Figure 2.25 • Diagram of a hypothetical receptor site, showing distances between functional groups.
25, 73, 102, and 147 have been identified as important either for binding the ligand to the site or for the receptor's intrinsic activity. Keep in mind that Figure 2.25 is a two-dimensional representation of a 3D image. Therefore, the distances between amino acid residues must take into account the fact that each residue is above or below the planes of the other three residues. For an artificial ligand to "dock," or fit into the site, six distances must be considered: (a) Lys-Glu, (b) Glu-Phe, (c) Phe-Ser, (d) Ser-Lys, (e) Glu-Phe, and (f) Lys-Phe. In reality, not all six distances may be important. In selecting potential ligands, candidates might include a positively charged residue (protonated amine), aromatic ring, hydrogen bond donor or acceptor (hydroxy, phenol, amine, nitro), and hydrogen bond acceptor or a negatively charged residue (carboxyl-ate) that will interact with the aspartate, phenylalanine, serine, and lysine residues, respectively. A template is constructed containing the appropriate residues at the proper distances with correct geometries, and the chemical database is searched for molecules that fit the template. A degree of fit or match is obtained for each "hit." Their biological responses are obtained, and the model for the receptor is further refined. New, better-defined ligands may be synthesized.
In addition to the interatomic distances, the chemical databases will contain important physicochemical values including partition coefficients, electronic terms, molar refrac-tivity, pKa's, solubilities, and steric values. Arrangements of atoms may be coded by molecular connectivity or other topological descriptors. The result is a "flood of data" that requires interpretation, large amounts of data storage, and rapid means of analysis. Compounds usually must fit within defined limits that estimate ADME.
Chemical databases can contain hundreds of thousands of molecules that could be suitable ligands for a receptor. But, no matter how good the fit is to the receptor, the candidate molecule is of no use if the absorption is poor or if the drug is excreted too slowly from the body. An analysis of 2,245 drugs has led to a set of "rules" called the Lipinski
Was this article helpful?