The chemical basis of pharmacology

Abstract

Molecular biology now dominates pharmacology so thoroughly that it is difficult to recall that only a generation ago the field was very different. To understand drug action today, we characterize the targets through which they act and new drug leads are discovered on the basis of target structure and function. Until the mid-1980s the information often flowed in reverse: investigators began with organic molecules and sought targets, relating receptors not by sequence or structure but by their ligands. Recently, investigators have returned to this chemical view of biology, bringing to it systematic and quantitative methods of relating targets by their ligands. This has allowed the discovery of new targets for established drugs, suggested the bases for their side effects, and predicted the molecular targets underlying phenotypic screens. The bases for these new methods, some of their successes and liabilities, and new opportunities for their use are described.

Figure 1

Information flow in molecular and classical pharmacology. (a) Central dogma of molecular biology and its sequelae in protein folding and protein function, illustrated through the structure of and ligand recognition by the β2-adrenergic receptor (63). (b) Ligand-to-target identification in classical pharmacology, illustrated by the classification of receptor subtypes for the β-adrenergic receptors. The differential activity of epinephrine, norepinephrine, and isoproterenol (1) on organ systems disentangled the α-adrenergic from the β-adrenergic receptors; the β-blocker propranolol was specific for β vs α receptors, and subsequently, atenolol and salbutamol were specific for the β1 and β2 subtypes, respectively.

Figure 2

Specific, receptor-classifying molecules could lead to drugs. In addition to the β-adrenergic acting drugs illustrated in Figure 1, others include (a) Buramide, the compound used to distinguish the gut-active H2 receptor from the H1 receptor, and Cimetidine, the anti-ulcer drug to which it led. (b) Tropisetron and Bemesetron defined the 5-HT3 subtype because of their specificity for it over the previously characterized 5-HT1 and 5-HT2 receptors. Tropisetron is an anti-nausea drug used after chemotherapy.

Figure 3

Receptors with high degrees of sequence similarity but little ligand similarity, and the converse. (a) Overall comparison of ligand similarity with sequence similarity for drug targets. Approximately 250 drug targets from the MDDR were compared against each other in a full matrix, first by a ligand similarity method [SEA (22)] and then by a protein sequence similarity method [PSI-BLAST (64)]. Where both methods agree, the matrix is white. Thus, both find that any given target pair on the diagonal, such as 5-HT2A vs itself, resembles itself. Where ligand similarity was stronger than sequence similarity the matrix is red; where the converse is true, it is dark gray. (b) Excerpt of the matrix in which the degree of ligand similarity is high but the degree of sequence similarity is low. This region includes enzymes and nuclear hormone receptors. (c) Except from the region with a high degree of sequence similarity (but a low degree of ligand similarity). These are often GPCRs. Reproduced from ref (22). Copyright 2007 Nature Publishing Group.

Figure 4

Representing molecules as topological fingerprints. (a) Encoding a molecule using Daylight fingerprints. Each atom-to-atom path across the molecule of increasing length is iteratively encoded as a bit string, and all of the bit strings are combined together into a final “fingerprint”. (b) Comparing fingerprints using a Tanimoto coefficient (Tc). The Tc calculates the number of on bits in common between the fingerprints divided by the total number of nonoverlapping on bits between fingerprints.

Figure 5

Varied approaches to organizing drug protein targets by their ligands. (a) Drug−target network linking FDA drugs (circles) to targets (rectangles) based on the known associations. Drugs are colored by their Anatomical Therapeutic Chemical Classification and target proteins by their Gene Ontology cellular component (24). (b) Target−target network, in which targets are linked if they bind one or more compound in common, within a preset affinity threshold. There are 486 targets, colored by gene family, linked by 3636 edges (23). (c) Predicted drug−target network. Each drug (gold) is linked to its known protein targets (cyan) by a gray edge. Red edges link drugs to their additionally predicted targets (41).