Optimizing phenylethylphosphonamidates for the inhibition of prostate-specific membrane antigen.

A series of eight N-2-phenylethylphosphonyl derivatives of glutamic acid was prepared to determine if the inhibitory potency of a phenylethylphosphonyl derivative of glutamic acid against prostate-specific membrane antigen (PSMA) could be improved through rational substitutions on the phenyl ring. The design of these eight analogs was based upon the Topliss batchwise approach. Of the inhibitors from the first generation, the 3,4-dichlorophenyl analog exhibited the greatest improvement over the lead compound which was an unsubstituted phenyl derivative, while the 4-methoxyphenyl analog was essentially void of inhibitory potency against PSMA in single-dose studies. From the potency ranking order of the first generation, the parameter most important to the pharmacophore was determined to be pi + sigma. Attempts to optimize further the potency of inhibitors by preparing a second generation of compounds did not result in structures with greater potency than that of the 3,4-dichlorophenyl analog from the first generation. Based upon K(i) values, the 3,4-dichlorophenyl analog represented a potency improvement of nearly one order of magnitude. These results confirm further the usefulness of the Topliss approach to analog development when large library synthesis cannot be achieved readily.