We report the results of a joint NMR and theoretical investigation devoted to the conformational properties of N-acetyl-N'-methylamides of aliphatic amino acids with side chains of increasing bulkiness: Gly, Ala, Leu, Ile, and tert.Leu. In this series, determination of the coupling constants 3JHNC alpha H together with the coupling constants 3JC'NC alpha H (thanks to specific carbon-13 labeling of the N-acetyl carbonyl group) led to the derivation of alternative A, B, and C parameters in a Karplus-type relation expressing the dependence of 3JC'NC alpha H upon the phi dihedral angle. The value of the latter is found to increase regularly following the increase of the side-chain bulkiness. The theoretical conformational analysis is performed by applying the SIBFA procedure, which uses empirical formulas based on ab initio SCF computations. The conformational energy maps illustrate the progressive distortion of the backbone conformation incurred in the series Gly to tert.Leu. Theoretical values computed for 3JHNC alpha H and 3JC'NC alpha H are found to be in a good quantitative agreement with the experimental ones.
The dipeptide, (DL)-alanyl-(DL)-norvaline, crystallizes in the monoclinic space group P2(1)/c, with a = 12.559(2)A, b = 5.265(1), c = 16.003(3), beta = 103.53(2) degrees, Z = 4. The structure was solved by direct methods and refined to an R-value of 0.054 for 871 reflections with I greater than 2 sigma. The molecule exists as a zwitterion in the crystal. The peptide unit is trans and shows significant deviations from planarity (delta omega = 12.4 degrees). The peptide backbone adopts an extended conformation. The unit cell contains D-Ala-L-norval and its enantiomer. The molecular conformation and packing features show a striking resemblance to those for D-Ala-L-Met (1), and leads to the speculation that norvaline might act as an analog of methionine.
The Gly3 residue in Gly2- and D-Ala2-enkephalins has been replaced by delta Ala3 and Ser3 in order to examine the effect on binding to the delta and mu opiate receptors. [D-Ala2, delta Ala3, Leu5]-enkephalin maintains most of its receptor binding affinities for both sites. It is suggested that the proper conjunctions of positions 2 and 3 of the enkephalin sequence are important to receptor preference and that position 3 may have a very specific interaction with the delta receptor. The in vivo analgesic and CNS activities of the peptides are also discussed.
Three Tic-containing (Tic = 1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid) model peptides were synthesized to assess the tendency of this constrained Phe analogue to fold into a beta-bend and a helical structure, and to adopt a preferred side-chain disposition. The results of the solution conformational analysis, performed by using Fourier transform infrared absorption and 1H nuclear magnetic resonance, indicate that in chloroform the -Aib-D-Tic-Aib-, -(Aib)2-D-Tic-(Aib)2-, and -L-Pro-D-Tic- sequences fold into intramolecularly H-bonded forms to a great extent. An X-ray diffraction analysis on p-BrBz-(Aib)2-DL-Tic-(Aib)2-OMe monohydrate and p-BrBz-L-Pro-D-Tic-NHMe allows us to conclude that, while the pentapeptide methylester forms an incipient (distorted) 3(10)-helix, the dipeptide methylamide adopts a type-II beta-bend conformation. In both cases, the D-Tic side-chain conformation is D, gauche(-). The implications for the use of the Tic residue in designing conformationally restricted analogues of bioactive peptides are briefly discussed.
The synthesis of a substrate for the endothelin converting enzyme using the continuous flow FMOC solid phase method has been accomplished. The allyl group, employed to protect Glu side chain, was cleanly removed during the synthesis in the presence of tert.-butyl based protecting groups to enable the introduction of the Edans moiety. The substrate was obtained in 31% yield after reverse phase hplc purification.
An improved synthesis of 4, 4-difluoro-L-proline is described. A key step in this synthesis involves the fluorination of Z-4-keto-L-proline benzyl ester, using diethylaminosulfur trifluoride (DAST), to give the corresponding Z-4, 4-difluoro-L-proline benzyl ester. Preparation of dipeptide derivatives containing 4, 4-difluoro-L-proline has been accomplished by initial synthesis of the corresponding 4-keto-L-proline depeptide derivatives, which were then fluorinated with DAST. A one-step synthesis of Boc-4-keto-L-proline from Boc-4-hydroxy-L-proline by chromium trioxide oxidation is reported. These synthetic procedures should facilitate the preparation of a variety of peptide analogues containing 4, 4-difluoro-L-proline and 4-keto-L-proline.
A chemotactic peptide CHO.Met.Leu.Phe.OH has been synthesized classically using the mixed anhydride procedure. The formyl group was introduced by coupling formic acid in the presence of dicyclohexylcarbodiimide to the partially protected triptide. The final product was obtained by treatment of the intermediate CHO.Met.Leu.Phe.OBzl with hydrogen fluoride. The ED50 of the peptide in the Boyden chamber assay was 7 x 10(-11) M; in the lysozyme release assay 2.4 x 10(-10) M and in the beta-glucuronidase release assay 2.6 x 10(-10) M. In a radioreceptor assay the ID50 of the peptide was 3.3 x 10(-10) M.
Solid-phase synthesis of peptide haptens containing 2-[2-(S-cysteinyl)ethanol has been achieved by solution-phase synthesis of a properly protected S-alkylated cysteine derivative and subsequent solid-phase incorporation.
Treatment of α-amanitin with I2-morpholine (1:1:3 mole ratio α-amanitin: I2: morpholine) affords 7' substitution of the α-amanitin with the N-atom of the morpholine rather than the expected 7'-iodination. The 7'-(N-morpholino)-α-amanitin is isolable in yields up to 95% and shows electronic spectra similar to the parent amatoxin but with distinct Λmax and Λmin at 243 and 237 nm, respectively. Spectrophotometric titration of the phenolic hydroxyl shows an increase in pKa from 9.72 (α-amanitin) to 10.24 for this product; an increase in pKa is expected on chemical grounds following ortho substitution with an amine. Proton magnetic resonance shows two aromatic protons (δ4H= 7.56 p.p.m.; δ5H= 6.88 p.p.m.; J = 9 Hz), thus establishing the extent and position of substitution at the 7' carbon in the tryptophan residue. The 7'-(N-morpholino)-α-amanitin inhibits RNA polymerase activity with half-maximal inhibition at 1 times 10-8M and 3 times 10-4M for class II and class III sea urchin RNA polymerase, respectively. Extension of this procedure with additional dialkylamines affords the preparation of a family of 7'-(N-dialkylamino)-α-amanitins analogous to the prototypical morpholine derivative. With tertiary and aryl amines no amatoxin derivative results, while with primary alkyl amines this procedure yields an unresolved product mixture containing an as yet undefined inhibitor of the class I as well as class II and III RNA polymerases. The title reaction thus provides a facile, efficient route to modified amatoxins that do not disrupt the amatoxin: RNA polymerase binary complex formation. A means toward reporter molecule analysis of class II RNA polymerase structure and function is implied by these results.
The solution conformations of thyrotrophin-releasing hormone (TRH), a biologically-active tripeptide, have been predicted computationally using a method which represents solution effects as a dielectric continuum surrounding the molecule throughout energy minimisation; this is described as the reaction field contribution. The predicted conformations have been compared with previous experimental and theoretical studies. Favoured conformations of TRH under other simulated conditions have also been examined, including the behaviour of the cis isomer, and have been related to possible interactions with intermediates and with specific receptors. The relevance of these predictions to the design of biologically-active analogues of TRH has been analysed.
Acylation of the hydroxyl groups in the side chains of serine, threonine or tyrosine occurs in coupling with active esters. This side reaction, which is quite pronounced in histidine-containing peptides, can be prevented with additives. From a series of compounds tested, 2,4-dinitrophenol and pentachlorophenol were the most effective.
L-Lysine D-glutamate crystallizes in the monoclinic space group P2(1) with a = 4.902, b = 30.719, c = 9.679 A, beta = 90 degrees and Z = 4. The crystals of L-lysine D-aspartate monohydrate belong to the orthorhombic space group P2(1)2(1)2(1) with a = 5.458, b = 7.152, c = 36.022 A and Z = 4. The structures were solved by the direct methods and refined to R values of 0.125 and 0.040 respectively for 1412 and 1503 observed reflections. The glutamate complex is highly pseudosymmetric. The lysine molecules in it assume a conformation with the side chain staggered between the alpha-amino and the alpha-carboxylate groups. The interactions of the side chain amino groups of lysine in the two complexes are such that they form infinite sequences containing alternating amino and carboxylate groups. The molecular aggregation in the glutamate complex is very similar to that observed in L-arginine D-aspartate and L-arginine D-glutamate trihydrate, with the formation of double layers consisting of both types of molecules. In contrast to the situation in the other three LD complexes, the unlike molecules in L-lysine D-aspartate monohydrate aggregate into alternating layers as in the case of most LL complexes. The arrangement of molecules in the lysine layer is nearly the same as in L-lysine L-aspartate, with head-to-tail sequences as the central feature. The arrangement of aspartate ions in the layers containing them is, however, somewhat unusual. Thus the comparison between the LL and the LD complexes analyzed so far indicates that the reversal of chirality of one of the components in a complex leads to profound changes in molecular aggregation, but these changes could be of more than one type.
The two Z-L-Ala-DL-(alpha Me)Trp-NH2 diastereomeric dipeptides were synthesized from (Z-L-Ala)2(O) and H-DL-(alpha Me)Trp-NH2. The latter racemate, prepared by phase-transfer catalyzed alkylation of the N alpha-benzylidene derivative of alanine amide followed by acidic hydrolysis of the resulting Schiff base, was characterized by X-ray diffraction. The molecular and crystal structure of Z-L-Ala-L-(alpha Me)Trp-NH2, separated from its diastereomer by silica-gel column chromatography, was determined by X-ray diffraction analysis. Both independent molecules in the asymmetric unit of the dipeptide adopt a type-II beta-bend conformation. However, only the more regularly folded conformation of molecule B is stabilized by a 1<--4 C = O...H--N intramolecular H bond. The present results indicate that: (i) the C alpha-methylated (alpha Me)Trp residue is a strong beta-bend and helix former, and (ii) the relationship between (alpha Me)Trp chirality and helix screw sense tends to be opposite to that of protein amino acids. The implications for the use of the (alpha Me)Trp residue in designing conformationally restricted analogs of bioactive peptides are briefly discussed.
A conformational study of two protected peptide segments, (1-10 and 11-28), spanning the entire sequence of thymosin alpha 1, in solvents of different polarity and capability of forming hydrogen bonds, is reported. By using infrared absorption and circular dichroism techniques the occurrence of the random coil conformation, the self-associated beta-structure, and the alpha-helix (the latter adopted only by the longer peptide) was established. The self-associated species of the two peptide segments were disrupted either by adding increasing amounts of hexamethylphosphoramide or by dilution. This structural transition was monitored by the disappearance of the amide-I C = O stretching band of strongly intermolecularly hydrogen-bonded molecules (near 1630 cm-1) in the infrared absorption spectra. The tendency of these peptides to aggregate is paralleled by a decrease in their solubility. The conformational findings are discussed in terms of the solvent-dependent product yields obtained in the reaction of segment (1-10) with the N alpha-deprotected (11-28) segment to give the fully protected thymosin alpha 1.
The synthesis, by fragment condensation, of protected peptides related to the N-terminal (positions 1-14) and C-terminal (positions 36-52) portions of the amino acid sequence of porcine pancreatic secretory trypsin inhibitor II (Kazal type) is described. The alpha-carboxyl function of the lysyl residue in position 14 was used as the free acid, or protected by the tert-butyloxycarbonylhydrazide. Two different synthetic pathways were used in the preparation of the 36-52 sequence of the inhibitor. The identity and purity of the synthesized peptide derivatives were established by amino acid analysis of acid hydrolysates, optical rotation and this layer chromatography in two solvent sytems. The final products were also evaluated, after partial deprotection with anhydrous hydrogen fluoride or aqueous 90% trifluoroacetic acid, by paper electrophoresis at different pH values and enzymic digestion with papain and aminopeptidase M followed by quantitative amino acid analysis.
N-9-Fluorenylmethoxycarbonyl-(Fmoc) amino-acid chlorides have been prepared by reaction of hydrogen chloride on purified mixed Fmoc-amino acid-monoalkyl carbonic acid anhydrides in dichloromethane. The products partially undergo subsequent conversion to the corresponding esters due to the presence of the liberated alcohol, the extent depending on the nature of the alkyl group. Esterification occurred to 5-20% when the alkyl group was isopropyl. Anhydrides of monoisopropenyl carbonic acid which liberate acetone instead of an alcohol gave products uncontaminated with ester. The three components in a reaction mixture could be determined as the reaction progressed by normal phase high-performance liquid chromatography of aliquots, which had been freed of excess hydrogen chloride, on a mu Porasil (underivatized silica) column using tert.-butanol-hexane (1.5:98.5) as solvent.
Dynorphin-(1-13) (Dyn-(1-13] and its analogs substituted by single introduction of Ala in positions 1-11 were synthesized by the solid-phase method and purified by high pressure liquid chromatography. Relative potencies of the synthetic compounds were determined by their ability to inhibit electrically-evoked contractions of the guinea pig ileum (GPI) and of the mouse vas deferens (MVD) and to compete with [3H]-etorphine for opiate receptors in rat brain homogenates. Introduction of Ala in positions 1 and 4 of Dyn-(1-13) provoked most important decreases in the activity of the molecule in the three assays (relative potency of 0.2% or less). Substitution of Ala in positions 2 or 5, but not 3, also severely decreased the potency of the peptide in the smooth muscle preparations (0.6-5.0% activity). However, the opiate receptor binding assay was less sensitive to the replacement of residue in position 2 (20% activity) than that in positions 3 or 5 (12% and 6% relative potencies, respectively). In the GPI assay and the opiate binding test, the other substitutions which greatly lowered the potency of the molecule were seen in positions 6, 7, 9 and 11, four basic residues. Among these, Arg6 and Arg7 were demonstrated to be the most important in the three biological tests. Finally, the replacement of Ile8 by Ala increased the relative potency of Dyn-(1-13) up to 191% and 900% in the MVD and the opiate binding tests, respectively.
Sixteen dermorphin analogues were synthesized and characterized for mu- and delta-opioid receptor binding properties using [3H]DAGO and [3H]DPDPE, respectively. The analogues included the following: substitutions at position 4 and/or the C-terminal residue; deletions of Gly4 or Pro6-Ser7; inclusion of Z or an acetyl group on the epsilon-amino group of Lys7; and the presence of either a C-terminal amide or free acid group. Two peptides, [Lys7-OH]- and [Lys7-NH2]dermorphin, had mu-affinities (Ki mu = 0.15-0.13 nM) and mu-selectivities (Ki delta/Ki mu = 1158-1482) higher than dermorphin (Ki mu = 0.28 nM; Ki delta/Ki mu = 295) and best fitted a one-site binding model similar to dermorphin. Significantly better (P < 0.0001) fits to a two-site binding model vs. a one-site model were observed with four dermorphin analogues: [Lys(Z)7-OH]heptapeptide, [des-Gly4(Tyr4,Pro5,Asn6-OH)]hexapeptide and two pentapeptides, [Tyr5-NH2] and [Trp4,Asn5-OH]. Our data revealed a complex binding pattern for dermorphin analogues to brain mu-receptors in which Hill coefficients less than 0.85 generally suggest heterogeneity of mu-receptors; however, only detailed analyses of the data derived from the non-linear regression fits for one- or two-components gave evidence for the possible existence of two separate [3H]DAGO binding sites. Eight of our dermorphin analogues had significantly better fits for a two-site model (P < 0.05), but only four seemed to have two distinct Ki values (P < 0.0001).
