Computational Determination of the Structure of Opioid Ligands
The structurally constrained cyclo(Dmt-Tic) was useful as a platform for computational modeling studies because of its conformation and that it displayed bioactivity. It enabled the development of a three-dimensional pharmacophore that described the position of the minimal physicochemical components needed to bind to the δ-opioid receptor. The pharmacophore consisted of a close proximity (5.4) and approximate parallel orientation of the aromatic rings, cis orientation of the peptide bond (N-C) and gauche- and gauche+ orientations around the Cα-Cβ bond of Dmt and Tic, respectively.
X-ray diffraction analysis of crystalline molecules containing the Dmt-Tic pharmacophore
Validation of the computational derived Dmt-Tic pharmacophore involved excellent agreement with X-ray crystal structures of N,N-(CH3)2-Dmt-Tic-OH (δ-antagonist). The x-ray crystal structures of two other Dmt-Tic peptides with C-terminal substitutions, H-Dmt-Tic-NH-1-adamantane (µ-opioid agonist/weak δ agonist), and N,N-(CH3)2-Dmt-Tic-NH-1-adamantane (δ-opioid antagonist/µ-opioid agonist) revealed a tilt in the upper portion of the Tic ring that was similar to the orientation of Tic in the TIPP analogues.
Molecular modeling of the Dmt-Tic pharmacophore
Molecular models of δ-opioid agonists H-Dmt-Tic-NH-CH2-Bid (Fig. 2), H-Dmt-Tic-Gly-NH-Bzl, H-Dmt-Tic-NH-(R)CH(CH2)-Bid, H-Dmt-Tic-NH-CH(CH2-Bid)-COOH were developed based on the crystalline crystal structure of H-Dmt-Tic-NH-1-adamantane to reflect the behavior of the upper ring of Tic in C-terminally substituted Dmt-Tic derivatives. Conformational searching and energy minimization paradigms were utilized to explore the conformational space accessible those C-terminal substitutions. Various low energy conformations with unique C-terminal orientations for each derivative were docked into proposed δ-opioid receptor binding sites. Furthermore, molecular models of δ-opioid antagonists H-Dmt-Tic-NH-CH2-CH2-Bid, H-Dmt-Tic-Gly-NH-CH2-Bid, H-Dmt-Tic-Gly-NH-CH2-Bzl, H-Dmt-Tic-NH-CH2-Bid(CH2-COOH) were developed based on the X-ray crystal structure of N,N-(CH3)2-Dmt-Tic-NH-1-adamantane. Several conformations for each derivative were docked into binding pockets of the δ-opioid receptor, and the study explored different points of interaction with the δ-opioid receptor for δ agonists and δ antagonists in the Dmt-Tic series.
Molecular models of endomorphin analogues
Molecular models were developed for Dmt-containing endomorphin analogues modified at the C-terminus based on J-coupling constants and nuclear Overhauser effect (NOE) cross peaks from two-dimensional 1H NMR analysis and used to evaluate µ- and δ-opioid agonist interaction with the δ-opioid receptor model. The model of the µ-opioid agonist [Dmt1,6-quinoyl4]endomorphin-2 [lacking δ-opioid receptor binding and bioactivity and contrasts to the 1-naphthyl and 5-quinoyl derivatives with bifunctional bioactivity (δ-/µ-opioid agonism)] suggested that the message and address domains of ligands may not be a linear, sequential array of amino acids in the peptide, but also involve the spatial orientation of the side-chains.
Differentiation between µ- and δ-opioid ligands by molecular modeling
Molecular modeling studies proposed that one major structural characteristic that differentiated µ- and δ-opioid agonists from δ-specific opioid antagonists was the intra-ring distance, which also affects their presumed solution conformation. Both µ- and δ-opioid agonists have extend topographies with aromatic rings separated by 9-13, which is twice that observed for δ-opioid antagonists. These data further corroborated the inability to obtain a bioactive pharmacophore of the bis-[Dmt-NH(CH2)n]-alkyl or - pyrazinone compounds from eight thousand low energy conformations that met the criteria of the 1 H NMR data; the flexibility and large number of possible conformations made it impossible to determine the significance of one conformer over another. Thus, modeling techniques alone cannot determine the bioactive form.
Docking of opioid ligands into a model of the δ-opioid receptor
After two years of building a model of the δ-opioid receptor based on bovine rhodopsin, sequence alignment, shifts and improvements, the initial and preliminary model is being further refined based on new technologies in the computational field and new results from mutagenesis of the receptor performed in other laboratories.
Validation of the computational derived Dmt-Tic pharmacophore involved excellent agreement with X-ray crystal structures of N,N-(CH3)2-Dmt-Tic-OH (δ-antagonist). The x-ray crystal structures of two other Dmt-Tic peptides with C-terminal substitutions, H-Dmt-Tic-NH-1-adamantane (µ-opioid agonist/weak δ agonist), and N,N-(CH3)2-Dmt-Tic-NH-1-adamantane (δ-opioid antagonist/µ-opioid agonist) revealed a tilt in the upper portion of the Tic ring that was similar to the orientation of Tic in the TIPP analogues.
Molecular models were developed for Dmt-containing endomorphin analogues modified at the C-terminus based on J-coupling constants and nuclear Overhauser effect (NOE) cross peaks from two-dimensional 1H NMR analysis and used to evaluate µ- and δ-opioid agonist interaction with the δ-opioid receptor model. The model of the µ-opioid agonist [Dmt1,6-quinoyl4]endomorphin-2 [lacking δ-opioid receptor binding and bioactivity and contrasts to the 1-naphthyl and 5-quinoyl derivatives with bifunctional bioactivity (δ-/µ-opioid agonism)] suggested that the message and address domains of ligands may not be a linear, sequential array of amino acids in the peptide, but also involve the spatial orientation of the side-chains.
Molecular modeling studies proposed that one major structural characteristic that differentiated µ- and δ-opioid agonists from δ-specific opioid antagonists was the intra-ring distance, which also affects their presumed solution conformation. Both µ- and δ-opioid agonists have extend topographies with aromatic rings separated by 9-13, which is twice that observed for δ-opioid antagonists. These data further corroborated the inability to obtain a bioactive pharmacophore of the bis-[Dmt-NH(CH2)n]-alkyl or - pyrazinone compounds from eight thousand low energy conformations that met the criteria of the 1 H NMR data; the flexibility and large number of possible conformations made it impossible to determine the significance of one conformer over another. Thus, modeling techniques alone cannot determine the bioactive form.
|