Study on the cyclization tendency of backbone cyclic tetrapeptides.

The cyclization kinetics of five backbone-cyclic tetrapeptides was investigated both experimentally and computationally. The aim was to both accurately measure the cyclization rates in solution and develop a method that efficiently estimates the relative cyclization tendencies computationally. Progression of the cyclization reaction was monitored directly, yielding the kinetics of changes in the amounts of the linear precursor and the products. These measurements were used to calculate the reaction rates; the results were consistent with a first-order reaction kinetics. In order to predict the cyclization rates computationally, the conformation space of the linear precursors was mapped and used to construct an approximate partition function. We assumed that the cyclization tendency was correlated with the relative probability of being found in a cyclization-prone conformation of the backbone, this probability was estimated from the partition function. The results supported this assumption and demonstrated that, within reasonable accuracy, we are able to predict the relative cyclization tendencies of the peptides measured.

Synthesis of N-carboxyalkyl and N-aminoalkyl functionalized dipeptide building units for the assembly of backbone cyclic peptides.

To improve the assembly of backbone cyclic peptides, N-functionalized dipeptide building units were synthesized. The corresponding N-aminoalkyl or N-carboxyalkyl amino acids were formed by alkylation or reductive alkylation of amino acid benzyl or tert-butyl esters. In the case of N-aminoalkyl amino acid derivatives the aldehydes for reductive alkylation were obtained from N,O-dimethyl hydroxamates of N-protected amino acids by reduction with LiAlH4. N-carboxymethyl amino acids were synthesized by alkylation using bromoacetic acid ester and the N-carboxyethyl amino acids via reductive alkylation using aldehydes derived from formyl Meldrums acid. Removal of the carboxy protecting group leads to free N-alkyl amino acids of very low solubility in organic solvents, allowing efficient purification by extraction of the crude product. These N-alkyl amino acids were converted to their tetramethylsilane-esters by silylation with N,O-bis-(trimethylsilyl)acetamide and could thus be used for the coupling with Fmoc-protected amino acid chlorides or fluorides. To avoid racemization the tert-butyl esters of N-alkyl amino acids were coupled with the Fmoc-amino acid halides in the presence of the weak base collidine. Both the N-aminoalkyl and N-carboxyalkyl functionalized dipeptide building units could be obtained in good yield and purity. For peptide assembly on the solid support, the allyl type protection of the branching moiety turned out to be most suitable. The Fmoc-protected N-functionalized dipeptide units can be used like any amino acid derivative under the standard conditions for Fmoc-solid phase synthesis.

A backbone-cyclic, receptor 5-selective somatostatin analogue: synthesis, bioactivity, and nuclear magnetic resonance conformational analysis.

Cyclo(PheN2-Tyr-D-Trp-Lys-Val-PheC3)-Thr-NH2 (PTR 3046), a backbone-cyclic somatostatin analogue, was synthesized by solid-phase methodology. The binding characteristics of PTR 3046 to the different somatostatin receptors, expressed in CHO cells, indicate high selectivity to the SSTR5 receptor. PTR 3046 is highly stable against enzymatic degradation as determined in vitro by incubation with rat renal homogenate and human serum. The biological activity of PTR 3046 in vivo was determined in rats. PTR 3046 inhibits bombesin- and caerulein-induced amylase and lipase release from the pancreas without inhibiting growth hormone or glucagon release. The major conformation of PTR 3046 in CD3OH, as determined by NMR, is defined by a type II' beta-turn at D-Trp-Lys and a cis amide bond at Val-PheC3.

Depsipeptide analogues of elastin repeating sequences: synthesis.

Depsipeptide analogues of peptide sequences can help in elucidating the role of specific hydrogen bonds in determining the conformation in peptides. The repeating pentapeptide and hexapeptide sequences of elastin have been suggested to contain a type II beta-turn with a 4----1 hydrogen bond. Depsipeptide analogues of the repeating sequences of elastin in which this 4----1 hydrogen bond cannot exist were synthesized. A fragment condensation approach was employed in which the depsipeptide ester bond was introduced early in the synthesis. This approach proved to be effective, although the increased lability of the depsipeptide ester bond resulted in side products and low yields in some reactions.

Depsipeptide analogues of elastin repeating sequences: conformational analysis.

In this work the effect of elimination of a specific hydrogen bond on the conformation of the repeating peptides of elastin was studied. These repeating sequences are the pentapeptide Val-Pro-Gly-Val-Gly and the hexapeptide Val-Ala-Pro-Gly-Val-Gly. These sequences have been proposed to occur in a beta-turn conformation with a hydrogen bond involving the amide NH of the internal valine residue and the carbonyl oxygen of the residue preceding proline. In the depsipeptide analogues studied in this work, this 4-1 beta-turn hydrogen bond cannot occur. We studied the depsipeptide sequences Val-Pro-Gly-Hiv-Gly and Val-Ala-Pro-Gly-Hiv-Gly (Hiv denotes S-alpha-hydroxyisovaleric acid, the hydroxy acid analogue of valine), as well as the peptide sequences Val-Pro-Gly-Val-Gly and Val-Ala-Pro-Gly-Val-Gly. Compounds studied included sequences with the Boc and benzyl ester protecting groups, derivatives with the acetyl and N-methylamide end groups and polymers of the above sequences. Our conclusions are based on a comparison of depsipeptides with analogous peptides. Conformational analysis was carried out by nmr, CD, and ir spectroscopy. We propose that in the repeating sequences of elastin an equilibrium exists between a gamma-turn structure and a beta-turn structure in the Pro-Gly segment resulting in a structure that combines flexibility with strong conformational preferences. The C7 involves the amide NH of the internal glycine and the carbonyl oxygen of the residue preceding proline. In the N-methylamide derivatives a similar equilibrium exists in the Gly-Val-Gly segment. In the depsipeptides the beta-turn cannot occur and only the gamma-turn is seen. In the polydepsipeptides the major conformational feature is a type I beta-turn involving Gly5 NH and Pro CO.

Cross-reactive asparagine-rich determinants shared between several blood-stage antigens of Plasmodium falciparum and the circumsporozoite protein.

From a Plasmodium falciparum cDNA expression library derived from mRNA of the asexual blood stages, we isolated and sequenced five different cDNA clones whose predicted protein products were unusually rich in asparagine (Asn). Two of the clones, R5 and G5, contain tandem imperfectly repeated sequences based on Asn-Asn-Thr (NNT) and Asn-Asn-Met (NNM) respectively. The other three, E4, C5 and R13, as well as G5, contain stretches of polyasparagine varying in length from 2 to 26 residues. Results of DNA blotting experiments with the individual cDNA sequences as probes suggest that each of the five clones corresponds to a different P. falciparum gene. The fragments of P. falciparum proteins expressed by the cDNA clones shared cross-reactive antigenic determinants which were present on multiple P. falciparum proteins. In immunoblotting experiments, owl monkey antibodies selected for binding to the polypeptide expressed by clone E4, C5 or G5 reacted with the expressed proteins from all 5 clones, and with at least 10 proteins from schizont infected erythrocytes. The cross-reactive epitopes could be modeled by two Asn-rich peptide structures: (1) (NNT)8, whose sequence was based on the R5 repeat; and (2) (NPNA)6, whose sequence was based on the Asn-rich repeat of the P. falciparum circumsporozoite protein (CSP). Antibodies that bound to each peptide were selected from sera of immune monkeys that had never been exposed to sporozoites. The selected antibodies bound all 5 expressed proteins in immunoblotting assays and also bound to several proteins from parasitized erythrocytes. Such cross reactivity between the CSP repeating unit and several blood-stage antigens has not been previously reported.

An evaluation of the advantages and effectiveness of picric acid monitoring during solid phase peptide synthesis.

The effectiveness of picric acid for the monitoring of coupling completion during solid phase peptide synthesis was evaluated. Each coupling step was monitored during the synthesis of 124 peptides by the method of simultaneous multiple peptide synthesis. Of the peptides tested (1622 picrate determinations), approximately 10% underwent a single low yield step, another 10% had 2 or 3 consecutive low yield steps and about 20% contained "humps" 4 to 8 steps long. In virtually all instances, the low yield steps occurred only at or after coupling step 7. High picrate values obtained after incorporation of methionine appear to be caused by a sulfonium salt in the side chain of methionine. Differences in average reactivities of amino acids correlated with structural elements of their side chains. Picric acid, which is currently seldom used for monitoring of coupling completion during solid phase peptide synthesis, was found to be straightforward and nondestructive. While picric acid monitoring is not as sensitive as the more commonly used ninhydrin method, it yields information which is not detectable with the use of ninhydrin and is clearly superior for certain applications and investigations.