Seven transmembrane receptors as shapeshifting proteins: the impact of allosteric modulation and functional selectivity on new drug discovery

Abstract

It is useful to consider seven transmembrane receptors (7TMRs) as disordered proteins able to allosterically respond to a number of binding partners. Considering 7TMRs as allosteric systems, affinity and efficacy can be thought of in terms of energy flow between a modulator, conduit (the receptor protein), and a number of guests. These guests can be other molecules, receptors, membrane-bound proteins, or signaling proteins in the cytosol. These vectorial flows of energy can yield standard canonical guest allostery (allosteric modification of drug effect), effects along the plane of the cell membrane (receptor oligomerization), or effects directed into the cytosol (differential signaling as functional selectivity). This review discusses these apparently diverse pharmacological effects in terms of molecular dynamics and protein ensemble theory, which tends to unify 7TMR behavior toward cells. Special consideration will be given to functional selectivity (biased agonism and biased antagonism) in terms of mechanism of action and potential therapeutic application. The explosion of technology that has enabled observation of diverse 7TMR behavior has also shown how drugs can have multiple (pluridimensional) efficacies and how this can cause paradoxical drug classification and nomenclatures.

Fig. 1.

Typical topology of 7TMRs. Shown is a representative structure of a family A 7TMR with a peptide ligand. The extracellular ectodomain is shown in the top of the lateral views, whereas the top views provide appreciation for the helical bundle domain. The cytosolic region is shown at the bottom of the lateral views. The transmembrane segments are colored blue to red from amino terminus toward carboxyl terminus of the receptor. Shown at the carboxyl-terminal end of TM7 is the extension into another helix 8 that lies adjacent to the cytosolic face of the lipid bilayer.

Fig. 2.

Allosteric systems. Energy is transmitted between loci of binding for the modulator and that of the guest. Energy transfer is reciprocal (Tränkle et al., 1999); therefore, the role of modulator and guest are operationally interchangeable. For an extracellular modulator binding site, the guest binding site can be extracellular, within the plane of the membrane, or facing the cytosol.

Fig. 3.

Two proposed modes of allostery. The allosteric “hot wire” proposes a preferred energy link between an allosteric binding site and the guest site; in the past, this has been an assumed mechanism. Global allosteric modulation predicts that changes at the guest allosteric site are part of global conformational variations within an ensemble of conformations.

Fig. 4.

Classification of allostery on the basis of number of sites. A, multiple site allostery involves the interaction of more than one guest with the modulator. Probe dependence results in different effects on each guest for a given modulator. The example shown shows the effect of the CCR5 modulator aplaviroc on CCL5 (no effect on binding) and HIV-1 (complete inhibition of binding) (Watson et al., 2005). B, probe dependence can also result from the interaction of different modulators at the same modulator binding site for the same guest. The different abilities of alkyltrimethylammonium compounds to initiate contraction of guinea pig ileum as mediated by G protein binding is a classic example (Stephenson, 1956). Both the modulator(s) and G proteins bind at a single site.

Fig. 5.

The vectorial nature of 7TMR allostery. Modulators can affect binding and function of other ligands cobinding to the receptor (classic guest allostery); of the cobinding of other proteins along the plane of the membrane (other receptors or membrane-binding proteins such as RAMPs); or of cytosolic signaling proteins, such as G proteins or β-arrestin. These effects can be simultaneous and all emanate from the same mechanism(s) of allosteric change within the conduit 7TMR protein.

Fig. 6.

Basic model for allosteric interaction of a modulator (A or B) interacting with a receptor (R) to affect the interaction of a guest (counterpart A or B). If A is an agonist, then agonism can emanate from the species AR and/or the allosterically modified species ARB. The modulator would then be B, which can affect the affinity of the agonist for the receptor (through α) or the efficacy of the receptor (relative efficacy of the species ARB/AR described as β). This can result in a range of new responses to ligand A from potentiation to antagonism. The binding species are from the model given by Ehlert (1988); the response components can be added through melding of that model with the operational model for receptor function (Black and Leff, 1983) to yield a general functional model of allosteric 7TMR function (Ehlert, 2005; Kenakin, 2005; Price et al., 2005).

Fig. 7.

Two types of allosteric system derived from receptor oligomers. A, the receptor dimer becomes a completely new conduit and ligands interact through this new conduit with revised interactive behaviors. For example, the dopamine antagonists quinpirole and sulpiride bind to a somatostatin SSTR5-dopamine receptor D2 heterodimer to block the binding of somatostatin (Rocheville et al., 2000). B, another receptor (or membrane bound protein such as RAMP) becomes the modulator that confers new reactivity toward a guest through a receptor. For example, the bradykinin-2 receptor can act as a modulator and bind to the angiotensin1 receptor to change the binding characteristics of angiotensin-1 (AbdAlla et al., 2000).

Fig. 8.

Two contrasting mechanisms of receptor antagonism. A, orthosteric antagonists block agonist activation of the receptor and all functions mediated by that activation are uniformly inhibited. For example, prostaglandin D₂ is the endogenous agonist for CRTH2 receptors causing activation of Gα_s protein and β-arrestin. An orthosteric antagonist blocking PDG₂ binding to the receptor would block both pathways. B, the allosteric modulator N-α-tosyltryptophan (N-α-T) blocks the Gα_s protein activation of CRTH2 through PDG₂ but still allows the agonist to active β-arrestin (Mathiesen et al., 2005).

Fig. 9.

Differences in relative agonist activity with changes in cellular background. A, calcium transient response for activation of human calcitonin receptors transfected into HEK293 cells by eel (●), porcine (○), and rat (▴) calcitonin and rat amylin (▵). B, responses to the same agonists in HEK293 cells transfected with Gα_s protein. Note that the relative potencies of the agonists change. The table shows the activity of the agonists in terms of their log transduction ratios [Log (τ/K_A)] values in the two cell hosts and their relative activity in terms of the reference agonist amylin [ΔLog(τ/K_A) values]. A measure of the relative effect of Gα_s-transfection on each agonist is given as the bias. Data from Watson et al. (2000).

Fig. 10.

Biased agonism at the angiotensin II receptor. The natural ligand angiotensin II produces activation of G proteins, which can produce a deleterious pressor response in hypertension and heart failure and also stimulates β-arrestin to initiate ERK, Akt, and phosphatidylinositol 3-kinase signals, which could be cytoprotective. The biased agonist SII produces only activation of β-arrestin to induce cytoprotection but not pressor response. Moreover, receptor occupancy of the receptor by SII prevents activation by endogenous angiotensin (the receptor is blocked), thereby further protecting against endogenous pressor response. Data from Violin and Lefkowitz (2007).