Synthesis of urea-based inhibitors as active site probes of glutamate carboxypeptidase II: efficacy as analgesic agents.

The neuropeptidase glutamate carboxypeptidase II (GCPII) hydrolyzes N-acetyl-L-aspartyl-L-glutamate (NAAG) to liberate N-acetylaspartate and glutamate. GCPII was originally cloned as PSMA, an M(r) 100,000 type II transmembrane glycoprotein highly expressed in prostate tissues. PSMA/GCPII is located on the short arm of chromosome 11 and functions as both a folate hydrolase and a neuropeptidase. Inhibition of brain GCPII may have therapeutic potential in the treatment of certain disease states arising from pathologically overactivated glutamate receptors. Recently, we reported that certain urea-based structures act as potent inhibitors of GCPII (J. Med. Chem. 2001, 44, 298). However, many of the potent GCPII inhibitors prepared to date are highly polar compounds and therefore do not readily penetrate the blood-brain barrier. Herein, we elaborate on the synthesis of a series of potent, urea-based GCPII inhibitors from the lead compound 3 and provide assay data for these ligands against human GCPII. Moreover, we provide data revealing the ability of one of these compounds, namely, 8d, to reduce the perception of inflammatory pain. Within the present series, the gamma-tetrazole bearing glutamate isostere 7d is the most potent inhibitor with a K(i) of 0.9 nM. The biological evaluation of these compounds revealed that the active site of GCPII likely comprises two regions, namely, the pharmacophore subpocket and the nonpharmacophore subpocket. The pharmacophore subpocket is very sensitive to structural changes, and thus, it appears important to keep one of the glutamic acid moieties intact to maintain the potency of the GCPII inhibitors. The site encompassing the nonpharmacophore subpocket that binds to glutamate's alpha-carboxyl group is sensitive to structural change, as shown by compounds 6b and 7b. However, the other region of the nonpharmacophore subpocket can accommodate both hydrophobic and hydrophilic groups. Thus, an aromatic ring can be introduced to the inhibitor, as in 8b and 8d, thereby increasing its hydrophobicity and thus potentially its ability to cross the blood-brain barrier. Intrathecally administered 8d significantly reduced pain perception in the formalin model of rat sensory nerve injury. A maximal dose of morphine (10 mg) applied in the same experimental paradigm provided no significant increase in analgesia in comparison to 8d during phase 1 of this pain study and modestly greater analgesia than 8d in phase 2. These urea-based inhibitors of GCPII thus offer a novel approach to pain management.