|Home | About | Journals | Submit | Contact Us | Français|
Opioid receptors, MORP, DORP and KORP, belong to the family A of G protein coupled receptors (GPCR), and have been found to modulate a large number of physiological functions, including mood, stress, appetite, nociception and immune responses. Exogenously applied opioid alkaloids produce analgesia, hedonia and addiction. Addiction is linked to alterations in function and responsiveness of all three opioid receptors in the brain. Over the last few years, a large number of studies identified protein-protein interactions that play an essential role in opioid receptor function and responsiveness. Here, we summarize interactions shown to affect receptor biogenesis and trafficking, as well as those affecting signal transduction events following receptor activation. This article also examines protein interactions modulating the rate of receptor endocytosis and degradation, events that play a major role in opiate analgesia. Like several other GPCRs, opioid receptors may form homo or heterodimers. The last part of this review summarizes recent knowledge on proteins known to affect opioid receptor dimerization.
Drug addiction is a complex disease, resulting from repeated exposure to a number of substances, including opiates, psychostimulants, nicotine and alcohol (1–3). Addiction involves several brain circuits, neurotransmitter systems, ion channels, and signal transduction molecules. For example, the spinal cord and periaqueductal grey neurons control analgesic responses, AMPA and dopamine receptors in mesocorticolimbic sites play a critical role in drug reward and craving, several nuclei in the amygdala modulate reward and relapse, while the noradrenergic nuclei of the locus coeruleus are highly activated during opiate withdrawal (1,4–6). Considerable effort has been put towards understanding the neuroanatomical and neurochemical mechanisms of addiction however the cell specific events that lead to addiction are not fully understood.
Among the most abused drugs, the opioid alkaloids heroin and morphine, are known to cause euphoria, dependence, analgesia and respiratory depression7). The use of opioid alkaloids as analgesics in the treatment of chronic pain is often limited by the serious side effects and the development of analgesic tolerance. Endogenous opiate peptides modulate mood, appetite and pain, by activating particular G protein coupled receptors, namely, mu (MOPR), delta (DOPR), and kappa (KOPR) (7). These receptors differ in tissue distribution and selectivity for endogenous ligands, and could differentially modulate the physiologic effects of opioid receptor activation (7). For example, the rewarding and analgesic effects of opioid alkaloids, are mostly mediated via MOPR, although DOPR and KOPR are also implicated in several opiate actions. Over the past decade, a number of in vitro and in vivo studies pointed to the importance of adaptive changes downstream of the receptors in the long-term effects of opioids (reviewed in 8). These adaptations may involve a switch in receptor signal transduction pathways, changes in receptor trafficking and recycling as well as changes in transcriptional activity. Emerging evidence suggests that dysfunctions of particular signal transduction molecules or molecules necessary for receptor localization and deactivation, are often the underlying causes of addictive behaviors (9). These same signal transduction molecules may offer new therapeutic targets for addiction and analgesia. In this review, we summarize recent reports on proteins that appear to modulate opiate receptor responsiveness via actions in receptor biosynthesis, localization, trafficking, signaling and deactivation (Tables 1–3).
While substantial effort has been put towards exploring the mechanisms modulating G protein-coupled receptor endocytosis and degradation (reviewed later in this article) relatively less effort has been put towards examining the mechanisms modulating receptor biogenesis and cell surface expression. Among the opioid receptor types, the plasma membrane expression of DOPR has been extensively studied. In recombinant systems, only a fraction of the newly synthesized receptor is ultimately transported to the cell surface (10), the rest being retained in the endoplasmic reticulum (ER) from where it is retrotranslocated to the cytosol, deglycosylated, ubiquitinated, and degraded by the proteasome (11). In the ER, DOPR interacts with the chaperone protein, calnexin (12). This interaction can be regulated by membrane-permeable opioid ligands that bind to the newly synthesized receptor (in the ER) causing it to dissociate from calnexin (12). This facilitates DOPR transport out of the ER and expression at the plasma membrane which, in turn, decreases the proportion of receptors that would otherwise be targeted for degradation (13). These compounds, that facilitate proper folding of the receptor and their escape from the ER quality control system, are called “pharmacological chaperones”.
In dorsal root ganglia and spinal cord neurons, DOPR is often found associated with substance P/calcitonin gene-related peptide (CGRP) positive neurons more specifically, in the regulated secretory pathway where it is sorted into large dense-core vesicles (14, 15). It is thought that in unstimulated axons, DOPR is principally retained in intracellular vesicular compartments. Upon stimulation, DOPR becomes incorporated into the membrane during exocytotic release of transmitters or peptide modulators (16). More recently, the mechanism underlying the sorting/retention of DOPR in the regulated secretory pathway has been uncovered (17). The 3rd luminal domain of DOPR interacts with protachykinin, the precursor molecule of the substance P neuropeptide, and this interaction is responsible for sorting DOPR to the large dense-core vesicles. Following stimulation, substance P is released from the central terminals of sensory neurons and this process leads to cell surface expression of DOPR (17). Compared to the proteins involved in the maturation of DOPR, relatively little is known about the maturation of other opioid receptors or their heterodimers. A study examining the ability of opiates to regulate the maturation of mutant MOPR that gets sequestered in an intracellular compartment (and degraded) supported the notion that opiates can function as pharmacological chaperones of MOPR. These studies suggest that similar to DOPR, MOPR may also associate with endogenous chaperones and this would help in receptor maturation and cell surface expression (18). In the case of KOPR, studies show that interaction with GEC-1, a member of the microtubule associated protein family, leads to an increase in KOPR expressed at the cell surface (19). GEC-1 interacts directly with the C-tail of KOPR in the ER/Golgi and this interaction could involve N-ethylmaleimide sensitive factor (NSF) regulated vesicle fusion dynamics (19). In addition, it has been shown that some opioid ligands like naloxone or naltrexone function as pharmacological chaperones in the ER promoting KOPR folding and maturation and enhancing cell surface expression of the receptor (20,21). In the case of opioid receptor heterodimers, it has been proposed that the MOR-DOR heterodimer could interact with Galphaz in the ER early after biosynthesis (22). However, a role for this interaction in receptor maturation has not been examined.
A number of studies have shown that following receptor activation the dynamics of receptor signaling can be modulated either by interaction with proteins regulating G-protein mediated signaling or by promoting receptor desensitization/resensitization. In the following sections we will describe some of the proteins that have been shown to modulate opioid receptor signaling.
RGS (Regulators of G protein signaling) proteins are among the most important modulators of receptor signaling (23,24). RGS proteins were first identified in yeast and nematode worms (25,26). Evidence for a role of RGS-like proteins in regulation of GPCR signaling came from studies on the sst2 and egl-10 genes (26–28). These studies revealed that RGS proteins act on GTP-bound Galphai and/or Galphaq proteins via a 120 amino acid domain (RGS domain), and regulate the duration of signaling by increasing the speed of GTP hydrolysis (23,24). The mammalian RGS family consists of about forty proteins categorized into nine subfamilies (23, 24). Beyond the common RGS domain, members of the RGS family vary in length, structure, regional and cellular localization (23,24). Growing evidence indicates that RGS proteins may have a more complex function, which involves interactions with scaffolding proteins, G protein subunits, and effector molecules (28).
Evidence for a role of RGS proteins in modulation of MOPR function comes from a number of in vitro and in vivo studies (24,29). Studies in a C6 glioma cell line stably expressing MOPR, have shown that endogenous RGS proteins have profound effects on maximal agonist response in a number of functional assays like inhibition of adenylyl cyclase activity, phosphorylation of mitogen-activated protein kinase, Akt phosphorylation and inhibition of calcium channels (30–33). However, these studies did not elucidate if specific RGS proteins were involved in the modulation of MOPR signaling. Interestingly, RGS4 a member of the B/R4 subfamily is thought to play a role in modulating MOPR activity in vivo. This is based on findings that suggest that RGS4 appears to be dynamically regulated at the transcriptional and translational level by morphine (34–36). In addition, RGS4 is thought to play a role in the development of opiate dependence, since over-expression of this protein decreases electrophysiological responses to morphine in locus coeruleus neurons (30). Co-immunoprecipitation assays indicate the presence of RGS17 (also known as RGSZ2) and RGS20 (also known as RGSZ1), members of the A/RZ subfamily, in MOPR immunoprecipitates from endogenous tissue (37). These observations suggest that RGS17 and RGS20 could play a role in modulating MOPR function.
The striatal enriched protein, RGS9-2, appears to potently modulate MOPR signaling. RGS9-2 belongs to the C/R7 family of RGS proteins characterized by the presence of at least two functional regions that facilitate interactions with GPCRs: The GGL (Ggamma-like) and the DEP (Disheveled, EGL-10, Pleckstrin) domains. The GGL domain forms a stable complex with the brain enriched G_eta5 protein. Cellular localization of C/R7 RGS proteins is mediated via their DEP domain and the DEP domain-associated syntaxin-like proteins called R7BP (RGS7 binding protein) and R9AP (RGS9 anchor protein) (38). Reports from genetic mouse models indicate that RGS9-2 is a major determinant of the sensitivity to the rewarding actions of psychostimulant and opioid drugs (35,39,40). Co-immunoprecipitation studies demonstrate that morphine treatment promotes the formation of complexes between RGS9-2, MOPR and beta-arrestin 2 (41). In addition, RGS9-2 functions as a negative modulator of MOR internalization and prevents opioid-induced phosphorylation of ERK1/2 (41). Interestingly, ablation of the RGS9 gene enhances the analgesic actions of morphine but does not affect the onset or duration of analgesia (41).
Most studies examining the interaction of RGS proteins with opioid receptors have focused on MOPR mainly because of the latter’s role in morphine-mediated antinociception. There is little information on the interactions between RGS proteins and DOPR or KOPR. Recent findings suggest that in HEK-293 cells MOPR mediated inhibition of adenylyl cyclase activity is significantly reduced following co-expression of either RGS1, RGS2, RGS4, RGS9, RGS10 or RGS19 (also known as Galpha interacting protein, GAIP) (42). However, DOPR mediated inhibition of adenylyl cyclase activity is decreased only upon co-expression of RGS9 (42). In the case of KOPR, studies show that co-expression of KOPR with RGS4 in Xenopus oocytes leads to a decrease in the basal levels of GIRK1/2 currents and to an increase in agonist-mediated K+-conductance (43). Another report has shown that treatment of PC12 cells with the KOPR agonist, U69,593, leads to an up-regulation on RGS4 mRNA levels, and this is blocked by the antagonist, norbinaltorphimine, or by pre-treatment with pertussis toxin (44). Therefore, further studies are needed to characterize the interactions of different RGS proteins with opioid receptors, to examine the functional consequences of these interactions, and to determine whether these interactions are cell- and/or tissue-specific.
Previous work documented that interactions between RGS proteins and effector molecules like adenylyl cyclase, GIRK channels, cGMP phosphodiesterase, guanylyl cyclase, phospholipase C, and calcium channels may control their function, localization and association with other signaling molecules (29, 45, 46). These studies suggest an important role for RGS proteins in modulating receptor activity perhaps by direct association; this needs to be further explored.
G protein-coupled receptor kinases (GRKs) play a critical role in attenuating receptor signaling leading to a desensitized state of the receptor upon continued exposure to the agonist (for review see 47). Specific GRKs phosphorylate the C-terminal tail of the receptors, which then associate with beta-arrestin. The binding of beta-arrestin to the phosphorylated receptor prevents the latter’s association with heterotrimeric G-proteins resulting in the disruption of G-protein mediated signal transduction (48).
To date seven GRKs have been identified in humans that are divided into three classes: GRK1-, GRK2- and GRK4-like. GRK1 (rhodopsin kinase) and the closely related GRK7 (iodopsin kinase) are found primarily in the retina where they regulate the function of opsin. GRK2 and the closely related GRK3 are widely expressed and share a C-terminal pleckstrin homology domain that binds PIP2 (phosphatidylinositol bisphosphate) and the Gbetagamma subunit. GRK4, GRK5 and GRK6 lack this Gbetagamma subunit binding domain but use direct PIP2 binding and/or covalent lipid modification with palmitate to reside primarily at the plasma membrane (49). Although there is no current evidence indicating the specificity of GRKs for opioid receptor subtypes, a number of studies in heterologous expression systems have shown that MOPR can be phosphorylated by GRK2 (48, 50), GRK3 (51,52) and GRK6 (52), DOPR can be phosphorylated by GRK2 (53, 54), GRK3 (51), and GRK5 (55) and KOPR by GRK2 (56) or GRK3 (57). However, in the case of KOPR, a study showed that agonist mediated phosphorylation of human KOPR was blocked by expression of a dominant negative GRK2 mutant, whereas rat KOPR was not phosphorylated by the same agonist even in the presence of GRK2 or GRK3 (58). These results suggest that there could be species differences regarding receptor phosphorylation by GRKs.
Few studies have looked at the in vivo consequences of the interactions between GRKs and opioid receptors. For example, studies examining the levels of MOPR in the brains of addicts that died due to opiate overdose observed a decrease in the levels of MOPR as well as GRK3, GRK6 and beta arrestin-2 (59). It addition it has been shown that a chronic infusion, in mice, of the non-selective opioid antagonist, naloxone, caused a dose-dependent increase in the levels of MOPR and a decrease in the levels of GRK2 (60). Although these studies demonstrate altered levels of receptors and GRKs, the functional implications of these changes have not been investigated. Interestingly in the case of KOPR, studies using mice lacking GRK3 indicated that the development of analgesic tolerance to kappa agonists could be due to prolonged receptor phosphorylation by GRK3 (61).
It is a dogma in the GPCR field that receptor phosphorylation ultimately leads to its endocytosis via recruitment of beta-arrestins. Studies demonstrating that interactions between opioid receptors and GRKs lead to endocytosis are described in section 5.
Opioid receptor phosphorylation by GRKs is thought to lead to the recruitment and binding of beta-arrestin to the phosphorylated C-terminal tail. This then leads to the attenuation of G-protein mediated signaling and receptor endocytosis. Studies describing the involvement of beta-arrestins in opioid receptor endocytosis are described in section 5. A number of studies have shown that beta-arrestins can play an important role in modulating receptor signaling. In the following paragraphs we will describe evidence for the involvement of beta-arrestins in the modulation of opioid receptor signaling.
To date four functional arrestin gene family members have been cloned. Two of these are expressed only in the retina (visual arrestin and cone arrestin) and regulate photoreceptor function. The non-visual arrestins, beta-arrestin 1 and beta-arrestin 2 are expressed ubiquitously in all cells and tissues and function in the desensitization of most GPCRs except rhodopsin. Beta-arrestin 1 and beta-arrestin 2 exhibit 78% homology and contain binding motifs for clathrin and the beta2-adaptin subunit of the AP-2 complex in their C-terminal tail which allows them to function as adaptor proteins and target GPCRs to clathrin coated pits for endocytosis (62, 63). The different opioid receptor subtypes exhibit different requirements for binding beta-arrestin 1 and beta-arrestin 2 (64) that could lead to different signaling outcomes (65, 66). GST pull down assays show that the 3rd intracellular loop and the C-terminal tail of DOPR and only the C-terminal tail of KOPR can interact with beta-arrestin 1 or beta-arrestin 2 (64). These studies did not observe any interaction between beta-arrestin 1 or beta-arrestin 2 and MOPR (64). However, studies in HEK-293 cells using beta-arrestin 2 tagged to GFP or in striatal neurons using dominant negative beta-arrestin 2 show that agonist activated MOPR can recruit beta-arrestin 2 (67, 68) although the efficacy of recruitment is agonist dependent (69). Interestingly, mice lacking beta-arrestin 2 exhibit a potentiation and increased duration of the analgesic effect of morphine underscoring the importance of beta-arrestin 2 in mediating MOPR function (70). In the case of DOPR, a BRET assay suggested that receptor phosphorylation promoted receptor selectivity for beta-arrestin 2 over beta-arrestin 1 without affecting the stability of the receptor-beta-arrestin complex (71). However, another study used fluorescence and co-immunoprecipitation to show that agonist treated DOPR bound and recruited beta-arrestin 1 and beta-arrestin 2 to the plasma membrane (72). In addition, it has been shown that over expression of beta-arrestin 1 leads to an attenuation of DOPR and KOPR but not MOPR mediated activation of G-proteins and inhibition of cAMP levels (73).
In addition to being involved in the attenuation of G protein mediated signaling, studies have shown that beta-arrestins can induce a sustained ERK phosphorylation that is distinct from the transient G-protein mediated ERK phosphorylation (74). A recent study showed that MOPR ligands such as etorphine and fentanyl, but not morphine or methadone, induced ERK phosphorylation via a beta-arrestin dependent pathway. This led to the translocation of phosphorylated ERK to the nucleus leading to an increase in the activity of Elk-1 and in the transcription of GRK2 and beta-arrestin 2 (75). More recently, a study showed that heterodimerization between MOPR and DOPR promotes the recruitment of beta-arrestin 2 to the plasma membrane leading to changes in the spatio-temporal dynamics of ERK-mediated signaling that are quite distinct from those observed with the MOPR homodimer (76). Taken together, these studies show that beta-arrestins play an important role in mediating opioid receptor signaling by serving as a switch between G protein dependent and independent signaling mechanisms.
In addition to RGSs, GRKs and beta-arrestins several other proteins have been shown to modulate opioid receptor signaling. MOPR has been shown to interact with actin and the intermediate filament-binding protein, periplakin. This interaction occurs between the fourth intracellular loop/helix VIII of the receptor and the C-terminal rod and linker region of periplakin. Periplakin reduces the coupling between MOPR and a Galphai fusion protein (77), suggesting a possible modulatory role of periplakin on MOPR function. MOPR function can also be regulated by interaction with synaptophysin. This complex facilitates MOPR endocytosis. This in turn accelerates the rate of receptor resensitization (78).
Another protein that has been shown to affect MOPR signaling is calmodulin (CAM). CAM is a ubiquitous Ca+2 sensitive regulatory protein, implicated in the regulation of a number of cytoplasmic enzymes including adenylyl cyclases, Ca2+/CAM-dependent kinases and phosphatases, ion channels, Ca+2-ATPases among others (79, 80). Several new studies show that calmodulin interacts directly with the 3rd intracellular loop of MOPR which results in reduced basal G protein coupling and attenuation of agonist mediated increase in GTPgammaS binding (81, 82).
There is also evidence that the scaffolding protein, spinophilin, interacts with alpha2 adrenergic alpha2AR), D2 dopamine and MOPR (83–85). Spinophilin contains a putative F-actin-binding domain at the amino terminus, a single PDZ domain, and a region predicted to form a coiled-coil structure at the carboxyl terminal (85, 86). A GPCR interacting domain lies between the actin binding and the PDZ domain (83, 84). In addition, spinophilin targets protein phosphatase 1 to specific substrates (83). Spinophilin competes with beta-arrestin 2 following activation of alpah2AR, and plays a critical role in alpha2AR stabilization at the cell surface. Recent data indicate that spinophilin is part of the signaling complexes involved in the modulation of MOPR signaling and endocytosis (85). Interestingly, spinophilin knockout mice exhibit reduced sensitivity to the analgesic effects of morphine, early development of tolerance but higher degree of physical dependence and increased sensitivity to the rewarding actions of the drug (85).
Another scaffold protein shown to interact with opioid receptors is Tamalin. This protein consists of multiple interacting domains, including PDZ, a leucine zipper region and a carboxy terminal PDZ binding motif (87). Tamalin interacts with PSD-95, PSD-95-associated proteins, ARFs (ADP-ribosylation factors), cytohesins, and Mint-2 and is required for neuritic localization of mGluR1a (87). In the nucleus accumbens, Tamalin appears to be part of signaling complexes that modulate the analgesic actions of morphine via association with MOPR (88).
A direct association with the plasma membrane localized phospholipid-specific phosphodiesterase, phospholipase D2 (PLD2), appears to modulate opioid receptor expression at the cell surface (89, 90). This interaction is mediated via the Phox homology domain in the N terminal region of PLD2 (91). In general, PLD2 regulates agonist dependent and agonist-independent endocytosis of MOPR, DOPR as well as CB1 cannabinoid receptors (90, 92). Taken together, these studies show that protein-protein interactions can play an important role in modulating opioid receptor signaling.
Receptor endocytosis is required for agonist induced turnover and resensitization of opioid receptors. As mentioned in the previous section following receptor activation by agonists the C-terminal tail of the receptors is phosphorylated by specific GRKs and this leads to the recruitment of beta-arrestin to the phosphorylated receptors leading to the disruption of G-protein mediated signal transduction (48). In addition, beta-arrestins tether the receptors to clathrin-coated pits (93, 94) leading to receptor endocytosis. Following endocytosis, opioid receptors are either recycled back to the cell surface to undergo another round of signaling and/or they are targeted to lysosomes for degradation. In the following section we will describe the proteins that have been implicated in the endocytosis and degradation of opioid receptors.
A number of studies show that the phosphorylation of specific residues in the C-terminal tail of opioid receptors is required for receptor internalization via clathrin-coated pits (56, 95–97). Present knowledge indicates that GPCR kinases (GRKs) are involved in the phosphorylation of activated opioid receptors. Confocal microscopy studies with enhanced green fluorescent protein (GFP) tagged opioid receptors and red fluorescent protein (RFP) tagged GRKs show that activated DOPR co-internalizes with GRK 2 and 3 while MOPR endocytosis is not accompanied by GRK endocytosis (55). In the case of KOPR, laser scanning microscopy using GFP tagged human KOPR and RFP tagged GRK2 or GRK3 suggests that GRK3 is better able to induce KOPR internalization than GRK2 (57).
A number of studies have demonstrated that beta-arrestin recruitment is required for opioid receptor endocytosis. In the case of MOPR, studies show that among most MOPR agonists morphine does not cause receptor endocytosis in heterologous expression systems; however, enhanced expression of recombinant beta-arrestin 2 and GRKs has been shown to increase morphine mediated MOPR internalization (50, 98). Interestingly, morphine can induce MOPR internalization in neurons, which is inhibited by dominant negative beta-arrestin 2 (68, 99, 100). In the case of DOPR, studies indicate that beta-arrestin 1 is required for phosphorylation dependent DOPR internalization while beta-arrestin 2 is required for phosphorylation–independent DOPR internalization (72). In the case of KOPR, selective peptide agonists (but not non-peptide agonists) were shown to induce endocytosis of rat KOPR (101). However, human KOPR has been shown to undergo endocytosis in response to the non-peptide agonist, U50,488H, but not etorphine by a mechanism involving GRKs, beta-arrestin 2 and dynamin I (102).
In addition to GRKs and beta-arrestins, a number of other proteins have been implicated in opioid receptor endocytosis. Studies show that RGS9-2, a protein that is enriched in striatum, delays agonist induced HA-MOPR internalization in transiently transfected PC12 cells (41). Coimmunoprecipitation assays reveal that morphine treatment enhances the interaction between RGS9-2 and MOPR (41). Another RGS, RGS14, has been found to associate with MOPR in the periaqueductal gray matter (PAG) neurons. Silencing of RGS14 leads to increased phosphorylation of S375 in the C-terminal tail of MOPR by morphine subsequently leading to increased receptor internalization and recycling to the membrane (103).
In the case of DOPR, a study has implicated RGS19 in its endocytosis. Immunofluorescence labeling and deconvolution analysis show that in the absence of agonist RGS19 is spatially segregated from Galphai3 and DOPR in clathrin-coated domains of the cell membrane while the Galphai3-YFP and DOPR are located in non-clathrin-coated microdomains of the plasma membrane (104). Upon receptor activation, Galphai3 partially colocalizes with RGS19 in clathrin-coated pits. Blocking of endocytosis with a dynamin mutant leads to a striking overlap in the distribution of DOPR, Galphai3-YFP and RGS19 in clathrin-coated pits (104). This suggests a mechanism where after agonist treatment DOPR and Galphai3 move together into clathrin-coated pits where Galphai3 and RGS19 meet and turn off G protein signaling. Subsequently Galphai3 returns to non-clathrin coated microdomains, RGS19 remains associated with clathrin-coated pits and DOPR is internalized via clathrin-coated vesicles (104).
Another protein implicated in MOPR endocytosis is p38MAPK. It has been shown that phosphorylation of 38MAPK is sufficient to trigger the constitutive internalization of MOPR even in the absence of agonist (105) since its inhibitor, SB203580, strongly impairs MOPR endocytosis. p38MAPK can phosphorylate the Rab5 effectors, EEA1 and Rabenosyn-5, on Thr-1392 and Ser-215 respectively and these phosphorylation events regulate the recruitment of the phosphorylated molecules to the plasma membrane ultimately leading to MOPR endocytosis (105). Interestingly, a phosphomimetic mutation of Thr-1392 in EEA1 can bypass the requirement for p38MAPK alphaisoform in MOPR endocytosis (105).
Another molecule implicated in GPCR endocytosis is phospholipase D, an enzyme that hydrolyzes phosphatidylcholine to generate choline and phosphatidic acid. Phospholipase D has been implicated in signal transduction, membrane trafficking and cytoskeletal reorganization (for review see 106). Two mammalian isoforms of this enzyme have been identified: phospholipase D1, present mostly in intracellular membranes, and phospholipase D2, associated mostly with the plasma membrane (107). Studies show that phospholipase D2, which has been implicated in the formation of endocytotic vesicles, plays a role in MOPR endocytosis. Yeast two-hybrid and coimmunoprecipitation assays demonstrate interactions between the C-terminal tail of MOPR and phospholipase D2. In addition, MOPR endocytosis can be blocked by a dominant negative mutant of phospholipase D2 and, by inhibition of phospholipase D2-mediated phosphatidic acid production (91). Agonist induced and ARF-dependent phospholipase D2 activation have been reported to be a prerequisite for MOPR endocytosis and recycling (92). Interestingly, endogenous opioids as well as DAMGO, but not morphine, have been shown to promote phospholipase D2 activation and rapid MOPR recycling (92). Receptor mediated phospholipase D2 activation has been shown to be also required for the endocytosis of DOPR (92).
Protein kinase C (PKC) has been implicated in agonist-independent endocytosis of DOPR (108). Activation of PKC by either PMA (phorbol 12-myristate 13 acetate), an alpha1AAR agonist or by ionomycin can lead to DOPR endocytosis that requires the phosphorylation of S344. This PKC mediated DOPR internalization involves a beta-arrestin and clathrin-dependent mechanism and does not require receptor phosphorylation by GRKs (108). A study showed that MOPR can also be endocytosed by a G-protein independent but tyrosine kinase-dependent mechanism, which can be blocked by genistein treatment (109).
Another protein that has been implicated in opioid receptor endocytosis is synaptophysin. Synaptophysin is an acidic Ca+2-binding glycoprotein of ~38kDa that is present mostly in synaptic vesicles. It has 4 transmembrane regions and is reported to be the major cholesterol binding protein in synaptic vesicles (110). Coimmunoprecipitation and BRET studies show that in HEK cells co-expressing MOPR and synaptophysin both proteins are constitutively associated and this leads to increased MOPR endocytosis. The authors postulated that this was due to the ability of synaptophysin to recruit dynamin to the plasma membrane thus facilitating the fission of clathrin-coated vesicles. This was supported by the observation that a synaptophysin mutant that no longer interacted with dynamin prevented agonist mediated MOPR endocytosis (78).
Another report showed that the C-terminal tail of human MOPR binds to the carboxyl terminal region of human filamin A (111). Filamin A is a homodimeric F-actin cross-linker with a high molecular weight (~280 kDa) that organizes actin filaments into three-dimensional arrays linking them to the cell membrane. In addition, Filamin A anchors a variety of transmembrane proteins to the actin cytoskeleton and provides a scaffold for many cytoplasmic and signaling molecules (112). Confocal microscopy studies revealed that MOPR internalization was greatly reduced in the absence of filamin A (113). Interestingly, chronic morphine treatment was found to upregulate MOPR levels in cells lacking filamin A although the mechanisms are not clearly understood (113).
MOPR also interacts with glycoprotein M6A (114). M6a is a member of the proteolipid protein family of tetraspan membrane proteins and is mainly expressed in neurons. The transmembrane domains 4, 5, and 6 of MOPR and the protein stretch/domain including transmembrane domains 3 and 4 of M6a are important regions for the MOPR-M6a interaction. This interaction leads to enhanced MOPR endocytosis and recycling, whereas a dominant negative M6a (truncated mutant) prevents agonist-induced MOPR internalization. M6A has been shown to also interact with DOPR and thus may act as a scaffold molecule in the regulation of opioid receptor endocytosis (114).
Following endocytosis receptors can either be dephosphorylated and recycled back to the cell surface or be targeted to lysosomes for degradation. An interesting study screened a library of C-terminal tails to 59 GPCRs fused to glutathione S- transferase for their ability to interact with 4 adapter proteins proposed to be involved in post-endocytotic sorting of receptors: ERM (ezrin-radixin-moesin)-binding phosphoprotein EB50, N-ethylmaleimide-sensitive factor, sorting nexin 1 (SNX1) and GASP. This study observed a strong interaction of DOPR with NSF (an adapter protein suggested to be involved in receptor recycling), SNX1 and GASP (adapter proteins suggested to be involved in lysosomal targeting of the receptors) (115).
Association with the G protein-coupled receptor associating protein (GASP-1) is thought to be involved in the trafficking and cell surface expression of DOPR (116). GASP-1 is found throughout the CNS and has been implicated in the sorting of a number of native GPCRs to the lysosome after endocytosis. GASP-1 interacts with the C terminal of several GPCRs from class A and class B subfamilies (117). The highest levels of binding for GASP-1 were observed with MOPR and with beta1AR (90,118,119). In fact, mutant forms of the MOPR with increased affinity to GASP show enhanced post endocytic receptor degradation (90, 117,118).
Ubiquitin, a 76 amino acid polypeptide, has been shown to be involved in the degradation of GPCRs. It attaches covalently to the epsilon-amino group of lysine residues present in GPCRs and targets the latter for proteasomal degradation (via lysine48-linked ubiquitination) or for receptor down-regulation in lysosomes (via lysine63-linked ubiquitination). In the case of DOPR, a majority of synthesized receptors are transported to the cytoplasmic site of the ER membrane via the Sec61 translocon, where they are deglycosylated and conjugated with ubiquitinin prior to degradation by the cytoplasmic 26 S proteasomes (10,11). There is some controversy regarding the involvement of ubiquitination in the degradation of DOPR endocytosed from the plasma membrane. A study showed that ubiquitination was not required for either the ligand-induced endocytosis of DOPR or for its post-endocytic trafficking to lysosomes since treatment with an inhibitor of proteasomal degradation or mutation of all cytoplasmic lysine residues in DOPR did not inhibit ligand-induced receptor endocytosis or the proteolytic degradation of endocytosed receptors (119). Another study showed that DOPR endocytosis was not affected by lysosomal protease inhibitors but was significantly attenuated by proteasome inhibitors leading to the accumulation of polyubiquitinated receptors (120). Interestingly, association with GASP1 has been shown to be able to modulate the trafficking of DOPR but not MOPR to lysosomes in the absence of ubiquitination (116). This ubiquitin-independent trafficking of DOPR to lysosomes utilizes some (Vps4 and Hrs) but not all (Tsg101) of the vacuolar protein sorting machinery that is needed for lysosomal sorting of ubiquitinated receptors (121). In the case of MOPR, although the endocytosed receptors can be sorted into lysosomes, most of them are thought to recycle rapidly to the plasma membrane due to the presence of a specific 17-amino acid sequence in the C-terminal tail that specifically promotes the sorting of receptors into a rapid recycling pathway (122). In addition, deletion of the “recycling signal” from the cytoplasmic tail of MOPR enhances its interaction with GASP leading to receptor degradation after endocytosis (118).
In the case of KOPR, there is evidence for interactions with Ezrin-radixin-moesin (ERM)-binding phosphoprotein-50/Na+/H+ exchanger regulatory factor (EBP50/NHERF), a PDZ domain-containing phosphoprotein. This interaction occurs between the PDZ domain I of EBP-50 and the C- terminal tail of KOPR and appears to block agonist-induced KOPR down-regulation by increasing the recycling rate of internalized receptors (123). Interestingly, a study showed that agonist-mediated endocytosis of human KOPR could target the receptors to lysosomes and proteasomes for degradation by a process requiring GRK2, beta-arrestin 2, dynamin I, and rab5 (124). In a later study the authors went on to show that the targeting of human KOPR to lysosomes for degradation also required the ubiquitination of three lysine residues in the C-terminal tail of KOPR via Lys 63-linked polyubiquitination (125).
Another level of protein-protein interaction leading to the modulation of receptor function is provided by receptor heterodimerization. In this case the receptor associates with another receptor type leading to the modulation of binding, signaling or trafficking properties. In the case of opioid receptors studies show that DOPR can heterodimerize with KOPR or MOPR leading to the formation of new receptor complexes that exhibit pharmacological and signaling properties that are distinct from each individual receptor (reviewed in 126,127). In addition, a number of studies have shown that opioid receptors can also heterodimerize with non-opioid receptors. For example studies show that MOPR can heterodimerize with either alpha2AAR, CB1 cannabinoid, ORL1, NK1, somatostatin 2A or chemokine 5 receptors leading to a modulation in the binding, signaling, trafficking or a combination of these (reviewed in 127). Also DOPR has been shown to heterodimerize with alpha2AAR, beta2AR, sensory neuron specific receptor or CXCR4 leading to modulation in the signaling or trafficking properties of individual receptors (reviewed in 127–128). In the case of KOPR, besides its heterodimerization with DOPR as mentioned above, it has been shown to heterodimerize with beta2AR and this drastically affects the trafficking properties of beta2AR (reviewed in 126, 127). Finally, recent studies support a role for dimerization in the maturation and trafficking of a variety of GPCRs including MOPR and DOPR (128).
The biosynthesis, cellular localization, trafficking, signaling and degradation of opioid receptors can be modulated by interactions with proteins as diverse as cytoskeletal proteins, signaling molecules, enzymes, kinases, etc. This would, in turn, modulate the responsiveness of opioid receptors following exposure to agonists. Although a great deal is known about proteins that modulate the trafficking of opioid receptors much less is known about proteins that interact with the receptors during biogenesis and help target them to the cell surface. Another area of research would involve the elucidation of the machinery involved in opioid receptor degradation. This would, in turn, contribute to our understanding of opioid receptor function and could provide insights into the development of tolerance to opiates.
In the last decade it has become increasingly apparent that opioid receptor function can be modulated by heterodimerization with other receptors. However, very little is known about the machinery involved in the biogenesis, trafficking and degradation of opioid receptor heterodimers. These studies would help us in understanding the role and regulation of opioid receptor heterodimerization in health and disease states as well as help in the design of novel heterodimer-specific reagents (ligands, allosteric modulators and antibodies).
This work was supported by NIDA grants DA008863 and DA019521 (L.A.D.) and by the Greek Secretariat for Research and Technology-PENED03 (V.Z.)