Protean behavior by agonists: agonist- directed trafficking of receptor signaling Receptor-binding drugs can behave as agonists (full or partial), neutral antagonists or inverse agonists, depending on how they modulate receptor function. In a 1995 TiPS article [1] Terry Kenakin proposed an explanation for the observed protean behavior of certain agonists that act as partial agonists in some systems and as inverse agonists in other systems, both being mediated via the same receptor. He based his explanation on the extended ternary complex model, which includes multiple receptor activation states (R refers to the inactive receptor state and R* refers to the active receptor state). If it is assumed that the unbound and drug-bound active receptor states (R* and AR*, respectively) are different entities and that they can couple with different affinity to G proteins, it is theoretically possible that R* couples stronger to the G protein than does AR*. In the latter case, the drug will act as a partial agonist in systems with a high R:R* ratio (little or no constitutive activity) and as an inverse agonist in systems with a lower R:R* ratio (significant constitutive activity). In fact, many different experimentally observed manifestations of agonism can be understood mechanistically in terms of binding to an array of active and inactive receptor conformations, each interacting differentially with G proteins [2,3]. Here, I would like to discuss briefly a special case of protean agonism. Recently, Pauwels and co-workers [4] proposed that the differences in G-protein content between cell types might also account for protean behavior of agonists. They showed that in transfected Chinese hamster ovary (CHO)-K1 cells, the α 2A -adrenoceptor ligand RX831003 acts as a partial agonist via a G α15 effector pathway (determined by measuring a Ca 2+ response) and also acts as an inverse agonist {determined by measuring [ 35 S]GTPγS binding to G αo Cys 351 Ile}, which suggests that RX831003 is a protean agonist at the α 2A -adrenoceptor. However, Pauwels and colleagues also suggested that it might be difficult to distinguish protean agonism from agonist-directed trafficking of receptor signaling (ADTRS), as proposed by Kenakin [5], when more than one G protein is involved to reveal protean behavior of the agonist. Importantly, examples of ADTRS have been reported for the α 2A -adrenoceptor [6,7] and other receptor systems such as the 5-HT 2A and 5-HT 2C receptors [8]. If one now assumes a three-state model of receptor activation [9] and in particular the extended version incorporating the ternary complex model [10], with R the inactive receptor conformation and R* and R** the active receptor conformations that couple to two distinct G-protein subtypes G 1 and G 2 , respectively (Fig. 1), then agonist-directed trafficking of receptor signaling will occur if, for example, agonist A preferentially binds R* or R**, thereby enriching one of the active conformations as AR* or AR**, respectively. Therefore, when one considers a case where a ligand binds R* ≥ R** > R (in order of affinity for receptor states), the drug will be an agonist via an isolated G 1 - or G 2 -mediated effector pathway. However, in an intact system containing both G 1 and G 2 the effector pathways are interdependent and a ligand driving the distribution strongly towards R* might also deplete R**, resulting in inverse agonism via the R** pathway [10]. To my knowledge a special scenario where the drug binds R* > R > R** (in order of affinity for receptor states) has not been discussed before. The drug will act as an agonist when measuring a response via G 1 (R*–AR* mediated) and as an inverse agonist when measuring a response via G 2 (R**–AR** mediated). This observation should be independent of whether the system containing G 1 and G 2 is intact or the G 1 and G 2 pathways are isolated. This would represent an example of where the protean behavior of an agonist results from ADTRS, which theoretically can also explain the results found by Pauwels and co-workers [4]. In conclusion, protean behavior by agonists, when measuring response via two distinct G proteins, can also be explained by the concept of ADTRS. Whereas the explanation of protean agonism by Kenakin [1] predicts that it is dependent on the magnitude of the equilibrium constant between the R and R* conformations in addition to the receptor:G-protein ratio, the current ADTRS explanation does not predict this dependency. Also the ADTRS explanation does not assume (or preclude) that R* and AR* couple differently to G proteins. ADTRS does, however, assume that different active receptor states (e.g. R* and R**) couple differently to distinct G-protein types and also that the drug binds differently to these respective active receptor states. It is important to note that ADTRS cannot explain protean behavior of an agonist when only one G protein is involved, in which case only the explanation of protean agonism by Kenakin [1] remains valid. Christiaan B. Brink Division of Pharmacology, Potchefstroom University for Christian Higher Education, P/B X6001, Potchefstroom, 2520, South Africa. e-mail: fklcbb@puknet.puk.ac.za TRENDS in Pharmacological Sciences Vol.23 No.10 October 2002 http://tips.trends.com 0165-6147/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0165-6147(02)02079-5 454 News & Comment Letters TRENDS in Pharmacological Sciences E 1 E 2 K 1 K G1(A) K G1 A + R*G 1 AR*G 1 A + R* AR* G 1 G 1 G 2 G 2 + + + + K 2 A + R**G 2 AR**G 2 K A * K A K G2(A) K G2 L (A) L A + R AR K A ** M (A) M A + R** AR** Fig. 1. Schematic representation of an extended three- state receptor model [10], where agonist A binds differentially to the inactive receptor conformation R and active receptor conformations R* and R**, and the distribution between the different states are determined by the equilibrium constants K 1 , K A *, K A , K A **, K 2 , K G1 , L, M, K G2 , K G1(A) , L (A) , M (A) and K G2(A) . Preferential binding of A to R* and/or R** (preferential to binding to R) enriches active receptor states and thereby promotes coupling to and activation of G proteins G 1 and/or G 2 , respectively, leading to corresponding responses E 1 and E 2 . Protean behavior of agonists can be explained by assuming that the agonist binds R* > R > R** (in order of affinity for receptor states).