Mitogen-Activated Protein Kinase Signaling Mediates Opioid-induced Presynaptic NMDA Receptor Activation and Analgesic Tolerance

Abstract

Opioid-induced hyperalgesia and analgesic tolerance can lead to dose escalation and inadequate pain treatment with μ-opioid receptor agonists. Opioids cause tonic activation of glutamate NMDA receptors (NMDARs) at primary afferent terminals, increasing nociceptive input. However, the signaling mechanisms responsible for opioid-induced activation of presynaptic NMDARs in the spinal dorsal horn remain unclear. In the present study, we determined the role of mitogen-activated protein kinase (MAPK) signaling in opioid-induced presynaptic NMDAR activation caused by chronic morphine administration. Whole-cell recordings of excitatory postsynaptic currents (EPSCs) were performed on dorsal horn neurons in rat spinal cord slices. Chronic morphine administration markedly increased the frequency of miniature EPSCs, increased the amplitude of monosynaptic EPSCs evoked from the dorsal root, and reduced the paired-pulse ratio of evoked EPSCs. These changes were fully reversed by an NMDAR antagonist and normalized by inhibiting extracellular signal-regulated kinase 1/2 (ERK1/2), p38, or c-Jun N-terminal kinase (JNK). Furthermore, intrathecal injection of a selective ERK1/2, p38, or JNK inhibitor blocked pain hypersensitivity induced by chronic morphine treatment. These inhibitors also similarly attenuated a reduction in morphine’s analgesic effect in rats. In addition, co-immunoprecipitation assays revealed that NMDARs formed a protein complex with ERK1/2, p38, and JNK in the spinal cord and that chronic morphine treatment increased physical interactions of NMDARs with these three MAPKs. Our findings suggest that opioid-induced hyperalgesia and analgesic tolerance are mediated by tonic activation of presynaptic NMDARs via three functionally interrelated MAPKs at the spinal cord level.

Graphical Abstract

Under the normal condition, NMDA receptors at the presynaptic terminals in the spinal cord are not phosphorylated or functionally active. However, presynaptic NMDA receptors in the spinal dorsal horn become tonically activated during chronic morphine treatment. Here we found that chronic treatment with morphine increases the association of NMDA receptors with three MAPKs (ERK1/2, p38, and JNK) in the spinal cord. This increased interaction leads to NMDA receptor phosphorylation at presynaptic terminals. As a result, presynaptic NMDARs are endogenously activated to augment synaptic glutamate release to postsynaptic neurons in the spinal dorsal horn to cause opioid-induced hyperalgesia and analgesic tolerance.

Introduction

Opioid analgesics are the mainstay of treatment of severe pain caused by cancer or tissue and nerve injury. However, repeated use of μ-opioid receptor (MOR) agonists causes loss of analgesic efficacy and opioid dose escalation, which contributes to the current epidemic of opioid abuse in the United States. Also, acute and chronic opioid use can paradoxically increase pain sensitivity, termed opioid-induced hyperalgesia, in both animals and humans (Celerier et al. 2000, Chia et al. 1999, Chen et al. 2007, Mao et al. 1994). Repeated morphine injections can progressively reduce the nociceptive threshold, which occurs in parallel with development of analgesic tolerance (Chen et al. 2007, Mao et al. 1994, Zhao et al. 2012). Concurrent development of hyperalgesia associated with opioid use can be perceived as the diminished opioid analgesic effect (i.e., analgesic tolerance). Furthermore, intrathecal administration of glutamate N-methyl-D-aspartate receptor (NMDAR) antagonists blocks both hyperalgesia and analgesic tolerance induced by chronic morphine use (Mao et al. 1994, Zhao et al. 2012), suggesting that increased NMDAR activity at the spinal cord level is essential for opioid-induced hyperalgesia and tolerance. Nevertheless, chronic morphine treatment reduces the activity of postsynaptic NMDARs in spinal dorsal horn neurons (Zhao et al. 2012).

NMDARs are also expressed at primary afferent terminals in the spinal dorsal horn (Liu et al. 1994). However, these NMDARs are not functionally active during the normal condition. After acute or chronic opioid exposure, presynaptic NMDARs become tonically activated, increasing glutamatergic input to spinal dorsal horn neurons (Zhao et al. 2012, Zhou et al. 2010, Zeng et al. 2006). Hence, tonic activation of presynaptic NMDARs can increase nociceptive input from primary afferent nerves to spinal dorsal horn neurons, resulting in hyperalgesia and opioid tolerance (Zhou et al. 2010, Zhao et al. 2012). However, how MOR stimulation leads to activation of presynaptic NMDARs in the spinal dorsal horn remains unclear.

Following MOR activation, MORs become phosphorylated by G protein receptor kinase, which initiates a β-arrestin—dependent desensitization process to terminate the MOR signaling (Bohn et al. 2004, Keith et al. 1996). Furthermore, β-arrestin may act as a scaffolding protein and promote the stable association of signaling proteins with MORs to activate the mitogen-activated protein kinase (MAPK) pathways (McDonald et al. 2000, Luttrell et al. 2001). Mice lacking β-arrestin 2 exhibit prolonged opioid analgesic effects and reduced tolerance to morphine (Bohn et al. 1999), suggesting that arrestin/MAPK signaling contributes to opioid tolerance development. The three major MAPKs, extracellular signal-regulated kinases 1/2 (ERK1/2), p38, and c-Jun N-terminal kinase (JNK), are activated in primary sensory neurons by nociceptive stimuli (Mizushima et al. 2005, Modol et al. 2015, Obata & Noguchi 2004). Although opioid-induced MAPK signaling is well documented (Belcheva et al. 2001, Fukuda et al. 1996), there is currently no evidence showing whether MAPK signaling contributes to opioid-induced activation of presynaptic NMDARs at the spinal cord level.

In the present study, we used a spinal cord slice preparation that has native MORs and preserves the intrinsic connection between primary afferents and dorsal horn neurons to define the role of ERK1/2, p38, and JNK in chronic morphine treatment-induced tonic activation of presynaptic NMDARs. We provide the first in vivo evidence that chronic morphine treatment potentiates the physical interaction between the NMDAR and ERK1/2, p38, and JNK in the spinal cord, leading to increased activity of presynaptic NMDARs at primary afferent terminals. This new information extends our understanding of signaling mechanisms involved in opioid-induced hyperalgesia and analgesic tolerance.

Materials and Methods

Animal Model and Morphine Treatment —

Adult male Sprague-Dawley rats (280–320 g; Harlan, Indianapolis, IN; RRID: RGD_5508397) were used in our experiments. All procedures and protocols were approved by The University of Texas MD Anderson Cancer Center (approval #882-RN01) and were performed in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The timeline of experimental procedures and # of animals used for each protocol was specified in Fig. 1. To induce opioid hyperalgesia and analgesic tolerance, morphine sulfate (West Ward Pharmaceuticals, Eatontown, NJ) was intraperitoneally injected into the rats at 5 mg/kg twice a day for 8 consecutive days (Zhao et al. 2012, Chen et al. 2007).

Figure 1. Flowchart diagrams show the timeline of experimental procedures used in the study.

Rats were treated with morphine or vehicle for 8 days and used for spinal cord slice recording, behavioral tests (during co-treatment with intrathecal injection of 3 MAPK inhibitors), or co-immunoprecipitation assays. The number of animals used in each group was indicated in parenthesis.

Intrathecal Catheterization and Drug Delivery —

A polyethylene PE10 catheter was implanted in the lumbar subarachnoid space in each rat under isoflurane-induced anesthesia according to the method described previously (Chen & Pan 2006a). Briefly, each animal was placed prone on a stereotaxic frame, and a small incision was made at the back of the neck. The catheter was inserted after a small puncture was made in the atlanto-occipital membrane of the cisterna magna. The caudal tip of the catheter was then advanced and reached the lumbar enlargement of the spinal cord. To minimize animal’s suffering, 1% lidocaine was injected around the incision site during the brief surgery. The animals were allowed to recover for 7 days before intrathecal injection, and no other medications were administered. Rats exhibiting neurological deficits (e.g., paralysis) or poor grooming were promptly killed via CO₂ asphyxiation. In 43 rats instrumented with intrathecal catheters, 4 rats were killed due to motor weakness and impairment. U0126 (5 μg; EMD-Calbiochem, La Jolla, CA), 4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)-1H-imidazole (SB203580, 5 μg; EMD-Calbiochem), and SP600125 (30 μg; Cayman Chemical, Ann Arbor, MI) were dissolved in DMSO and intrathecally injected 20 min before morphine administration on each testing day (days 1–8).

Spinal Cord Slice Preparation and Electrophysiological Recordings of Dorsal Horn Neurons —

The lumbar spinal cords were rapidly removed from the rats via laminectomy under isoflurane-induced anesthesia. Each spinal cord was immediately placed in ice-cold sucrose artificial cerebrospinal fluid containing (in mM) 3.6 KCl, 1.2 MgCl₂, 2.5 CaCl₂, 1.2 NaH₂PO₄, 25.0 NaHCO₃, 234 sucrose, and 12.0 glucose presaturated with 95% O₂ and 5% CO₂. The spinal cord was then placed in a shallow groove formed in an agar block and glued onto the stage of a vibratome (Leica, Wetzlar, Germany). Transverse slices of the spinal cords (400 μm thick) were cut in ice-cold sucrose artificial cerebrospinal fluid and preincubated in Krebs solution containing (in mM) 3.6 KCl, 1.2 MgCl₂, 117.0 NaCl, 2.5 CaCl₂, 1.2 NaH₂PO₄, 11.0 glucose, and 25.0 NaHCO₃ oxygenated with 95% O₂ and 5% CO₂ at 34 °C for at least 1 h before being transferred to a recording chamber. The spinal cord slice was placed in an incubation chamber and continuously perfused with Krebs solution at 5 ml/min and 34 °C maintained by an inline solution heater and a temperature controller.

The spinal cord lamina II, a translucent region in the superficial dorsal horn, was identified in each slice under an upright fixed-stage microscope (BX50WI; Olympus, Tokyo, Japan) with differential interference contrast/infrared illumination. Neurons in the lamina II outer zone were visualized and selected for whole-cell recordings because they are important for nociceptive transmission (Pan & Pan 2004, Wang et al. 2018). A previous study shows that most lamina II neurons are glutamate-releasing excitatory interneurons (Santos et al. 2007), and these neurons form networks that play a critical role in transmitting nociceptive information from the primary afferents. The impedance of the glass electrode was 4–7 MΩ when a pipette was filled with an internal solution containing (in mM) 135 potassium gluconate, 0.5 CaCl₂, 5 KCl, 2.0 MgCl₂, 5.0 HEPES, 5.0 ATP-Mg, 0.5 Na-GTP, 5.0 EGTA, and 10 QX314 (adjusted to pH 7.25 with 1.0 M KOH, 280–300 mOsm).

Excitatory postsynaptic currents (EPSCs) were recorded at a holding potential of −60 mV using whole-cell voltage-clamp techniques. The input resistance was monitored, and the recording was abandoned if it changed by more than 15%. EPSCs were recorded using an amplifier (MultiClamp 700A, Axon Instruments, Foster City, CA), filtered at 1–2 kHz, and digitized at 10 kHz. Miniature EPSCs (mEPSCs) were recorded in the presence of 1 μM tetrodotoxin (Chen et al. 2014, Zhou et al. 2010). Because our whole-cell recording was done at a holding potential of −60 mV and in the presence of 1.2 mM Mg²⁺ (in the Krebs solution), the postsynaptic NMDARs were largely closed. To evoke glutamate release from primary afferent nerves, a bipolar tungsten electrode connected to a stimulator (0.2 ms, 0.3–0.6 mA, 0.1 Hz; Grass Instruments, Quincy, MA) was used to electrically stimulate the dorsal root, and monosynaptic EPSCs were identified on the basis of the constant latency and absence of conduction failure of evoked EPSCs in response to 20-Hz electrical stimulation (Zhou et al. 2010, Xie et al. 2017b). To calculate the paired-pulse ratio (PPR), two EPSCs were evoked by a pair of stimuli at 50-ms intervals. The PPR was expressed as the ratio of the amplitude of the second synaptic response to the amplitude of the first synaptic response (Xie et al. 2017b, Zhao et al. 2012). All the drugs were prepared in artificial cerebrospinal fluid immediately before the experiments and delivered via syringe pumps to reach their final concentrations. AP5 and tetrodotoxin were purchased from Hello Bio (Princeton, NJ).

Behavioral Assessment of Nociception —

To quantify the mechanical nociceptive threshold in the rats, a paw pressure test on a hindpaw was conducted using an analgesiometer (Ugo Basile, Varese, Italy). The device was activated by pressing a foot pedal to generate a constantly increasing force. When the animal displayed pain by either withdrawing its paw or vocalizing, the device was immediately inactivated, and the animal’s withdrawal threshold was read on the scale (Chen et al. 2007, Zhao et al. 2012). A maximum of 400 g of pressure was used as a cutoff to avoid tissue injury to the rats.

To assess thermal sensitivity, rats were placed on the glass surface of a thermal testing apparatus (IITC Life Science, Woodland Hills, CA) and allowed to acclimate for 30 min before testing. The temperature of the glass surface was maintained constant at 30 °C. A mobile radiant heat source located under the glass was focused onto a hindpaw of each rat. The paw withdrawal latency was recorded using a timer, and the hindpaw was tested twice to obtain the average. A cut-off of 30 s was set to prevent tissue damage (Li et al. 2016, Chen & Pan 2006b).

Co-immunoprecipitation —

After treatment with morphine and vehicle for 8 days, the dorsal quadrants of lumbar spinal cord tissues were quickly removed from rats anesthetized with 3% isoflurane. The spinal tissues were dissected and homogenized in ice-cold immunoprecipitation buffer (50 mM Tris [pH7.4], 250 mM NaCl, 20 mM NaF, 1 mM Na₃VO₄, 10% glycerol, 0.5% NP-40, 10 mM N-ethylmaleimide, 1 mM phenylmethylsulfonyl fluoride, and 2 mM benzamide) containing cocktails of protease and phosphatase inhibitors (Sigma-Aldrich, St. Louis, MO). Samples were then put on ice for 30 min with constant shaking. Lysates were centrifuged at 13,000 × g for 30 min at 4 °C. The supernatant was carefully collected, and the protein concentration was measured using a DC Protein Assay Kit (Bio-Rad, Hercules, CA). The soluble fraction was incubated at 4 °C overnight with Protein A/G beads (#16–266, Millipore, Darmstadt, Germany) prebound to a mouse anti-GluN1 antibody (#75–272, 1:1,000, NeuroMab, Davis, CA; RRID: AB_11000180). Protein A/G beads prebound to mouse IgG were used as controls. Samples were washed 3 times with an IP buffer and then immunoblotted. The following antibodies were selected for immunoblotting: rabbit anti-GluN1 (#G-8913, 1:1,000, Sigma-Aldrich; RRID: AB_259978), rabbit anti-ERK1/2 (#4695, 1:1,000, Cell Signaling Technology, Danvers, MA; RRID: AB_390779), rabbit anti-p38 (#8690, 1:1,000, Cell Signaling Technology; RRID: AB_10999090), and rabbit anti-SAPK/JNK (#9252, 1:1,000, Cell Signaling Technology; RRID:AB_2250373). The specificity of these antibodies was validated in previous studies (Rudhard et al. 2003, Brady et al. 2017, Li et al. 2017, Sun et al. 2017). For immunoblots, lysates were separated on a 4–12% gradient SDS-polyacrylamide gel and transferred to a polyvinylidene fluoride membrane. The membrane was blocked by 5% blotting-grade buffer (#170–6404, Bio-Rad, Hercules, CA). After the antibody incubation, the membrane was washed 3 times with Tris-buffered saline with Tween 20 (#9997S, Cell Signaling Technology). An ECL kit (#PI34580, Fisher Scientific, Lenexa, KS) was used to detect the protein band, which was visualized and quantified using an Odyssey Fc Imager (LI-COR Biotechnology, Lincoln, NE).

Study Design and Data Analysis —

Data were expressed as means ± SEM. No statistical methods were used to predetermine sample sizes for the studies, but our sample sizes were similar to those generally employed in the field. The animals were assigned (1:1 allocation) to the control and treatment groups, and no randomization methods were used. We did not exclude any animals from data analysis and did not use any test for outliers on the data. The investigators performing electrophysiological and behavioral experiments were blinded to drug treatment. In electrophysiological experiments, only one neuron was recorded in each tissue slice, and at least three rats were used in each group. The amplitude of 10 consecutive evoked EPSCs was analyzed and averaged using the Clampfit 10.0 software program (Axon Instruments). mEPSCs were analyzed off-line using the MiniAnalysis peak detection program (Synaptosoft, Leonia, NJ). D’Agostino-Pearson normality test was used to assess the normality of data. The Student’s t test was used to compare two groups, and one-way analysis of variance followed by Dunnett’s post hoc test was used to compare more than two groups. Two-way analysis of variance followed by Bonferroni’s post hoc test was used to determine the difference in development of hyperalgesia and opioid tolerance between the vehicle- and morphine-treated rats. Statistical analyses were performed using Prism software (version 7, GraphPad Software Inc., La Jolla, CA). P <0.05 was considered to be statistically significant.

Results

ERK1/2 Mediates Chronic Morphine-induced Increases in Presynaptic NMDAR Activity in the Spinal Cord —

We first determined whether ERK1/2 plays a role in increased presynaptic NMDAR activity in the spinal cord induced by chronic morphine treatment. Presynaptic ERK signaling plays a role in synaptic plasticity in the hippocampus (Vara et al. 2009), and presynaptic ERK1/2 is important for the control of central glutamatergic synapses (Sindreu et al. 2011). We recorded miniature excitatory postsynaptic currents (mEPSCs) in spinal cord lamina II neurons, which reflect quantal glutamate release from presynaptic terminals (Xie et al. 2017a, Chen et al. 2014). Bath application of 50 μM AP5, a specific NMDAR antagonist, rapidly normalized the increased frequency of mEPSCs in lamina II neurons obtained from morphine-treated rats (baseline, 6.45 ± 0.60 Hz; AP5, 4.17 ± 0.45 Hz; n = 11 neurons, Fig. 2, A and B). Bath application of AP5 does not affect the frequency or amplitude of mEPSCs in lamina II neurons from control rats (Zhao et al. 2012, Zhou et al. 2010). These results suggest that chronic morphine potentiates presynaptic NMDAR activity of spinal dorsal horn neurons, a finding similar to that in our previous study (Zhao et al. 2012). Also, U0126 is a highly selective ERK1/2 inhibitor used to study ERK1/2-mediated signaling (Favata et al. 1998, Arana-Argaez et al. 2010). Pretreatment with U0126 (10 μM for 1 h) significantly reduced the baseline frequency (3.00 ± 0.26 Hz vs. 6.45 ± 0.60 Hz), but not the amplitude, of mEPSCs in lamina II neurons in spinal cord slices obtained from morphine treated rats (n = 10 neurons, 6.45 ± 0.60 Hz vs. 3.00 ± 0.26 Hz, P < 0.05; Fig. 2, B and D). In these 10 neurons, subsequent application of 50 μM AP5 had no effect on the frequency or amplitude of mEPSCs (Fig. 2, C and D).

Figure 2. ERK1/2 contributes to increased glutamatergic input to spinal dorsal horn neurons induced by chronic morphine treatment.

A, original recording traces and cumulative plots show the baseline and the effect of bath application of 50 μM AP5 on the frequency and amplitude of mEPSCs of a lamina II neuron from a morphine-treated rat. B, summary data show the baseline and the effect of 50 μM AP5 on the mean frequency and amplitude of mEPSCs in morphine-treated rats (n = 11 neurons from 4 rats). C, representative recording traces and cumulative plots show a lack of effect of 50 μM AP5 on the frequency or amplitude of mEPSCs in a lamina II neuron pretreated with 10 μM U0126 from a spinal cord slice obtained from a morphine-treated rat. D, group data show the effect of 50 μM AP5 on the mean frequency and amplitude of mEPSCs (n = 10 neurons from 4 rats) in spinal cord slices pretreated with 10 μM U0126 in morphine-treated rats. Data are presented as means ± SEM. *p < 0.05 when compared with the respective baseline control. #p < 0.05 when compared with the baseline in morphine + vehicle group (one-way ANOVA followed by Dunnett’s post hoc test).

The synapse formed by central terminals of primary afferent neurons and spinal dorsal horn neurons is essential for nociceptive transmission and modulation by opioids (Chen & Pan 2006a, Zhou et al. 2010, Zhou et al. 2008). We thus determined whether ERK1/2 is involved in morphine-induced increases in the activity of NMDARs expressed at primary afferent nerve terminals in the spinal dorsal horn. Monosynaptic EPSCs of spinal dorsal horn neurons were evoked from the dorsal root, which represents elicited glutamate released from primary afferent terminals (Xie et al. 2017b, Zhou et al. 2010). Bath application of 50 μM AP5 considerably reduced the amplitude of monosynaptic EPSCs (363.7 ± 21.6 pA vs. 441.4 ± 23.9 pA) and increased the paired-pulse ratio (PPR, 0.76 ± 0.06 vs. 0.65 ± 0.05) of monosynaptically evoked EPSCs of lamina II neurons obtained from morphine-treated rats (n = 10 neurons, Fig. 3, A – C). These data are consistent with those in our previous study demonstrating opioid-induced tonic activation of NMDARs at primary afferent terminals (Zhao et al. 2012). After pretreatment of lamina II neurons from morphine-treated rats with U0126 (10 μM for 1 h), bath application of 50 μM AP5 no longer had an effect on the amplitude (n = 10 neurons) or PPR of monosynaptic EPSCs (n = 9 neurons, Fig. 3, D and F). Taken together, these data indicate that ERK1/2 plays a critical role in opioid-induced tonic activation of presynaptic NMDARs in the spinal cord.

Figure 3. ERK1/2 mediates chronic morphine-induced potentiation of presynaptic NMDAR activity at primary afferent terminals in spinal dorsal horn neurons.

A, original recording traces and summary data show the effect of 50 μM AP5 on evoked monosynaptic EPSCs in a lamina II neuron from a morphine-treated rat. B, original recording traces show that AP5-produced inhibition of evoked EPSC amplitude in morphine-treated rats was associated with an increased paired-pulse ratio (PPR). C, summary data show the effect of 50 μM AP5 on the amplitude (n = 10 neurons from 4 rats) and PPR (n = 10 neurons from 4 rats) of monosynaptic EPSCs in lamina II neurons evoked from the dorsal root in morphine-treated rats. D, original traces show a lack of effect of 50 μM AP5 on the amplitude of monosynaptically evoked EPSCs in a lamina II neuron in spinal cord slices pretreated with 10 μM U0126 in a morphine-treated rat. E, original recording traces show the lack of the effect of AP5 on paired-pulse ratio of monosynaptic EPSCs of a lamina II neuron from spinal cord slices pretreated with 10 μM U0126 in morphine-treated rats. F, group data show a lack of effect of 50 μM AP5 on the amplitude (n = 10 neurons from 4 rats) and PPR (n = 9 neurons from 4 rats) of monosynaptic EPSCs of lamina II neurons pretreated with 10 μM U0126 in spinal cord slices from morphine-treated rats. Data are presented as means ± SEM. *p < 0.05 when compared with respective baseline controls (one-way ANOVA followed by Dunnett’s post hoc test).

P38 Is Involved in Morphine-induced Presynaptic NMDAR Activation in the Spinal Cord —

We next determined whether the p38 MAPK is involved in chronic morphine-induced increases in presynaptic NMDAR activity in the spinal cord. p38 is expressed presynaptically (Nakata et al. 2005), and SB203580 is a pyridinyl imidazole inhibitor commonly used to study p38 activity in vitro and in vivo (Arana-Argaez et al. 2010, Jin et al. 2013). Pretreatment with SB203580 (5 μM for 1 h) caused a large reduction in the baseline frequency of mEPSCs in spinal lamina II neurons obtained from morphine-treated rats (3.63 ± 0.41 Hz vs. 6.45 ± 0.60 Hz; n = 10 neurons, P < 0.05; Fig. 2B and Fig. 4B). In these neurons, further treatment with AP5 failed to affect the frequency or amplitude of mEPSCs (n = 10 neurons, Fig. 4, A and B).

Figure 4. p38 is involved in chronic morphine-induced increases in presynaptic NMDAR activity in the spinal cord.

A, original recordings and cumulative plots show the lack of effect of 50 μM AP5 on the frequency and amplitude of mEPSCs in a lamina II neuron from spinal cord slices pretreated with 5 μM SB203580 in morphine-treated rats. B, group data show a lack of effect of AP5 on the frequency and amplitude of mEPSCs in lamina II neurons in spinal cord slices pretreated with 5 μM SB203580 from morphine-treated rats (n = 10 neurons from 3 rats). C, representative current traces show the lack of effect of 50 μM AP5 on the amplitude of monosynaptic EPSCs of a lamina II neuron from spinal cord slices pretreated with SB203580 in a morphine-treated rat. D, original current traces show a lack of effect of AP5 on the PPR of evoked monosynaptic EPSCs in spinal cord slices pretreated with SB203580 in morphine-treated rats. E, summary data show the lack of effect of 50 μM AP5 on the amplitude (n = 10 neurons from 3 rats) and PPR (n = 9 neurons from 3 rats) of monosynaptic EPSCs in lamina II neurons from spinal cord slices pretreated with SB203580 in morphine-treated rats. Data are presented as means ± SEM.

We also examined whether p38 inhibition affects morphine-induced activation of NMDARs at primary afferent terminals. After treatment of spinal lamina II neurons obtained from morphine-treated rats with SB203580 (5 μM for 1 h), subsequent bath application of 50 μM AP5 had no effect on the amplitude of evoked monosynaptic EPSCs (n = 10 neurons, Fig. 4, C and E) or the PPR of monosynaptic EPSCs (n = 9 neurons, Fig. 4, D and E). These results suggest that p38 mediates opioid-induced tonic activation of NMDARs at primary afferent terminals.

JNK Participates in Morphine-induced Increases in Presynaptic NMDAR Activity in the Spinal Cord —

We then examined whether JNK is involved in increased presynaptic NMDAR activity in the spinal dorsal horn in morphine-treated rats. JNK2 is expressed presynaptically and plays a role in NMDAR-mediated glutamate release in the cortex (Nistico et al. 2015). SP600125 is an anthrapyrazolone capable of inhibiting JNK1, JNK2, and JNK3 with high affinity (Kocab et al. 2015, Bennett et al. 2001). In lamina II neurons obtained from morphine-treated rats, pretreatment with SP600125 (40 μM for 1 h) significantly reduced the frequency (4.40 ± 0.44 Hz vs .6.45 ± 0.60 Hz), but not the amplitude, of mEPSCs (n = 10 neurons, Fig. 5). In the same neurons, subsequent bath application of AP5 did not significantly affect the frequency or amplitude of mEPSCs (Fig. 5, A and B).

Figure 5. JNK contributes to chronic morphine-induced activation of presynaptic NMDARs at primary afferent terminals.

A, representative recordings and cumulative plots show a lack of effect of 50 μM AP5 on the frequency and amplitude of mEPSCs in a lamina II neuron (the spinal cord slice was pretreated with 40 μM SP600125 for 1 h) in a morphine-treated rat. B, mean changes in the frequency and amplitude of mEPSCs in lamina II neurons show a lack of effect of AP5 on the frequency and amplitude of mEPSCs in spinal cord slices pretreated with SP600125 from morphine-treated rats (n = 10 neurons from 3 rats). C, representative current traces show a lack of effect of 50 μM AP5 on the amplitude of evoked monosynaptic EPSCs in a lamina II neuron from spinal cord slices pretreated with SP600125 in a morphine-treated rat. D, original recording traces show the lack of effect of AP5 on the PPR of evoked EPSCs in spinal cord slices pretreated with SP600125 in a morphine-treated rat. E, summary data show the lack of effect of 50 μM AP5 on the amplitude (n = 10 neurons from 3 rats) and PPR (n = 9 neurons from 3 rats) of monosynaptic EPSCs in lamina II neurons from spinal cord slices pretreated with SP600125 in morphine-treated rats. Data are presented as means ± SEM.

In addition, after treatment of lamina II neurons from morphine-treated rats with SP600125, bath application of AP5 (50 μM) did not have a significant effect on the amplitude of monosynaptically evoked EPSCs (n = 10 neurons, Fig. 5, C and E) or the PPR of monosynaptic EPSCs (n = 9 neurons, Fig. 5, D and E). These findings suggest that increased JNK activity critically contributes to opioid-induced tonic activation of presynaptic NMDARs in the spinal dorsal horn.

MAPK Signaling at the Spinal Cord Level Mediates the Development of Opioid-induced Hyperalgesia and Tolerance —

The electrophysiological results in this study suggest that ERK1/2, p38, and JNK are all involved in opioid-induced tonic activation of presynaptic NMDARs, which can increase nociceptive input to spinal dorsal horn neurons. To determine whether these three MAPKs are similarly involved in opioid-induced hyperalgesia and tolerance, we determined the effects of intrathecal administration of individual MAPK inhibitors on hyperalgesia and tolerance caused by chronic morphine administration. We induced morphine analgesic tolerance in the rats via daily intraperitoneal injection of morphine (5 mg/kg twice a day) for 8 consecutive days (Chen et al. 2007, Zhao et al. 2012). We also intrathecally injected 5 μg of U0126, 5 μg of SB203580, 30 μg of SP600125, or a vehicle control (in a volume of 10 μl) 20 min before each morphine administration in separate groups of rats. The intrathecal doses of MAPK inhibitors were derived from previous studies (Chen et al. 2008, Daulhac et al. 2006, Tsai et al. 2009). We measured their withdrawal thresholds in response to noxious pressure and thermal stimuli 30 min before (baseline) and 30 min after morphine injection daily. In vehicle-treated rats, daily morphine administration caused a gradual reduction in the baseline pressure and thermal withdrawal thresholds, indicating the presence of mechanical and thermal hyperalgesia (n = 9 rats, Fig. 6, A and B). In contrast, treatment with U0126 (n = 7 rats), SB203580 (n = 7 rats), or SP600125 (n = 8 rats) blocked the decrease in the baseline withdrawal thresholds induced by chronic morphine administration (Fig. 6, A and B). Furthermore, when 5 μg of U0126, 5 μg of SB203580, and 30 μg of SP600125 were combined for intrathecal injection, no additive effect on morphine-induced mechanical and thermal hyperalgesia was observed compared to a single MAPK inhibitor (n = 8 rats, Fig. 6, A and B).

Figure 6. Inhibition of ERK1/2, p38, or JNK activity at the spinal cord level attenuates both opioid-induced hyperalgesia and analgesic tolerance caused by chronic morphine treatment.

A, time course of changes in the baseline nociceptive withdrawal threshold in response to a noxious pressure stimulus of the hindpaw in rats given systemic morphine plus intrathecal injections of a vehicle (n = 9 rats), 5 μg U0126 (n = 7 rats), 5 μg SB203580 (n = 7 rats), 30 μg SP600125 (n = 8 rats), or combination of 5 μg U0126, 5 μg SB203580, and 30 μg SP600125 (n = 8 rats). B, time course of changes in the baseline withdrawal latency measured with a noxious heat stimulus in rats given systemic morphine plus intrathecal injections of a vehicle (n = 9 rats), 5 μg U0126 (n = 7 rats), 5 μg SB203580 (n = 7 rats), 30 μg SP600125 (n = 8 rats), or combination of 5 μg U0126, 5 μg SB203580, and 30 μg SP600125 (n = 8 rats). C, time course of the analgesic effect of morphine on the mechanical nociceptive withdrawal threshold in rats given intrathecal injections of a vehicle (n = 9 rats), 5 μg U0126 (n = 7 rats), 5 μg SB203580 (n = 7 rats), 30 μg SP600125 (n = 8 rats), or combination of 5 μg U0126, 5 μg SB203580, and 30 μg SP600125 (n = 8 rats). D, time course of the analgesic effect of morphine on the heat withdrawal latency in rats given intrathecal injections of a vehicle (n = 9 rats), 5 μg U0126 (n = 7 rats), 5 μg SB203580 (n = 7 rats), 30 μg SP600125 (n = 8 rats), or combination of 5 μg U0126, 5 μg SB203580, and 30 μg SP600125 (n = 8 rats). The baseline withdrawal threshold was measured before morphine injection each day, and the analgesic effect of morphine was tested 30 min after each morphine injection (5 mg/kg, intraperitoneal). Data are presented as means ± SEM. *p < 0.05 when compared with the corresponding baseline value or morphine effect in the vehicle control group (two-way ANOVA followed by Bonferroni’s post hoc test).

Also, in vehicle-treated rats, intraperitoneal injection of 5 mg/kg morphine on day 1 produced robust increases in the pressure and thermal withdrawal thresholds. Daily morphine administration caused gradual attenuation of the analgesic effect of morphine. By day 6, morphine produced no analgesic effect, indicating the development of analgesic tolerance (n = 9 rats, Fig. 6, C and D). In comparison, treatment with U0126 (n = 7 rats), SB203580 (n = 7 rats), or SP600125 (n = 8 rats) largely attenuated the reduction in the analgesic effect of morphine (Fig. 6, C and D). Even by day 8, intraperitoneal injection of 5 mg/kg morphine still markedly increased the mechanical and thermal withdrawal thresholds in rats given any one of the three MAPK inhibitors. Additionally, when U0126, SB203580, and SP600125 were combined for intrathecal injection, no additive on the development of morphine-induced analgesic tolerance effect was detected compared to a single MAPK inhibitor (n = 8 rats, Fig. 6, C and D). These data suggest that increased activity of ERK1/2, p38, and JNK at the spinal level is required for the development of opioid-induced hyperalgesia and analgesic tolerance.

Chronic Morphine Treatment Increases the Physical Interaction between MAPKs and NMDARs in the Spinal Cord —

Because our electrophysiological and behavioral data unexpectedly showed that ERK1/2, p38, and JNK are similarly involved in opioid-induced presynaptic NMDAR activation, hyperalgesia, and analgesic tolerance, we reasoned whether chronic morphine treatment promotes the association of these three functionally redundant MAPKs with NMDARs at the spinal cord level. To determine the interaction between the NMDAR and ERK1/2, p38, and JNK in vivo, we conducted co-immunoprecipitation (co-IP) assays using total protein extracts of rat dorsal spinal cords obtained from vehicle- and chronic morphine-treated rats. We used an anti-GluN1 antibody for co-IP assays because GluN1 is an obligatory subunit of NMDARs. Co-IP analysis showed that the anti-GluN1 antibody, but not an irrelevant anti-immunoglobulin G (anti-IgG) antibody, precipitated ERK1/2, p38, and JNK proteins in spinal cord protein extracts (Fig. 7, A and B). Furthermore, co-IP analysis demonstrated that the anti-GluN1 antibody precipitated considerably more ERK1/2, p38, and JNK proteins in the spinal protein extracts from chronic morphine-treated than from vehicle-treated rats (P < 0.05, n = 6 rats per group; Fig. 7, A and B). However, the total protein amounts (i.e., input controls) for GluN1, ERK1/2, p38, and JNK in the spinal cord did not differ significantly between the vehicle-treated and morphine-treated groups (Fig. 7, A and B). These results indicate that NMDARs form a large protein complex with ERK1/2, p38, and JNK at the spinal cord level and that chronic opioid treatment promotes the association of NMDARs with the three MAPKs.

Figure 7. The physical interaction of NMDARs between ERK1/2, p38, and JNK in the spinal cord is enhanced by chronic morphine treatment.

A, Co-IP analysis shows the protein-protein interaction between the NMDAR subunit GluN1 and ERK1/2, p38, and JNK in the total protein extracts of dorsal spinal cords tissues obtained from rats given a vehicle (V) or morphine (M) for 8 days. Proteins were initially immunoprecipitated with a mouse anti-GluN1 or anti-IgG antibody. Western immunoblotting was then performed using rabbit anti-GluN1, rabbit anti-ERK1/2, rabbit anti-p38, or rabbit anti-JNK antibodies. IgG and input (tissue lysates only, without immunoprecipitation) were used as negative and positive controls, respectively. The molecular marker (mm) is shown on the left. B, Summary data show the effect of morphine treatment on the amounts of GluN1-precipitated protein complexes containing ERK1/2, p38, and JNK in the dorsal spinal cord (n = 6 rats/group). The amount of ERK1/2, p38, and JNK protein was normalized according to that of GluN1 in the same samples, and the mean value of ERK1/2, p38, JNK, and GluN1 protein levels in vehicle-treated rats was considered to be 1. Data are presented as means ± SEM. *p < 0.05 when compared with the vehicle-treated group (paired Student’s t-test).

Discussion

Our current understanding of the dynamic opioid signaling mechanism remains fragmented, and the in vitro opioid signaling is often not directly linked with the in vivo physiological or pharmacological effects of opioids. MORs at the spinal cord level are essential for the analgesic effect of systemically administered opioids (Chen & Pan 2006a). Activation of MORs typically causes G protein-mediated inhibition of voltage-activated calcium channels, resulting in attenuation of nociceptive glutamatergic transmission in the spinal dorsal horn, producing an analgesic effect (Wu et al. 2004, Zhang et al. 2016). On the other hand, β-arrestin-mediated signaling is involved in rapid attenuation of opioid analgesic actions and opioid tolerance (Pierce & Lefkowitz 2001, Bohn et al. 1999). Arrestin may function as MAPK scaffolds and facilitate activation of MAPKs (Lefkowitz & Shenoy 2005, Song et al. 2009). ERK1/2, p38, and JNK are the most extensively studied MAPKs. In general, ERK1/2 is preferentially stimulated in response to exposure to growth factors and phorbol esters, whereas JNK and p38α/β are more responsive to stress stimuli, including osmotic shock and cytokine stimulation (Pearson et al. 2001). It has been recognized that phosphorylated arrestin-bound MOR complexes not only inactivate MOR’s effects but also recruit MAPK signaling (Lefkowitz & Shenoy 2005). In the present study, we showed that inhibiting the activity of ERK1/2, p38, or JNK at the spinal cord level not only preserved the analgesic effect of morphine but also abolished morphine-induced hyperalgesia. Consistent with our results, morphine tolerance is blocked in JNK2-knockout mice (Melief et al. 2010). Our findings suggest that ERK1/2, p38, and JNK play a critical role in the development of opioid-induced hyperalgesia and analgesic tolerance.

A striking finding of our study is that increased activity of ERK1/2, p38, and JNK is required for morphine-induced tonic activation of presynaptic NMDARs in the spinal cord. NMDARs at the spinal cord level play an essential role in both the development and maintenance of opioid-induced hyperalgesia and tolerance (Mao et al. 1994, Trujillo & Akil 1991, Zhao et al. 2012). We showed that inhibiting ERK1/2, p38, or JNK activity completely blocked opioid-induced activation of presynaptic NMDARs in the spinal dorsal horn. We previously reported that even brief stimulation of MORs can trigger long-lasting NMDAR-mediated glutamate release from primary afferent terminals in the spinal cord (Zhou et al. 2010). Chronic opioid treatment can sustain activation of presynaptic NMDARs to increase nociceptive input to the spinal dorsal horn, thereby producing prolonged hyperalgesia and opposing the opioid’s analgesic effects (i.e., causing tolerance). Notably, chronic morphine treatment increases the phosphorylation of ERK1/2, p38, and JNK in primary sensory neurons (Ma et al. 2001, Chen et al. 2008). Also, ERK activation can stimulate NMDAR activity in the brain (Nateri et al. 2007), and presynaptic JNK2 regulates NMDAR-dependent glutamate release in the brain (Nistico et al. 2015). Both GluN2A- or GluN2B-containing NMDARs are involved in the increased presynaptic NMDAR activity by chronic morphine treatment (Zhao et al. 2012). The increased presynaptic NMDAR activity after chronic morphine exposure is abolished by inhibition of PKC and MAPKs, suggesting that increased NMDAR phosphorylation may enhance presynaptic NMDAR function by promoting NMDAR trafficking and surface expression and by reducing the Mg²⁺ block (Zhao et al. 2012, Lan et al. 2001, Chen & Huang 1992, Chen et al. 2018). The sources of endogenous glutamate and glycine that cause tonic activation of presynaptic NMDARs during chronic morphine treatment are not clear. Glutamate may originate from the same primary afferent nerve terminals on which the NMDARs are expressed or from excitatory interneurons in the superficial dorsal horn (Wang et al. 2018). Also, glutamate and glycine may diffuse from an adjacent active terminal or reduced glutamate/glycine reuptake due to reduced glutamate/glycine transporter activity (Mao et al. 2002). Our present findings reveal a critical role of presynaptic MAPK signaling in opioid-induced stimulation of NMDARs in the spinal cord.

Another salient finding of our study is that NMDARs and ERK1/2, p38, and JNK can form a protein complex in the spinal cord and that chronic morphine treatment substantially increases the interaction of presynaptic NMDARs with the three MAPKs. This increased association may lead to presynaptic NMDAR phosphorylation to potentiate NMDAR activity in the spinal dorsal horn. It is uncertain how the three major groups of distinctly regulated MAPK cascades converge to NMDARs in the spinal cord after MOR stimulation, although the increased presynaptic NMDAR phosphorylation seems to be the common downstream signaling mechanism for the three MAPKs. The signaling cascade and/or crosstalk among the three MAPKs may be necessary for potentiation of presynaptic NMDAR phosphorylation and activity in the spinal dorsal horn during the development of opioid-induced hyperalgesia and tolerance. In cell lines, phosphorylated p38 is capable of forming a complex with ERK1/2 (Zhang et al. 2001). In addition, PKC inhibition can block opioid-induced presynaptic NMDAR activation, hyperalgesia, and tolerance (Zhao et al. 2012). PKC is considered an early signaling component in the opioid-induced MAPK signaling pathway (Belcheva et al. 2001, Fukuda et al. 1996). In cell lines, PKC activity is required for both MOR-induced JNK activation and ERK1/2 phosphorylation (Kuhar et al. 2015, Zheng et al. 2008). Although all MOR agonists activate ERK signaling, in vitro experiments suggest that morphine uses the PKC pathway to induce ERK phosphorylation, whereas etorphine and fentanyl use the β-arrestin pathway (Zheng et al. 2008). Further studies are warranted to determine whether β-arrestin and PKC are required for the recruitment of MAPKs to NMDARs in the spinal cord during chronic opioid treatment.

In summary, our study provides important new evidence that chronic opioid treatment enhances the physical association of NMDARs with MAPKs, including ERK1/2, p38, and JNK, in the spinal cord. These three functionally interrelated MAPKs critically contribute to tonic activation of presynaptic NMDARs in the spinal cord, which is responsible for opioid-induced hyperalgesia and analgesic tolerance. This new information is important to understanding the underlying signaling mechanisms of opioid-induced hyperalgesia and analgesic tolerance. Clinically used NMDAR antagonists are effective in reducing opioid tolerance in patients (Mercadante et al. 2000, Xuerong et al. 2008). However, these drugs often produce intolerable adverse effects. MAPKs may be targeted for inhibiting opioid-induce abnormal presynaptic NMDAR activation, thereby improving opioid analgesic efficacy and minimizing opioid side effects.

Open Science Badges.

This article has received a badge for *Open Materials* because it provided all relevant information to reproduce the study in the manuscript. The complete Open Science Disclosure form for this article can be found at the end of the article. More information about the Open Practices badges can be found at https://cos.io/our-services/open-science-badges/.

Abbreviations:

AP5

2-amino-5-phosphonopentanoic acid

DRG

dorsal root ganglion

EPSC

excitatory postsynaptic current

mEPSC

miniature excitatory postsynaptic current

NMDAR

N-methyl-D-aspartate receptor

PPR

paired-pulse ratio

ERK1/2

extracellular signal-regulated kinase 1/2

JNK

c-Jun N-terminal kinase

MOR

μ-opioid receptor

MAPK

mitogen-activated protein kinase

IP

immunoprecipitation

RRID

Research Resource Identifiers