Fourth Smissman Award Address. The long search for valid structure-action relationships in drugs

In 1869, Crum Brown discovered the first structure-activity link by showing that alkaloids, even convulsive ones, were converted by N-methylation to muscle relaxants resembling curarine (itself a quaternary amine). This led to an attempt to link every type of drug action to its own cluster of atoms. This quest was jolted when Loewi (1926) found that a quaternary amine (acetylcholine) was the principal activator of muscle! Suddenly it was seen that a chemical group could be either an agonist or antagonist, depending on its molecular setting. That the agonists were smaller molecules suggested operation of a steric factor. Moreover, Cushny (1926) had focused attention on optical enantiomers: usually only one member of each pair had biological activity, although both were identical in all other properties. The stage was now set for physical properties to play the leading role in relating structure to activity. People recalled the demonstration by Overton and Meyer (1900) that the depressant action of a drug was linked to its lipophilicity. Unhappily, further physical correlations were slow to appear. My colleagues and I, who had been studying (from 1941) the antimicrobial action of aminoacridines and hydroxyquinolines, established quantitatively the role of ionization and chelation (two electronic influences) in the action of drugs. Today most people would agree that the most important properties in determining the action of drugs are not some particular nucleus or substituent but a trio of physical properties: lipophilicity, electron distribution, and shape. Although these properties govern the activity of drugs, their selectivity is due, as I have long maintained, to another trio of properties, namely, comparative distribution (not necessarily lipophilic), comparative biochemistry, and comparative cytology.