Design and models for estimating antagonist potency (pA2, Kd and IC50) following the detection of antagonism observed in the presence of intrinsic activity

The antagonist activity of the metabotropic glutamate receptor ligand (2S,1'S,2'S)-2-methyl-2(carboxycyclopropyl)glycine (MCCG) was examined using the [35S]GTPgammaS binding and forskolin (FSK)-stimulated adenosine 3':5'-cyclic monophosphate (cAMP) assays with recombinant Chinese hamster ovary (CHO) cells expressing the G protein-coupled human subtype 2 metabotropic glutamate (hmGlu2) receptor. Whereas MCCG proved to be a partial agonist in the GTPgammaS binding assay, it not only antagonized the agonist effect of (IS,3R)-ACPD in the cAMP assay but further produced an anomalous increase of the cAMP level relative to baseline. The anomalous MCCG response was also observed following treatment of the cells with MCCG in the absence of added agonist. Determination of the glutamate concentration in the incubate at the start and end of the cAMP reaction revealed the existence of micromolar concentrations of cellularly released glutamate throughout the course of the assay, reaching levels which exceeded its reported affinity for the mGlu2 receptor. Considering MCCG's partial agonist effect in the GTPgammaS binding assay and its pseudo-inverse agonist effect in the cAMP assay, available methods of estimating its antagonist potency were inappropriate since the classical Schild method and the alternative model suggested by Waud both assume the antagonist to lack a concentration-response relationship. We derived an alternate design and models that permit estimation of the pA2 (pAx), Kd and IC50 for antagonists which produce a concentration related effect when applied by themselves. With their use, the data acquired in both assays support the designation of MCCG as a competitive antagonist of the hmGlu2 receptor and provide similar pA2 estimates between assays. In addition, the newly derived models and design permit the determination of antagonist potency for partial and inverse agonists so characterized in studies employing the Schild design.