESY) were recorded and analyzed for each molecule. Interspatial distance constraints were derived from the intensity of the NOESY connectivities. The formation of hydrogen bonding was monitored from the temperature dependence of the NH chemical shifts. Molecular models calculated by means of distance geometry, simulated annealing and energy minimization, using the NMR constraints, strongly suggested a global elongated structure for the formylated analogs exhibiting biological activity, and a folded arrangement for the unformylated derivatives. Homology comparisons allowed the identification of a beta-turn-like folding of the C-terminal segment Asp18-Trp21 as a probable key-factor for activity.

The development of potent peptide agonists and antagonists for the endothelin receptors

The endothelins (ETs), sarafotoxins (SRTXs), vasoactive intestinal contractor (VIC), and bibrotoxin are a family of potent vasoconstrictor peptides. All peptides in this family possess 21 amino acids arranged in a unique bicyclic motif formed between cystine bridges in the 1-15 and 3-11 positions. Since the discovery of endothelin-1 (ET-1) in 1988, significant effort has been focused on the understanding of its structure-activity relationships. The identification of endothelin receptor subtypes has led to the discovery/design of potent peptide agonists and antagonists, along with nonpeptide antagonists of endothelin with varying levels of potency and receptor subtype selectivity. In keeping with the theme of this journal, this review will focus only on the development of peptidic-based agonists and antagonists of endothelin in addition to their applications in understanding the physiological and/or pathophysiological role of endothelin and its isopeptides.