Opioids : Basic Mechanisms

The Binding Site

Depending on the ligand type, the binding site of GPCRs is either located in extracellular domains (e.g., thrombin receptor) or buried within the seven-helix bundle (small biogenic amine receptors). Alternatively, the binding site can overlap both the external and transmembrane regions, as is the case for peptidic GPCRs, including opioid receptors.

The extracellular loops (e1, e2, and e3) in opioid receptors establish first contact with the ligand approaching the binding site and are important for mu/delta/kappa selectivity. The study of chimeric mu receptors incorporating delta, kappa, or even angiotensin II receptor domains led to the proposal that e1 and e3 are important determinants in the mu receptor for the high-affinity binding of mu-selective compounds. E3 in the delta receptor was dissected by both loss-of-function and binding rescue experiments and was shown to be the most critical site in this receptor for high-affinity binding of delta-selective ligands. In kappa receptors, acidic amino acid residues of e2 seem to represent a unique feature that would favor the binding of basic dynorphin peptides, and e3 also contributes to the recognition of small non-peptidic kappa-selective compounds. Altogether, the extracellular domains of opioid receptors are considered to be both anchoring points for large opioid ligands and gates filtering opioid entry into the binding pocket.

In contrast to extracellular domains, transmembrane (Tm) domains are highly conserved and form an opioid binding pocket that is similar across mu, delta, and kappa receptors. Three-dimensional computer models have highlighted a binding pocket penetrating the upper half of the helical bundle that consists of two subsites: a large hydrophobic domain is formed by aromatic residues spanning Tm3–Tm7, whereas a hydrophilic area lies over Tm3 and Tm7. Amino acid residues were tested by site-directed mutagenesis, and data from studies on the delta receptor, for example, confirmed the implication that the hydrophobic pocket is formed by Y129 (Tm3); W173 (Tm4); F222 (Tm5); W274, I277, I278, H279, and W284 (Tm6); and L300, I304, and Y308 (Tm7). The hydroxyl groups in Y129 (Tm3) and Y308 (Tm7), as well as the carboxyl group of D128 (Tm3), delimit the hydrophilic part of the site, and D128 was proposed to be the counter ion for the universal protonated amine present in every opioid ligand (for discussion, see ). The binding crevice of all three receptors was investigated by cysteine accessibility scanning of Tm6. The outward half of the helix was found to be water accessible in each case, consistent with the notion of a binding pocket penetrating the helical bundle halfway.

Receptor Activation

Several site-directed mutagenesis experiments provided the first hints on activation determinants in opioid receptors. Most interesting are mutations inducing constitutive activation of the receptor (i.e., ligand-independent activity). In the delta receptor, D128 (Tm3) replaced by Q, A, K, or H enhances spontaneous activity of the receptor ( ). Furthermore, the Y308F mutant (Tm7) is also a constitutively activated mutant (CAM) receptor. Three-dimensional modeling indicates a possible hydrogen bond between D128 and Y308, which suggests that a Tm3–Tm7 interhelical interaction could contribute to maintain the receptor in an inactive conformation ( ). Interestingly, the mutation of a conserved S in Tm4 unexpectedly transformed the classic antagonists—in particular, naloxone—into agonists in mu, delta, and kappa receptors, thus suggesting a role of Tm4 in the activation process ( ). These site-directed mutagenesis studies, however, remain limited to agonist binding domains of the receptor.

Recently, a random mutagenesis approach was used in an attempt to visualize the entire activation process without any preconceived model-guided assumption ( ). A collection of about 3000 delta receptor mutants was generated randomly and screened for constitutive activity with a reporter gene–signaling assay. Thirty CAM receptors were isolated and point mutations were found distributed throughout the receptor protein, thus indicating that many receptor domains could contribute to receptor activation. Mutations within the helical bundle, as well as e3, were analyzed on a receptor three-dimensional model. Strikingly, activating mutations clustered in space and revealed an activation path throughout the receptor protein. From this study, a mechanism was proposed in which the opioid ligand would bind to e3 and destabilize Tm6–Tm7 interactions at the extracellular face of the receptor. Entering the binding pocket, the amphiphilic agonist would disrupt the strong hydrophobic and hydrophilic interactions that maintain Tm3–Tm6–Tm7 tightly packed in the inactive receptor. Tm3 would move toward Tm4, whereas Tm6 and Tm7 would separate from each other. This helical movement would propagate to the cytoplasmic face of the receptor and break an ionic lock between Tm6 and Tm7. The resulting structural modifications of i3 and C-terminal domains proximal to the membrane would favor G-protein activation. The latter step is consistent with the previous observation that peptides competing with i3, but not i2, impair delta receptor coupling ( ). Many of the residues found mutated in the random mutagenesis study are conserved in mu, delta, and kappa receptors, as well as in other GPCRs, thus suggesting that this mechanism may apply broadly.

Receptor Signaling

Opioid receptors are coupled to Go/Gi inhibitory proteins, and modulation of many G-protein effectors has been demonstrated in both transfected cells and native tissue (for review, see ). Opioids inhibit voltage-dependent calcium channels or activate inwardly rectifying potassium channels, thereby decreasing neuronal excitability. Opioids also inhibit the cyclic adenosine monophosphate pathway and activate mitogen-activated protein kinase cascades, both activities affecting cytoplasmic events and transcriptional activity of the cell. Finally, crosstalk between the mu opioid receptor and the insulin receptor was recently demonstrated ( ), thus broadening the panel of opioid receptor–associated signaling cascades. Consistent with their highly homologous intracellular loops, all three opioid receptors show similar coupling properties, although some differences in the ratios of mu-, delta-, or kappa-activated Go/Gis have been observed in heterologous expression systems ( ). Opioid receptors also stimulate G protein-independent signaling pathways, notably via β–arrestins ( ), and activate phosphorylation cascades that ultimately modify gene transcription and durably affect cell physiology. Overall opioid receptor activation leads to inhibit neuronal activity ( Fig. 30-1 B) and a main goal in the field of opioid receptor signaling, as for GPCR signaling in general, is the identification of signaling pathways that indeed operate in vivo and control specific behavioral responses and drug effects. Whether this is the case in neuronal networks remains to be clarified.

Research over the last decade has indicated that G protein-coupled receptors exist in multiple conformations and that agonists can stabilize different active states. The distinct receptor conformations induced by ligands result in distinct receptor-effector complexes, which produce varying levels of activation or inhibition of subsequent signaling cascades. As a consequence, the agonist-receptor-effector complex–rather than the receptor itself–is the key determinant for subsequent cellular and in vivo signaling ( Fig. 30-1 B). This concept, referred to ligand-directed signaling or biased agonism, has important biological and therapeutic implications ( ). Many in vitro studies have shown biased agonism at mu, delta, and kappa receptors (see Fig. 30-1 C for mu). Importantly, in vivo consequences of this phenomenon have recently been demonstrated (reviewed in ). For the mu and delta opioid receptors, agonist-specific signaling and trafficking events are observed in response to both acute and repeated drug administration in vivo. The ability of mu agonists to internalize the receptor influences drug analgesic efficacy and tolerance, as well as addictive behaviors ( ). Similarly, internalizing properties of delta agonists strongly influence the development of tolerance, which evolves into either generalized tolerance or pain-specific tolerance ( , Pradhan et al 2011). For the kappa opioid receptor, ligand-specific signaling responses have been proposed to differentially mediate analgesic and dysphoric effects, which potentially are dissociable at the level of downstream effectors ( ).

Overall, biased agonism is one of the several mechanisms by which opioid ligands can produce diverse physiological effects and may underlie the highly complex opioid pharmacological heterogeneity. Future studies will indicate whether the concept of ligand-dependent responses for a given receptor translates into tailor-made pharmacotherapies where advantageous drug effects are selectively targeted over adverse effects. Most important to cell physiology is rapid termination of receptor signaling, and several regulatory processes are known to follow agonist-induced receptor activation. Such processes include phosphorylation of intracellular receptor determinants by a number of protein kinases, binding of arrestin to the phosphorylated domains, uncoupling of receptors from G proteins, rapid receptor endocytosis, and receptor recycling or down-regulation. All these phenomena lead to desensitization of receptor signaling. Receptor truncation or point mutagenesis experiments in opioid receptors have demonstrated the important role of the C-terminal tail in all the regulatory events (for delta receptor, see ). The correlation between these distinct events has been further investigated, particularly in mu receptor mutants (for review, see ), and data indicate that phosphorylation is not obligatory for internalization or that down-regulation does not necessarily correlate with desensitization. Therefore, many distinct molecular mechanisms concomitantly contribute to the modulation of opioid receptor activities, all of which involve interaction of intracellular receptor domains with specific cytoplasmic proteins. These clearly differ from cell to cell and, for a large part, remain to be discovered ( ). Obviously, the large differences in the C-terminal structures of opioid receptors should lead to distinct mu, delta, and kappa receptor physiology despite their similar binding and transduction abilities ( Fig. 30-1 C). A good example is the observation that after agonist-induced internalization, mu receptors efficiently recycle to the cell surface, whereas delta receptors are committed to lysosomal degradation, two distinct endocytic fates that can be modified by C-terminal swapping ( ).