Key differences in regulation of opioid receptors localized to presynaptic terminals compared to somas: relevance for novel therapeutics

Abstract

Opioid receptors are G protein-coupled receptors (GPCRs) that regulate activity within peripheral, subcortical and cortical circuits involved in pain, reward, and aversion processing. These receptors are expressed in both presynaptic terminals where they inhibit neurotransmitter release and postsynaptic locations where they act to hyperpolarize neurons and reduce activity. Agonist activation of postsynaptic receptors at the plasma membrane signal via ion channels or cytoplasmic second messengers. Agonist binding initiates regulatory processes that include phosphorylation by G protein receptor kinases (GRKs) and recruitment of beta-arrestins that desensitize and internalize the receptors. Opioid receptors also couple to effectors from endosomes activating intracellular enzymes and kinases. In contrast to postsynaptic opioid receptors, receptors localized to presynaptic terminals are resistant to desensitization such that there is no loss of signaling over the same time scale. Thus, the balance of opioid signaling in circuits expressing pre- and postsynaptic opioid receptors is shifted toward inhibition of presynaptic neurotransmitter release. The functional implication of this shift is not often acknowledged in behavioral studies. This review covers what is currently understood about regulation of opioid/nociceptin receptors, with an emphasis on opioid receptor signaling in pain and reward circuits. Importantly, the review covers regulation of presynaptic receptors and the critical gaps in understanding this area, as well as the opportunities to further understand opioid signaling in brain circuits.

1. Introduction

GPCRs are the targets of greater than 35% of currently available therapeutic drugs (Sriram and Insel, 2018). These receptors are highly dynamic with regulated expression at both transcriptional and translational levels, as well as tightly regulated localization and function. Opioid receptors comprise a family of GPCRs that have been extensively studied in terms of their regulation and the resulting changes in behavior. This review will primarily cover opioid receptor regulation with an emphasis on known differences in mechanisms regulating opioid receptors localized to presynaptic versus those localized to postsynaptic locations in neurons. Examples of differential regulation of pre- versus postsynaptic receptors are given in different circuits. The goal is to stress that further defining these differences will yield additional targets for therapeutic intervention.

Opioid receptors are encoded by 3 main genes, mu opioid (OPRM1), delta opioid (OPRD1) and kappa opioid receptors (OPRK1), and the related gene for the nonclassical nociception/Orphanin FQ receptor (OPRL1), encoding MOR, DOR, KOR and NOP, respectively. These receptors are members of the Class A subclass of GPCRs that share significant homology to the light-activated rhodopsin receptor (Surratt and Adams, 2005). All four receptors are predominately coupled to Gα_i/o G proteins and signaling pathways that inhibit neuronal firing and/or activity. Activation of G proteins and the release of the Gα from the Gβγ from the receptors can stimulate or inhibit signaling simultaneously from multiple effectors. These signaling pathways differ depending on the local environment surrounding the receptors. For example, postsynaptic MORs release Gα to inhibit adenylyl cyclase (AC) activity while Gβγ subunits directly activate G protein-coupled inwardly-rectifying potassium (GIRK) channels (Sadja et al., 2003; Wickman and Clapham, 1995) or inhibit voltage-gated Ca²⁺ channels (Herlitze et al., 1996; Rusin et al., 1997). MORs localized to presynaptic terminals inhibit voltage-gated Ca²⁺ channels (Endo and Yawo, 2000) but are also coupled to voltage-dependent potassium channels (Vaughan et al., 1997b), but not GIRK (Luscher et al., 1997). Presynaptic MOR-activated Gβγ subunits directly bind to SNARE proteins involved in vesicle release (Blackmer et al., 2005; Gerachshenko et al., 2005; Zurawski et al., 2019a; Zurawski et al., 2019b). Importantly, regulation of opioid receptors is different from postsynaptic opioid receptors in several brain areas (Fyfe et al., 2010; Lowe and Bailey, 2015; Pennock et al., 2012; Pennock and Hentges, 2011).

2. Canonical regulation of postsynaptic opioid receptors

Opioid agonists bind to opioid receptors and promote conformational changes in the receptor that activate Gα_i/o and subsequent release of the Gα and Gβγ subunits from the receptor. This release allows G protein receptor kinase (GRK) and other kinases to phosphorylate the receptor and recruit binding of arrestins that internalize receptors (Fig. 1) (Chen et al., 2013; Miess et al., 2018; Williams et al., 2013). This regulation has been demonstrated for all 4 receptor subtypes and contributes to receptor desensitization and tolerance to opioid agonists; MOR (Arttamangkul et al., 2018; Miess et al., 2018; Schulz et al., 2004), DOR (Pei et al., 1995; Pradhan et al., 2016), KOR (Abraham et al., 2018; McLaughlin et al., 2004), and NOP (Thakker and Standifer, 2002). For many GPCRs, it is widely understood that sustained agonist activation can lead to receptor phosphorylation and internalization (Pierce and Lefkowitz, 2001). Internalization of the receptors and trafficking through various endosome and lysosome pathways are dependent on complex protein-protein interactions which are different for the various opioid receptors. For example, MORs are readily returned to the plasma membrane whereas DORs are rapidly degraded upon internalization (Finn and Whistler, 2001; Whistler et al., 2001). In addition, signaling via arrestin complexes have been shown for several GPCRs (Shukla et al., 2011) and more recently, intracellular signaling by internalized opioid receptors was documented for opioid receptors (Stoeber et al., 2018).

Figure 1. Summary of postsynaptic opioid receptor signaling.

Agonist stimulation leads to coupling of opioid receptors (MOR, KOR, DOR, and NOP) to heterotrimeric Gi/o proteins (1), resulting in G protein activation and dissociation of Gα and Gβγ subunits (2) which in turn activate separate effectors. The Gβγ complex is responsible for inhibition of Ca²⁺ channels, and activation of GIRK channels (2a) while Gα inhibits adenylyl cyclase activity and activates MAPK, ERK1/2, JNK, and p38 signaling pathways (2b). Subsequently, the receptor is phosphorylated by GRK and other kinases, including PKC (3), resulting in desensitization, recruitment of β-arrestins (4), and internalization of these receptors (5). At this point, the receptor is degraded or recycled to the membrane. β-Arrestin-dependent endosomes also signal to various effectors (6). Activation of these signaling pathways (2b and 6) increases phosphorylation of the transcription factor CREB in the nucleus. cAMP, cyclic adenosine monophosphate; ERK, extracellular signal regulated kinase; JNK, c-jun N-terminal kinase; MAPK, mitogen-activated protein kinases; P, phosphorylation.

2.1. MOR

MORs are critical for morphine-induced antinociception, reward and withdrawal (Matthes et al., 1996). This has led to a strong focus on regulation of MORs which has dominated the literature since cloning of these receptors. Early studies using opioid receptors expressed in various cell lines provided evidence that MOR activation of pertussis-toxin-sensitive G proteins affect multiple signaling pathways within cells, including MOR inhibition of adenylyl cyclase activity, inhibition of Ca²⁺ channels, and activation of GIRK channels (Childers, 1991). Multiple signaling pathways are involved with regulation of MORs, including PKC, phosphoinositide turnover, MAP kinase pathways and CREB (Al-Hasani and Bruchas, 2011; Bailey et al., 2009b; Williams et al., 2013) (Fig 1).

Agonist activation of MORs in postsynaptic locations often results in membrane-delimited activation of GIRK channels through direct binding of Gβγ subunits (Huang et al., 1997) to produce hyperpolarization (Sadja et al., 2003; Wickman and Clapham, 1995). Outward currents elicited by GIRK channels quickly desensitize (within minutes) due to GRK-mediated phosphorylation of the C-terminus tail of MORs (Blanchet and Luscher, 2002; Fox and Hentges, 2017). Desensitization and internalization were thought to be the result of beta-arrestin recruitment to receptors because these processes occur on the same timescale. However, desensitization of MOR-activated GIRK currents occurs normally in beta-arrestin knock-out (KO) mice (Arttamangkul et al., 2006), indicating that these processes are independent (Arttamangkul et al., 2018; Birdsong et al., 2015). These studies clearly established that desensitization and internalization processes are separable. Agonist-induced MOR desensitization of GIRK currents depends on beta-arrestin and ERK1/2 signaling in the locus coeruleus (Dang et al., 2009).

Opioid ligands engage signaling pathways with different efficacies, termed “functional selectivity” (Urban et al., 2007). In addition, opioid agonists differentially engage opioid signaling pathways, desensitization and internalization pathways (Borgland et al., 2003; Johnson et al., 2006; Kunselman et al., 2021b; McPherson et al., 2010; Narita et al., 2006; Whistler et al., 1999; Whistler and von Zastrow, 1998), as well as kinases such as PKC (Bailey et al., 2004; Bailey et al., 2006). The development of beta-arrestin knockout mice by Bohn and colleagues (Bohn et al., 2002; Bohn et al., 1999) that displayed increased morphine antinociception with decreased development of tolerance and opioid-dependent side effects led to extensive research focused on G protein-dependent and beta-arrestin-dependent signaling. The hope was that opioid ligands with a preference for G protein pathways would provide adequate antinociception without substantial opioid-induced side effects. The rich array of opioid ligands allowed for detailed studies of opioid ligand bias for the receptors (Faouzi et al., 2020; Kelly, 2013; Malfacini et al., 2015; McPherson et al., 2010; Miess et al., 2018; Thompson et al., 2015). Most of these studies used biochemical assays that test compounds on cell lines expressing the receptors, and thus focus on postsynaptic-like environments. Unfortunately, the therapeutic promise of opioid bias is more complicated as G protein-biased agonists were shown to produce significant opioid-mediated side effects, including respiratory depression, one of the main causes of opioid-induced death (Gillis et al., 2020a; Gillis et al., 2020b; Gillis et al., 2020c; Kliewer et al., 2020; Kliewer et al., 2019). One thing that is clear from these studies is that the expressed levels of opioid receptors, GRK, G proteins and cell environments determine the level of bias observed. Much less is known about biased signaling of opioid receptors in presynaptic locations.

Interesting adaptations to MOR signaling and regulation occur with repeated or continued opioid agonist administration (Williams et al., 2013). In neurons throughout the brain, continued morphine exposure results in decreased signaling by MORs to many effector pathways, termed “cellular tolerance” in both postsynaptic (Bagley et al., 2005a; Bailey et al., 2009a; Christie et al., 1987; Dang et al., 2011; Dang and Williams, 2004; Kim et al., 2008; Quillinan et al., 2011; Sim et al., 1996) and presynaptic locations (Fyfe et al., 2010; Hack et al., 2003), although cellular tolerance is less pronounced in some coupling pathways such as Ca²⁺ channels (Connor et al., 1999a). These results are consistent with early reports of decreased GTPyS binding following chronic morphine administration in expression systems (Puttfarcken and Cox, 1989; Puttfarcken et al., 1988). Paradoxically, “cellular tolerance” may be the end result of an increase in agonist affinity and/or efficacy at the receptor after prolonged administration of morphine or other opioid agonists (Birdsong et al., 2013; Ingram et al., 1998), promoting enhanced desensitization, internalization and disrupted resensitization of receptors at the plasma membrane (Williams et al., 2013). After chronic morphine treatment, different kinases play a role in desensitization and trafficking of MORs, including PKC (Bailey et al., 2009a; Bailey et al., 2009b; Gabra et al., 2008; Hull et al., 2010; Leff et al., 2020), PKA (Bagley et al., 2005a; Ingram et al., 1998), cJun-N-Terminal kinase (Leff et al., 2020), beta arrestins (Connor et al., 2015; Dang et al., 2011; Quillinan et al., 2011) and ERK1/2 (Dang et al., 2009; Macey et al., 2009; Macey et al., 2015). These adaptations can differ depending on several factors. First, the administration paradigm is important as MOR signaling differs following intermittent repeated injections of morphine compared to chronic administration of morphine via pellets or slow-release preparation (Bobeck et al., 2014; Ingram et al., 2008; Ingram et al., 1998) over the same time-period of several days. Second, different agonists can elicit antinociception via different signaling pathways. There are many examples of opioid ligand bias for specific signaling pathways (Gillis et al., 2020b; Kelly, 2013; Kelly et al., 2022). In the PAG, fentanyl-induced antinociception is reduced by a blocker of GIRK channels, while morphine-induced antinociception is reduced by a blocker of presynaptic Kv channels (Morgan et al., 2020). In addition, ERK1/2 inhibition blocks tolerance to repeated DAMGO administration but not fentanyl (Bobeck et al., 2016). Another important factor is cell type. MOR activation of ERK1/2 is dependent on GRK 3 in striatal neurons (Macey et al., 2006), whereas GRK2 is important in peripheral sensory neurons (Zhang and Jeske, 2020). These multiple factors make it difficult to identify the roles each of these signaling pathways play in the observed behaviors but represent possible targets that could differentiate between different behavioral effects of opioids. Regulation and signaling of the other opioid receptors subtypes are described below.

2.2. KOR

The kappa opioid receptor (KOR) is widely expressed in the CNS (Chen et al., 2020; Mansour et al., 1994b) in both pre- and postsynaptic locations (Ford et al., 2007; Ford et al., 2006; Margolis et al., 2003, 2005; Margolis et al., 2006). KORs modulate pain perception, stress responses, and affective reward state (Al-Hasani and Bruchas, 2011; Bruchas and Roth, 2016; Chavkin, 2011). Postsynaptic KORs are coupled to GIRK channels in midbrain dopamine neurons (Ford et al., 2006; Margolis et al., 2003) and serotonin neurons (Lemos et al., 2012) resulting in hyperpolarization of the neurons. KOR agonists promote G protein activation, beta-arrestin2 recruitment, and coupling to multiple downstream signaling pathways (Bruchas and Chavkin, 2010), including ERK1/2 (Bruchas et al., 2008), c-Jun N-terminal kinase (Schattauer et al., 2019) and p38 (Bruchas et al., 2007; Bruchas et al., 2006; Ehrich et al., 2015), as well as CREB, which is involved in controlling release of the endogenous KOR agonist dynorphin (Carlezon et al., 2006; Kreibich and Blendy, 2004). Coupling of KORs to these kinase pathways has also been described in astrocytes (Bruchas et al., 2006; Xu et al., 2007). KORs are desensitized by canonical pathways (see Fig. 1) with continued agonist exposure and with stress that promotes release of dynorphin, the endogenous KOR ligand (Appleyard et al., 1999; Bruchas and Chavkin, 2010; Lemos et al., 2012).

KOR biased signaling has been documented with downstream kinase pathways depending on either G protein or beta-arrestin signaling (Brust et al., 2016; Schmid et al., 2013). The physiological effects of KOR activation result from these different signaling cascades, with analgesia primarily associated with G protein activation and desensitization/side effects mediated through beta-arrestin2 recruitment, suggesting possible therapeutic potential of KOR agonists as analgesics. The anti-pruritic efficacy of KOR agonists is retained in β-arrestin 2 knockout mice (Morgenweck et al., 2015) suggesting that G protein signaling is key for those beneficial effects. Thus, biased KOR agonists that primarily signal through G proteins without significant beta-arrestin recruitment are also being investigated as novel analgesic drugs.

2.3. DOR

DOR mRNA and protein are widely expressed throughout the brain, spinal cord and dorsal root ganglia (DRG) (Cahill et al., 2001; Erbs et al., 2015; Mansour et al., 1994a). DORs regulate important physiological processes such as thermal and mechanical hyperalgesia, chronic inflammatory pain, anxiety and depression, migraine, locomotion, seizures, emotions, learning and memory, as well as addiction and tolerance development. Postsynaptic DORs couple to similar signaling pathways as MORs and KORs (Law and Loh, 1993; North et al., 1987; Piros et al., 1996). DOR agonists produce antinociception, although they are not as efficacious as MOR agonists. DORs are coupled to GIRK channels in supraspinal brain areas associated with nociceptive pathways, at least in some species. In the vlPAG, DORs are coupled to GIRK channels in a subpopulation of neurons in mice (Vaughan et al., 2003) but not in rats (Chieng and Christie, 1994a). DORs are coupled to GIRK channels in a subpopulation of rat RVM neurons as well (Marinelli et al., 2005). In contrast to MORs, DORs are not coupled to Ca²⁺ channels in the vlPAG (Connor and Christie, 1998; Connor et al., 1999b). However, DOR-mediated inhibition of Ca²⁺ channels in dorsal root ganglion neurons plays a role in antinociception (Pradhan et al., 2013), again demonstrating that coupled effectors are different depending on the cells expressing the receptors.

Continued administration of DOR agonists in DOR expressing cell lines results in GRK2-dependent desensitization and internalization, as with other opioid receptors (Guo et al., 2000; Pei et al., 1995). One of the main differences between MOR and DOR is the fate of the receptors after internalization; MORs are readily recycled to the plasma whereas DORs are sent to lysosome pathways (Whistler et al., 2002; Whistler et al., 2001). DOR agonists induce desensitization and internalization to different degrees (Pradhan et al., 2013; Pradhan et al., 2016; Pradhan et al., 2010; Pradhan et al., 2006; Scherrer et al., 2006; Vicente-Sanchez et al., 2018). For example, single injections of a high- (SNC80) and a low (AR-M100390)- internalizing compound in vivo produce equal levels of analgesia in a model of inflammatory pain. SNC80 induced strong desensitization and internalization of DORs such that a second drug injection had no effect. However, the low-internalizing agonist AR-M100390 did not produce internalization, G-protein uncoupling or acute desensitization (Pradhan et al., 2009). Tolerance to these agonists also differed (Pradhan et al., 2009; Pradhan et al., 2010).

One interesting difference noted between DORs compared to MORs and KORs is their ability to be upregulated and/or trafficked to the plasma membrane in response to stress and inflammatory conditions (Bie et al., 2009a; Cahill et al., 2003; Chang et al., 2019; Morinville et al., 2003; van Rijn et al., 2012). In addition, chronic morphine treatment promotes movement of DORs to the cell surface in the dorsal horn of the rat spinal cord and hippocampus (Erbs et al., 2016), as well as the vlPAG (Hack et al., 2005) and RVM (Ma et al., 2006). Thus, expression patterns of DOR appear to be regulated by stressful stimuli and injury.

2.4. NOP

NOPs are also expressed throughout the CNS in both pre- and postsynaptic locations (Houtani et al., 1996; Nothacker et al., 1996; Vaughan et al., 2001; Winters et al., 2019; Zaveri, 2016). Activation of NOPs produce anti-opioid antinociceptive behavioral effects (Mogil et al., 1996b; Mogil et al., 1996c; Morgan et al., 1997b). That is, they oppose the actions of MORs in supraspinally-mediated analgesia (Grisel et al., 1996b). The anti-opioid effect is the result of NOP actions at the level of circuits given that they are Gα_i/o-coupled receptors that activate the same effectors as opioid receptors in the same cells. The receptors are coupled to GIRK channels when expressed in postsynaptic locations resulting in reduced firing rates (Connor et al., 1996; Ikeda et al., 1997; Nazzaro et al., 2010; Vaughan et al., 1997a). The main difference between opioid receptors and NOP are the cell types that express the receptors. In the vlPAG, only a subpopulation of neurons are hyperpolarized by MORs coupled to GIRK channels (Vaughan et al., 2003); however, NOPs coupled to GIRK channels are expressed on all vlPAG neurons (Chiou, 2001; Vaughan et al., 1997a). Similarly, NOPs are coupled to inhibition of Ca²⁺ channels currents in all vlPAG neurons but again MORs only inhibit Ca²⁺ currents in a subpopulation of vlPAG neurons (Connor and Christie, 1998). The differential expression likely explains the “anti-opioid” actions of NOP agonists on analgesia. MOR agonists suppress GABA release onto vlPAG-RVM output neurons, exciting them (Lau and Vaughan, 2014). Since all vlPAG neurons express NOP, NOP agonists directly inhibit vlPAG-RVM output neurons, thus negating actions of opioids on the circuit.

Activation of NOP receptors with N/OFQ produces rapid desensitization (Pennock et al., 2012; Pu et al., 1999) and robust receptor internalization (Baiula et al., 2013; Corbani et al., 2004; Donica et al., 2013; Spampinato and Baiula, 2006; Spampinato et al., 2007; Spampinato et al., 2001). Phosphorylation of NOP by GRK3 results in desensitization, β-arrestin2-recruitment, internalization and arrestin-dependent JNK MAPK signaling (Hawes et al., 1998; Zhang et al., 2012). This finding is consistent with previous data suggesting that arrestin binding to GPCRs may enable MAPK activation and act to regulate receptor signaling (Bohn et al., 2004; Bruchas et al., 2006; Shenoy and Lefkowitz, 2011). PKC plays a role in GRK translocation from the cytosol to the membrane and subsequent NOP desensitization (Mandyam et al., 2002).

Thus, all 3 opioid receptors and NOP are coupled to similar effector systems that are determined by the local environment and the cell types that express them. They display functional selectivity to different agonists and are desensitized and internalized following GRK activation and beta-arrestin recruitment with chronic or repeated administration of agonists. Coupling to effectors is regulated by protein kinases that are activated with continued agonist administration, injury, inflammation, and stress. These regulatory pathways are well worked out for postsynaptic GPCRs. Studies of the regulation of presynaptic GPCRs are not as detailed, even though clear differences in signaling pathways have been documented depending on the synapse.

3. Regulation of presynaptic opioid receptors

Opioid and NOP receptors are expressed in presynaptic terminals and act to inhibit neurotransmitter release (Fig. 2). Vesicle release is strongly dependent on Ca²⁺ levels so inhibition of Ca²⁺ channels is an effective strategy to inhibit neurotransmitter release (Katz and Miledi, 1967; Littleton et al., 1993). Multiple signaling pathways have been identified that result in decreased Ca²⁺ levels in the presynaptic terminals, including direct inhibition of voltage-gated Ca²⁺ channels (Bardoni et al., 2014b; Heinke et al., 2011), inhibition of release from Ca²⁺ stores (Velazquez-Marrero et al., 2014) and activation of voltage-gated K⁺ channels (Barral et al., 2003; Vaughan et al., 1997b). However, opioid inhibition in the hippocampus was not dependent on Ca²⁺ or K⁺, so direct inhibition of vesicle release was proposed (Capogna et al., 1993). Later studies have indeed found that Gβγ subunits bind directly to synaptotagmin to inhibit vesicle release machinery (Blackmer et al., 2005; Blackmer et al., 2001). Opioid receptor-effector coupling is not dependent on the neurotransmitter being released at the synapse, so it is critical to study specific synapses within each circuit under study.

Figure 2. Summary of presynaptic opioid receptor signaling.

Opioid receptors couple to multiple effectors, some in common with receptors at postsynaptic sites, but the primary action is to inhibit the release of neurotransmitter. Stimulation of opioid receptors activates potassium channels (1) and inhibit calcium channels (2), thereby reducing neurotransmitter release. In addition, there is inhibition of neurotransmitter release via direct interaction of βγ subunits with the synaptotagmin-SNARE complex that controls vesicle fusion to the membrane (3). Phosphorylation of the receptors is required for internalization of presynaptic opioid receptors (4) but a role of β-arrestin recruitment has not been demonstrated. Lateral diffusion of a reserve pool of receptors allows a continuous/sufficient supply of opioid receptors at the synapse (5), thus explaining the lack of desensitization of presynaptic opioid and NOP receptors.

Regulation of presynaptic opioid GPCRs is significantly different from postsynaptic GPCRs because they are typically resistant to acute desensitization (Blanchet and Luscher, 2002; Fyfe et al., 2010; Jullie et al., 2020; Lowe and Bailey, 2015; Pennock et al., 2012; Rhim et al., 1993). That is, with continued agonist administration, little to no decrease in the ability to decrease neurotransmitter release is observed. Similar results are observed with other presynaptic GPCRs (Bouchet et al., 2022; Pennock et al., 2012; Wetherington and Lambert, 2002). However, chronic treatment with morphine (over days) results in diminished opioid receptor coupling to effectors in some presynaptic terminals (Fyfe et al., 2010; Lowe and Bailey, 2015) and enhanced receptor coupling in others (Ingram et al., 1998; Manzoni and Williams, 1999). It is remarkable, given the focus on biased opioid agonists as therapeutics, that regulation of presynaptic opioid receptors by GRKs, beta-arrestins and trafficking processes has only been studied in a handful of synapses. Here we will focus on opioid signaling pathways and mechanisms regulating presynaptic opioid receptors in brain circuits involved in pain and opioid reward.

3.1. Descending pain modulatory circuit

Opioid receptors are expressed throughout the descending pain modulatory circuit from the ventrolateral PAG to the rostral ventromedial medulla (RVM) to the spinal cord (Fig. 3). Activation of vlPAG output neurons by opioids results in antinociception (reviewed in (Heinricher and Ingram, 2020). All four receptors are expressed in mouse vlPAG (Vaughan et al., 2003), with little evidence for DOR signaling in the rat under normal conditions (Lau et al., 2020; Vaughan et al., 2003), although there is immunohistochemistry data indicating that DORs are localized to the cytoplasmic compartment of axon terminals within the rat PAG (Commons, 2003).

Figure 3. Opioid receptors in the descending pain modulatory circuit.

Left panel: MOR, KOR and NOP receptors are localized on presynaptic GABA and glutamate terminals in the PAG. DORs are localized to intracellular compartments under normal conditions. Some PAG to RVM projection neurons have postsynaptic MORs and are hyperpolarized by opioids (not shown). Right panel: RVM output neurons to the spinal cord are either pain inhibitory (OFF-cells) or pain facilitatory (ON-cells). Some RVM OFF cells are directly inhibited by KOR agonists, which also inhibit release from both GABA and glutamate terminals in the RVM. MOR and NOP agonists inhibit both GABA and glutamate release onto ON- and OFF cells. DORs are localized to intracellular compartments but are functional on the plasma membrane after chronic morphine treatment. ON cells are directly inhibited by MOR and both cell types are inhibited by NOP.

3.1.1. Presynaptic MOR signaling

Presynaptic MOR signaling has been studied extensively in the vlPAG because activation of vlPAG output neurons occurs through inhibition of GABA release in the vlPAG, termed disinhibition (Lau and Vaughan, 2014). MORs inhibit both glutamate and GABA release from presynaptic terminals impinging on vlPAG neurons (Chieng and Christie, 1994b; Vaughan and Christie, 1997; Vaughan et al., 1997b). Presynaptic MORs also disinhibit RVM neurons via inhibition of GABA release (Pan et al., 1990). Interestingly, MOR coupling to phospholipase A2 and activation of voltage-gated potassium (Kv) channels is responsible for inhibition of GABA release but this signaling pathway is not responsible for MOR-mediated inhibition of glutamate release (Vaughan et al., 1997b; Zhang and Pan, 2010). In fact, the coupling pathway in glutamate terminals has not been elucidated but may involve direct coupling of Gβγ subunits to release machinery.

As mentioned previously, acute opioid administration does not desensitize presynaptic opioid receptor-mediated inhibition of neurotransmitter release in the vlPAG (Fyfe et al., 2010). Continuous opioid exposure (min-hours) induces a shift to the left in the concentration-response curve indicating an increase in agonist affinity that is dependent on activation of adenylyl cyclase and PKA (Ingram et al., 1998), an effect replicated in MOR-expressing cell lines (Birdsong et al., 2013). Chronic morphine treatment (days) results in a decrease in the maximal inhibition of GABA release onto vlPAG neurons; however, there is no evidence for acute desensitization but instead a decrease in receptor reserve (Fyfe et al., 2010; Jullie et al., 2022). Therefore, there are some questions as to whether MORs in presynaptic terminals are regulated by similar processes to those characterized in postsynaptic locations, namely phosphorylation by GRKs, beta-arrestin binding and internalization/trafficking mechanisms. To get at this critical question, Jullie and colleagues used cultured neurons from a variety of brain areas with transfected tagged-MORs in presynaptic terminals to follow internalization and trafficking in response to agonists (Jullie et al., 2020). They observed GRK-dependent internalization that was rapid and reversible indicating for the first time that regulated endocytosis is also evident in presynaptic terminals. Interestingly, they also observed lateral movement of MORs from axons into synapses giving a mechanistic explanation for why there is no observed MOR desensitization in presynaptic terminals (Fig. 2). Further studies with chronic treatment (18 hours incubation in cultured striatal neurons) with the high efficacy agonist DAMGO showed GRK2/3-dependent internalization and loss of cell surface receptors explaining the decrease in presynaptic MOR function observed with long-term agonist treatment (Fyfe et al., 2010; Jullié et al., 2022). The role for beta-arrestin in regulation of MORs in presynaptic terminals is not as clear. An early study in the locus coeruleus and PAG examining MOR inhibition of neurotransmitter release in beta-arrestin KO mice observed enhanced presynaptic inhibition of GABA release, but also increased vesicular release in those animals due to impaired phosphodiesterase (PDE4) activity (Bradaia et al., 2005). Further studies are necessary to define the roles of beta-arrestin in regulation of MORs in presynaptic terminals.

Another adaptation with chronic morphine administration in the PAG and RVM is an increase in cAMP levels and PKA activation (Bie and Pan, 2005; Duman et al., 1988; Ingram et al., 1998). In fact, presynaptic MOR signaling in the PAG is dependent on PKA with chronic morphine treatment, with no effect of Kv channel blockers that are required for MOR inhibition of GABA release under normal conditions (Ingram et al., 1998). Thus, it is important to note that MOR coupling to effectors can be changed with continuous exposure to morphine.

3.1.2. Presynaptic KOR signaling

Presynaptic KORs inhibit GABA and glutamate release in the PAG (Lau et al., 2020; Vaughan and Christie, 1997), as well as in RVM (Bie and Pan, 2003). KORs also inhibit RVM GABA/glycine terminals in the spinal cord (Otsu and Aubrey, 2022). Postsynaptic KORs are localized to RVM OFF-cells and act to block morphine analgesia (Meng et al., 2005).

The signaling pathways involved in presynaptic KOR-mediated inhibition of neurotransmitter release in either the PAG or RVM have not been determined. In the BNST, KOR inhibition of GABA release is blocked with an ERK1/2 inhibitor, one of the first demonstrations that ERK signaling affects presynaptic opioid receptor signaling (Li et al., 2012). This study assessed the dependence on Ca²⁺, p38 and ERK signaling but did not assess cAMP, or K⁺ channel signaling pathways. Stimulation of ERK signaling in presynaptic terminals is interesting since MAPK signaling pathways are typically linked to communication with the nucleus. Since axon terminals are typically located at substantial distances from the nucleus, the cellular role of MAPK and other kinases activated by receptors localized to presynaptic terminals is a question that needs to be studied.

3.1.3. Presynaptic DOR signaling

DORs are expressed in presynaptic terminals in multiple brain areas (Arvidsson et al., 1995; Bardoni et al., 2014a; Commons et al., 2001; He et al., 2021; Pradhan et al., 2009; Pradhan et al., 2010; Scherrer et al., 2006; Yamada et al., 2021). Presynaptic DOR coupling to voltage-gated K⁺ channels has been observed in RVM (Zhang and Pan, 2010, 2012) and inhibition of Ca²⁺ channels in hippocampal synapses (He et al., 2021) and spinal cord dorsal horn (Bardoni et al., 2014a) under normal conditions.

As mentioned earlier, there is little functional evidence for DOR signaling in presynaptic terminals in the PAG (Vaughan et al., 2003) but DOR has been visualized in intracellular compartments (Commons, 2003). Chronic morphine treatment upregulates presynaptic DORs in several brain regions (Hack et al., 2005; Ma et al., 2006; Zhang and Pan, 2010) where they inhibit neurotransmitter release in a cAMP and PKA-dependent manner, similarly to MORs with chronic morphine administration. Interestingly, presynaptic DOR signaling in animals pretreated with chronic morphine and the induced DOR signaling is dependent on MORs and beta-arrestin signaling (Hack et al., 2005) suggesting that internalization processes are critical for the emerging DOR activity. Unlike MORs, trafficking of DORs in presynaptic terminals is not dependent on GRK2/3 phosphorylation or phosphorylation of the C-terminal tail (Jullié et al., 2022). However, trafficking of vesicles may be required to get DORs to the plasma membrane. Further studies are necessary to understand the cellular mechanisms underlying upregulation of DORs with chronic opioid treatment. DORs in vlPAG are also internalized after swim stress (Commons, 2003) indicating additional regulatory processes exist. Visualization of DOR trafficking in presynaptic terminals showed that while both MORs and DORs undergo internalization, DORs recycle less efficiently, resulting in a greater degree of tolerance following 18 hours of agonist administration (Jullié et al., 2022).

3.1.4. Presynaptic NOP regulation

Presynaptic NOP receptors are expressed in many of the same brain areas as opioid receptors; however, their distribution on GABAergic and glutamatergic synapses differ in many instances. For example, within the PAG, NOPs inhibit both evoked GABAergic and glutamatergic synaptic currents in approximately 50% of neurons throughout the PAG, except for the vlPAG where NOP agonists inhibit evoked GABAergic synaptic currents in all neurons (Vaughan and Christie, 1997). Within the RVM, N/OFQ inhibits evoked GABAergic synaptic currents, but not evoked glutamatergic synaptic currents (Vaughan et al., 2001), similar to MORs (Pan et al., 1990). In the spinal cord dorsal horn, NOPs are expressed on glutamatergic terminals in the spinal cord dorsal horn and inhibit glutamate release, but do not affect GABA release (Ahmadi et al., 2001a; Ahmadi et al., 2001b; Liebel et al., 1997; Zeilhofer et al., 2000). The NOP-mediated inhibition of GABA release in the vlPAG and RVM should result in antinociception, similar to opioids. Interestingly, NOPs are expressed postsynaptically on most vlPAG and RVM neurons and hyperpolarize these neurons (Vaughan et al., 2001; Vaughan et al., 1997a), reducing activity of output neurons. Thus, NOP agonists are characterized as having anti-opioid behavioral effects (Grisel et al., 1996a; Mogil et al., 1996a; Morgan et al., 1997a).

A few studies have examined the signaling pathways used by NOPs in presynaptic terminals. NOPs inhibit N-type Ca²⁺ channels in vestibular afferents (Sesena et al., 2020) and N- and P-type Ca²⁺ channels in the hypothalamus (Gompf et al., 2005). Within the RVM, N/OFQ, unlike opioids, does not affect spontaneous miniature inhibitory synaptic currents frequency under basal conditions (Vaughan et al., 2001). However, N/OFQ reduces the frequency of miniature events when external K⁺ is elevated via a Ca²⁺-dependent mechanism suggesting that NOPs are coupled to Ca²⁺ channels in RVM. Presynaptic NOP inhibition of neurotransmitter release does not desensitize during prolonged agonist administration (Pennock et al., 2012), similar to other presynaptic opioid receptors.

As with the presynaptic opioid receptors, relatively little is known concerning the cellular mechanisms underlying regulation of NOPs in terminals. Studies of GRK and beta-arrestin knockout animals have provided some insight into the roles of these proteins in the actions of opioid and NOP agonists but did not directly examine presynaptic actions of NOP in these models.

3.2. VTA reward circuits

Opioids injections into the ventral tegmental area (VTA) and nucleus accumbens increase dopamine (DA) release that contributes to opioid reward (Koob et al., 1998; Wise, 1988). Opioids induce reward through disinhibition of VTA DA neurons (Fields and Margolis, 2015; Johnson and North, 1992). Multiple opioid receptors are expressed in presynaptic terminals impinging on VTA DA neurons, as well as on DA terminals in projection areas (Fig. 4).

Figure 4. Reward circuitry and opioid receptors.

Schematic depicting opioids and NOP receptors in a brain region involved in reward circuitry, the ventral tegmental area (VTA). At the presynaptic terminals, MOR receptors are located on glutamatergic and GABAergic synapses, whereas DOR and NOP receptors are found only on GABA terminals, allowing for disinhibition of DA neurons. At the postsynaptic level, KOR and NOP receptors are present in DA neurons, while KOR receptors are also found in terminals projecting to the nucleus accumbens (NAc).

3.2.1. Presynaptic MORs in VTA

Presynaptic MOR signaling is important in GABAergic (Bull et al., 2017a; Johnson and North, 1992; Lowe and Bailey, 2015; Matsui et al., 2014; Matsui and Williams, 2011; St Laurent et al., 2020) and glutamatergic synapses (Manzoni and Williams, 1999; Margolis et al., 2005) in the VTA. Inhibition of GABA inputs onto DA neurons disinhibits dopamine release in the VTA (Bosse and Kuschinsky, 1976; Gysling and Wang, 1983; Teuchmann and Kania, 1977) which is involved in the rewarding effects of opioids administered directly into the VTA (Bozarth and Wise, 1984). MOR signaling in VTA GABAergic terminals is via beta-arrestin and c-Src (Bull et al., 2017a). This signaling cascade had been shown to mediate MOR coupling to voltage-activated calcium channels in dorsal root ganglion neurons (Raingo et al., 2007; Walwyn et al., 2007). Inhibition of Src kinase has no effect on morphine antinociception but blocks the development of tolerance (Bull et al., 2017b). In contrast, MOR activation of a phospholipase A2/Kv channel signaling pathway results in inhibition of glutamate release (Manzoni and Williams, 1999). Thus, these data indicate that MOR signaling in these different axon terminals are different and can be pharmacologically dissociated from one another.

Presynaptic MOR inhibition of GABA release does not desensitize with prolonged agonist administration; however, activation of PKC induces desensitization of MORs in VTA terminals, at least to morphine (Lowe and Bailey, 2015). In addition, MOR inhibition of GABA release onto VTA dopamine neurons is reduced in beta-arrestin KO mice (Bull et al., 2017b) even though opioid-induced DA release in the VTA is potentiated (Bohn et al., 2003) in these mice indicating that beta-arrestins have different roles in GABA compared to dopamine release mechanisms. Further studies are necessary to characterize the role of beta-arrestins in presynaptic terminals in this circuit as well.

Chronic opioid administration potentiates the inhibition of glutamate release in VTA synapses through a cAMP/PKA-dependent mechanism, but MOR-KV channel coupling is decreased (Manzoni and Williams, 1999). Uncoupling of MORs from specific effectors occurs with chronic morphine, but MORs signal efficiently through other effectors in some synapses. Thus, whether presynaptic desensitization occurs with long-term opioid administration can depend on the signaling pathway that is measured.

Not all inputs to any single brain area are sensitive to opioid inhibition. The VTA receives GABA inputs from multiple brain areas, as well as from interneurons within the VTA. However, GABAergic inputs from the RMTg are more strongly inhibited by MOR agonists than those from the nucleus accumbens or local interneurons (Matsui et al., 2014). Accordingly, MORs in RMTg terminals have reduced function following chronic morphine treatment, although the signaling pathways have not been studied in detail.

3.2.2. Presynaptic KORs in the VTA

Presynaptic KORs inhibit GABA release within the VTA (Ford et al., 2007), but do not modulate glutamate release (Manzoni and Williams, 1999) or dopamine release measured with microdialysis (Devine et al., 1993b). Postsynaptic KORs directly hyperpolarize a subpopulation of VTA DA neurons and are associated with aversive behaviors (Margolis et al., 2003). Interestingly, DA neurons that project to the NAc are not hyperpolarized by KOR agonists but KORs localized to presynaptic terminals in the NAc inhibit neurotransmitter release (Margolis et al., 2006), at least in the rat. In the mouse, VTA DA neurons are hyperpolarized and somatodendritic dopamine release is inhibited by KOR agonists as measured by dopamine D2 receptor activation of GIRK channels (Ford et al., 2006). Few studies have assessed KOR signaling pathways involved in presynaptic terminals. One study using a conditional p38 KO transgenic mouse found that p38 KO does not affect DA release in the NAc directly but blocks KOR-mediated tyrosine phosphorylation of a GIRK channel subunit, resulting in decreased DA release and aversion behaviors (Ehrich et al., 2015). These actions of KOR likely contribute to the aversive effects of KOR agonists.

3.2.3. Presynaptic DORs in the VTA

DOR agonists hyperpolarize subpopulations of VTA neurons (Margolis et al., 2017); however, also potentiate release of DA in the striatum (Devine et al., 1993a; Devine et al., 1993b) suggesting that DORs disinhibit DA neurons that project to the NAc. Indeed, rats will self-administer both MOR and DOR selective agonists into the VTA, although MOR agonists are more effective (Devine and Wise, 1994). DORs do not affect glutamate release onto DA neurons (Manzoni and Williams, 1999).

3.2.3. Presynaptic NOPs in the VTA

In the ventral tegmental area (VTA), postsynaptic NOPs directly inhibit VTA neurons by activating GIRK channels (Zheng et al., 2002), but also indirectly inhibit VTA neurons through potentiation of GABA_A receptor currents (Driscoll et al., 2020). This contrasts with the disinhibition of VTA dopamine neurons induced by presynaptic MORs on GABAergic terminals within the VTA. Interestingly, NOPs are also localized to GABAergic terminals within the VTA and inhibit release (Zheng et al., 2002). Thus, the actions of nociceptin acting at NOPs in this circuit are complex.

4.0. Physiological consequences of differential regulation of pre- vs. postsynaptic receptors

The main difference between regulation of pre- and postsynaptic opioid and NOP receptors is the fact that presynaptic receptors are functional with high doses of agonist without desensitization (Fyfe et al., 2010; Lowe and Bailey, 2015; Pennock et al., 2012; Pennock and Hentges, 2011), mainly due to a reserve pool and local trafficking of the receptors into synapses, at least for MORs and DORs (Jullie et al., 2022). There is evidence for phosphorylation of the receptors by GRK in both locations and hints that presynaptic receptors are internalized by beta-arrestin-dependent mechanisms similarly to postsynaptic receptors. However, the functional consequences of these regulatory mechanisms are quite different in presynaptic terminals in terms of signaling and coupling to signaling pathways. In fact, agonist activation of presynaptic receptors in GRK and beta-arrestin knockout mice is functionally equivalent to wildtype mice. Although opioid-induced antinociception and other behaviors are altered in these knockout mouse models, an alternative interpretation to GRK and beta-arrestin signaling being involved in the behavioral changes is that there is an imbalance between pre- and postsynaptic opioid receptor signaling. Many of the studies to date using GRK and beta-arrestin knockouts have documented changes in antinociception, tolerance, withdrawal, dependence, and opioid-dependent side effects. The majority of these studies assess localization and trafficking of postsynaptic MORs in a few brain areas, but less is known about the role of presynaptic MORs in these behaviors in the transgenic models.

4.1. Acute opioid administration

In the case of supraspinal analgesia, presynaptic MORs in the vlPAG are important for initiating antinociception via descending pain circuits (Lau and Vaughan, 2014; Morgan et al., 1991; Morgan et al., 1992). In thinking about a single injection of morphine (or other opioids), postsynaptic MORs undergo desensitization within minutes with no change in presynaptic MOR signaling (Fig. 5). Resensitization of MORs at the plasma membrane occurs over ~45 min (Arttamangkul et al., 2012). Although postsynaptic MOR desensitization reaches a new steady-state, the balance of presynaptic and postsynaptic signaling is shifted to presynaptic signaling. In addition, opioid agonists are generally more potent in presynaptic terminals (Fyfe et al., 2010; Pennock and Hentges, 2011). Phosphorylation-deficient MORs show a decrease in rapid internalization of presynaptic MORs (Jullie et al., 2022) and transgenic mice expressing these receptors have potentiated antinociception (Kliewer et al., 2019). This is consistent with data from beta-arrestin KO mice where dose-response curves for morphine antinociception are shifted to the left (Bohn et al., 1999). However, it is clear that these adaptations are not the result of blocking MOR desensitization in the presynaptic terminals, but are more likely to be the result of other adaptations in the terminals affecting vesicular release (Bradaia et al., 2005). Further, blockade of GRK2/3 has no effect on MOR inhibition of GABA release in the vlPAG (Bouchet et al., 2022). These results confirm that GRK and beta arrestin-mediated internalization of MORs in presynaptic terminals do not decrease MOR signaling to effectors in presynaptic terminals, at least with single injection or drug administration on a minute to hour time scale (Fig. 5).

Figure 5. Time course of pre- and postsynaptic receptor signaling.

General schematic of pre- and postsynaptic opioid receptor regulation after a single injection of morphine (sec-hours) or prolonged exposure over days that result in tolerance (left and right panel, respectively). Single injection of morphine: After opioid ligand binding, each step corresponds to the time course (logarithmic scale). Below, graph represents the analgesia following opioid exposure over time. A single exposure to opioid agonist leads to the coupling of opioid receptors with its ligand, internalization and possibly endosomal signaling to kinase cascades, such as the MAPK pathway. The signaling cascade starts with the rapid phosphorylation of the receptor by GRK (<20 seconds) while arrestin binding occurs within 5 minutes. Finally, receptor endocytosis is observed within minutes, and receptor recycling/resensitization in 45 minutes - 1 hour. Thus, at the peak of the antinociceptive effect, many postsynaptic opioid receptors are desensitized. However, in presynaptic terminals, lateral diffusion of receptors prolongs presynaptic signaling and maintains a receptor reservoir at the synaptic cleft. Presynaptic signaling is maintained throughout the antinociceptive response. Repeated subchronic injections of morphine (right panel): Repeated or prolonged administration of morphine induces adaptations that decrease both pre- and postsynaptic opioid receptor responses. These adaptations are caused by opioid receptor coupling to different effectors, including kinases.

As shown in Figs. 3 and 4, multiple opioid receptors are expressed on presynaptic terminals and postsynaptic cell bodies in the vlPAG, RVM and VTA. The redundancy suggests that endogenous opioid signaling is intricately regulated. Endogenous opioid peptides are released with temporal and spatial specificity (Bagley and Ingram, 2020; Kissiwaa et al., 2020; Winters et al., 2017). Peptides, even within the same family, activate receptors coupled to different effectors in different cell compartments (Kunselman et al., 2021a). In addition, opioid peptides are processed within vesicles and rapidly degraded by peptidases (Fricker et al., 2020; Gupta et al., 2014). Opioid biosensors are currently being developed and used to visualize endogenous opioid peptide activation of opioid receptors (Massengill et al., 2022; Razlansari et al., 2022; Stoeber et al., 2018). Similarly, biosensors for signaling pathways are also being developed, such as for cAMP (Massengill et al., 2022) and PKA (Ma et al., 2022). These tools will help to dissect the role of the different opioid peptides in different circuits.

Although studies to date provide evidence that GABAergic and glutamatergic terminals are inhibited by multiple opioid receptor subtypes (Figs. 3 and 4), the majority of studies can not discern whether the expression of these receptors are co-localized to the same terminals. Thus, although the schematics show that there is redundancy of receptor subtypes on inputs to a single brain area, the receptors could be expressed selectively on inputs from different circuits. Given that opioid peptides are synthesized and processed within specific neuron populations, they will be released during activation of different circuits. Optogenetic studies are already characterizing projections that are selectively inhibited by specific opioid receptors. For example, MORs and DORs target different terminals in the stratum from the medial thalamus and anterior cingulate cortex (Birdsong et al., 2019). These results also suggest that specific circuits could be targeted by novel therapeutics in the future.

4.2. Chronic opioid administration

Chronic or repeated agonist exposure (> 12 hours to days) produces a loss in effector coupling (cellular tolerance) in both pre- and postsynaptic locations. However, opioid receptors can couple to additional effectors after chronic opioid administration and the functional consequences can be either a reduction in inhibition of neurotransmitter release (Fyfe et al., 2010) or enhanced inhibition of neurotransmitter release (Manzoni and Williams, 1999), depending on the synapses. MOR phosphorylation and trafficking directly impact opioid-mediated pain behaviors and tolerance. Loss of MOR phosphorylation abolishes behavioral tolerance to agonists (Arttamangkul et al., 2018; Grecksch et al., 2011) indicating that phosphorylation of MOR is intricately involved with opioid behaviors and that there are differences between agonists. Generally, these mutations result in greater antinociception and interfere with the development of tolerance and dependence, similar to results observed for GRK (Gluck et al., 2014) and beta-arrestin KO mice (Bohn et al., 1999). Mutating all of the serine and threonine residues in the MOR C-terminal tail potentiates MOR-mediated inhibition of vesicle endocytosis but abolishes rapid internalization and the development of tolerance with chronic treatment of the MOR agonist DAMGO (Jullie et al., 2022). Thus, phosphorylation of MORs negatively regulates MOR activity and positively regulates downstream adaptations that lead to the development of tolerance. In addition, knockout of beta-arrestin unmasks other signaling pathways, including JNK (Mittal et al., 2012). The long-term adaptations that result from knockdown or knockout of the critical proteins involved in GPCR regulation make it difficult to fully delineate the roles that these proteins play in behavior. The results with phosphorylation-deficient mice generally recapitulate the enhanced antinociception observed with GRK and beta-arrestin KOs effects of knocking out GRK and beta-arrestin but also clearly show an increase of opioid-mediated side effects (Kliewer et al., 2019).

One interesting observation is the fact that chronic opioid treatment results in ERK1/2, PKA and JNK kinase activation. Increasing PKA activity induced by chronic morphine leads to sustained neurotransmitter release (Fyfe et al., 2010; Zhang and Pan, 2010). Other opioid-dependent adaptations also increase neurotransmitter release (Bie et al., 2009b; Fyfe et al., 2010; Ma et al., 2006; Madhavan et al., 2010; Manzoni and Williams, 1999; Zhang and Pan, 2010), increased firing of MOR-expressing neurons (Bagley et al., 2005b), and changes in GABAA receptor function (Bobeck et al., 2014).

These adaptations can be the result of signaling via endosomes (Stoeber et al., 2018) or prolonged signaling from the membrane in situations where endocytosis is blocked (Madhavan et al., 2010). Activation of these kinases are dependent on beta-arrestin recruitment and/or the increase in vesicle trafficking for exocytosis is dependent on beta-arrestin-dependent trafficking processes. The impact of activating these kinases is not fully understood, especially in presynaptic terminals. In fact, the time course of activation of these kinase effectors is not known because many kinase assays are run under steady-state conditions for 30 min or longer. The new biosensors that are being developed will probably provide new information in this area. Many of these kinase cascades are known to induce signaling to the nucleus but it is not known whether kinase activation in presynaptic terminals also communicate with the nucleus, located some distance away. Alternatively, kinase activation in axon terminals may modulate other processes, such as mitochondrial function (Jaferi et al., 2009).

5.0. Concluding remarks

The regulatory processes of presynaptic opioid and NOP receptors are understudied relative to postsynaptic receptors, especially with repeated and long-term administration of opioids. The number of signaling pathways that are recruited and that modulate neurotransmitter release in different synapses represent possible therapeutic targets for pain and addiction. As the field moves toward using unbiased transcriptomic and proteomic datasets to identify new targets for development of pain and addiction therapeutics, it is important to note that these techniques underestimate regulatory processes in presynaptic terminals because many opioid-modulated inputs arise from neurons located outside the brain area collected for the analyses. Recent data suggesting that there are differences in translational regulation in soma and dendrites/terminal fields are particularly intriguing (Biever et al., 2019; Glock et al., 2017; Perez et al., 2021; Perez and Schuman, 2022; Scarnati et al., 2018). Newly generated transgenic models, such as the phosphorylation-deficient MOR mice (Kliewer et al., 2019) and other conditional KO models in specific cell types and functional electrophysiological studies examining presynaptic opioid signaling will provide more information on how opioids regulate specific circuits. The breadth of opioid agonists and biased signaling profiles of the agonists will help to identify new avenues for development of drugs for pain with reduced abuse liability.

Highlights.

• Opioid receptors are expressed throughout the neuraxis and their activity is highly regulated through desensitization, internalization and recycling processes.

• Presynaptic opioid and nociceptin receptors are resistant to desensitization compared to postsynaptic receptors.

• The functional consequences of desensitization and internalization is dependent on the pool of receptors in the local environment and coupling to effectors.

• Opioid receptor-effector coupling is altered with continued administration of agonist.