Search

We have performed a ligand coordinate analysis to monitor the movement of the inhibitor SC-558 from the active site of the COX-2 protein to the exterior using molecular dynamics techniques. This study provides an insight into the intermolecular interactions formed by the ligand during this journey. The published crystal structure of COX-2 with SC-558 in the active site (1cx2) was taken, and the ligand was moved incrementally in 13 steps. At each of these points on the path, exhaustive minimization and dynamics calculations were performed. The role of water was found to be important in these computations. An average structure was obtained from 250 conformations at each point and minimized. At each point on the path, the 10 lowest-energy conformations were also selected; a consideration of the average and lowest conformations provides fine details on the consistency of specific and strong interactions, and also on the geometry of the ligand. The movement of the ligand through the protein may be divided into three stages that are distinguished from each other because of energy and geometry discontinuities in both the ligand and the protein. The first of these covers the region between the active site and the point at 5.8 A from it. The second, which covers the distance between 8.2 and 10.0 A and is associated with maximum energetic and structural instability, is of critical importance. The third stage covers the distance between 10.5 A and the exterior and represents a stage of increasing hydration and expulsion of the ligand from the protein. Our results provide a confirmation for the existence of a shallow cavity near the protein surface in which the ligand is bound reversibly. By examining the residues that show maximum mobility, one obtains an idea of the gating mechanism that governs the entry and exit of the protein into or from the deep pocket that contains the active site. We note, however, that the variation of the root-mean-square deviation of all residues begins to increase almost as soon as the ligand leaves the active site, and even before there are any changes in the gate inter-residue distances. This loosening of the protein even before the gate opens might be a part of the enthalpy-entropy balance that accompanies the ligand's passage through the protein. Our results provide an energy profile of the ligand during its entry/exit into/from the protein and can, in principle, enable one to assess the residence time, which in turn may be associated or indirectly correlated with adverse cardiovascular side effects of nonsteroidal anti-inflammatory drugs. We believe that similar analyses for other selected COX-2-specific inhibitors will provide a measure (or prediction) of possible toxicity effects.