vasopressin and oxytocin receptors

Molecular docking simulations are now fast developing area of research. In this work we describe an effective procedure of preparation of the receptor-ligand complexes. The amino-acid residues involved in ligand binding were identified and described.


METHODS
Starting models of new ligands and receptors were constructed using homology modeling.This part of our work involved usage of the available amino-acid sequences [1] in conjunction with a search within crystallographic database CSDS [2].Three-dimensional (3D) models of our receptors: V 1a R, V 2 R and oxytocin receptor, were built using the human opiate-g receptor model proposed by Pogozheva and coworkers [3].The missing 2 nd and 3 rd cytosolic loops and the extracellular Nand C-termini were built using the SYBYL [4] suite of programs.For verification of our 3D theoretical model of V 1a , V 2 and OT receptors (OTR) we have calculated the root mean square (RMS) deviation between our theoretical 3D structures and the X-ray crystal structure of rhodopsin at 2.8 Å resolution, which was published recently [5].The result: 2.66 Å positively verifies our computer-modeled 3D structures.The next steps were: 1. preparation of full-atom models with atomic charges, 2. their relaxation and minimization (stand-alone and in the complexes with V 1a R, V 2 R and OTR).All non-standard aminoacid residues included in both docked ligands were parameterized in accordance with the recommendations in the AMBER 5.0 [6] manual.Atomic net charges were optimized by fitting to the ab initio molecular electrostatic potentials (using the 6-31G* basis set in the GAMESS [7] molecular program package).Minimization of these models was done in AMBER 5.0 force field using constraints for transmembrane domains in V 1a R, V 2 R and OTR models based on the opiate receptor models proposed by Pogozheva et al. [3].The starting models of ligand-receptor complexes were prepared using the AutoDock 3.0 [8] program.These computer-docked and minimized models were subsequently used as the starting point in the discussion of the properties of newly-designed ligands and for the characterization of the binding sites of our receptors.Final complexes were selected based on the criterion of the internal ligand energy of the minimized receptor-ligand complexes.The amino-acid residues involved in ligand binding were identified and described.The docking procedure involved the AutoDock 3.0 program and its new and promising hybrid search technique that implements an adaptive global optimizer with local search.The global search method is an implementation of a modified genetic algorithm, with 2-point crossover and random mutation.The local search method is based on the optimization algorithm of Solis & Wets [9], which has the advantage that it does not require gradient information in order to proceed -as was the case in previous versions of AutoDock.It also uses fixed variances for the determination of the probabilistic way of a change of a particular state variable, like the x-translation.These variances are either doubled or halved during the search, depending on the number of consecutive successful or failed moves.Success is a drop in energy.Receptor-ligand complexes were relaxed and minimized by the consecutive use of the minimization and constrained simulated annealing (CSA) protocols in vacuo (in accordance to the AMBER 5.0 manual [6]) with all but the transmembrane-domain C a atoms free to move.This was done in the AMBER 5.0 force field.Sample complexes of docked ligands are presented in Fig. 1.

RESULTS AND DISCUSSION
For CSA refinement only those complexes were retained whose receptor-ligand interaction energy was about 1000 kcal/mole or less above the absolute minimum.The AutoDock force field used in the docking procedure is very limited in its functionality -it uses electrostatic interactions and van der Waals potential.This 1000 kcal/ mole-criterion eliminated all complexes which were not properly minimized by AMBER.For efficient relaxation of the remaining part of complexes we used CSA with heating the environment up to 1200 K (1 ps), keeping this temperature constant (2 ps) and re-cooling to low temperatures (12 ps of CSA) -as shown in Fig. 2. Having this done we have selected, based on the energetic criterion, 5-7 complexes of [Mca 1 , Tyr(Me) 2 ]-AVP and DesGly 9 -[Mca 1 ]AVP in each of the V 1a R, V 2 R and OTR receptors.Their final energies are presented in Table 1.Receptor amino-acid residues responsible for binding new ligands were identified on the basis of the distance criterion.The residues whose any atom is not further than 4.5 Å away from all accepted conformations of our lig-ands are shown in Table 2 and arranged so that the more often a particular residue is involved in binding the ligand the higher is the position it occupies in the table.
Not in line with our expectations both ligands were docked relatively shallow in the binding site of the V 1a receptor -close to the extracellular loops.Their minimization and relaxation did not radically change their position, which suggests W(304,285,288), F(308,289,291), F(309,290,292); and TM7: A(334),M(311,315) on their bottom.Both our ligands being rather small molecules (93 and 97 atoms) had the advantage of free movements within the receptors' pockets but their positions did not change much even after a relatively hot (up to 1200 K) CSA, despite allowing unconstrained flexibility to all TM side chains, both ligands and the receptors' loops.This suggests a very good implementation of the genetic algorithm in the docking procedure even though the ligand-receptor interaction energies were extremely high.Slight differences between the conformations of the two ligands complexed to the three receptors studied (especially between V 1a and V 2 receptors) show that the CSA protocol caused good relaxation of ligands and found very good minima.
Figure 1.Stereoview of ligand Mca 1 ,Tyr(Me) 2 ]AVP (panel A) and DesGly 9 -[Mca 1 ]AVP (panel B) docked in human V 2 receptor.The amino-acid residues responsible for ligand binding are colored according to their chemical properties.Colors of polar amino-acid residues are brighter than non-polar.We used standard RasMol (Sayle, R., RasMol V2.6, Molecular Visualisation Program, Glaxo Wellcome Research and Development, Stevenage, Hertfordshire, U.K.) color coding of all receptor amino-acid residues.

Table 1 .Figure 2 .
Figure 2. Changes in temperature during the constrained simulated annealing (CSA).This sample drawing presents changes in temperature during CSA of DesGly 9 -[Mca 1 ]AVP ligand in OT receptor.