Binding and action of cecropin and cecropin analogues: antibacterial peptides from insects.

The mechanism of action of cecropin was studied by using liposomes as a model system. The bilayer was efficiently destroyed if the liposome net charge was zero or negative. Cecropin analogues with an impaired N-terminal helix had reduced membrane disrupting abilities that correlate with their lower antibacterial activity. The reduced bactericidal activity of the analogues was rationalized in terms of reduced binding to bacteria. The stoichiometry of cecropin killing of bacteria suggests that amounts of cecropin sufficient to form a monolayer strongly modify the bacterial membrane. Although some bacteria were resistant to cecropin they did bind large amounts in a non-productive manner. In contrast, mammalian erythrocytes achieve resistance by avoiding the binding of cecropin.

Effects on electrophoretic mobility and antibacterial spectrum of removal of two residues from synthetic sarcotoxin IA and addition of the same residues to cecropin B.

Cecropin B and cecropin IA (sarcotoxin IA) are 35- and 39-residue antibacterial peptides from a silk moth and a meat fly, respectively. Using solid phase synthesis we have made these peptides as well as two 37-residue analogs, one containing a deletion of leucine and lysine (residues 2a and 2b) as compared to cecropin IA, the other containing an insertion of leucine and lysine at the corresponding place in cecropin B. This addition and removal of a lysine residue did not cause the expected change in electrophoretic mobility. When tested for antibacterial spectra, the insertion analog was found to be as active as the parent compound while the deletion analog had lost most of its antibacterial capacity. In addition it was shown that the C-terminal amide contributes to the broad spectrum properties of the cecropins.

Antibacterial and antimalarial properties of peptides that are cecropin-melittin hybrids.

Solid phase synthesis was used to produce 5 hybrid peptides containing sequences from the antibacterial peptide, cecropin A, and from the bee venom toxin, melittin. Four of these chimeric peptides showed good antibacterial activity against representative Gram-negative and Gram-positive bacterial species. The best hybrid, cecropin A(1-13)-melittin(1-13) was 100-fold more active than cecropin A against Staphylococcus aureus. It was also a 10-fold better antimalarial agent than cecropin B or magainin 2. Sheep red cells were lysed by melittin at low concentrations, but not by the hybrid molecules, even at 50 times higher concentrations.

Shortened cecropin A-melittin hybrids. Significant size reduction retains potent antibiotic activity.

We have earlier reported two 26-residue antibacterial peptides made up from different segments of cecropin A (CA) and melittin (M). We now report a substantial reduction in size at the C-terminal section of the highly active hybrid CA(1-8)M(1-18), leading to a series of 20-, 18- and 15-residue analogs with antibiotic properties similar to the larger molecule. In particular, the 15-residue hybrids CA(1-7)M(2-9), CA(1-7)M(4-11) and CA(1-7)M(5-12) are the shortest cecropin-based peptide antibiotics described so far, with antibacterial activity and spectra similar or better than cecropin A and a 60% reduction in size. Their reduced size and highly alpha-helical structure require an alternative mechanism for their interaction with bacterial membranes.

On the primary structures of lysozyme, cecropins and attacins from Hyalophora cecropia.

Diapausing pupae of Cecropia respond to a bacterial infection by the selective synthesis of RNA and 15-20 hemolymph proteins. Of these we have purified lysozyme and two classes of antibacterial proteins called cecropins and attacins. The primary structure has been determined for the lysozyme, one attacin and five cecropins. We have also prepared a cDNA bank, isolated and sequenced clones corresponding to the lysozyme, the two main attacins and one cecropin. The results of these structural studies are briefly summarized. Finally we review the solid phase synthesis of cecropin A and B and 9 analogs of cecropin A.

Biological activities of des-His1[Glu9]glucagon amide, a glucagon antagonist.

Hyperglycemia in diabetes mellitus is generally associated with elevated levels of glucagon in the blood. A glucagon analog, des-His1[Glu9]glucagon amide, has been designed and synthesized and found to be an antagonist of glucagon in several systems. It has been a useful tool for investigating the mechanisms of glucagon action and for providing evidence that glucagon is a contributing factor in the pathogenesis of diabetes. The in vitro and in vivo activities of the antagonist are reported here. The analog bound 40% as well as glucagon to liver membranes, but did not stimulate the release of cyclic AMP even at 10(6) higher concentration. However, it did activate a second pathway, with the release of inositol phosphates. In addition, the analog enhanced the glucose-stimulated release of insulin from pancreatic islet cells. Of particular importance were the findings that the antagonist also showed only very low activity (less than 0.2%) in the in vivo glycogenolysis assay, and that at a ratio of 100:1 the analog almost completely blocked the hyperglycemic effects of added glucagon in normal rabbits. In addition, it reduced the hyperglycemia produced by endogenous glucagon in streptozotocin diabetic rats. Thus, we have an analog that possesses properties that are necessary for a glucagon antagonist to be potentially useful in the study and treatment of diabetes.

Efficient synthesis of 1,2,4-dithiazolidine-3,5-diones [dithiasuccinoyl-amines] from bis(chlorocarbonyl)disulfane plus bis(trimethylsilyl)amines.

The 1,2,4-dithiazolidine-3,5-dione heterocycle, also referred to as a dithiasuccinoyl (Dts)-amine, serves as a readily removable amino protecting group for building blocks used in syntheses of peptides, glycopeptides, and PNA; it is also useful as a masked isocyanate and (inversely) as a sulfurization reagent for trivalent phosphorus. Bis(chlorocarbonyl)disulfane, the two-sulfur analogue of succinyl chloride, has been envisioned as a reagent for facile single-step elaboration of the heterocycle. However, reactions of bis(chlorocarbonyl)disulfane directly with primary amines fail to yield Dts-amines for reasons that are discussed. Inspired by several precedents from the organosilicon chemistry literature that a trimethylsilyl group may serve as a "large proton," a successful, high-yield preparation of Dts-amines through reactions of bis(chlorocarbonyl)disulfane with bis(trimethylsilyl)amines has been developed. Studies aimed at elucidating mechanistic reasons for these observations are also presented.

Concept of internal structural controls for evaluation of inactive synthetic peptide analogs: synthesis of [Orn13,14]apamin and its guanidination to an apamin derivative with full neurotoxic activity.

The importance of arginine residues 13 and 14 in the bee venom neurotoxin, apamin, was teste by the synthesis of replacement analogs. [13,14-di-Ndelta-trifluoroacetylornithine]Apamin was synthesized by the solid phase method on a benzhydrylamine resin. It was deprotected to [13,14-diornithine]apamin, which was then guanidinated to produce the 4-homoarginine-13,14-diarginine analog, [Har4]apamin. Neither the trifluoroacetylornithine analog nor the ornithine analog produced any detectable symptoms when injected intravenously into mice. However, the synthetic [Har4]apamin exhibited the full neurotoxic activity of native apamin and of [Har4]apamin derived from the natural toxin. This provided an internal structural control for the correctness of the primary structure of the inactive synthetic analogs and strengthened the conclusion that one, or both, of the arginine residues plays an important role in the action of apamin.

The role of serine-123 in the activity and specificity of ribonuclease. Reactivation of ribonuclease 1-118 by the synthetic COOH-terminal tetradecapeptide, ribonuclease 111-124, and its O-methylserine and alanine analogs.

The COOH-terminal tetradecapeptide of ribonuclease A, Glu-Gly-Asn-Pro-Tyr-Val-Pro-Val-His-Phe-Asp-Ala-Ser-Val, and two analogs, [Ser(Me)-123]-RNase 111-124 and [Ala-123]-RNase 111-124, were synthesized by the solid phase method and were purified to chromatographic and electrophoretic homogeneity. Methods are described for the hydrolysis and quantitative amino acid analysis of peptides containing O-methylserine. The peptides were combined noncovalently with RNase 1-118 and examined for ability to regenerate enzymatic activity in the presence of the substrates C greater than p, U greater than p, poly(C) poly(U), and poly(AF). The dissociation constants of the peptide-protein complexes, and the Michaelis constants for C greater than p and U greater than p with the reconstituted enzymes were determined. The data were used to test hypotheses, drawn from x-ray crystallographic and other studies, for the role of serine-123 in the binding of substrates by ribonuclease. It was found that Ser-123- and Ala-123-containing peptides were equally active for the hydrolysis step when measured with C greater than p as substrate and for the transphosphorylation step as measured in the assays with poly(C). The serine and alanine analogs were also equally active for the transphosphorylation step when poly AF was the substrate. With U greater than p as substrate the alanine analog was 4 times less active than the serine derivative and with poly U it was 2 times less active. The semisynthetic enzyme composed of RNase 1-118 and [Ala-123]-RNase 111-124, therefore, shows appreciable selectivity for substrates containing cytosine. It was concluded that a hydrogen bond between the hydroxyl of serine-123 and the C4 amino group of cytidine or the C-7 amino group of formycin is not important for substrate binding and catalytic activity. In contrast, the hydrogen bond between the hydroxyl of serine 123 and the C-4 carbonyl oxygen of uridine contributes significantly to substrate binding and catalytic activity. The data with serine-O-methyl ether at position 123 in the tetradecapeptide were less clear because it was difficult to separate steric effects from the contributions of hydrogen bonding. Substrate binding to ribonuclease was rationalized in terms of a binding energy equivalent to a total of two hydrogen bonds per pyrimidine.

Solid phase synthesis of gastrin I. Comparison of methods utilizing strong acid for deprotection and cleavage.

A successful synthesis of human gastrin I in 60% overall yield based on the first residue attached to a benzhydrylamine-resin was achieved by the stepwise solidphase method. The synthesis was carried out on a 1% crosslinked polystyrene support, using conventional benzyl-based side chain protecting groups and final deprotection with different acidic protocols. Several improvements in this general approach were applied, including new scavengers, new resin attachment and especially a new technique that allows the strong acid reactions to occur by an SN2 mechanism.

A synthetic approach to structure-function relationships in the murine epidermal growth factor molecule.

Murine epidermal growth factor, a 53-amino acid peptide that is mitogenic for a number of cell types, has been synthesized by the solid-phase method. The synthetic peptide is identical to the natural material in amino acid composition, chromatographic behavior, receptor binding, and stimulation of DNA synthesis. Fragments of the EGF molecule corresponding to residues 42-53, 32-53, and 15-53 were constructed as well as the methionine sulfoxide derivative of EGF, [Met(O)21]EGF-(1-53), and a polymeric form of EGF. [Met(O)21]EGF-(1-53) was slightly less active than EGF in receptor binding and stimulation of DNA synthesis. Polymeric EGF was 1/100th as active as EGF, while EGF-(15-53) was less potent than EGF-(1-53) by a factor of 10(4). EGF-(32-53) was even less active and EGF-(43-53) was inactive.

Glucagon antagonists. Synthesis and inhibitory properties of Asp3-containing glucagon analogs.

In an effort to find analogs of glucagon that would bind to the glucagon receptor of the rat liver membrane but would not activate membrane-bound adenyl cyclase, several hybrid molecules were synthesized which contained sequences from both glucagon and secretin. [Asp3, Glu9]Glucagon and [Asp3, Glu9, Arg12]glucagon were inactive in the adenyl cyclase assay even at high concentrations but retained some binding affinity for the receptor. They were able to displace 125I-glucagon completely from its receptor and could completely inhibit the activation of adenyl cyclase by natural or synthetic glucagon. The inhibition index [I/A]50 was approximately 110 for both analogs. [Asp3]Glucagon, [Glu3]glucagon and [Asp3, Lys17, 18, Glu21]glucagon were weak partial agonists, while [Asp3, Glu21]glucagon was inactive and a poor inhibitor. The peptides were synthesized by solid-phase methods and purified to homogeneity by reverse-phase high-performance liquid chromatography on C18 silica columns. These are the first fully synthetic competitive glucagon antagonists to be reported.

Identification of an essential serine residue in glucagon: implication for an active site triad.

Several glucagon analogs containing substitutions for serine have been synthesized to assess the role of the four serine residues in the hormone. The strategic importance of His1 has been confirmed, and we have previously identified an aspartic acid critical for activity at position 9. While these findings have led to a series of pure glucagon antagonists, the details of specific glucagon-receptor interactions that switch on the ensuing signaling events are still not readily apparent. The requirement for serine was tested by the chemical synthesis of a series of analogs containing substitutions for the hydrophilic hydroxyl group in each of the highly conserved serine residues at positions 2, 8, 11, and 16 of glucagon. The resulting analogs were analyzed in rat hepatocyte membranes for their receptor-binding affinities as well as their abilities to stimulate adenylate cyclase. Positions 2 and 8 were the most sensitive to modification, where both binding and activity were adversely affected. This is consistent with the notion that although the sequence responsible for transduction lies in the amino-terminal half of glucagon, some residues at that end also contribute to binding affinity. Modifications at position 11 generated high-binding-affinity derivatives that were full or moderate agonists. In contrast, position 16 replacement analogs maintained significant receptor binding affinities while the agonist properties were almost completely lost, thus separating binding and transduction functions. Therefore, Ser16 is a third critical residue that determines glucagon activity. It is postulated, but not proven, that a serine residue, together with His1 and Asp9, may participate in the putative active center of glucagon, which, upon initial recognition and binding to receptor, leads to transduction of the biological signal. A dependence of the glucagon action on a three-residue cooperative mechanism might be analogous to the charge-relay scheme of serine proteases. It is suggested that, after binding to its receptor, glucagon becomes activated and functions like a coenzyme in catalyzing the specific hydrolysis of a peptide bond in the receptor, generating new amino and carboxyl end groups, and that one of these exposed chains may contact the GTP-binding protein and activate it for further interaction with adenylate cyclase. This idea was supported by inhibition experiments with 4-amidinophenylmethanesulfonyl fluoride (APMSF), a specific and irreversible inhibitor of serine proteases, which at a concentration of 5 mM completely suppressed cAMP formation by glucagon in liver membranes. cAMP formation was not affected if either glucagon or membranes were separately pretreated with APMSF and then assayed.

Solid-phase synthesis of thymosin alpha 1 using tert-butyloxycarbonylaminoacyl-4-(oxymethyl)phenylacetamidomethyl-resin.

Thymosin alpha 1 and its desacetyl analogue were synthesized by the solid-phase method. Use of aminoacyl-4-(oxymethyl)phenylacetamidomethyl-resin resulted in an improved yield and allowed the synthetic products to be purified by simple ion-exchange and gel filtration chromatography. Success of the synthesis was largely due to enhanced stability of the peptide-resin linkage to trifluoroacetic acid and to the elimination of hydroxy functions on the resin. This improved quality of the solid support helps eliminate chain loss and chain termination during the synthesis. The purified synthetic peptides were found to be homogeneous by paper electrophoresis, isoelectric focusing in polyacrylamide gel, and thin-layer chromatography. They also had biological activity in the azathioprine-sensitive rosette assay. Use of the new 9-(2-sulfo)fluorenylmethyloxycarbonyl chloride reagent for purification of protected peptides was also demonstrated and discussed.

Solid-phase synthesis of crystalline glucagon.

Mammalian glucagon was synthesized by a stepwise solid-phase method. The support was an alkoxybenzyl alcohol resin, and the biphenylylisopropyloxycarbonyl group was used for temporary alpha-amino protection. After purification by gel filtration and ion-exchange chromatography, the 29-residue hormone was readily crystallized from water at alkaline pH. The product was homogeneous and indistinguishable from natural bovine glucagon by gel electrophoresis, ion-exchange chromatography, reverse-phase high-pressure liquid chromatography, fluorescence spectroscopy, and amino acid analysis. The synthetic hormone was fully active in the rabbit hyperglycemia assay.

Chemical mechanism to account for artifactual formation of shortened peptides with free alpha-amino groups in solid phase peptide synthesis.

The formation of terminated peptides with free alpha-amino groups has often been observed in stepwise solid phase peptide synthesis. This has been attributed to variable accessibility in regions of the swollen crosslinked resin supports. It is now shown that impurities in the amino acid reagents are responsible for these by-products. Thus, sec.-butyloxycarbonylamino acids were isolated from tert.-butyloxycarbonylamino acids after treatment with trifluoroacetic acid under standard deprotection conditions for the removal of the tert.-butyloxycarbonyl (Boc) group. Direct reverse phase HPLC analysis of Boc-amino acids from commercial sources also showed the sec.-Boc-amino acids as impurities present at varying levels. The sec.-Boc group was stable to treatment at room temperature with trifluoroacetic acid in dichloromethane (1:1, v/v) (half-life 7 years), but was removed by HF-anisole under the standard conditions of cleavage and deprotection of assembled peptides. In model syntheses, the level of terminated free peptides corresponded to the level of preexisting sec.-Boc-amino acid impurities present in the Boc-amino acid reagents. Use of Boc-amino acids with no detectable sec.-Boc resulted in negligible levels (less than 0.05%) of terminated peptides. The problem is thus readily overcome by the use of pure Boc-amino acid starting materials and is not a reflection of a shortcoming inherent to the polymer supported nature of solid phase syntheses as has been previously suggested.