Viewpoint Cite This: J. Med. Chem. XXXX, XXX, XXX−XXX
pubs.acs.org/jmc
A Surprising Recipe for Designing Biased Ligands Sid Topiol* 3D-2Drug, LLC, P.O. Box 184, Fair Lawn, New Jersey 07410, United States ABSTRACT: The determination of the potential value of receptor trafficking at melanocortin receptors has been hampered by the absence of known biased ligands. Heterobivalent MC4R ligands linking agonist to antagonist small peptidic moieties were designed and found to act as Gαs enhancers while minimally activating β-arrestin recruitment. The strategy invoked offers intriguing promise as a surprising approach that is possibly broadly applicable to the challenge of designing biased ligands at other GPCRs.
T
arrestin recruitment. Biased ligands have been reported for other targets with resultant improved activity profiles. However, no biased ligands had been available to explore their possible advantages for the MCRs. The need to identify biased MC4R compounds as tools to evaluate their therapeutic potential as well as subsequent drug discovery is therefore compelling. Faced with the general challenges for generating biased ligands described above and the goal of doing so for the MC4 receptor, the report by Lensing et al.3 in this issue describes a novel approach. Specifically, the work presents a hypothesis that creating a bivalent ligand by linking together an agonist and antagonist for MC4R would yield compounds with increased Gαs and diminished β-arrestin activities. Although at first blush, this sounds almost obvious, upon some reflection, it clearly defies the general model for biased ligand activity as outlined above, wherein a single pharmacophore acts on a receptor to tune its conformation to selectively interact with one effector. Rather, the hypothesis assumes that the two pharmacophores of the bivalent ligand occupy separate monomers of a dimer (and/or higher-order multimers) of the receptor. While this may be a more palatable model of the binding mode(s) of such compounds and is certainly plausible within the knowledge base that exists regarding the propensity for GPCR dimerization, it is not obvious that the net effect of such binding would result in the desired biased signaling. The study builds on earlier reports by the same authors and conceptually similar studies by others, wherein both moieties of bivalent MC4R ligands comprised agonist pharmacophores. Dimerization of GPCRs is well-established as is the recognition that such dimerization can result in allosteric interactions between the monomers leading to enhanced or reduced ligandmediated activity. As before, the monomers used are small peptidic pharmacophores coupled through inert linkers. Unlike their earlier work linking only agonist pairs or antagonist pairs of peptidic pharmacophores, the authors now link agonist peptides (e.g., His-DPhe-Arg-Trp) with antagonist peptides (e.g., His-DNal(2′)-Arg-Trp) via inert linkers (e.g., PEDG20). These heterobivalent dimers are termed unmatched bivalent
he design of GPCR-receptor-mediating small molecules with optimal therapeutic value has become an increasingly more complex task because of our enhanced understanding of the finer details of receptor activity and the consequences of the differences in these details. Beyond compounds acting simply as on/off switches resulting in agonism or antagonism, compounds may have more complex behaviors such as partial agonism, allosteric modulation, and biased activity. These receptor activity characteristics often translate into therapeutic differences in compound activity such as maintaining endogenous tone with positive allosteric modulators or a more tailored downstream effector path via biased compounds leading to potential reductions in side effects.1,2 Our understanding of the atomic-level details of the molecular mechanisms of GPCR activity has advanced significantly over the past 10 years, largely fueled by the growing number of X-ray structures of GPCR protein complexes. These reports have demonstrably facilitated our ability to rationally develop compounds with higher affinity and, with many recent reports of newly identified binding sites, have opened opportunities for more effective design of allosteric modulators. In contrast, despite the increasingly more appreciated need to impart receptor bias on newly design compounds, rational design thereof remains particularly difficult, albeit more elaborate screening efforts for determining ligand bias have grown significantly. In the usual model, ligands binding at the orthosteric region of a GPCR control its differential interaction at its intracellular region with different effectors (Figure 1A). Within this paradigm, our structural understanding of this area has generally been limited to efforts at rationalizing the binding differences between differently biased ligands as well as investigations of protein structural and dynamics processes which may be involved in receptor trafficking. For melanocortin receptors, the challenge to find biased ligands is further exacerbated by the absence of corresponding GPCR structures. The melanocortin system has been implicated in multiple therapeutic areas including oncology, various CNS disorders, and metabolic diseases. For the melanocortin 4 (MC4) receptor, compounds have been identified and shown to have clinically favorable metabolic effects but undesirable side effects. The five melanocortin receptor subtypes (MC1−5R) signal through the Gαs signaling pathway. MC4R agonists signaling through this pathway activate cAMP production and trigger β© XXXX American Chemical Society
Special Issue: Allosteric Modulators Received: May 14, 2018
A
DOI: 10.1021/acs.jmedchem.8b00764 J. Med. Chem. XXXX, XXX, XXX−XXX
Journal of Medicinal Chemistry
Viewpoint
Figure 1. Schematic comparison of models for biased agonism. (A) Common model for biased ligand action wherein a ligand binds to a monomeric unit of a GPCR such as MC4R and thereby induces the conformational changes in the GPCR to signal increased interaction with one effector (Gαs) and diminished interaction with another effector (arrestin). (B) Proposed model3 for biased ligand action wherein each of the two pharmacophores of a heterobivalent ligand interacts with a separate monomer of a GPCR homodimer such that the interactions of each pharmacophore with its respective GPCR monomer and/or the interactions between the GPCR monomers induce the conformational changes in the GPCR dimers to signal increased interactions with one effector (Gαs) and diminished interactions with another effector (arrestin).
recipe to develop other, more advanced compounds that need not be peptide-based. The plausibility of a more favorable in vivo profile of activity is well founded. More generally the underlying strategy, and likely operative mechanism, can be broadly applied to any GPCR target. This would shift the identification of biased compounds to an up-front design process and away from an after-the-fact testing process. Any advantages of such ligand bias, if successfully achieved, would have to be determined for each target. In addition, the underlying principles here are of interest in their own right. Allosteric modulators for GPCRs are generally operative via actions on alternative locations within a given GPCR monomer rather than by acting on another (GPCR) protein. Conceptually, inhibitors of transporters of GPCR ligands have similar consequences to those of allosteric modulators of their corresponding GPCRs. While these “allosteric” effects are also based on ligands acting on a separate protein (the transporter), that protein is not directly interacting with the GPCR. The biased activity of the compounds reported by Lensing et al.3 in this issue act via the allosteric cross talk and resulting conformational changes in the GPCR monomers that interact directly. This interesting model involving different effectors interacting (favorably or unfavorably) with each of the monomers of a GPCR homodimer adds to the potential complexity of the many stoichiometry relationships that have been described for GPCR/effector systems.4 Many variants on the model(s) discussed in the report by Lensing et al.3 (higherorder multimers, etc.) and/or other models may need to be considered. While the veracity of the model for the underlying mechanism may be questioned, the heterobivalent-ligand approach implemented in the cited report has succeeded in generating the first biased MC4R ligands, is conceptually simple to implement, and can be expected to have broad applicability.
ligands (UmBLs) with analogous nomenclature for melanocortin UmBLs (MUmBLs), biased UmBLs (BUmBLs), etc. The results can be illustrated as follows. cAMP signaling in live HEK293 cells stably expressing human (h)MC4R was assessed via ALPHAScreen cAMP assay technology. Homobivalent agonists as well as heterobivalent agonist + antagonist ligands all had low nanomolar activity, whereas homobivalent antagonists had minimal activity. Equal mixtures of monovalent agonists and antagonists also had low nanomolar activity. βArrestin recruitment was evaluated using the PRESTO-Tango assay. Homobivalent agonists exhibited high signals, whereas homobivalent antagonists as well as heterobivalent agonist + antagonist ligands exhibited minimal β-arrestin recruitment. BRET studies that had previously shown the dimerization of (h)MC4R indicate conformational changes in the dimers that are largest for the homobivalent agonist ligands, intermediate for the heterobivalent ligands, and least for the homobivalentantagonist ligands. Several possible explanations for the nature of these conformational changes are suggested by the authors, with the proposed and perhaps most intriguing and plausible one being that which assumes that the agonist pharmacophore of a heterobivalent UmBL binds to one GPCR monomer of a dimer that is in a (Gαs) active conformation, while the antagonist pharmacophore binds to the second GPCR monomer that is in a (β-arrestin) inactive mode (Figure 1B). This could be achieved by first binding the agonist pharmacophore to one MC4R monomer in the active state, which then allosterically induces the inactive state of the second GPCR monomer to which the antagonist pharmacophore binds. (Other binding sequences yield the same net result.) These results have profound significance for MC4R (and probably other MCRs) therapeutics specifically, as well as for the design and discovery of biased ligands for GPCRs more generally. For MC4R, these findings provide the first tool compounds with Gαs vs β-arrestin bias with which the exploration of therapeutic implications can begin, as well as a B
DOI: 10.1021/acs.jmedchem.8b00764 J. Med. Chem. XXXX, XXX, XXX−XXX
Journal of Medicinal Chemistry
■
Viewpoint
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Sid Topiol: 0000-0002-2574-4204
■
REFERENCES
(1) Rockwell, K. L.; Alt, A. Positive Allosteric Modulators of Opioid Receptors. In Allosterism in Drug Discovery; RSC Drug Discovery Series 56; Doller, D, Ed.; The Royal Society of Chemistry: Cambridge, U.K., 2017; pp 194−219. (2) Manglik, A.; Lin, H.; Aryal, D. K.; McCorvy, J. D.; Dengler, D.; Corder, G.; Levit, A.; Kling, R. C.; Bernat, V.; Hübner, H.; Huang, X.P.; Sassano, M. F.; Giguère, P. M.; Löber, S.; Duan, D.; Scherrer, G.; Kobilka, B. K.; Gmeiner, P.; Roth, B. L.; Shoichet, B. K. Structurebased discovery of opioid analgesics with reduced side effects. Nature 2016, 537, 185−190. (3) Lensing, C. J.; Freeman, K. T.; Schnell, S. M.; Speth, R. C.; Zarth, A. T.; Haskell-Luevano, C. Developing a Biased Unmatched Bivalent Ligand (BUmBL) Design Strategy to Target the GPCR Homodimer Allosteric Signaling (cAMP over β-Arrestin 2 Recruitment) Within the Melanocortin Receptors. J. Med. Chem. 2018, DOI: 10.1021/ acs.jmedchem.8b00238. (4) Gurevich, V. V.; Gurevech, E. V. GPCRs and Signal Transducers: Interaction Stoichiometry. Trends Pharmacol. Sci. 2018, DOI: 10.1016/ j.tips.2018.04.002.
C
DOI: 10.1021/acs.jmedchem.8b00764 J. Med. Chem. XXXX, XXX, XXX−XXX
