Mu-Opioid Receptors Expressed in Glutamatergic Neurons are Essential for Morphine Withdrawal

Abstract

Although opioids still remain the most powerful pain-killers, the chronic use of opioid analgesics is largely limited by their numerous side-effects, including opioid dependence. However, the mechanism underlying this dependence is largely unknown. In this study, we used the withdrawal symptoms precipitated by naloxone to characterize opioid dependence in mice. We determined the functional role of mu-opioid receptors (MORs) expressed in different subpopulations of neurons in the development of morphine withdrawal. We found that conditional deletion of MORs from glutamatergic neurons expressing vesicular glutamate transporter 2 (Vglut2⁺) largely eliminated the naloxone-precipitated withdrawal symptoms. In contrast, conditional deletion of MORs expressed in GABAergic neurons had a limited effect on morphine withdrawal. Consistently, mice with MORs deleted from Vglut2⁺ glutamatergic neurons also showed no morphine-induced locomotor hyperactivity. Furthermore, morphine withdrawal and morphine-induced hyperactivity were not significantly affected by conditional knockout of MORs from dorsal spinal neurons. Taken together, our data indicate that the development of morphine withdrawal is largely mediated by MORs expressed in Vglut2⁺ glutamatergic neurons.

Electronic supplementary material

The online version of this article (10.1007/s12264-020-00515-5) contains supplementary material, which is available to authorized users.

Introduction

Opioids have been used in pain management for a long time, but their chronic use has been limited by numerous side-effects. Among these, opioid-induced dependence causes drug craving and prolongs medication, leading to a public health problem. Multiple maladaptive behaviors induced by the termination of opioid drugs, termed opioid withdrawal, are often used to characterize opioid dependence. Mu-opioid receptors (MORs), encoded by Oprm1 [1–3], are responsible for opioid withdrawal, as MOR-knockout mice do not express naloxone-precipitated opioid withdrawal behavior [4]. Like other opioid receptors, MORs are coupled with inhibitory G proteins, which in turn activate several intracellular effectors [5–7]. Repeated opioid administration impairs the homeostasis of the nervous system, as reflected by adaptive changes at both the cellular and synaptic levels [8]. These adaptations can lead to dysfunction of excitatory and inhibitory synaptic transmission [8–10].

Recent studies have started to examine the contribution of MORs expressed in different neuronal populations to morphine analgesia and other effects, including withdrawal, using genetic approaches [11–13]. As forebrain GABAergic neurons are an important substrate for opioid reward and drug addiction [14], several studies have examined the functional role of MORs in forebrain GABAergic neurons [15–17], and found that these MORs play essential roles in locomotor and motivational effects, as well as alcohol reward [15, 16]. Although MORs in striatal GABAergic neurons play an important role in opioid reward, these MORs are not involved in morphine withdrawal [17]. Thus, MORs expressed in forebrain GABAergic neurons are not the key component for the aversive effects in opioid adaptation. The involvement of excitatory transmission during opioid dependence has also been examined in several studies. It has been shown that the excitatory projection from the paraventricular nucleus of the thalamus to the nucleus accumbens [18], and the projection from the lateral habenula to the raphe nucleus [10] are engaged in aversive behaviors during morphine withdrawal. The brain areas with both dense expression of MORs and essential roles in aversive behaviors are potentially involved in naloxone-precipitated withdrawal. A recent study found that MORs expressed in the medial habenula significantly contribute to aversive effects in opioid withdrawal, but not the analgesic, locomotor, or reward effects of morphine [19]. Although the involvement of multiple neuronal populations in morphine withdrawal has been investigated, the mechanism underlying opioid dependence and the precise contribution of MORs in different neuronal subpopulations to withdrawal symptoms need to be further investigated.

MORs are widely expressed in the dorsal spinal cord [20], and they are important for morphine analgesia [21, 22]. Previous studies have focused on microglial adaptation in the spinal cord during opioid dependence, and found chronic morphine-induced activation [23–26]. In addition, morphine withdrawal also induces long-term synaptic facilitation in the superficial dorsal spinal cord [27], which is a result of the ATP release induced by activation of the spinal microglial pannexin-1 channel [28]. However, the potential involvement of MORs expressed in dorsal spinal neurons in morphine withdrawal is largely unknown.

To characterize the functional role of MORs expressed in different neuronal populations in morphine withdrawal, we investigated the contribution of MORs in distinct neuronal subpopulations to morphine withdrawal in MOR conditional-knockout mouse lines.

Materials and Methods

Animals

Vglut2-ires-Cre (JAX #016963, referred to as Vglut2-Cre), Vgat-ires-Cre (JAX #016962, referred to as Vgat-Cre), and Lbx1-Cre [29] mice were used for experiments. We generated an MOR conditional-knockout line, referred to as Oprm1^fl/fl, in which two loxP sites flanking Oprm1 exon 2 and exon 3 were inserted to an Oprm1 allele of Oprm1^fl/fl mice (Fig. 1A), allowing for conditional deletion of MORs in neurons with Cre recombinase. Oprm1^fl/fl mice were crossed with Vglut2-Cre, Vgat-Cre or Lbx1-Cre mice to obtain the MOR conditional-knockout mice, referred to as Vglut2-Cre/Oprm1^fl/fl, Vgat-Cre/Oprm1^fl/fl, and Lbx1-Cre/Oprm1^fl/fl mice. Male mice (8 weeks–12 weeks old) were used for behavioral experiments. All mice were raised under a 12-h light/dark cycle (lights on at 07:00) with ad libitum food and water. All procedures were approved by the Animal Care and Use Committee of the Institute of Neuroscience, Chinese Academy of Sciences, Shanghai, China.

Fig. 1.

Generation and verification of the MOR conditional-knockout mouse line. A Targeting strategy for generating the Oprm1^fl/fl mouse line. B Representative images of Oprm1 expression in Oprm1^fl/fl and Vglut2-Cre/Oprm1^fl/fl mice determined by in situ hybridization (scale bar, 500 μm). C Representative images of MOR expression in Oprm1^fl/fl and Vglut2-Cre/Oprm1^fl/fl mice determined by fluorescent immunostaining (scale bar, 200 μm). Cpu, striatum; NAc, nucleus accumbens; AcbC, accumbens nucleus, core; MHb, medial habenula; PV, paraventricular thalamic nucleus; CM, central medial thalamic nucleus; CeA, central amygdala; BLA, basolateral amygdala; IPN, interpeduncular nucleus; VTA, ventral tegmental area; PBN, parabrachial nucleus; scp, superior cerebellar peduncle; LPB, lateral parabrachial nucleus.

Tissue Preparation

Each mouse was anesthetized with pentobarbital sodium (0.1 mg/g) and perfused transcardially with saline followed by 4% paraformaldehyde (PFA; Sigma-Aldrich, St. Louis, MO). The brain and spinal cord (lumbar segments) were dissected. The brain was weighed and post-fixed overnight in 4% PFA at 4°C, followed by cryoprotection in 30% sucrose in phosphate-buffered saline (PBS) at 4°C for further experiments.

In Situ Hybridization

In situ hybridization was performed as previously described [30]. The Oprm1 probe was made using digoxigenin (DIG)-labeled nucleotides. Brain sections were cut at 40 μm on a cryostat (Leica CM 1950, Buffalo Grove, IL) and floated in diethyl pyrocarbonate (DEPC)-PBS. The sections were washed once in DEPC-PBS and then twice in DEPC-PTw (PBS-Tween, 0.1% Tween-20 in PBS) for 5 min each. After immersion in 2 × saline sodium citrate (SSC) with 0.5% Triton for 30 min and washing twice in DEPC-PTw for 5 min, the sections were treated with proteinase K (4 μg/mL) for 5 min, and then fixed in 4% PFA for 10 min. After washing once in DEPC-PTw, followed by acetylation in triethanolamine (0.1 mol/L) with 0.25% acetic anhydride for 10 min, sections were washed twice in DEPC-PBS for 5 min. Sections were pre-hybridized for 3 h at 65°C in hybridization buffer, then hybridized in hybridization buffer containing 1 μg/mL Oprm1 probe for 12 h–16 h at 65°C. A series of washes were performed sequentially as follows: pre-hybridization buffer and pre-hybridization/TBST (1% Tween-20, 500 mmol/L NaCl, 20 mmol/L Tris) at 65°C, each for 30 min; 2 × 5 min in TBST at room temperature; one rinse in TBST/Tris-acetate-EDTA (TAE); and 3 × 5 min in TAE at room temperature. To remove the probe, sections were collected into wells in 2% agarose/TAE gel, and electrophoresed at 60 V for 2 h. After washing twice in TBST, the sections were next incubated with sheep anti-DIG-alkaline phosphatase (AP; 1:2000 in 0.5% blocking reagent) overnight at 4°C. The next day, after 3 × 30 min washes in TBST and 2 × 30 min in AP buffer (100 mmol/L Tris-HCl, 100 mmol/L NaCl, 0.1% Tween-20, 50 mmol/L MgCl₂), sections were visualized in AP buffer containing 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium and fixed in 4% PFA after colorization.

MOR Immunofluorescent Staining

Brain sections were cut at 40 μm on a cryostat for immunofluorescent staining. The sections were blocked for 1 h at room temperature in PBST (0.3% Triton X-100) with 5% normal donkey serum, followed by incubation with primary antibody at 4°C overnight and secondary antibody at room temperature for 2 h in PBST (0.3% Triton X-100) with 1% normal donkey serum. The primary antibody used was anti-MOR [17, 31] (rabbit, 1:500, ab134054 Abcam, Cambridge, MA), and the secondary antibody was donkey anti-rabbit IgG-Cy3 (1:200, 711-165-152, Jackson ImmunoResearch Laboratories, West Grove, PA).

Behavioral Procedure for Morphine Withdrawal

The procedure for morphine withdrawal was adapted from previous studies [26, 28]. Briefly, mice were randomly divided into groups and received escalating doses of morphine (7.5 mg/kg–50 mg/kg, i.p.) or saline as the control twice daily for four consecutive days with an 8-h interval (day 1, 7.5 and 15 mg/kg; day 2, 20 and 25 mg/kg; day 3, 30 and 35 mg/kg; and day 4, 40 and 45 mg/kg). On day 5, the mice received a single morphine (50 mg/kg) or saline injection and were challenged with naloxone (2 mg/kg, i.p.) 2 h after the last injection to induce acute withdrawal. Next, each mouse was placed in a transparent box and somatic signs of withdrawal were videotaped and scored across 30 min by an observer blinded to the treatment. Jumping, teeth-chartering, headshakes, wet-dog shakes, and grooming behaviors were evaluated and scored from 0 to 3 based on behavioral bouts (0 = absent; 1, 1–3 bouts; 2, 4–6 bouts; 3, ≥7) at 5-min intervals. Piloerection, paw tremor, and twitching were also evaluated; one point was given when each sign appeared in the 5-min interval. The total bouts of each behavior and the global withdrawal scores were calculated.

Open Field Test

Locomotor activity was evaluated in the open field test. Mice were placed in the testing room for ~1 h for habituation. In the baseline test, each mouse was placed in an open field box (40 × 40 × 40 cm³) and videotaped for 10 min. Then the mouse was removed from the box and subcutaneously injected with 10 mg/kg morphine. The test was run again 40 min after morphine injection. Total travel distance and average speed in a 10-min period were recorded. The track was automatically analyzed using LabState (AniLab, Ningbo, China) or EthoVision XT (Noldus, Wageningen, Netherlands) software.

Elevated Plus Maze Test

The elevated plus maze test was used to assess anxiety-like behavior [32]. The maze contained two opposing closed arms (30 cm × 5 cm× 15 cm), two open arms (30 cm × 5 cm), and a center zone (5 cm × 5 cm). The platform was 50 cm above the floor. At the beginning of the test, each mouse was placed in the central zone, facing one of the open arms. In the subsequent 5 min, the total time spent in the open arms, closed arms, and center zone, and the number of entries into the open arms and closed arms were recorded. The track was automatically analyzed using EthoVision XT (Noldus) software.

Tail Suspension Test

The tail suspension test was used to assess depression-like behavior [33]. During the test, each mouse was suspended by tape ~1.5 cm from the tail tip, and ~20 cm above the floor. The mice were videotaped for 6 min, and the time spent immobile was scored manually. The suspension test was repeated once if a mouse climbed up its tail.

Novel Object Test

The novel object test was used to evaluate the exploration behaviors and anxiety levels [34]. A circular wooden cone (base diameter, 5 cm; height, 8 cm) was placed in the center of an open field arena (40 × 40 cm²) and the entries into and time spent exploring the object in a 13.3 × 13.3 cm² square in the center of the arena were recorded in a 10-min period. The track was automatically analyzed using EthoVision XT (Noldus) software.

Sucrose Preference Test

The sucrose preference test was used to assess anhedonia behavior. We adapted the procedure from previous studies [35, 36]. Mice were housed individually and habituated before the test. The experimental mice were habituated with two identical tubes, one containing 1% sucrose and the other containing water in their home cages for 1 day. The positions of the tubes were switched once to reduce position preference. Then, these mice were deprived of both water and food for 24 h. During the test, two identical tubes containing 1% sucrose and water were placed in the cage and left for 2 h, and the positions of the tubes were switched and weighed after 1 h. The consumption of each fluid in the 2-h interval was measured by weight, then the sucrose preference was calculated as the percentage of sucrose intake over the total fluid intake in the first hour and the whole 2 h.

Image Acquisition and Statistical Analysis

Images were captured using an Olympus VS120 microscope (Olympus, Tokyo, Japan). Statistical analysis was performed using Prism 6 (GraphPad Software, San Diego, CA). The data were analyzed using two-way ANOVA, one-way ANOVA followed by Bonferroni post hoc analysis, and the two-tailed unpaired t-test. The cut-off for significance was set at P = 0.05. All data are presented as the mean ± SEM.

Results

MORs Expressed in Vglut2⁺ Glutamatergic Neurons are Essential for Morphine Withdrawal

MORs are widely expressed in the nervous system and play an important role in mediating opioid dependence [4]. Opioid dependence is commonly characterized by the withdrawal symptoms at the termination of drug use. We investigated the functional role of MORs expressed in different subpopulations of neurons in morphine withdrawal by selectively deleting MOR in specific neuronal subpopulations using the conditional–knockout strategy. To achieve this, we generated a Oprm1^fl/fl mouse line, in which two loxP sites were inserted to flank exon 2 and exon 3 of the Oprm1 gene (Fig. 1A), allowing for conditional deletion of MOR with Cre recombinase. We first investigated the functional role of MORs expressed in glutamatergic neurons, most of which express Vglut2. We thus crossed Vglut2-Cre mice with Oprm1^fl/fl mice to obtain Vglut2-Cre/Oprm1^fl/fl mice, which would lead to deletion of MORs from most glutamatergic neurons. We found that both the body weight and brain weight were comparable in Oprm1^fl/fl and Vglut2-Cre/Oprm1^fl/fl littermates (Fig. S2A, B). Next, we assessed Oprm1 expression in the brains of Vglut2-Cre/Oprm1^fl/fl mice using in situ hybridization, and found that its expression in the thalamus, medial habenula, and parabrachial nucleus (PBN), where most neurons are glutamatergic, was largely abolished compared to that in Oprm1^fl/fl mice. In contrast, expression of Oprm1 in the striatum, lnucleus accumbens (NAc), central amygdala (CeA), and interpeduncular nucleus (IPN), where most neurons are GABAergic, was comparable to that in Oprm1^fl/fl mice (Figs. 1B and S1A, B). We also used immunohistochemistry to confirm the selective deletion of MORs in Vglut2-Cre/Oprm1^fl/fl mice. The immunostaining results also showed that the expression of MORs was selectively abolished in the glutamatergic neuron-enriched brain regions of Vglut2-Cre/Oprm1^fl/fl mice (Figs. 1C and S2C, D). These results showed that the expression of MORs was selectively abolished from glutamatergic neurons in Vglut2-Cre/Oprm1^fl/fl mice.

We next examined the functional role of MORs expressed in glutamatergic neurons in morphine withdrawal using Vglut2-Cre/Oprm1^fl/fl mice and littermate Oprm1^fl/fl mice. The mice were given escalating doses of morphine (7.5 mg/kg–50 mg/kg, i.p.) for 5 consecutive days to establish dependence, and another batch of mice was given saline as controls with the same schedule (Fig. 2A). Acute morphine withdrawal was induced by injection of the opioid receptor antagonist, naloxone, 2 h after the last morphine/saline injection. The mice were video-taped for 30 min after naloxone injection for behavioral analysis (Fig. 2A). Consistent with previous reports [28], naloxone induced robust withdrawal signs. The withdrawal score of the Oprm1^fl/fl mice in the morphine group was significantly higher than that in the control group (Fig. 2B, C). In contrast, the withdrawal score of Vglut2-Cre/Oprm1^fl/fl mice treated with morphine was comparable to that of mice treated with saline. In addition, Vglut2-Cre/Oprm1^fl/fl mice treated with morphine showed a significantly lower global withdrawal score than Oprm1^fl/fl littermates treated with morphine (Fig. 2B, C), indicating that opioid dependence was largely abolished after selective deletion of MORs in glutamatergic neurons. Specifically, naloxone significantly increased the jumping, teeth-chattering, headshaking, piloerection, and twitching behaviors compared to the saline-treated control group (Fig. 2D, E, G, I, K), and the paw tremor behavior in Oprm1^fl/fl mice with morphine treatment was slightly elevated (Fig. 2J). However, the changes of withdrawal symptoms in Vglut2-Cre/Oprm1^fl/fl mice were variable. Vglut2-Cre/Oprm1^fl/fl mice exhibited a significant reduction in jumping, teeth-chattering, headshaking, and paw tremor (Fig. 2D, E, G, J), while the expression of wet-dog shakes, piloerection, and twitching were not significantly affected (Fig. 2F, I, K). The number of grooming bouts was slightly higher in Vglut2-Cre/Oprm1^fl/fl mice than in Oprm1^fl/fl littermates (Fig. 2H), and the decrease of grooming behavior in Oprm1^fl/fl mice might have resulted from the increases in other withdrawal symptoms.

Fig. 2.

Deletion of MORs in glutamatergic neurons alleviates naloxone-precipitated withdrawal symptoms. A Experimental timeline for chronic morphine delivery and naloxone-precipitated withdrawal. B, C Time course of scores and total scores for naloxone-precipitated withdrawal in Oprm1^fl/fl and Vglut2-Cre/Oprm1^fl/fl mice (n = 6–8 mice/group). D–K Bar graphs showing bouts and scores for each withdrawal symptom in Oprm1^fl/fl and Vglut2-Cre/Oprm1^fl/fl mice. Jumping, teeth-chattering, wet-dog shakes, headshakes, grooming, piloerection, paw tremor, and twitching behaviors were analyzed from recorded videos. L, M Effects of morphine on locomotor activity of Oprm1^fl/fl and Vglut2-Cre/Oprm1^fl/fl mice. Travel distance and average speed were measured in the open field test before and after morphine administration (n = 9–12 mice/group). *P < 0.05, **P < 0.01, ***P < 0.001, n.s., not significant (one-way ANOVA in C–K; two-way ANOVA between genotypes or morphine treatment in L and M, followed by a post hoc Bonferroni test). Data are presented as the mean ± SEM.

As morphine can also lead to locomotor hyperactivity in mice, we evaluated the functional role of MORs expressed in glutamatergic neurons on morphine-induced locomotor activity using Vglut2-Cre/Oprm1^fl/fl mice. In the open field test, Oprm1^fl/fl and Vglut2-Cre/Oprm1^fl/fl mice showed comparable basal locomotor activity (Fig. 2L, M). Morphine significantly elevated the speed of movement and travel distance in Oprm1^fl/fl mice, whereas the morphine-induced locomotor hyperactivity was abolished in Vglut2-Cre/Oprm1^fl/fl mice (Fig. 2L, M). We also assessed anxiety/depression-associated behaviors in Vglut2-Cre/Oprm1^fl/fl mice. In the novel object test, the frequency and time spent on novel object exploration by Vglut2-Cre/Oprm1^fl/fl mice were comparable to those in Oprm1^fl/fl mice (Fig. S2E). There was also no difference between Oprm1^fl/fl and Vglut2-Cre/Oprm1^fl/fl mice in the elevated plus maze, tail suspension, and sucrose preference tests (Fig. S3F–H), suggesting that Vglut2-Cre/Oprm1^fl/fl mice maintained normal anxiety-like behaviors indistinguishable from those of their Oprm1^fl/fl littermates. Taken together, these data indicate that morphine withdrawal and morphine-induced locomotor hyperactivity largely rely on MORs expressed in Vglut2⁺ glutamatergic neurons.

MORs Expressed in GABAergic Neurons Play a Negligible Role in Morphine Withdrawal

To determine the involvement of MORs expressed in GABAergic neurons in morphine withdrawal, we generated Vgat-Cre/Oprm1^fl/fl mice by crossing Vgat-Cre mice with Oprm1^fl/fl mice to conditionally knockout MORs from GABAergic neurons. We found that the body weight of Vgat-Cre/Oprm1^fl/fl mice was slightly higher than their Oprm1^fl/fl littermates, while their brain weights were comparable (Fig. S3A, B). We assessed both the Oprm1 and MOR expression in Vgat-Cre/Oprm1^fl/fl mice, and found that they were largely abolished in GABAergic neuron-enriched brain areas such as the striatum, NAc, CeA, and IPN, while the MOR expression in non-GABAergic regions, such as the thalamus and PBN, was intact (Figs. 3, S1A, C, S3C, D). Thus, the expression of MORs was selectively abolished from GABAergic neurons in Vgat-Cre/Oprm1^fl/fl mice.

Fig. 3.

MOR expression is abolished in multiple brain regions of Vgat-Cre/Oprm1^fl/fl mice. A Representative images of Oprm1 expression in Oprm1^fl/fl and Vgat-Cre/Oprm1^fl/fl mice (scale bar, 500 μm). B Representative images of MOR expression in Oprm1^fl/fl and Vgat-Cre/Oprm1^fl/fl mice determined by fluorescent immunostaining (scale bar, 200 μm). NAc, nucleus accumbens; MHb, medial habenula; CeA, central amygdala; IPN, interpeduncular nucleus; PBN, parabrachial nucleus. Cpu, striatum; NAc, nucleus accumbens; AcbC, accumbens nucleus, core; MHb, medial habenula; PV, paraventricular thalamic nucleus; CM, central medial thalamic nucleus; CeA, central amygdala; BLA, basolateral amygdala; IPN, interpeduncular nucleus; VTA, ventral tegmental area; PBN, parabrachial nucleus; scp, superior cerebellar peduncle; LPB, lateral parabrachial nucleus.

We next examined the functional role of MORs in GABAergic neurons in opioid dependence using the withdrawal procedure in Vgat-Cre/Oprm1^fl/fl mice (Fig. 4A). Naloxone administration induced robust withdrawal signs in Oprm1^fl/fl mice of the morphine group compared to the saline-treated control group (Fig. 4B–K). Deletion of MORs from GABAergic neurons had no significant effect on the naloxone-precipitated withdrawal symptoms, as evidenced by the comparable global withdrawal scores in Oprm1^fl/fl and Vgat-Cre/Oprm1^fl/fl mice (Fig. 4B, C). Most withdrawal behaviors were consistent with each other, as Vgat-Cre/Oprm1^fl/fl mice showed almost intact withdrawal symptoms relative to Oprm1^fl/fl mice except for decreased jumping bouts (Fig. 4D–K). Although the jumping bouts in Vgat-Cre/Oprm1^fl/fl mice were significantly decreased, they were not totally blocked (Fig. 4D). In addition, morphine-induced locomotor hyperactivity was slightly but significantly lower in Vgat-Cre/Oprm1^fl/fl mice than in Oprm1^fl/fl mice, while the two groups showed similar basal locomotor activity (Fig. 4L, M). We also applied anxiety/depression-associated behavioral tests in Vgat-Cre/Oprm1^fl/fl mice, and found that the behavioral expression of Oprm1^fl/fl and Vgat-Cre/Oprm1^fl/fl mice were comparable (Fig. S3E–H), indicating normal anxiety-like behaviors in the Vgat-Cre/Oprm1^fl/fl mice. These results suggest that MORs expressed in GABAergic neurons play a minor role in mediating the development of morphine withdrawal and morphine-induced locomotor hyperactivity.

Fig. 4.

Deletion of MORs in GABAergic neurons has limited effects on naloxone-precipitated withdrawal symptoms. A Experimental timeline for chronic morphine delivery and naloxone-precipitated withdrawal. B, C Time course of scores and total scores for naloxone-precipitated withdrawal in Oprm1^fl/fl and Vgat-Cre/Oprm1^fl/fl mice (n = 6–8 mice/group). D–K Bar graphs showing bouts and scores for each withdrawal symptom in Oprm1^fl/fl and Vgat-Cre/Oprm1^fl/fl mice. Jumping, teeth-chattering, wet-dog shakes, headshakes, grooming, piloerection, paw tremor, and twitching behaviors were analyzed from recorded videos. L, M Effects of morphine on the locomotor activity of Oprm1^fl/fl and Vgat-Cre/Oprm1^fl/fl mice. Travel distance and average speed were measured in the open field test before and after morphine administration (n = 9–10 mice/group). **P < 0.01, ***P < 0.001, n.s., not significant (one-way ANOVA in C–K; two-way ANOVA between genotypes or morphine treatment in L and M, followed by a post hoc Bonferroni test). Data are presented as the mean ± SEM.

MORs Expressed in Dorsal Spinal Neurons Are Not Involved in Morphine Withdrawal

MORs are also highly expressed in the dorsal spinal cord, which exhibits adaptive changes during withdrawal [25, 28]. We thus explored the functional role of MORs expressed in the dorsal spinal cord in morphine withdrawal. To achieve this, we generated Lbx1-Cre/Oprm1^fl/fl mice by crossing Lbx1-Cre mice with Oprm1^fl/fl mice (Fig. 5A), which led to the selective deletion of MORs from dorsal spinal neurons. The expression of Oprm1 was assessed by in situ hybridization, and the results showed that the Oprm1 signal was lost in the dorsal spinal cord of Lbx1-Cre/Oprm1^fl/fl mice (Fig. 5B, C). Next, we determined the morphine withdrawal in Lbx1-Cre/Oprm1^fl/fl mice (Fig. 5D), and found significant naloxone-precipitated withdrawal symptoms in both Oprm1^fl/fl and Lbx1-Cre/Oprm1^fl/fl mice with chronic morphine treatment compared to the control group (Fig. 5E, F). Furthermore, the behavioral expression of Lbx1-Cre/Oprm1^fl/fl mice was indistinguishable from Oprm1^fl/fl littermates, as both the global withdrawal score and the behavioral symptoms were comparable in these groups (Fig. 5E–N). In addition, the reduction of grooming bouts in the two morphine groups might be a result of the appearance of other withdrawal behaviors (Fig. 5K). We also examined the functional role of these MORs in morphine-induced locomotor hyperactivity, and found that Oprm1^fl/fl and Lbx1-Cre/Oprm1^fl/fl mice showed comparable travel distances and average speeds in the open field test both at baseline and after morphine injection (Fig. 5O, P). These results indicate that MORs expressed in dorsal spinal neurons are not involved in the development of morphine withdrawal and morphine-induced locomotor hyperactivity.

Fig. 5.

Deletion of MORs in the dorsal spinal cord has no significant effect on naloxone-precipitated withdrawal symptoms. A Schematic of the protocol for deleting MORs from the dorsal spinal cord. Lbx1-Cre mice were crossed with Oprm1^fl/fl mice, and exons 2 and 3 of Oprm1 were excised. B, C Representative images of Oprm1 expression in the dorsal cord of Oprm1^fl/fl and Lbx1-Cre/Oprm1^fl/fl mice (scale bar, 200 μm). D Experimental timeline for chronic morphine delivery and naloxone-precipitated withdrawal. E, F Time course of scores and total scores for naloxone-precipitated withdrawal in Oprm1^fl/fl and Lbx1-Cre/Oprm1^fl/fl mice (n = 6–7 mice/group). G–N Bar graphs showing bouts and scores for each withdrawal symptom in Oprm1^fl/fl and Lbx1-Cre/Oprm1^fl/fl mice. Jumping, teeth-chattering, wet-dog shakes, headshakes, grooming, piloerection, paw tremor, and twitching behaviors were analyzed from recorded videos. O, P Effect of morphine on locomotor activity of Oprm1^fl/fl and Lbx1-Cre/Oprm1^fl/fl mice. Travel distance and average speed were measured in the open field test before and after morphine administration (n = 6 mice/group). *P < 0.05, **P < 0.01, ***P < 0.001, n.s., not significant (one-way ANOVA in F–N; two-way ANOVA between genotypes or morphine treatment in O and P, followed by a post hoc Bonferroni test). Data are presented as the mean ± SEM.

Discussion

Due to the wide distribution of opioid receptors, their activation gives rise to numerous side-effects besides analgesia, including opioid dependence. Both spinal and supraspinal mechanisms have been implicated in opioid withdrawal [9]. In this study, we determined the involvement of MORs expressed in different neuronal populations during morphine withdrawal using genetic approaches. We demonstrated that MORs expressed in Vglut2⁺ glutamatergic neurons are essential for the development of morphine withdrawal and the morphine-induced locomotor hyperactivity, while those expressed in GABAergic neurons are not.

Using genetic approaches, we investigated the functional roles of MORs expressed in glutamatergic and GABAergic neurons in the opioid dependence that is characterized by morphine withdrawal. Our results showed that deletion of MORs expressed in Vglut2⁺ glutamatergic neurons almost abolished the naloxone-precipitated withdrawal symptoms, indicating that these MORs play an important role in opioid dependence. The reduction of behaviors associated with morphine withdrawal in Vglut2-Cre/Oprm1^fl/fl mice is likely due to less neuronal adaptation. In wild-type mice, the absence of opioid receptor activation induced by naloxone unmasks the neuronal adaptation resulting from continuous opioid application, leading to acute naloxone-precipitated withdrawal. Accordingly, the activation of MORs in glutamatergic neurons by morphine results in long-term inhibition, and termination of the action of morphine may result in rebound activation of these neurons. Thus, the absence of MORs in glutamatergic neurons dramatically eliminates the withdrawal symptoms. Our results are consistent with a recent study showing that deletion of MORs from the MHb, in which most neurons are glutamatergic, decreases the withdrawal symptoms [19]. Thus, the reduction of withdrawal symptoms in our study may be partially due to the loss of MORs in the MHb. Interestingly, the influences of loss of MORs in Vglut2⁺ neurons on withdrawal symptoms are divergent. It is likely that different withdrawal symptoms are mediated by MORs expressed in brain structures that are differentially involved in aversive or anxiety circuits [37]. Thus, further studies are needed to dissect the role of MORs expressed in distinct glutamatergic neuronal subpopulations and neural circuits. MORs are also likely expressed in glutamatergic neurons labeled with other molecular markers, such as Vglut1⁺ and Vglut3⁺. However, the contribution of those MORs is likely to be minimal, as most withdrawal symptoms were largely reduced in mice lacking MORs in Vglut2⁺ glutamatergic neurons. Although several previous studies have demonstrated the involvement of the excitatory circuits for aversive behaviors during morphine withdrawal [10, 18], further studies are still necessary to further dissect the circuit mechanism underlying opioid withdrawal. In contrast, our study demonstrated that MORs expressed in GABAergic neurons play a negligible role in mediating morphine withdrawal, which is consistent with a previous study showing that selective expression of MORs in striatal neurons does not rescue the morphine withdrawal symptoms [17]. This has also been further confirmed in several recent studies showing that deletion of MORs from forebrain GABAergic neurons does not change the behavioral expression during morphine withdrawal [15]. Here, we deleted MORs from all GABAergic neurons, thus extending investigations of the functional role of MORs expressed in all GABAergic neurons in morphine withdrawal. In addition, some of the withdrawal symptoms were intact with both the MOR conditional-knockout of Vglut2⁺ glutamatergic and GABAergic neurons, such as piloerection, suggesting that MORs expressed in other cell populations are also engaged during withdrawal.

Our results also demonstrated that MORs expressed in the dorsal spinal cord are not involved in morphine withdrawal. We selectively abolished their expression in the dorsal cord, and found that naloxone-precipitated withdrawal symptoms were not affected, suggesting that MORs expressed in the dorsal cord are not involved in the development of morphine withdrawal. As MORs are densely expressed in the ventral spinal cord, we cannot exclude their possible role in morphine withdrawal. However, recent studies have indicated that the spinal cord exhibits plasticity induced by ATP release from microglia during morphine withdrawal [28, 38], and the expression of MORs in microglia is very limited [13]. Further investigation is needed to address the functional role of spinal microglia in morphine withdrawal, including the action of opioids on microglia and neuron–glia interactions.

In summary, our study revealed the major neuronal subpopulation that mediates morphine withdrawal. These results provide important insights into the mechanism of morphine withdrawal, and lay a foundation for further dissection of the circuit mechanisms of opioid dependence. A better understanding of the mechanisms underlying opioid dependence will benefit the development of novel opioid analgesics with fewer side-effects.

Electronic supplementary material

Below is the link to the electronic supplementary material.