What Can We Learn from Molecular Recognition in Protein–Ligand Complexes for the Design of New Drugs?

Abstract

The understanding of noncovalent interactions in protein-ligand complexes is essential in modern biochemistry and should contribute toward the discovery of new drugs. In the present review, we summarize recent work aimed at a better understanding of the physical nature of molecular recognition in protein -ligand complexes and also at the development and application of new computational tools that exploit our current knowledge on structural and energetic aspects of ~~ ~ protein-ligand interactions in the design of novel ligands. These approaches are based on the exponentially growing amount of information about the geometry of protein structures and the properties of small organic molecules exposed to a structured molecular environment. The various contributions that determine the binding affinity of ligands toward a particular receptor are discussed. Their putative binding site conformations are analyzed, and some predictions are attempted. The similarity of ligands is examined with respect to their recognition properties. This information is used to understand and propose binding modes. In addition, an overview of the existing methods for the design and selection of novel protein ligands is given.

Method for Including the Dynamic Fluctuations of a Protein in Computer-Aided Drug Design

“…30 The method was quickly embraced and is now a standard practice in computational drug design. 4,5,7,[31][32][33][34] MUSIC is the multiple-copy method that is employed in the development of the dynamic and static pharmacophore models; it is a procedure available in the Monte Carlo program BOSS. 28 Methanol molecules have been used to describe binding sites that complement the catalytic residues D64 and D116 in the active site.…”

“…Current methods to develop receptor-based models usually rely on a single representation of the protein conformation. [4][5][6][7][8][9][10][11][12] Our previous study demonstrated that a "static" pharmacophore model, based on the crystal structure used to initiate the MD study, exhibited a poorer performance than the dynamic model when fitting known inhibitors of the integrase. 2 HIV-1 integrase is an interesting test case for developing dynamic methods because it has an active site that is shallow, solvent exposed, and minimally restricted in conformational sampling.…”

“…4,5,[7][8][9][10][11][12] Because dynamic behavior is very important for regulating function in many protein systems, this method has been proposed to accommodate such information in drug design. The goal of developing a dynamic pharmacophore model is to identify compounds that complement the protein while causing minimal disruption of the conformation and flexibility of the active site, potentially reducing the entropic penalties 4,5 incurred by the protein upon binding a ligand. While rigid versions of the ligands can be synthesized to reduce such penalties on the part of the ligand, this is the first method introduced to reduce similar entropic penalties incurred by the receptor.…”

Chiral Separation for Enantiomeric Determination in the Pharmaceutical Industry

“…All these interactions are usually non-covalent in nature. The experimentally determined binding constants, K i , are typically in the range of 10 -2 to 10 -12 M -1 , which corresponds to a negative free energy of binding of 10 to 80 kJmol -1 [408]. The interactions between ligands and proteins include a wide range of interactions including electrostatic, van der Waals, steric, hydrophobic and hydration forces [409] related to the active sites of the protein.…”

“…In the course of binding, the ligand and protein have to adapt to each other to achieve a successful recognition process. The small molecule ligand is usually the more flexible partner, and thus can adopt a large variety of different low energy conformations upon interaction with a protein [408]. The enantioselective recognition of molecules with one chiral centre requires a protein to interact with a minimum of three substrate locations, while stereoselectivity towards a substrate with two or three stereocentres requires interactions with a minimum of four or five substrate locations, respectively [410].…”