Molecular mechanisms of constitutive activity: mutations at position 111 of the angiotensin AT1 receptor.

A possible molecular mechanism for the constitutive activity of mutants of the angiotensin type 1 receptor (AT1) at position 111 was suggested by molecular modeling. This involves a cascade of conformational changes in spatial positions of side chains along transmembrane helix (TM3) from L112 to Y113 to F117, which in turn, results in conformational changes in TM4 (residues I152 and M155) leading to the movement of TM4 as a whole. The mechanism is consistent with the available data of site-directed mutagenesis, as well as with correct predictions of constitutive activity of mutants L112F and L112C. It was also predicted that the double mutant N111G/L112A might possess basal constitutive activity comparable with that of the N111G mutant, whereas the double mutants N111G/Y113A, N111G/F117A, and N111G/I152A would have lower levels of basal activity. Experimental studies of the above double mutants showed significant constitutive activity of N111G/L112A and N111G/F117A. The basal activity of N111G/I152A was higher than expected, and that of N111G/Y113A was not determined due to poor expression of the mutant. The proposed mechanism of constitutive activity of the AT(1) receptor reveals a novel nonsimplistic view on the general problem of constitutive activity, and clearly demonstrates the inherent complexity of the process of G protein-coupled receptor (GPCR) activation.

3D model for TM region of the AT-1 receptor in complex with angiotensin II independently validated by site-directed mutagenesis data.

A three-dimensional model of the complex of angiotensin II (AII) with the transmembrane (TM) region of the angiotensin II receptor of type 1 (the AT-1 receptor) was obtained by molecular modeling procedures employing structural homology to the X-ray structure of rhodopsin. Since the modeling procedure considered only steric and energy considerations without prior knowledge of the experimental results of site-directed mutagenesis, the results with receptor mutants could be used for independent validation of the model. Indeed, the model brings in contact the residues of AII responsible for agonistic activity, Tyr(4), His(6), and Phe(8), with many residues of AT-1 involved in signal transduction according to site-directed mutagenesis. The model also predicts the existence of several possible conformational pathways for transferring the binding signal through the TM region of AT-1 to the intracellular loops interacting with the G-protein.

Novel approach to computer modeling of seven-helical transmembrane proteins: current progress in the test case of bacteriorhodopsin.

G-protein coupled receptors (GPCRs) are thought to be proteins with 7-membered transmembrane helical bundles (7TM proteins). Recently, the X-ray structures have been solved for two such proteins, namely for bacteriorhodopsin (BR) and rhodopsin (Rh), the latter being a GPCR. Despite similarities, the structures are different enough to suggest that 3D models for different GPCRs cannot be obtained directly employing 3D structures of BR or Rh as a unique template. The approach to computer modeling of 7TM proteins developed in this work was capable of reproducing the experimental X-ray structure of BR with great accuracy. A combination of helical packing and low-energy conformers for loops most close to the X-ray structure possesses the r.m.s.d. value of 3.13 A. Such a level of accuracy for the 3D-structure prediction for a 216-residue protein has not been achieved, so far, by any available ab initio procedure of protein folding. The approach may produce also other energetically consistent combinations of helical bundles and loop conformers, creating a variety of possible templates for 3D structures of 7TM proteins, including GPCRs. These templates may provide experimentalists with various plausible options for 3D structure of a given GPCR; in our view, only experiments will determine the final choice of the most reasonable 3D template.

Ab initio modeling of small, medium, and large loops in proteins.

This study presents different procedures for ab initio modeling of peptide loops of different sizes in proteins. Small loops (up to 8--12 residues) were generated by a straightforward procedure with subsequent "averaging" over all the low-energy conformers obtained. The averaged conformer fairly represents the entire set of low-energy conformers, root mean square deviation (RMSD) values being from 1.01 A for a 4-residue loop to 1.94 A for an 8-residue loop. Three-dimensional (3D) structures for several medium loops (20--30 residues) and for two large loops (54 and 61 residues) were predicted using residue-residue contact matrices divided into variable parts corresponding to the loops, and into a constant part corresponding to the known core of the protein. For each medium loop, a very limited number of sterically reasonable C(alpha) traces (from 1 to 3) was found; RMSD values ranged from 2.4 to 5.9 A. Single C(alpha) traces predicted for each of the large loops possessed RMSD values of 4.5 A. Generally, ab initio loop modeling presented in this work combines elements of computational procedures developed both for protein folding and for peptide conformational analysis.

Molecular modeling and design of regioselectively addressable functionalized templates with rigidified three-dimensional structures.

Extensive conformational analysis of a series of beta-alkyl substituted cyclopeptides-cyclo(Pro(1)-Xaa(2)-Nle(3)-Ala(4)-Nle(5)-Pro(6)-Xaa(7)- Nle(8)-Ala(9)-Nle(10)) and cyclo[Pro(1)-Xaa(2)-Nle(3)-(Cys(4)- Nle(5)-Pro(6)-Xaa(7)-Nle(8)-Cys(9))-Nle(10)] as well as their corresponding unsubstituted core structures cyclo(Pro(1)-Xaa(2)-Ala(3)-Ala(4)-Ala(5)-Pro(6)-Xaa(7)-Ala(8)-Ala(9)- Ala(10)) and cyclo(Pro(1)-Xaa(2)-Ala(3)-Cys(4)- Ala(5)-Pro(6)-Xaa(7)-Ala(8)-Cys(9)-Ala(10)) has been performed employing both the ECEPP/2 and the MAB force fields (Xaa = Gly, L-Ala, D-Ala, Aib, and D-Pro). Results show that (a) possible three-dimensional structures of the cyclo(Pro(1)-Gly(2)-Lys(3)-Ala(4)-Lys(5)-Pro(6)-Gly(7)-Lys(8)-Ala(9)- Lys(10)) molecule are not limited to a single extended "rectangular" conformation with all Lys side chains oriented at the same side of the molecule; (b) conformational equilibrium in monocyclic analogues obtained by replacements of conformationally flexible Gly residues for L-Ala, D-Ala, Aib, or D-Pro is not significantly shifted towards the target "rectangular" conformational type; and (c) introduction of disulfide bridges between positions 4 and 9 is a very powerful way to stabilize the target conformations in the resulting bicyclic molecules. These findings form the basis for further design of rigidified regioselectively addressable functionalized templates with many application areas ranging from biostructural to diagnostic purposes.

Isolated transmembrane helices arranged across a membrane: computational studies.

A computational procedure for predicting the arrangement of an isolated helical fragment across a membrane was developed. The procedure places the transmembrane helical segment into a model triple-phase system 'water-octanol-water'; pulls the segment through the membrane, varying its 'global' position as a rigid body; optimizes the intrahelical and solvation energies in each global position by 'local' coordinates (dihedral angles of side chains); and selects the lowest energy global position for the segment. The procedure was applied to 45 transmembrane helices from the photosynthetic reaction center from Rhodopseudomonas viridis, cytochrome c oxidase from Paracoccus denitrificans and bacteriorhodopsin. In two thirds of the helical fragments considered, the procedure has predicted the vertical shifts of the fragments across the membrane with an accuracy of -0.15 +/- 3.12 residues compared with the experimental data. The accuracy for the remaining 15 fragments was 2.17 +/- 3.07 residues, which is about half of a helix turn. The procedure predicts the actual membrane boundaries of transmembrane helical fragments with greater accuracy than existing statistical methods. At the same time, the procedure overestimates the tilt values for the helical fragments.

Studies of conformational isomerism in alpha-melanocyte stimulating hormone by design of cyclic analogues.

Results of energy calculations for alpha-MSH (alpha-melanocyte stimulating hormone, Ac-Ser1-Tyr2-Ser3-Met4-Glu5-His6-Phe7-Arg8-Trp9- Gly10-Lys11-Pro12-Val13-NH2) and [D-Phe7] alpha-MSH were used for design of cyclic peptides with the general aim to stabilize different conformational isomers of the parent compound. The minimal structural modifications of the conformationally flexible Gly10 residue, as substitutions for L-Ala, D-Ala, or Aib (replacing of hydrogen atoms by methyl groups), were applied to obtain octa- and heptapeptide analogues of alpha-MSH(4-11) and alpha-MSH(5-11), which were cyclized by lactam bridges between the side chains in positions 5 and 11. Some of these analogues, namely those with substitutions of the Gly10 residue with L-Ala or Aib, showed biological activity potencies on frog skin comparable to the potency of the parent tridecapeptide hormone. Additional energy calculations for designed cyclic analogues were used for further refinement of the model for the biologically active conformations of the His-Phe-Arg-Trp "message" sequence within the sequences of alpha-MSH and [D-Phe7]alpha-MSH. In such conformations the aromatic moieties of the side chains of the His6, L/D-Phe7, and Trp9 residues form a continuous hydrophobic "surface," presumably interacting with a complementary receptor site. This feature is characteristic for low-energy conformers of active cyclic analogues, but it is absent in the case of inactive analogues. This particular spatial arrangement of functional groups involved in the message sequence is very close for alpha-MSH and [D-Phe7]alpha-MSH, as well as for biologically active cyclic analogues despite differences of dihedral angle values for corresponding low-energy conformations.

A novel, non-statistical method for predicting breaks in transmembrane helices.

We have developed a novel, non-statistical procedure for predicting possible breaks in transmembrane helices based on energy calculations. The procedure consists of stepwise elongation of the 'core' helical fragment determined by consensus results of several available prediction procedures. At each step, we calculate the conformational energies corresponding to the regular 'frozen' helical conformer of the 'core' fragment elongated by two flanking residues, E(alpha), as well as those to several options for the fragment to enter or exit the helix by changing conformations of the flanking residues, Ei. The minimal values out of Ei - E(alpha), delta(k), can be viewed as a profile of relative energies, where each minimum of delta(k) is a signal to start or to stop transmembrane helix. We suggest that boundaries of the transmembrane helix would be determined by the signals closest to the 'core' sequence in the delta(k) profiles. Our procedure was applied to prediction of the N- and C-termini for 45 transmembrane helices from the photosynthetic reaction center from Rhodopseudomonas viridis, bacteriorhodopsin and the cytochrome c oxidase from Paracoccus denitrificans. The results clearly showed that it is significantly more probable that a prediction accuracy within an error of +/- 2 residues will be obtained by our procedure than by three different statistical approaches.

The use of topographical constraints in receptor mapping: investigation of the topographical requirements of the tryptophan 30 residue for receptor binding of Asp-Tyr-D-Phe-Gly-Trp-(N-Me)Nle-Asp-Phe-NH2 (SNF 9007), a cholecystokinin (26-33) analogue that binds to both CCK-B and delta-opioid receptors.

The cholecystokinin (26-33) [CCK (26-33)] octapeptide analog Asp-Tyr-D-Phe-Gly-Trp(N-Me)-Nle-Asp-Phe-NH2 (SNF 9007) is a potent and selective ligand for both the CCK-B and delta-opioid receptors. Pharmacological studies of SNF 9007 suggest a relationship between the ligand requirements of CCK-B and delta-opioid receptors, which further implies a possible structural relationship between these receptors. We have utilized topographical constrainment of the important Trp30 residue to investigate structural features of SNF 9007 that would distinguish between binding requirements in this region for the CCK-B and delta-opioid receptors. Thus, the four optically pure isomers of beta-MeTrp were substituted for L-Trp30 of SNF 9007. Receptor binding results suggest that the preferred topography of the Trp30 residue for CCK-B receptor binding may be the 2S,3S (erythro-L) configuration whereas for the delta-opioid receptor it may be the 2S,3R (threo-L) configuration. Molecular modeling studies of these ligands further support the recently revised receptor-bound model for CCK-B octapeptide ligands (Kolodziej et al. J. Med. Chem. 1995, 38, 137-149) and are in good agreement with the DPDPE-delta opioid receptor "template" model (Nikiforovich et al. Biopolymers 1991, 31, 941-955).

Novel cyclic analogs of angiotensin II with cyclization between positions 5 and 7: conformational and biological implications.

To study the conformational features of molecular recognition of angiotensin II (Asp1-Arg2-Val3-Tyr4-Val/IIe5-His6-Pro7-Phe8, AII), the synthesis and biological testing of several cyclic analogs of AII cyclized between positions 5 and 7 have been performed. The synthesized analogs were Sar1-Arg2-Val3-Tyr4-cyclo(Cys5-His6-Pen7)-Phe8 (3), Sar1-Arg2-Val3-Tyr4-cyclo(Asp5-His6-Apt7)-Phe8 (4), Sar1-Arg2-Val3-Tyr4-cyclo(Glu5-His6-Apt7)-Phe8 (5), Sar1-Arg2-Val3-Tyr4-cyclo-(Cys5-His6-Mpt7)-Phe8 (6), Sar1-Arg2-Val3-Tyr4-cyclo(Cys5-His6-Mpc7)-Phe8 (7), Sar1-Arg2-Val3-Tyr4-cyclo(Hcy5-His6-Mpt7)-Phe8 (8), and Sar1-Arg2-Val3-Tyr4-cyclo(Hcy5-His6-Mpc7)-Phe8 (9), where Apt stands for 4-amino-trans-proline, and Mpt and Mpc for 4-mercapto-trans- and -cis-prolines, respectively. Compound (9) showed good affinity at AT-1 receptors, namely a KD = 20 nM. In functional assays, it showed the characteristics of a weak partial agonist with a relative affinity of 0.26% of that for AII and an intrinsic efficacy, alpha E, of 0.42. Molecular modeling suggested a possible explanation for this finding: the relatively strong binding and the weak partial agonistic activity of compound 9 are due to interaction with AT-1 receptor of only two functionally important groups, namely, the side chains of the His6 and Phe8 residues.

Conformational analysis of beta-methyl-para-nitrophenylalanine stereoisomers of cyclo[D-Pen2, D-Pen5]enkephalin by NMR spectroscopy and conformational energy calculations.

Solution conformations of beta-methyl-para-nitrophenylalanine4 analogues of the potent delta-opioid peptide cyclo[D-Pen2, D-Pen5]enkephalin (DPDPE) were studied by combined use of nmr and conformational energy calculations. Nuclear Overhauser effect connectivities and 3JHNC alpha H coupling constants measured for the (2S, 3S)-, (2S, 3R)-, and (2R, 3R)-stereoisomers of [beta-Me-p-NO2Phe4]DPDPE in DMSO were compared with low energy conformers obtained by energy minimization in the Empirical Conformational Energy Program for Peptides (ECEPP/2) force field. The conformers that satisfied all available nmr data were selected as probable solution conformations of these peptides. Side-chain rotamer populations, established using homonuclear (3JH alpha H beta) and heteronuclear (3JH alpha C gamma) coupling constants and 13C chemical shifts, show that the beta-methyl substituent eliminates one of the three staggered rotamers of the torsion angle chi 1 for each stereoisomer of the beta-Me-p-NO2Phe4. Similar solution conformations were suggested for the L-Phe4-containing (2S, 3S)- and (2S, 3R)-stereoisomers. Despite some local differences, solution conformations of L- and D-Phe4-containing analogues have a common shape of the peptide backbone and allow similar orientations of the main delta-opioid pharmacophores. This type of structure differs from several models of the solution conformations of DPDPE, and from the model of biologically active conformations of DPDPE suggested earlier. The latter model is allowed for the potent (2S, 3S)- and (2S, 3R)-stereoisomers of [beta-Me-p-NO2Phe4]DPDPE, but it is forbidden for the less active (2R, 3R)- and (2R, 3S)-stereoisomers. It was concluded that the biologically active stereoisomers of [beta-Me-p-NO2Phe4]DPDPE in the delta-receptor-bound state may assume a conformation different from their favorable conformations in DMSO.

Conformationally readdressed CCK-B/delta-opioid peptide ligands.

The sequence of a cholecystokinin (CCK) related peptide was modified to obtain analogues, which interact selectively either with CCK-B, or with delta-opioid receptors. Two kinds of peptides were designed, namely, the cyclic peptides of the H-Tyr-cyclo (D-Pen-Gly-Trp-L/D-3-transmercaptoproline)-Asp-Phe-NH2 sequence (compounds 1a and 1b, respectively), and the linear peptides of the H-Tyr-D-Val-Gly-Trp-L/D-3-trans-methylmercaptoproline-Asp-Phe- NH2 sequence (compounds 2a and 2b, respectively). The only difference between the chemical structures of the linear analogues compared to the cyclic ones is that one covalent bond has been eliminated and a sulfur atom is replaced by a methyl group. Molecular modeling showed that, among low-energy conformers of cyclic compounds 1, there are three-dimensional structures compatible to the model for delta-receptor-bound conformer, suggested earlier [G. V. Nikiforovich, V.J. Hruby, O. Prakash, and C.A. Gehrig (1991) Biopolymers, vol. 31, pp. 941-955]. Results of binding assays fully supported the rationale for the design of compounds 1 and 2. The cyclic analogue 1a has Ki values of 4.5 and > 5000 nM at delta- and mu-opioid receptors, respectively; and IC50 values of 1.6 and > 10,000 nM for CCK-A and CCK-B receptors, respectively. The results of this study demonstrate a possibility to redirect a peptide sequence that interacts with one type of receptors (CCK-B receptors) toward interaction with another type (delta-opioid receptors) belonging to a different physiological system. This redirection could be performed by changing the conformational properties of the peptide with very minimal changes in its chemical structure.

Design, synthesis, biology, and conformations of bicyclic alpha-melanotropin analogues.

Seven side chain-constrained bicyclic alpha-melanotropin (alpha-MSH) analogues were designed and synthesized, their conformations analyzed, and their biological properties examined in the frog skin and lizard skin bioassays. The structure of these analogues is based on the central sequence Ac-Cys4-Xaa5-His6-DPhe7-Arg8-Trp9-Cys10-Lys11 -NH2 (Xaa5 = Asp or Glu) and has been extended on the N-terminal with the amino acids Ser1-Tyr2-Ser3 and on the C-terminal with Pro12-Val13 to more closely resemble the native hormone alpha-MSH. The analogue Ac-Cys4-Asp5-His6-DPhe7-Arg8-Trp9-Lys10-Cys11 -NH2 also was synthesized, and its conformational and biological properties were examined. Design of these analogues was based upon the previously identified superpotent monocyclic peptides [Cys4,DPhe7,Cys10]alpha-MSH(4-10)-NH2 and [Nle4,Asp5,DPhe7,Lys10]alpha-MSH(4-10)-NH2 with the rationale of increasing conformational constraints to restrict the available backbone conformations as a means to identify the conformations that facilitate biological activity. Computer-assisted conformational analysis of the central tetrapeptide residues 6-9 identified beta-turns which varied with respect to the residue in the i + 1 position. Each highly constrained peptide contains D-Phe7 and a 23-membered ring which has previously been identified as crucial to produce prolonged acting peptides with superagonistic activities. The bicyclic peptides reported in this study are full agonists and are 25-400-fold less potent than alpha-MSH in the frog and lizard skin bioassays.

Ac-[3- and 4-alkylthioproline31]-CCK4 analogs: synthesis and implications for the CCK-B receptor-bound conformation.

It has been reported that substitution of the Met31 residue in Boc-CCK4 (Boc-Trp30-Met31-Asp32-Phe33-NH2, CCK33 numbering) by trans-3-propyl-L-proline yields a highly potent and selective CCK-B agonist. To further explore the structural requirements of the Met31 side chain in the receptor-bound conformation of CCK4, we have synthesized several Ac-CCK4 analogs containing substitution of Met31 by 3- and 4-(alkylthio)-substituted proline derivatives. To this end we have developed novel synthetic routes to enantiomerically pure N-Boc-4-cis- and -trans-(methylthio)prolines and racemic N-Boc-3-cis and -trans-[(4-methylbenzyl)thio]prolines. The protected mercaptoprolines were incorporated into Ac-CCK4 analogs using SPPS and were alkylated using various electrophiles following cleavage from the solid support. Binding assays reveal that 3-(alkylthio)prolines analogs have higher affinities at the CCK-B receptor than the corresponding 4-(alkylthio)proline analogs, and that trans-3-(alkylthio)proline analogs had higher affinities than corresponding cis-3-(alkylthio)proline analogs. Within both the cis- and trans-3-(alkylthio)proline series, the order of potency was found to be Me < Et < n-Pr. The trans-3-(n-propylthio)-L-proline analog demonstrates a higher affinity than that reported for Boc-CCK4[trans-3-propyl-L-Pro31]. Comparison of the low-energy structures calculated for several high-affinity Ac-CCK4 analogs reveal a common geometry which we propose to be the CCK-B receptor-bound conformation. This model shows grouping of the hydrophobic side chains of Trp, Met, and Phe at one side of the molecule and the hydrophilic side chain of Asp and the C-terminal carboxamide at the other side.

Computational molecular modeling in peptide drug design.

The review concentrates on practical applications of computer molecular modeling in peptide drug design. The examples of the predictions (successful or not) made by computational modeling before synthesis of peptide analogs, not the explanations provided after synthesis and biological testing of peptides, are discussed. The review spans over 20 years of predictions made by computer molecular modeling for bradykinin, angiotensin, thyrotropin-releasing factor, tuftsin, substance P, CCK-related peptides, luliberin, alpha-melanotropin and opioid peptides. The described examples are discussed in terms of finding the optimal way to use computer modeling for peptide design. The step-by-step 'technology' of peptide design is outlined in detail.

Conformational analysis of two cyclic analogs of angiotensin: implications for the biologically active conformation.

Conformations of two cyclic analogs of angiotensin (Asp1-Arg2-Val3-Tyr4-Val/Ile5-His6-Pro7-Phe8, AT), cyclo[Sar1, Cys3, Mpt5]-AT and cyclo[Sar1, HCys3, Mpt5]-AT, were studied, independently employing two complementary techniques, energy calculations and NMR measurements in DMSO solution. NMR data were indicative of well-defined solution conformations for the cyclic moieties of cyclo[Sar1, Cys3, Mpt5]-AT and cyclo[Sar1, HCys3, Mpt5]-AT, including the phi values for the Cys3/HCys3 and Tyr4 residues, as well as the chi 1 value for the Tyr4 residue. Solution conformations for the exocyclic linear parts of both molecules cannot be described by the NMR data with the same precision. At the same time, independent energy calculations revealed the same conformations of cyclic moieties of cyclo[Sar1, Cys3, Mpt5]-AT and cyclo[Sar1, HCys3, Mpt5]-AT among low-energy conformers for both peptides. Moreover, the same conformations are compatible with the model of AT receptor-bound conformation (Nikiforovich & Marshall, 1993), which assumes the particular spatial arrangement of aromatic moieties of Tyr4, His6, and Phe8 residues and the C-terminal carboxyl. These conformers of cyclo[Sar1, Cys3, Mpt5]-AT and cyclo[Sar1, HCys3, Mpt5]-AT contain "an open turn" in the backbone of the Tyr4-Val5 residues, instead of the earlier proposed beta-like reversal, thus confirming the suggestion that the conformation(s) ensuring binding of AT analogs with specific receptors should not be described in terms of a unique backbone conformer.

Sequence-dependent modification of Trp by the Pmc protecting group of Arg during TFA deprotection.

The extent of transfer of the Pmc protecting group from the guanidino group of arginine to the side chain of tryptophan depends on the spacial distance of these side chains. When these two amino acids are separated by one amino acid, the transfer of the Pmc protecting group is the most pronounced, and it cannot be completely prevented by the use of currently utilized scavenger mixtures. The extent of this side reaction also depends on the amino acid separating the arginine and tryptophan residues and position of tryptophan within the peptide chain as well as on the type of the solid-phase carrier.

Conformation-function relationships in LHRH analogs. I. Conformations of LHRH peptide backbone.

A systematic conformational build-up procedure was performed for the LHRH molecule, pGlu1-His2-Trp3-Ser4-Tyr5-Gly6-Leu7-Arg8-Pro9-Gly10-NH 2. The results showed a very high flexibility of the LHRH backbone, with 300 conformers being regarded as having low energy. At the same time, the conformational flexibility of LHRH differs among the fragments of the molecule. The most rigid fragments of LHRH are the Ser4-Tyr5-Gly6-Leu7 and Tyr5-Gly6-Leu7-Arg8 central tetrapeptides, the latter possessing only eight different types of low-energy backbone conformers. These eight conformer types belong to different kinds of chain reversals which are stabilized by different systems of intramolecular hydrogen bonds. Some of them resemble the beta-II' turn, which was derived as the LHRH structure from energy calculations by others. The results obtained are in good agreement with the experimental data on LHRH flexibility in solution.

Conformation-function relationships in LHRH analogs. II. Conformations of LHRH peptide agonists and antagonists.

Systematic energy calculations were performed for a series of LHRH analogs including five agonists with substitutions of D- or N-Me-amino acid residues in positions 4, 6 and 7, and five antagonists with substitutions of D-, N-Me- or alpha-Me-amino acid residues in positions 1, 2, 3, 6, 7 and 10, as well as a bicyclic LHRH antagonist. The geometrical shapes of the calculated low-energy backbone structures for each compound were compared to those of LHRH itself. It appeared that the beta-II' turn at the Tyr5-Gly6-Leu7-Arg8 central tetrapeptide is the common structure for all LHRH agonists considered. LHRH antagonists also possess a common chain reversal in the central tetrapeptide, but it is different from that for LHRH agonists. The LHRH agonists share a similar low-energy conformer at the level of the entire peptide backbone. A characteristic feature of this conformer is a 'surface' formed by a 'polygon' with hydrophobic moieties of pGlu1, Trp3, Tyr5, Leu7 and Pro9 in the corners and with the side chain of the His2 residue in the middle, the latter being crucial for a manifestation of LHRH agonistic activity. Since the N-terminal tripeptide of LHRH presumably participates in a direct interaction with specific receptors, it is legitimate to suggest that the beta-II' turn in the central tetrapeptide maintains the proper spatial arrangement of the N-terminal tripeptide. On the other hand, LHRH antagonists considered in this study were shown to possess low-energy structures, with the spatial arrangement of the residues in the N-terminal tripeptide similar to that of agonists. This would suggest a new approach to the design of LHRH antagonists, namely by stabilizing this specific arrangement, rather than the beta-II' turn in the central tetrapeptide.