α2δ-1–Bound N-Methyl-d-aspartate Receptors Mediate Morphine-induced Hyperalgesia and Analgesic Tolerance by Potentiating Glutamatergic Input in Rodents

The μ-opioid receptor agonists remain the gold standard for the treatment of cancer pain and severe pain caused by tissue and nerve injury. However, over time, opioid use can paradoxically cause hyperalgesia and the loss of analgesic efficacy leading to rapid opioid dose escalation, a significant clinical problem in the treatment of pain with opioids. The cellular and molecular mechanisms responsible for opioid-induced hyperalgesia and analgesic tolerance are not well understood. Although it has been shown that morphine-induced hyperalgesia, but not tolerance, requires μ-opioid receptor–dependent expression of P2X4 receptors in microglia in the spinal cord,¹  there is a strong link between opioid-induced hyperalgesia and analgesic tolerance. For example, both the analgesic effect of opioids and opioid-induced hyperalgesia and analgesic tolerance are predominantly mediated by μ-opioid receptors in the dorsal root ganglion and the spinal dorsal horn.^2–6  Morphine-induced hyperalgesia and tolerance largely results from stimulation of µ-opioid receptors expressed on TRPV1-expressing dorsal root ganglion neurons.^5,7  Additionally, the glutamate N-methyl-d-aspartate (NMDA) receptors at the spinal cord level are critically involved in opioid-induced hyperalgesia and analgesic tolerance.^4,8–11  Conventional NMDA receptors are located postsynaptically, but both acute and chronic opioid treatments predominantly affect presynaptic NMDA receptors in the spinal cord.^4,12,13  Chronic morphine administration diminishes postsynaptic NMDA receptor activity in the spinal dorsal horn.⁴  NMDA receptors are expressed at the central terminals of primary sensory neurons in the spinal dorsal horn,¹⁴  although they are not functionally active under physiologic conditions.^4,13,15  Chronic opioid treatment induces tonic activation of NMDA receptors at primary afferent terminals, which augments nociceptive input to spinal dorsal horn neurons.^4,12  Blocking NMDA receptors with ketamine can potentiate the opioid analgesic effect or reduce opioid consumption in patients with opioid tolerance.^16,17  Nevertheless, it remains unclear how opioids lead to increased presynaptic NMDA receptor activity at the spinal cord level. Understanding the molecular and signaling mechanism involved could lead to improved opioid analgesic efficacy.

α2δ-1 (encoded by Cacna2d1) is generally known to be a subunit of voltage-gated Ca²⁺ channels and is expressed in the dorsal root ganglion and spinal superficial dorsal horn.^18–20  However, quantitative proteomic analysis has revealed that α2δ-1 only weakly interacts with voltage-gated Ca²⁺ channels in brain tissues.²¹  Moreover, voltage-gated Ca²⁺ channel currents in brain neurons are similar in wild-type and Cacna2d1 knockout mice.²²  Our recent findings indicate that α2δ-1, through its C-terminal domain, forms a heteromeric complex with NMDA receptors to promote their synaptic/surface trafficking.²³  Although α2δ-1–bound NMDA receptors are primarily involved in NMDA receptor hyperactivity in pathologic conditions, such as neuropathic pain and hypertension,^23,24  there is no evidence that they play a role in tonic activation of presynaptic NMDA receptors associated with opioid-induced tolerance and hyperalgesia.

In this study, we determined the role of α2δ-1–bound NMDA receptors in tonic activation of presynaptic NMDA receptors associated with opioid-induced hyperalgesia and tolerance. In this study, we provide substantial new evidence showing that α2δ-1 contributes critically to the increase in presynaptic NMDA receptor activity in the spinal cord induced by opioids. This new information extends our understanding of the role of α2δ-1–bound NMDA receptors at the spinal cord level in opioid-induced analgesic tolerance and hyperalgesia.

All experimental procedures and protocols were approved by the Animal Care and Use Committee of the University of Texas MD Anderson Cancer Center (Houston, Texas) and were performed in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Adult male Sprague–Dawley rats (9 to 11 weeks of age; Harlan, USA) were used in most of the experiments.

Repeated intrathecal injections in rats were performed via the intrathecal catheter. Single intrathecal injection in rats was performed using a lumbar puncture technique as previously described.²⁵  For intrathecal catheter insertion, the rats were placed under isoflurane-induced anesthesia and positioned prone on a stereotaxic frame. A small puncture was made in the atlantooccipital membrane of the cisterna magna. A catheter (PE-10 tubing) was then inserted such that the caudal tip reached the lumbar enlargement of the spinal cord.^3,7  We then exteriorized the rostral end of the catheter and closed the wound with sutures. The animals were allowed to recover for at least 5 days before intrathecal injections. In 20 rats with intrathecal catheters, two rats displayed neurologic deficits (e.g., paralysis) or poor grooming and were promptly killed with CO₂ inhalation.

The generation of conventional Cacna2d1 knockout mice (C57BL/6 genetic background) was described previously.²⁶  Two breeding pairs of Cacna2d1^+/− mice were purchased from the Medical Research Council (United Kingdom). Cacna2d1^−/− (knockout) mice and Cacna2d1^+/+ (wild-type) littermates were obtained by breeding the Cacna2d1^+/− heterozygous mice. The animals were ear-marked at the time of weaning (3 weeks after birth), and tail biopsies were used for genotyping. Both male and female adult mice (8 to 10 weeks of age) were used for final electrophysiologic and behavioral studies. All experiments were conducted between 9:00 am and 6:00 pm.

To induce opioid analgesic tolerance, morphine sulfate (West Ward Pharmaceuticals, USA) was injected intraperitoneally at a dose of 5 mg/kg (in rats) or 10 mg/kg (in mice) twice a day for 8 consecutive days.^4,7  The α2δ-1Tat peptide and scrambled control peptide were synthesized by Bio Basic Inc. (Canada) and validated by using liquid chromatography and mass spectrometry. The α2δ-1Tat peptide and Tat-fused scrambled control peptide were dissolved in saline and injected intrathecally, followed by a 10-µl saline flush 20 min before morphine administration on each testing day. Gabapentin (Tocris Bioscience, USA) was dissolved in saline and injected intraperitoneally before morphine administration on each testing day.

Western blotting was used to quantify the α2δ-1 protein level in the dorsal spinal cord and dorsal root ganglion. Spinal cord and dorsal root ganglion tissues at the L5–L6 levels were collected and homogenized in 300 μl of radioimmunoprecipitation assay buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1 mM Na₃VO₄, 1 mM EDTA, 1 mM NaF, 1% Nonidet P-40, and 0.25% sodium deoxycholate) with a protease inhibitor cocktail (Sigma–Aldrich, USA). The samples were homogenized with lysis buffer on ice for 30 min, and centrifuged at 13,000g for 30 min at 4°C. The supernatant was carefully collected, and its protein concentration was measured. The protein samples extracted from spinal cord and dorsal root ganglion tissues were subjected to 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane. The blots were probed with a rabbit anti–α2δ-1 antibody (1:500; #ACC-015; Alomone Labs, Israel) or rabbit anti-GAPDH antibody (1:5,000; #14C10; Cell Signaling Technology, USA). The protein bands were detected with an ECL kit (Thermo Fisher Scientific, USA), and protein band intensity was visualized and quantified using an Odyssey Fc Imager (LI-COR Biosciences, USA).

Spinal cord tissues at the L5 and L6 levels were collected and homogenized in ice-cold hypotonic buffer (20 mM Tris [pH 7.4], 1 mM MgCl₂, and 1 mM CaCl₂) containing a protease inhibitor cocktail (Sigma–Aldrich) for extracting membrane proteins. The unbroken cells and nuclei were removed by centrifugation at 300g for 5 min, and the supernatant was centrifuged for 20 min at 21,000g. The pellets were resuspended and solubilized in immunoprecipitation buffer (50 mM Tris [pH 7.4], 250 mM NaCl, 1 mM Na₃VO₄, 10 mM N-ethylmaleimide, 20 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 2 mM benzamide, 10% glycerol, and 0.5% Nonidet P-40) containing a protease inhibitor cocktail (Sigma–Aldrich), and the soluble fraction was incubated at 4°C overnight with protein A/G beads (#16-266; Millipore, Germany) prebound to mouse anti-GluN1 antibodies (1:1,000; #75-272; NeuroMab, USA). Protein A/G beads prebound to mouse immunoglobulin G were used as a control. All samples were washed three times with immunoprecipitation buffer and then immunoblotted. The following antibodies and concentrations were used for immunoblotting: rabbit anti–α2δ-1 (1:1,000; #C5105, Sigma–Aldrich; and #ACC-015; Alomone Labs) and rabbit anti-GluN1 (1:1,000; #G8913; Sigma–Aldrich).

The spinal cord tissues at the L5 and L6 levels were collected and homogenized using a glass–Teflon homogenizer in 10 volumes of ice-cold HEPES-buffered sucrose (0.32 M sucrose, 4 mM HEPES, and 1 mM EGTA at pH 7.4) containing a protease inhibitor cocktail (Sigma–Aldrich). The homogenate was centrifuged at 1,000g at 4°C for 10 min to remove sediment including large debris and nuclei. To obtain the crude synaptosomal fraction, the supernatant was centrifuged at 10,000g for 15 min. The synaptosomal pellet was lysed in 9 volumes of ice-cold HEPES buffer with the protease inhibitor cocktail for 30 min. The lysate was centrifuged at 25,000g at 4°C for 20 min to obtain the synaptosomal membrane fraction. After the protein concentration was measured, 30 μg of proteins were used for Western blotting. Postsynaptic density 95-kDa protein (PSD-95), a known synaptic protein, was used as an internal loading control for synaptosomes. The amount of α2δ-1 and GluN1 in the spinal cord synaptosomes was normalized to the amount of PSD-95 in the same gels.²³  The following antibodies and concentrations were used for immunoblotting: rabbit anti–α2δ-1 (1:1,000; #ACC-015; Alomone Labs), rabbit anti-GluN1 (1:1,000; #G8913, Sigma–Aldrich), and mouse anti-PSD95 (1:1,000; #75-348; NeuroMab).

The lumbar spinal cord was rapidly removed through laminectomy from animals that had been anesthetized with isoflurane. The spinal cord tissues were immediately placed in ice-cold sucrose artificial cerebrospinal fluid presaturated with 95% O₂ and 5% CO₂. The fluid contained 234 mM sucrose, 3.6 mM KCl, 1.2 mM MgCl₂, 2.5 mM CaCl₂, 1.2 mM NaH₂PO₄, 25.0 mM NaHCO₃, and 12.0 mM glucose. The spinal cord tissue was then placed in a shallow groove formed in an agar block and glued to the stage of a vibratome (Leica, Germany). Transverse slices (400 µm thick) of the spinal cords were cut in ice-cold sucrose artificial cerebrospinal fluid and preincubated in Krebs solution oxygenated with 95% O₂ and 5% CO₂ at 34°C for at least 1 h before being transferred to the recording chamber. The Krebs solution contained 117.0 mM NaCl, 3.6 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgCl₂, 1.2 mM NaH₂PO₄, 11.0 mM glucose, and 25.0 mM NaHCO₃. The spinal cord slices in a recording chamber were perfused with Krebs solution at 5.0 ml/min at 34°C. The lamina II outer neurons were visualized and selected for recording because they predominantly receive nociceptive input and are mostly glutamate-releasing excitatory neurons.^27–29 

Excitatory postsynaptic currents were recorded using whole cell voltage-clamp techniques. The impedance of the glass electrode was 4 to 7 MΩ when the pipette was filled with an internal solution containing 135.0 mM potassium gluconate, 2.0 mM MgCl₂, 0.5 mM CaCl₂, 5.0 mM KCl, 5.0 mM HEPES, 5.0 mM EGTA, 5.0 mM ATP-Mg, 0.5 mM Na-GTP, and 10 mM QX314 (280 to 300 mosM, adjusted to pH 7.25 with 1.0 M KOH). Excitatory postsynaptic currents were evoked from the dorsal root using a bipolar tungsten electrode connected to a stimulator (0.5 ms, 0.6 mA, and 0.1 Hz). Monosynaptic excitatory postsynaptic currents were recorded on the basis of the constant latency and the absence of conduction failure of evoked excitatory postsynaptic currents in response to a 20-Hz electrical stimulation, as we described previously.^27,28  To determine the treatment presynaptic effect, two excitatory postsynaptic currents were evoked by a pair of electrical stimulating pulses (at 50-ms intervals) applied to the dorsal root. The paired-pulse ratio of excitatory postsynaptic currents was calculated as the ratio of the amplitude of the second synaptic response to the amplitude of the first synaptic response from each trial.²⁸  Miniature excitatory postsynaptic currents were recorded at a holding potential of –60 mV in the presence of 10 μM bicuculline, 2 μM strychnine, and 1 μM tetrodotoxin. The input resistance was monitored, and the recording was abandoned if the input resistance changed by more than 15%. All signals were recorded using an amplifier (MultiClamp700B; Axon Instruments Inc., USA), filtered at 1 to 2 kHz, and digitized at 10 kHz.

All drugs were prepared in artificial cerebrospinal fluid before the recording and delivered via syringe pumps to reach their final concentrations. Tetrodotoxin citrate and 2-amino-5-phosphonopentanoic acid were purchased from Hello Bio Inc. (USA).

To quantify the mechanical nociceptive threshold in rats and mice, we conducted the paw pressure (Randall–Selitto) test on the left hind paw using an analgesiometer (Ugo Basile, Italy). To activate the device, we pressed a foot pedal that applied a constantly increasing force on a linear scale. When the animal withdrew the paw or vocalized, the pedal was immediately released, and the scale of the withdrawal threshold was recorded.³⁰  A maximum pressure of 400 g (in rats) or 200 g (in mice) was used to avoid potential tissue injury to the animals.

To assess the thermal sensitivity of the hind paw, the rats or mice were lightly restrained in a small enclosure placed on the glass surface maintained constant at 30°C (IITC Life Sciences, USA). We allowed them to acclimate for 30 min before testing. A mobile radiant heat source located under the glass was focused onto the hind paw of each animal. The paw withdrawal latency was recorded with a timer, and the hind paw was tested twice to obtain the average. A cutoff of 30 s was used to prevent potential tissue damage.^2,31 

Rotarod performance test was conducted using a rotarod accelerator treadmill (Med Associates Inc., USA) to determine motor function of rats as previously described.³²  The falling latency was measured from the start of the acceleration until the rat fell off the drum.

Data are expressed as means ± SD. No statistical methods were used to predetermine sample sizes for the studies, but our sample sizes were based on our previous experience with similar studies and were similar to those generally employed in the field. The animals were randomly assigned (1:1 allocation) to receive saline, α2δ-1Tat peptide, control peptide, or gabapentin. At least three animals were used for each recording protocol, and only one neuron was recorded from each spinal cord slice. The amplitude of the evoked excitatory postsynaptic currents was quantified by averaging six consecutive excitatory postsynaptic currents using Clampfit software (version 10.0; Axon Instruments). The miniature excitatory postsynaptic currents were analyzed off-line using the peak detection program MiniAnalysis (Synaptosoft, USA). The Kolmogorov–Smirnov test was used to compare the cumulative probability of the amplitude and interevent interval of miniature excitatory postsynaptic currents. The primary outcomes of electrophysiologic experiments are changes in the frequency of miniature excitatory postsynaptic currents and the amplitude of evoked excitatory postsynaptic currents. In biochemical experiments, the primary outcome is the target protein level. The primary outcome of the behavioral experiments is the altered withdrawal threshold and latency. For the biochemical and electrophysiologic data, two-tailed paired t tests were used to compare two groups, and one-way ANOVA followed by Tukey’s post hoc test was used to evaluate differences among more than two groups. Two-way ANOVA followed by Tukey’s post hoc test was used for the comparison of differences in behavioral data between groups/subjects (vs. vehicle/control peptide/wild-type groups) and the differences within subjects (vs. the pretreatment baseline control). The investigators performing the behavioral and electrophysiologic experiments were blinded to the treatment and genotypes. Outliers were not evaluated, and no data were excluded from statistical analysis. All statistical analyses were performed using Prism software (version 7; GraphPad Software Inc., USA). P values of less than 0.05 were considered to be statistically significant.

To determine whether chronic administration of morphine alters α2δ-1 expression levels, we measured α2δ-1 protein amounts in dorsal root ganglion and dorsal spinal cord tissues from morphine- and vehicle-treated rats. Morphine (5 mg/kg) or vehicle was injected intraperitoneally twice a day for 8 consecutive days. Immunoblotting analysis using total proteins showed that α2δ-1 levels in both the dorsal root ganglion (P = 0.039, t₍₁₀₎ = 2.37, n = 6 rats/group; fig. 1A) and dorsal spinal cord (P = 0.014, t₍₁₀₎ = 2.97; n = 6 rats/group; fig. 1B) were significantly higher in morphine-treated rats than in vehicle-treated rats.

To determine whether chronic morphine treatment affects the physical interaction between α2δ-1 and NMDA receptors in vivo, we conducted coimmunoprecipitation analyses using membrane extracts of dorsal spinal cords obtained from vehicle-treated and morphine-treated rats. Using specific antibodies, GluN1, an obligatory subunit of NMDA receptors, was coprecipitated with α2δ-1 in the dorsal spinal cord. The amounts of the α2δ-1–GluN1 protein complex in the dorsal spinal cord were significantly higher in morphine-treated rats than in vehicle-treated rats (P = 0.009, t₍₁₀₎ = 3.213, n = 6 rats/group; fig. 1C).

α2δ-1 primarily increases synaptic NMDA receptor activity via promoting NMDA receptor trafficking.²³  We therefore determined whether chronic morphine exposure increases synaptic targeting of α2δ-1 and NMDA receptors. Immunoblotting analysis using synaptosomes isolated from the dorsal spinal cord showed that both GluN1 (P < 0.001, t₍₁₀₎ = 6.056; n = 6 rats/group; fig. 1D) and α2δ-1 (P < 0.001, t₍₁₀₎ = 9.285, n = 6 rats/group; fig. 1D) protein amounts were significantly higher in morphine-treated rats than in vehicle-treated rats. These results indicate that chronic opioid treatment causes α2δ-1 upregulation and increases the α2δ-1–NMDA receptor interaction at spinal cord synapses.

The C terminus of α2δ-1 is required for its interaction with NMDA receptors,²³  and a Tat (YGRKKRRQRRR)-fused 30–amino-acid peptide (VSGLNPSLWSIFGLQFILLWLVSGSRHYLW) effectively disrupts the α2δ-1–NMDA receptor interaction in vivo.^23,33  To determine the ability of α2δ-1Tat peptide to disrupt the α2δ-1–NMDA receptor interaction in the spinal cord of morphine-treated rats, we conducted coimmunoprecipitation analyses using membrane extracts of spinal cord tissue sections treated with 1 µM Tat-fused scrambled control peptide (FGLGWQPWSLSFYLVWSGLILSVLHLIRSN) or 1 µM α2δ-1Tat peptide for 30 min. Treatment with α2δ-1Tat peptide significantly reduced the level of the α2δ-1–GluN1 protein complex in the dorsal spinal cord, compared with that treated with the control peptide (P = 0.008, t₍₁₀₎ = 3.331, n = 6 rats/group; fig. 1E).

Chronic morphine treatment increases presynaptic NMDA receptor activity but diminishes postsynaptic NMDA receptor activity in the spinal dorsal horn.⁴  Gabapentin binds primarily to α2δ-1^34,35  and is a clinically used α2δ-1 inhibitory ligand. Thus, we used gabapentin to determine whether α2δ-1 contributes to the increased presynaptic NMDA receptor activity in the spinal cord induced by chronic morphine treatment. We recorded glutamatergic miniature excitatory postsynaptic currents, which reflect spontaneous quantal release of glutamate from presynaptic terminals.⁴  In vehicle-incubated spinal cord slices from morphine-treated rats, bath application of 50 µM 2-amino-5-phosphonopentanoic acid, a specific NMDA receptor antagonist, reversed the increased frequency of miniature excitatory postsynaptic currents in dorsal horn neurons (4.95 ± 1.06 Hz vs. 6.74 ± 1.09 Hz, P = 0.007, F_(5,57) = 10.84, n = 11 neurons; fig. 2, A and B). These results suggest that chronic morphine exposure increases the activity of presynaptic NMDA receptors in the spinal dorsal horn, as shown previously.⁴  In spinal cord slices from morphine-treated rats, gabapentin pretreatment (100 µM for 60 min) substantially reduced the baseline frequency (4.44 ± 1.29 Hz vs. 6.74 ± 1.09 Hz, P < 0.001, F_(5,57) = 10.84), but not the amplitude, of miniature excitatory postsynaptic currents in dorsal horn neurons (n = 10 neurons; fig. 2, C and D). In these neurons, subsequent bath application of 50 µM 2-amino-5-phosphonopentanoic acid had no effect on the frequency or amplitude of miniature excitatory postsynaptic currents (fig. 2, C and D).

Furthermore, to determine the role of α2δ-1 in NMDA receptor–mediated synaptic glutamate release from central terminals of primary afferent nerves, we examined the effect of gabapentin on the amplitude of monosynaptic excitatory postsynaptic currents evoked from the dorsal root. In vehicle-incubated spinal cord slices from morphine-treated rats, bath application of 50 µM 2-amino-5-phosphonopentanoic acid significantly reduced the amplitude of monosynaptic excitatory postsynaptic currents (375.0 ± 85.3 pA vs. 511.3 ± 70.0 pA, P < 0.001, F_(5,57) = 8.73) and increased the paired-pulse ratio (0.61 ± 0.28 vs. 0.85 ± 0.40, P = 0.038, F_(5,57) = 4.20) of monosynaptically evoked excitatory postsynaptic currents of lamina II neurons (n = 10 neurons; fig. 3). These data are consistent with our previous findings.⁴  Gabapentin incubation (100 µM for 60 min) of spinal cord slices from morphine-treated rats considerably reduced the baseline amplitude of evoked excitatory postsynaptic currents of lamina II neurons (398.4 ± 44.6 pA vs. 511.3 ± 70.0 pA, P = 0.003, F_(5,57) = 8.73; fig. 3, C and F). After gabapentin incubation, further bath application of 2-amino-5-phosphonopentanoic acid had no significant effect on the amplitude or paired-pulse ratio of evoked excitatory postsynaptic currents of lamina II neurons (n = 10 neurons; fig. 3, D–F). These findings suggest that α2δ-1 plays a crucial role in the increase in presynaptic NMDA receptor activity in the spinal cord induced by chronic opioid treatment.

To validate the critical role of α2δ-1 in the opioid-induced increase in presynaptic NMDA receptor activity, we used Cacna2d1 knockout (Cacna2d1⁻/⁻) mice. Spinal dorsal horn neurons from morphine-treated Cacna2d1 knockout mice had lower baseline miniature excitatory postsynaptic current frequencies than did neurons from morphine-treated wild-type mice (4.81 ± 0.54 Hz vs. 6.21 ± 1.17 Hz, P = 0.007, F_(5,57) = 10.55; fig. 4). Bath application of 2-amino-5-phosphonopentanoic acid (50 µM) significantly reduced the baseline frequency of miniature excitatory postsynaptic currents in dorsal horn neurons from morphine-treated wild-type mice (4.10 ± 1.00 Hz vs. 6.21 ± 1.17 Hz, P < 0.001, F_(5,57) = 10.55, n = 11 neurons; fig. 4, A and B). In contrast, in dorsal horn neurons from morphine-treated Cacna2d1 knockout mice, application of 2-amino-5-phosphonopentanoic acid had no significant effect on the frequency or amplitude of miniature excitatory postsynaptic currents (n = 10 neurons; fig. 4, C and D).

Furthermore, the baseline amplitude of monosynaptic excitatory postsynaptic currents evoked from the dorsal root in spinal cord slices from morphine-treated wild-type mice was significantly higher than in those from morphine-treated Cacna2d1 knockout mice (449.1 ± 87.0 pA vs. 368.0 ± 61.3 pA, P = 0.04, F_(5,60) = 5.58; fig. 5). In dorsal horn neurons from morphine-treated wild-type mice, bath application of 50 µM 2-amino-5-phosphonopentanoic acid significantly reduced the amplitude of monosynaptic excitatory postsynaptic currents (327.1 ± 80.3 pA vs. 449.1 ± 87.0 pA, P = 0.003, F_(5,60) = 5.58) and increased the paired-pulse ratio (0.57 ± 0.26 vs. 0.77 ± 0.30, P = 0.02, F_(5,60) = 5.04) of evoked excitatory postsynaptic currents (n = 11 neurons; fig. 5, A–C). However, in dorsal horn neurons from morphine-treated Cacna2d1 knockout mice, 2-amino-5-phosphonopentanoic acid had no effect on the amplitude or the paired-pulse ratio of monosynaptic excitatory postsynaptic currents (n = 11 neurons; fig. 5, D–F). These data provide unequivocal evidence that α2δ-1 is essential for the opioid-induced increase in presynaptic NMDA receptor activity at the spinal cord level.

We next used α2δ-1Tat peptide to determine the role of α2δ-1–bound NMDA receptors in the opioid-induced increase in presynaptic NMDA receptor activity. In dorsal horn neurons from morphine-treated rats, the frequency, but not the amplitude, of miniature excitatory postsynaptic currents was significantly lower in slices incubated with the α2δ-1Tat peptide (1 µM for 60 min) than in slices incubated with a Tat-fused scrambled control peptide (1 µM for 60min; 4.10 ± 0.78 Hz vs. 6.60 ± 0.92 Hz, P < 0.001, F_(5,57) = 20.87; fig. 6). Furthermore, bath application of 50 µM 2-amino-5-phosphonopentanoic acid significantly reduced the frequency of miniature excitatory postsynaptic currents in control peptide-incubated dorsal horn neurons from morphine-treated rats (4.65 ± 0.91 Hz vs. 6.60 ± 0.92 Hz, P < 0.001, F_(5,57) = 20.87, n = 10 neurons; fig. 6, A and B). In contrast, 2-amino-5-phosphonopentanoic acid had no effect on the frequency of miniature excitatory postsynaptic currents in α2δ-1Tat peptide-incubated dorsal horn neurons from morphine-treated rats (n = 11 neurons; fig. 6, C and D).

We also examined whether the α2δ-1Tat peptide affects the morphine-induced activation of NMDA receptors at primary afferent terminals in spinal cord slices obtained from morphine-treated rats. The amplitude of monosynaptic excitatory postsynaptic currents of dorsal horn neurons was significantly higher in the Tat-fused control peptide-incubated group than in the α2δ-1Tat peptide–incubated group (503.0 ± 34.3 pA vs. 394.2 ± 51.5 pA, P < 0.001, F_(5,60) = 15.74; fig. 7). Bath application of 50 µM 2-amino-5-phosphonopentanoic acid markedly reduced the amplitude (388.5 ± 39.4 pA vs. 503.0 ± 34.3 pA, P < 0.001, F_(5,60) = 15.74) and increased the paired-pulse ratio (0.63 ± 0.25 vs. 0.87 ± 0.27, P = 0.002, F_(5,60) = 4.30) of monosynaptic excitatory postsynaptic currents in control peptide–incubated dorsal horn neurons from morphine-treated rats (n = 11 neurons; fig. 7, A–C). In contrast, in spinal cord slices treated with α2δ-1Tat peptide (1 µM for 60 min), further application of 2-amino-5-phosphonopentanoic acid had no significant effect on the amplitude of evoked monosynaptic excitatory postsynaptic currents (n = 11 neurons; fig. 7, D and F) or the paired-pulse ratio of evoked excitatory postsynaptic currents (n = 11 neurons; fig. 7, E and F). These findings indicate that α2δ-1–bound NMDA receptors are indispensable for opioid-induced activation of presynaptic NMDA receptors in the spinal dorsal horn.

Having shown the importance of α2δ-1–bound NMDA receptors in the opioid-induced increase in presynaptic NMDA receptor activity, we next determined whether α2δ-1 also contributes to opioid-induced hyperalgesia and analgesic tolerance. We administered intraperitoneal morphine (5 mg/kg, twice a day) to rats for 8 consecutive days.^4,7  We examined the withdrawal thresholds in response to noxious pressure and heat stimuli 30 min before (baseline) and 30 min after morphine injection (5 mg/kg) each day. We injected 100 mg/kg gabapentin (or vehicle) intraperitoneally or 1 µg of α2δ-1Tat peptide or control peptide intrathecally 20 min before each morphine treatment in separate groups of rats. Gabapentin and α2δ-1Tat peptide do not affect the acute nociceptive thresholds in naïve animals.^23,36,37 

Daily morphine injection in vehicle-treated (n = 8 rats) or control peptide–treated (n = 10 rats) rats caused a gradual reduction in the baseline withdrawal thresholds, indicating the presence of mechanical and thermal hyperalgesia (fig. 8, A–D). These rats also showed a gradual decrease in the antinociceptive effect of morphine. By day 6, intraperitoneal injection of 5 mg/kg morphine failed to produce a significant effect on withdrawal thresholds, indicating the development of analgesic tolerance (fig. 8, A–D). In contrast, cotreatment with gabapentin (n = 8 rats) or α2δ-1Tat peptide (n = 10 rats) completely blocked the reduction in baseline withdrawal thresholds induced by chronic morphine injections (fig. 8, A–D). Furthermore, cotreatment with gabapentin or α2δ-1Tat peptide substantially attenuated the reduction in the analgesic effect of morphine. Even at day 8, injection of 5 mg/kg morphine still significantly increased the nociceptive mechanical and thermal withdrawal thresholds in cotreated rats (fig. 8, A–D). To determine whether gabapentin treatment affects motor performance, we conducted rotarod tests in rats 30 min after intraperitoneal injection of 100 mg/kg gabapentin or 5 mg/kg morphine. The falling latency was not significantly affected by gabapentin (157.0 ± 26.5 s vs. 153.6 ± 21.2 s) or morphine (155.9 ± 18.3 s vs. 154.4 ± 24.1 s) treatment.

We then used Cacna2d1 knockout mice to validate the role of α2δ-1 in opioid-induced hyperalgesia and analgesic tolerance. In wild-type mice, twice-daily intraperitoneal injection of morphine (10 mg/kg) for 8 consecutive days gradually reduced the baseline withdrawal thresholds in response to noxious pressure and heat stimuli and diminished the antinociceptive effect of morphine (n = 8 mice; fig. 8, E and F). However, in Cacna2d1 knockout mice, daily morphine injections did not significantly affect the baseline withdrawal threshold. In addition, the antinociceptive effect of morphine was largely preserved at the end of 8-day morphine treatment in Cacna2d1 knockout mice (n = 8 mice; fig. 8, E and F).

In addition, we determined whether gabapentin or α2δ-1Tat peptide can reverse the established hyperalgesia and analgesic tolerance caused by chronic morphine treatment. Rats were first treated with morphine (5 mg/kg, twice a day) for 8 days and then tested after a single intraperitoneal injection of gabapentin (100 mg/kg) or vehicle or a single intrathecal injection of 1 µg of α2δ-1Tat peptide or control peptide with or without intraperitoneal administration of morphine (5 mg/kg). Treatment with gabapentin or α2δ-1Tat peptide (n = 8 rats) significantly reduced baseline mechanical and thermal hyperalgesia induced by chronic morphine administration (n = 8 rats in each group; fig. 9, A–D). Moreover, cotreatment with gabapentin and morphine or α2δ-1Tat peptide and morphine significantly potentiated on antinociceptive effect of morphine (n = 8 rats in each group; fig. 9, A–D). Together, these findings indicate that α2δ-1–bound NMDA receptors at the spinal cord level play a crucial role in the development of opioid-induced hyperalgesia and analgesic tolerance.

Our knowledge of the molecular mechanisms underlying opioid-induced hyperalgesia and analgesic tolerance is still fragmentary. Hyperalgesia and analgesic tolerance, two seemingly unrelated phenomena, may share common neural substrates that act at NMDA receptors, which are involved in the development of both opioid-induced hyperalgesia and opioid-induced analgesic tolerance.^4,8,9  In the present study, chronic morphine treatment not only increased α2δ-1 protein levels but also potentiated the physical association between α2δ-1 and NMDA receptors in the spinal cord. It has been shown that protein kinase C plays a major role in opioid-induced hyperalgesia and analgesic tolerance and in presynaptic NMDA receptor hyperactivity in the spinal cord.^4,8  However, whether there is a potential link between α2δ-1–bound NMDA receptors and protein kinases in regulating NMDA receptor activity is still uncertain. Because increased protein phosphorylation can strengthen protein-protein binding complexes,³⁸  it is possible that protein kinase C potentiates phosphorylation of α2δ-1 and/or NMDA receptor proteins to promote their physical interactions by changing their physicochemical conformation. We also found that chronic morphine exposure increased the prevalence of α2δ-1–NMDA receptor complexes at spinal cord synapses. Therefore, opioid-induced hyperalgesia and analgesic tolerance are associated with increased synaptic expression of α2δ-1–NMDA receptors at the spinal cord level.

A major finding of our study is that α2δ-1–bound NMDA receptors are essential for the opioid-induced increase in presynaptic NMDA receptor activity at primary afferent terminals. In our study, the α2δ-1Tat peptide not only abolished aberrant presynaptic NMDA receptor activity in the spinal dorsal horn but also diminished the hyperalgesia and analgesic tolerance induced by chronic morphine treatment. Because we recorded miniature excitatory postsynaptic currents in the presence of tetrodotoxin (a sodium channel blocker), voltage-gated Ca²⁺ channels were not open in this recording condition. Therefore, α2δ-1 likely enhances the presynaptic NMDA receptor activity of spinal dorsal horn neurons independent of voltage-gated Ca²⁺ channels. We have shown that the increased synaptic glutamate release to spinal dorsal horn neurons by chronic morphine administration is mediated by GluN2A- and GluN2B-containing NMDA receptors on primary afferent terminals.⁴  In addition to promoting synaptic trafficking of NMDA receptors, α2δ-1 also reduces the Mg²⁺ block of GluN2A-containing NMDA receptors,²³  which may contribute to the potentiation of NMDA receptor activity by opioids. Because neither α2δ-1 genetic ablation nor α2δ-1Tat peptide had an effect on basal NMDA receptor activity in normal conditions,^23,39  α2δ-1–bound NMDA receptors seem to be preferentially responsible for opioid-induced presynaptic NMDA receptor hyperactivity. Our results therefore show that the interaction with NMDA receptors accounts for the crucial role of α2δ-1 in opioid-induced aberrant presynaptic NMDA receptor activity.

Our findings also provide strong evidence showing that gabapentin reduces opioid-induced hyperalgesia and analgesic tolerance primarily by targeting α2δ-1–bound NMDA receptors. Gabapentinoids can attenuate opioid-induced analgesic tolerance and hyperalgesia in animal models,^40–42  and gabapentin may also reduce opioid-induced hyperalgesia and opioid consumption in patients^43,44  through a largely unknown mechanism. Although the therapeutic action of gabapentinoids, including gabapentin and pregabalin, is mediated by α2δ-1 binding,^23,26,35  gabapentinoids have no effect on the interaction between α2δ-1 and voltage-gated Ca²⁺ channel α1 subunits or voltage-gated Ca²⁺ channel activity in neurons and cell lines.^23,45,46  Gabapentin treatment (up to 7 days) does not affect voltage-gated Ca²⁺ channel trafficking or voltage-gated Ca²⁺ channel-dependent neurotransmitter release.⁴⁷  Also, the interaction between Cav2.2 and α2δ-1 is not disrupted by gabapentin,⁴⁸  which is confirmed by our recent study.²³  Our recent study reveals that gabapentinoids primarily target α2δ-1–bound NMDA receptors to normalize the nerve injury-induced increase in synaptic NMDA receptor activity in the spinal dorsal horn.²³  We showed in the present study that gabapentin fully reversed the NMDA receptor–mediated increase in the frequency of miniature excitatory postsynaptic currents and the amplitude of excitatory postsynaptic currents monosynaptically evoked from the dorsal root in chronically morphine-exposed rats. Our findings thus indicate that gabapentin alleviates opioid-induced hyperalgesia and analgesic tolerance by diminishing abnormal NMDA receptor activity at primary afferent terminals in the spinal cord. Nevertheless, our conclusion is based solely on the data from rodent models, and the effectiveness of gabapentinoids in reducing opioid-induced hyperalgesia and analgesic tolerance should be further validated in clinical studies.

In conclusion, our study provides new evidence that chronic opioid treatment causes upregulation of α2δ-1 and enhances the association between α2δ-1 and NMDA receptors in the spinal cord. Our findings support the notion that α2δ-1–bound NMDA receptors are critically involved in opioid-induced tonic activation of presynaptic NMDA receptors in the spinal cord, which augments glutamatergic input to spinal dorsal horn neurons to cause hyperalgesia and analgesic tolerance. This information is important for understanding the mechanisms of opioid-induced synaptic plasticity and suggests new strategies for improving the analgesic efficacy of opioids by eliminating aberrant presynaptic NMDA receptor activation. We demonstrated that gabapentin and α2δ-1Tat peptide not only prevented opioid-induced hyperalgesia and analgesic tolerance but also reversed established hyperalgesia and tolerance induced by chronic opioid administration. Gabapentinoids and α2δ-1 C terminus–interfering peptides do not affect physiologic α2δ-1–free NMDA receptors and could therefore be used to avoid the adverse effects associated with the use of general NMDA receptor antagonists, such as ketamine. Thus, targeting α2δ-1–bound NMDA receptors is a more desirable approach to managing opioid-induced hyperalgesia and analgesic tolerance than blocking total NMDA receptors with nonselective NMDA receptor antagonists.

Supported by National Institutes of Health (Bethesda, Maryland) grant No. R01 DA041711 (to Dr. Pan) and by the N. G. and Helen T. Hawkins Endowment (Houston, Texas; to Dr. Pan).

The authors declare no competing interests.