A Detailed Description of Conformational Space of Native Enkephalines and their Cyclic Analogues



Zdenek Kriz(a), Jaroslav Koca(a), Per H. J. Carlsen(b)

(a) Laboratory of Computational Chemistry and Molecular Modeling, Department of Organic Chemistry, Masaryk University, 611 37 Brno, Czech Republic

(b) Institute of Organic Chemistry, University of Trondheim, Norwegian Institute of Technology, N-7034 Trondheim-NTH, Norway





Introduction

The discovery of two endogenous pentapeptides with morphine-like action, Met-enkephalin (Tyr-Gly-Gly-Phe-Met) and Leu-enkephalin (Tyr-Gly-Gly-Phe-Leu), have induced considerable interest as to the activities and conformational properties of opioid peptides, which are of interest as possible substitutes for alkaloid opiate drugs and for their biological importance as endogenous analgesies. Because of their varied and important biological roles, the structural features of enkephalines and their analogues have been extensively investigated over the past 20 years by various methods including spectroscopy (NMR, CD, IR, UV), X-ray crystallography and molecular modeling [1]. These linear small flexible molecules that, due largely to this flexibility, can interact with several distinct receptor types, each requiring different ligand conformations, to initiate different physiological events. One method reducing the flexibility of these peptides is cyclization. This not only greatly reduces the flexibility, but has also led to the synthesis of many active and receptor selective analogues. For our study we have chosen:


1. linear enkephalines

  • Leu-enkephalin (Tyr-Gly-Gly-Phe-Leu)
  • Met-enkephalin (Tyr-Gly-Gly-Phe-Met)

2. cyclic analogues of enkephalines

  • DLFE (Tyr-c(D-Lys-Gly-Phe))
  • DPDPE (Tyr-c(D-Pen-Gly-Phe-D-Pen))




Methodology

We have performed a detailed conformational study using the ROSE[2] , CICADA[3] and PANIC[4] software. For energy computation we have used the PMMX program, which is the MMX[5] program specially parametrized for peptides[6]. The interface between the programs is schematically illustrated in the followed scheme.

Scheme 1



Program ROSE is used for generation of the starting conformations by a combinatorial algorithm. Program CICADA performs travelling along the PES. Energy of the structures are computed by program PMMX which is MMX program specially parametrized for peptides. Results from the program CICADA are used for PES and flexibility analysis which is performed by program PANIC. Results from the programs CICADA and PANIC serve for the simulation of the conformational travelling along the PES which is performed by program COMBINE. Program ANALYSE is used to extract geometry parameters of structures obtained by COMBINE run. The data could be visualized by the graphics interface.



Results

Computed relative and absolute flexibilities[7] of the all investigated molecules are summarized in Table 1.


Table 1

It is shown, that both relative and absolute flexibilities[7] of linear enkephalines are comparable with flexibility numbers of cyclic analogues. Flexibility numbers of the side chain torsions are similar to flexibility numbers of the backbone torsions, which is different than computed flexibilities of single amino acids[9]. Torsions of the backbone of the peptides are more rigid than the same of the amino acids due to the arrangement in peptides. The major part of flexibility is concentrated on the outside backbone torsions of the leucine and methionine of the linear enkephalines and on the torsions of the glycine and phenylalanine of the cyclic enkephalin analogues. The most rigid from the relative and absolute flexibilities point of view is Met-enkephalin. Computed flexibilities reflect the numbers of conformers found in various energy cutoff which are summarized in next the table. The number of points is the total number of nuclear configurations saved by CICADA travelling along the PES. In order to allow for comparison of results, the calculation was stop ed on the same number of points for each molecule.

Table 2



It is shown, that the total numbers of conformations calculated for the same number of points are comparable for all molecules. The number of conformers in 10 kcal/mol energy window are for linear peptides almost two times bigger then for the cyclic analogues. The most rigid molecule , met-enkephalin, has the lowest number of conformer in energy windows under 2 kcal/mol. It is interesting that DPDPE, which is the most flexible molecule from the flexibility numbers point of view, has the lowest number of conformers in 10 kcal/mol energy window. This is in contradiction with the assumption, that a flexible molecule has a large number of low energy conformers. Explanation of this phenomenon is shown in the next table. We have performed the clustering of the conformations found into conformational families[8]. As a criteria of the distribution, we have included the differences between backbone torsions.

Table 3



It is shown that the cyclic analogues fit just one very large conformational family with Boltzmann probability around 50 % and two smaller families. The conformations of the linear peptides are spread into four (leu-enkephalin) and seven (met-enkephalin) families with even distributed Boltzmann probability around 15 %. The total numbers of families found are as follows: leu-enkephalin - 335, met-enkephalin - 382, DLFE - 44 and DPDPE - 34. It is shown, that conformations of linear peptides are distributed in large number of families with large distances and with large energy barriers separating them, whilst conformations of the cyclic analogues were found in low number of families with conformations near by each other and with small interconversion energy barriers. The computed flexibilities reflect this phenomenon. Our assumption is confirmed by RMS analysis of the dynamics simulation based on interconversion phenomenon, which is pictured in the following figures.

The graphs of RMS analysis are available
by clicking on the name of the molecule in the following line
Leu-enkephalin Met-enkephalin DLFE DPDPE


Figures show development of RMS differences during the simulation of conformational movement within the step window 0-1000 at the thermodynamic temperature 300 K. All the structures on the trajectory are compared with the lowest energy conformer. It has been observed that conformational movement of the linear peptides is accompanied by much larger RMS differences than that of the cyclic analogues.


Conclusions


Conformational potential energy hypersurface for linear enkephalines and their cyclic analogues have been analyzed by means of molecular mechanics in combination with programs ROSE, CICADA, PANIC, COMBINE and ANALYSE.




References





Acknowledgements


This work has partially been supported by the Grant Agency of the Czech Republic, grant No. 203/94/0522, and by Chem-Consult, Hundhamaren, Norway. This financial support is gratefully acknowledged. We would also like to thank Academic Supercomputer Center in Brno for providing us access to the computer facilities. The academic license provided by Biosym/MSI, Inc., for Insight II software is also acknowledged.