Gα_q signaling in primary sensory neurons shifts opioid analgesia to NMDA receptor–driven tolerance and hyperalgesia

Abstract

Opioids relieve pain by activating μ-opioid receptors (MORs), which inhibit communication between pain-sensing neurons (nociceptors) and the spinal cord. However, prolonged opioid use can paradoxically lead to increased pain sensitivity (hyperalgesia) and reduced analgesic efficacy (tolerance), partly because of the activation of NMDA-type glutamate receptors (NMDARs) at the central terminals of primary sensory neurons in the spinal cord. Here, we identified a critical role for the G protein Gα_q in this paradox. Pharmacological inhibition of Gα_q in rats reversed morphine-induced increases in NMDAR phosphorylation, synaptic trafficking, and activity at sensory neuron terminals and reduced morphine-induced excitatory nociceptive input to spinal dorsal horn neurons. Morphine enhanced Gα_q coupling specifically to metabotropic glutamate receptor 5 (mGluR5) dimers in the spinal cord. Furthermore, targeted knockdown of Gα_q in dorsal root ganglion neurons in mice normalized NMDAR-related changes and prevented NMDAR-mediated synaptic potentiation triggered by MOR activation. In addition, either pharmacological or genetic disruption of Gα_q signaling enhanced morphine’s analgesic effects while reducing hyperalgesia and tolerance. These findings reveal that Gα_q signaling contributes to opioid-induced NMDAR hyperactivity at nociceptor central terminals by promoting MOR-mGluR5 cross-talk. Targeting this pathway may improve the safety and efficacy of opioid-based pain management.

INTRODUCTION

Agonists of the μ-opioid receptor (MOR) are among the most potent analgesics for managing moderate to severe pain associated with cancer, surgery, and serious injuries. However, their use is often complicated by adverse effects such as hyperalgesia and analgesic tolerance, which can necessitate escalating doses and increase the risk of dependence and addiction. MORs are widely present in the dorsal root ganglion (DRG), spinal dorsal horn, and specific brain regions (1, 2). MORs in primary sensory neurons and their central terminals are critical for both the analgesic and hyperalgesic effects of systemically administered opioids (3–5). The MOR is a G protein–coupled receptor (GPCR) primarily coupled to inhibitory Gα_i/_o proteins (6, 7). The analgesic effects of MOR agonists largely arise from their suppression of primary afferent input to spinal dorsal horn neurons through the inhibition of voltage-gated Ca²⁺ channels (VGCCs) (8–10). In contrast, glutamate NMDA (N-methyl-d-aspartate) receptors (NMDARs) at the spinal cord level play a key role in opioid-induced hyperalgesia and tolerance (11, 12). Both short- and long-term treatments with MOR agonists potentiate presynaptic NMDAR activity at primary sensory nerve terminals in the spinal dorsal horn (12–15), leading to augmented nociceptive transmission and hyperalgesia. Furthermore, this presynaptic NMDAR hyperactivity diminishes the analgesic effect of opioids because MOR agonists have little effect on NMDAR-mediated nociceptive hypersensitivity (16–18). It is important to understand the signaling pathways linking MOR activation to presynaptic NMDAR hyperactivity.

Heterotrimeric G proteins—comprising Gα, Gβ, and Gγ subunits—are central to GPCR signaling. Gαβγ heterotrimers associate with unstimulated GPCRs through the Gα subunit until ligand binding induces a conformational change in the receptor. This leads to the release of the Gβ and Gγ subunits, which form a stable complex, and frees the Gα subunit to interact with downstream effectors. Gα subunits are classified into four families—Gα_s, Gα_i/_o, Gα_q/₁₁, and Gα₁₂/₁₃—on the basis of sequence similarity. Gα_i/_o and Gβγ subunits released from activated MORs are involved in inhibition of VGCCs (19). However, Gβγ subunits also activate phospholipase Cβ3 (PLCβ3), which counteracts opioid-induced analgesia (20, 21). PLC activation leads to the hydrolysis of phosphatidylinositol 4,5-bisphosphate, generating inositol trisphosphate and diacylglycerol, which promote intracellular Ca²⁺ release and activate protein kinase C (PKC) (22). PKC activation enhances NMDAR phosphorylation, synaptic trafficking, and activity in the spinal cord (23). The inhibition of PKC reduces opioid-induced NMDAR hyperactivity in the spinal cord and analgesic tolerance (11–13). Thus, it is critical to define the mechanisms by which inhibitory Gα_i/_o-coupled MORs activate excitatory PKC- and NMDAR-mediated pathways.

In contrast with the antinociceptive action of Gα_i/_o-coupled GPCRs, the activation of Gα_q/₁₁-coupled GPCRs, such as metabotropic glutamate receptor 5 (mGluR5), typically produces pronociceptive effects (24–26). Among the Gα_q and Gα₁₁ family members, Gα_q predominantly mediates nociceptor sensitization (27). The intrathecal administration of YM-254890, a Gα_q inhibitor, enhances the acute analgesic effects of morphine (28). Furthermore, the ablation of mGluR5 in DRG neurons attenuates opioid-induced presynaptic NMDAR hyperactivity, hyperalgesia, and tolerance (29). Moreover, in human embryonic kidney (HEK) 293 cells and NIH-3T3 cells, Gβγ subunits released from Gα_i-coupled GPCRs activate PLCβ3 only in the presence of active Gα_q (21, 30, 31), suggesting potential cross-talk between Gα_q and Gα_i signaling pathways. Nevertheless, the involvement of Gα_q in the development of opioid-induced hyperalgesia and tolerance remains largely unexplored.

Here, we investigated the role of Gα_q signaling in opioid-induced NMDAR hyperactivity, hyperalgesia, and tolerance. Our findings reveal that Gα_q signaling in DRG neurons promoted NMDAR phosphorylation, synaptic expression, and hyperactivity in response to opioids. These insights advance our understanding of the signaling mechanisms driving opioid-induced hyperalgesia and tolerance.

RESULTS

Gα_q at the spinal cord level is required for morphine-induced NMDAR phosphorylation in the spinal cord

Increased NMDAR phosphorylation is closely associated with opioid-induced presynaptic NMDAR hyperactivity in the spinal cord (12, 29, 32). NMDARs are heterotetramers of GluN1, GluN2, and GluN3 subunits, and GluN1 is the obligatory subunit of a functional NMDAR (33). We first determined whether Gα_q activity is involved in NMDAR phosphorylation in the spinal cord enhanced by prolonged opioid treatment. To this end, we collected dorsal lumbar spinal cords from rats treated with vehicle, morphine, or morphine plus FR900359 twice per day for 7 consecutive days. FR900359 is a highly potent and specific inhibitor of Gα_q (34, 35). Total proteins were extracted for immunoprecipitation using an antibody specific for phosphoserine (pSer). The input proteins and proteins precipitated by the antibody were subjected to immunoblotting analysis of GluN1. Morphine treatment significantly increased the amounts of GluN1 protein in the pSer immunoprecipitates (Fig. 1, A to C), indicating enhanced GluN1 serine phosphorylation. Cotreatment with FR900359 prevented the morphine-induced increase in pSer-GluN1 (Fig. 1, A to C). Neither morphine treatment alone nor morphine and FR900359 cotreatment had any effect on the total amounts of GluN1 in the spinal cord (Fig. 1, A to C). These findings suggest that Gα_q is critically involved in increased serine phosphorylation of NMDARs in the spinal cord induced by opioid treatment.

Fig. 1. Inhibiting Gα_q activity reduces morphine treatment–induced GluN1 serine phosphorylation and synaptic trafficking of α2δ-1–GluN1 complexes in the spinal cord.

(A to C) Representative blotting images (A) and quantification of GluN1 (B) and serine-phosphorylated GluN1 [pSer-GluN1 (C)] in the spinal cords of rats treated with saline, morphine plus vehicle (M + Veh), or morphine plus FR900359 (M + FR). Total proteins (input) and pSer immunoprecipitates (IP) were blotted for pSer and GluN1. n = 6 rats per group. (D to F) Representative blotting images (D) and quantification of GluN1 and α2δ-1 in total (E) and GluN1-immunoprecipitated (F) symaptosomal proteins from the spinal cords of rats treated as indicated. n = 6 rats per group. Rats were intraperitoneally injected with morphine (M, 5 mg/kg, twice per day) or saline for 7 consecutive days. FR900359 (5 μg) or vehicle was intrathecally injected 15 min before each morphine injection daily. Input proteins and immunoprecipitates were blotted for α2δ-1 and GluN1. β-Tubulin and PSD95 were used as loading controls. Data are means ± SEM. *P < 0.05 and ***P < 0.001 (one-way ANOVA followed by Tukey’s post hoc test).

Gα_q at the spinal cord level promotes α2δ-1–NMDAR interaction and synaptic trafficking of α2δ-1–bound NMDARs in the spinal cord

Independently of its role as a subunit in VGCCs, α2δ-1 can interact physically with phosphorylated NMDARs to promote synaptic trafficking of NMDARs in the spinal cord (23, 36). Opioid treatment increases the physical interaction between NMDARs and α2δ-1 and synaptic trafficking of α2δ-1–bound NMDARs in the spinal dorsal horn (32). Therefore, we determined whether Gα_q activity plays a role in the accumulation of α2δ-1–bound NMDARs at spinal cord synapses. Synaptosomal proteins were extracted from the dorsal spinal cords of rats treated with vehicle, morphine, or morphine plus FR900359 for 7 consecutive days. The total proteins and α2δ-1 proteins precipitated by an antibody specific for GluN1 were analyzed by immunoblotting. Compared with the vehicle group, morphine treatment significantly increased the amounts of GluN1 and α2δ-1 proteins in the synaptosomal protein input. Cotreatment with FR900359 largely normalized the amounts of GluN1 and α2δ-1 proteins in spinal synaptosomes increased by morphine treatment (Fig. 1, D to F). Furthermore, morphine treatment substantially increased the amounts of α2δ-1 protein in the GluN1 immunoprecipitates. Cotreatment with FR900359 largely blocked the increase in α2δ-1–GluN1 complexes in spinal synaptosomes induced by repeated morphine treatment (Fig. 1, D to F). These data suggest that Gα_q activity is required for opioid treatment to increase the synaptic accumulation of α2δ-1–bound NMDARs in the spinal cord.

Morphine treatment does not affect Gα_q abundance or induce direct coupling of Gα_q to MORs in the DRG or spinal cord

To determine whether prolonged morphine treatment altered the amounts of Gα_q in the DRG and spinal cord, we collected lumbar DRGs and dorsal spinal cords from rats subjected to repeated treatment with morphine or saline for 7 consecutive days. Immunoblotting analysis of extracted total and synaptosomal protein fractions showed no significant differences in Gα_q protein amounts in the DRGs or spinal cords between morphine-treated and saline-treated rats (Fig. 2, A and B).

Fig. 2. Morphine treatment increases the coupling of Gα_q-mGluR5 dimers but does not affect Gα_q abundance in the DRG or spinal cord.

(A and B) Representative blotting images (A) and quantification of Gα_q (B) in total proteins in the DRGs (n = 9 rats per group) and total and synaptosomal (Syn) proteins in the spinal cords (n = 6 rats per group) of rats treated with saline (Sal) or morphine (Mor). (C) Representative blotting for Gα_q and MOR in total proteins (input) and Gα_q immunoprecipitates (IPs) from the DRGs and spinal cords of rats treated with saline or morphine. Data are representative of n = 2 independent experiments. (D) Representative blotting images show Gα_q and mGluR5 in total proteins and Gα_q immunoprecipitates from the DRGs of rats treated with saline or morphine. Data are representative of n = 4 independent experiments. (E and F) Representative blotting for Gα_q and mGluR5 (E) and quantification of mGluR5 (F) in total proteins and Gα_q immunoprecipitates from spinal cords of rats treated with saline or morphine. Arrowheads in (E) indicate mGluR5 dimers (top) and monomers (bottom). n = 6 rats per group. Rats were intraperitoneally injected with morphine (5 mg/kg, twice per day) or saline for 7 consecutive days. Gα_q immunoprecipitates were blotted for MOR, mGluR5, and Gα_q. β-Tubulin or PSD95 was used as loading control. Data are means ± SEM. **P < 0.01 and ***P < 0.001 (two-tailed Student’s t test).

We next determined whether prolonged morphine treatment induced coupling between MORs and Gα_q in the DRG and spinal cord. Proteins extracted from the DRG and dorsal spinal cord were immunoprecipitated with an antibody specific for Gα_q. No MOR proteins were detected in the Gα_q immunoprecipitates (Fig. 2C). These results indicate that opioid treatment did not induce direct Gα_q coupling to MORs in the DRG or spinal cord.

We also determined whether prolonged morphine treatment affected Gα_q coupling to mGluR5, a Gα_q-coupled receptor, in the DRG and dorsal spinal cord. Consistent with our previous study (29), immunoblotting showed that only mGluR5 monomers (130 kDa) were detected in the DRG tissue. However, no mGluR5 protein was present in the Gα_q immunoprecipitates from the DRG (Fig. 2D), indicating no direct interaction between Gα_q and mGluR5 monomers in the DRG. By contrast, both mGluR5 monomers and dimers (260 kDa) were detected in the spinal cord tissues from morphine-treated and saline-treated rats (Fig. 2E). Morphine treatment significantly increased the total amounts of mGluR5 dimers, but not of mGluR5 monomers, in the spinal cord (Fig. 2, E and F). Furthermore, mGluR5 monomers and dimers in the Gα_q immunoprecipitates were significantly greater in the morphine group than in the saline group (Fig. 2, E and F). Although both forms of mGluR5 were detected in the spinal cord Gα_q immunoprecipitates, the dithiothreitol used in the assay can disrupt mGluR5 oligomerization by breaking disulfide bonds between monomers (29). This suggests that only mGluR5 dimers coupled to Gα_q are functional receptors in the spinal cord, the activity of which is potentiated by opioid treatment.

Gα_q mediates morphine-induced presynaptic NMDAR hyperactivity in the spinal dorsal horn

Opioid treatment increases presynaptic NMDAR activity to augment glutamatergic input to spinal dorsal horn neurons, which causes hyperalgesia and diminishes opioid analgesic efficacy (12, 13, 29, 37). We thus determined whether Gα_q was involved in the augmented presynaptic NMDAR activity in the spinal dorsal horn induced by morphine treatment. We recorded miniature excitatory postsynaptic currents (mEPSCs), which measure spontaneous quantal release of glutamate from presynaptic terminals, in lamina II neurons in spinal cord slices obtained from rats treated with morphine or saline for 7 days. The baseline frequency of mEPSCs was much higher in samples from morphine-treated rats than in saline-treated rats (Fig. 3, A to C), whereas the amplitude of mEPSCs did not differ significantly between the two groups (Fig. 3, A, B, and D). Bath application of 2-amino-5-phosphonopentanoic acid (AP5), a specific NMDAR antagonist, for 6 min rapidly decreased the frequency of mEPSCs in spinal cord slices from morphine-treated rats but had no such effect in slices from saline-treated rats. Treatment of spinal cord slices with FR900359 for 30 min largely reversed the increased baseline frequency of mEPSCs in lamina II neurons from morphine-treated rats to a frequency similar to that in saline-treated rats (Fig. 3, A to C). In these spinal cord slices treated with FR900359, bath application of AP5 had no further effect on the frequency of mEPSCs. These data suggest that morphine treatment induces tonic activation of presynaptic NMDARs in the spinal dorsal horn through Gα_q.

Fig. 3. Gα_q mediates morphine treatment–induced hyperactivity of presynaptic NMDARs at primary afferent terminals in the spinal dorsal horn.

(A and B) Representative recording traces (A) and cumulative plots (B) show the effect of bath application of 50 μM AP5 (an NMDAR antagonist) on the frequency and amplitude of mEPSCs in lamina II neurons from rats that had been treated with saline or morphine (10 mg/kg, twice per day for 7 days). Spinal cord slices were untreated or treated with FR900359 (0.1 μM) or vehicle for 30 min immediately before recording. (C and D) Summary data show baselines and the effect of AP5 on the mean frequency (C) and amplitude (D) of mEPSCs in lamina II neurons from rats treated with saline (n = 12 neurons from four rats), morphine + vehicle (n = 12 neurons from four rats), or morphine + FR900359 (n = 11 neurons from four rats). (E and F) Representative recording traces show the effect of bath application of 50 μM AP5 on dorsal root–evoked monosynaptic EPSCs (E) and the evoked paired-pulse EPSCs (F) in lamina II neurons from rats treated with saline or morphine. Spinal cord slices were treated with FR900359 (0.1 μM) or vehicle for 30 min immediately before recording. (G) Summary data show baselines and the effect of AP5 on the amplitude of evoked EPSCs in lamina II neurons from rats treated with saline (n = 12 neurons from five rats), morphine + vehicle (n = 13 neurons from five rats), or morphine + FR900359 (n = 9 neurons from four rats). (H) Mean data show baselines and the effect of AP5 on the PPR of evoked EPSCs in lamina II neurons from rats treated with saline (n = 10 neurons from four rats), morphine + vehicle (n = 10 neurons from four rats), or morphine + FR900359 (n = 9 neurons from four rats). Data are shown as means ± SEM. *P < 0.05 and ***P < 0.001 (two-way ANOVA followed by Tukey’s post hoc test).

Gα_q participates in morphine treatment–induced hyperactivity of NMDARs at primary afferent terminals in the spinal cord

To specifically determine the role of Gα_q in morphine treatment–induced activation of NMDARs at primary afferent central terminals, we examined the effect of FR900359 on the amplitude and paired-pulse ratio (PPR) of EPSCs monosynaptically evoked from dorsal root stimulation in lamina II neurons explanted from morphine- or saline-treated rats. The baseline amplitude of evoked EPSCs in explanted lamina II neurons was significantly higher in morphine-treated rats than in saline-treated rats (Fig. 3, E to H). Bath application of AP5 for 6 min quickly reversed the increased amplitude of evoked EPSCs in lamina II neurons from morphine-treated rats but had no such effect on evoked EPSCs in saline-treated rats (Fig. 3, E and G). Furthermore, treatment of spinal cord slices with FR900359 normalized the increased baseline amplitude of evoked EPSCs in lamina II neurons from morphine-treated rats. In spinal cord slices treated with FR900359, subsequent application of AP5 had no significant effect on the amplitude of evoked EPSCs in these neurons (Fig. 3, E and G).

We also examined the effect of FR900359 on the PPR of evoked EPSCs, which measures the probability of neurotransmitter release from the presynaptic terminal based on responses to a pair of stimuli (14, 32). The baseline PPR of monosynaptically evoked EPSCs was significantly smaller in lamina II neurons from morphine-treated rats than in saline-treated rats (Fig. 3, F and H). Bath application of AP5 inhibited the first evoked EPSCs more than it did the second evoked EPSCs, resulting in an increase in the PPR in morphine-treated rats. However, AP5 had no such effect on the PPR of evoked EPSCs in saline-treated rats. Treatment of spinal cord slices with FR900359 reversed the decreased baseline PPR of evoked EPSCs in lamina II neurons from morphine-treated rats (Fig. 3, F and H). Also, subsequent bath application of AP5 had no significant effect on the PPR of evoked EPSCs in these neurons treated with FR900359. These results suggest that Gα_q is involved in opioid-induced tonic activation of NMDARs at primary afferent central terminals.

Gα_q at the spinal cord level contributes to opioid-induced hyperalgesia and analgesic tolerance

We next determined whether Gα_q signaling at the spinal cord level played a role in both hyperalgesia and analgesic tolerance induced by prolonged morphine treatment. FR900359 or vehicle were administered intrathecally 15 min before each morphine injection to rats over 7 consecutive days. In the vehicle group, morphine treatment caused a gradual reduction in the baseline threshold for paw withdrawal from mechanical or thermal stimuli, indicative of mechanical and thermal hyperalgesia (Fig. 4, A to C). In addition, these rats displayed a gradual decline in the analgesic effect of morphine, consistent with the development of analgesic tolerance. Cotreatment with FR900359 significantly mitigated the reduction in baseline withdrawal thresholds and the attenuation of morphine’s analgesic effect (Fig. 4, A to C).

Fig. 4. Inhibiting Gα_q activity at the spinal cord level reduces morphine treatment–induced hyperalgesia and tolerance.

(A to C) Time course of baseline paw withdrawal thresholds (hyperalgesia) and the morphine analgesic effect (tolerance). Rats were intraperitoneally injected with morphine (5 mg/kg) twice per day for 7 consecutive days. FR900359 (5 μg) or vehicle were intrathecally injected 15 min before each morphine injection daily. Tactile (A), pressure (B), and heat (C) thresholds were tested before (baseline) and 30 min after the first morphine injection every day. n = 9 rats per group. (D to F) Time course of the acute analgesic effect of morphine in rats cotreated with vehicle or FR900359. Tactile (D), pressure (E), and heat (F) nociceptive thresholds were tested every 30 min after the first morphine injection on day 1 and day 7. n = 9 rats per group. Data are expressed as means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001 versus day 1 or time 0. #P < 0.05, ##P < 0.01, and ###P < 0.001 versus respective vehicle control at the same time point (two-way ANOVA followed by Tukey’s post hoc test).

Also, we determined whether cotreatment with FR900359 potentiated the acute analgesic effect of morphine on both the first and seventh days of morphine administration. The withdrawal thresholds were assessed at the time of the first morphine injection on both days. Intrathecal injection of FR900359 alone had no significant effect on tactile, pressure, or heat withdrawal thresholds (Fig. 4, D to F). Morphine treatment induced a transient analgesic effect, as evidenced by increased mechanical and heat withdrawal thresholds on both days 1 and 7. This analgesic effect peaked at 30 min postinjection and gradually returned to baseline (Fig. 4, D to F). Notably, the acute analgesic effect of morphine was much smaller on day 7 than it was on day 1. Cotreatment with FR900359 15 min before morphine injection potentiated the increases in withdrawal thresholds induced by morphine on both days 1 and 7 (Fig. 4, D to F). These findings suggest that Gα_q activity at the spinal cord level antagonizes the opioid analgesic effect and contributes to the development of opioid-induced hyperalgesia and tolerance.

Validation of lentiviral vectors expressing Gnaq-specific gRNA in vitro and in vivo

Because intrathecally injected agents readily access both the DRG and spinal cord (38), it is difficult to distinguish the role of Gα_q in the DRG or spinal cord in opioid-induced hyperalgesia and tolerance in this way. To specifically determine the role of Gα_q in primary sensory neurons in the development of opioid-induced hyperalgesia and tolerance, we used a CRISPR-Cas9 approach to produce mice with conditional Gα_q (encoded by the Gnaq gene) knockdown in DRG neurons. To this end, we cloned two Gnaq-specific guide RNAs (gRNAs), each of which targeted a different location in the mouse Gnaq gene, into lentiGuide-Puro plasmids. These gRNA-containing plasmids were then cotransfected with a pcDNA6 plasmid containing the complete mouse Gnaq coding sequence and a lentiCas9 plasmid into HEK293FT cells. After 3 days of coincubation, cells were collected for immunoblotting analysis of Gα_q. Both gRNAs markedly reduced Gα_q protein amounts in the presence of Cas9 (Fig. 5A). The #1 Gnaq gRNA was slightly more effective than the #2 Gnaq gRNA, so it was selected for subsequent experiments.

Fig. 5. Validation of Gnaq-specific gRNA and CRISPR/Cas9–induced Gα_q ablation in the DRG and spinal cord.

(A) Representative blotting images show the effect of two different Gnaq-specific gRNAs on Gα_q protein amounts in HEK293FT cells. Proteins were extracted from cells cotransfected with a Gnaq-pcDNA6 plasmid, a lenti-Cas9 plasmid (Cas9), and lentiGuide-Puro plasmids containing two different Gnaq-specific gRNAs or empty lentiGuide-Puro plasmid (Cont). n = 3 independent experiments. (B) Representative blotting images show the effect of the lentivirus expressing Gnaq-specific gRNA on Gα_q protein amounts in HEK293FT cells. Cells were first cotransfected with Gnaq-pcDNA6 plasmid and lenti-Cas9 plasmid and then treated with the lentivirus expressing Gnaq-specific gRNA or the control virus. n = 3 independent experiments. (C to E) Representative blotting for Cas9 and Gα_q [(C) and (D)] and quantification of Gα_q (E) in total proteins from the DRG and in total and synaptosomal (syn) proteins from the spinal cords of mice expressing Gnaq-specific gRNA lentivirus. The lentivirus expressing Gnaq-specific gRNA was intrathecally injected into Avil^Cre+::Cas9^Flox+ mice and Avil^Cre−/−::Cas9^Flox+ mice. n = 7 mice per group. The lumbar DRGs and dorsal spinal cords were obtained 20 days after virus injection. β-Tubulin and PSD95 were used as loading controls. Data are means ± SEM. ***P < 0.001 (two-tailed Student’s t test).

We packaged the lentiGuide-Puro plasmid containing the #1 Gnaq-specific gRNA (sequence and location in mouse Gnaq: gene ID: 14682, 702-GATCAACGACGAGATCGAG-720) into the lentivirus through cotransfection with packaging plasmids in HEK293FT cells. Three days later, the medium containing the virus was collected. To determine the efficacy of the Gnaq-specific gRNA–containing lentivirus, we cotransfected Gnaq-pcDNA6 and lentiCas9 plasmids into new HEK293FT cells to express Gα_q and Cas9. Five hours later, the cells were washed, and the medium was replaced with fresh medium. The Gnaq-specific gRNA lentivirus–containing medium was directly added to these cells. After 2 days of coculture, the cells were collected to assess Gα_q abundance. The lentivirus expressing Gnaq gRNA, but not the negative control virus, substantially reduced Gα_q amounts (Fig. 5B). These results confirm the successful construction of the lentivirus expressing Gnaq gRNA.

We then determined the effect of the Gnaq gRNA–expressing lentivirus on Gα_q abundance in the DRG and spinal cord in vivo. Cas9 proteins were detected in the DRGs, but not the spinal cords, obtained from Avil^Cre+::Cas9^Flox+ mice (Fig. 5, C and D). Thus, the Avil^Cre+::Cas9^Flox+ mice were considered to have conditional Cas9 knockin in primary sensory neurons. We intrathecally injected the Gnaq gRNA–expressing lentivirus into Avil^Cre−/−::Cas9^Flox+ and Avil^Cre+::Cas9^Flox+ mice at the lumbar level (29). Twenty days later, the lumbar DRGs and dorsal spinal cords were obtained from these mice, and the total and synaptosomal proteins were extracted for immunoblotting analysis. The Gnaq gRNA–expressing lentivirus reduced Gα_q amounts in the DRGs in Avil^Cre+::Cas9^Flox+ mice compared with Avil^Cre−/−::Cas9^Flox+ mice (Fig. 5, C to E). In the spinal cords, Gα_q abundance did not differ significantly between the two types of mice. Furthermore, injection of the Gnaq gRNA–expressing virus reduced Gα_q amounts in spinal cord synaptosomes in Avil^Cre+::Cas9^Flox+ mice compared with Avil^Cre−/−::Cas9^Flox+ mice (Fig. 5, D and E). These data indicate that intrathecal delivery of Gnaq gRNA–expressing lentiviruses in conditional Cas9 knockin mice efficiently ablates Gα_q expression in DRG neurons and their central terminals in the spinal cord.

Gα_q in DRG neurons mediates morphine treatment–induced NMDAR phosphorylation in the spinal cord

We examined whether Gα_q conditional knockdown (cKD) in DRG neurons diminished the NMDAR phosphorylation in the spinal cord that was enhanced by morphine treatment. We intrathecally injected Gnaq-specific gRNA–expressing lentiviruses into Avil^Cre+::Cas9^Flox+ mice and Avil^Cre−/−::Cas9^Flox+ mice to produce Gα_q-cKD and wild-type (WT) control mice, respectively. These mice were administered morphine for 7 days. Total proteins were extracted from lumbar spinal cord tissues and used for coimmunoprecipitation. The total proteins in the input and proteins precipitated by the pSer-specific antibody were subjected to immunoblotting analysis of GluN1. Morphine treatment significantly increased the amount of GluN1 proteins in the precipitates obtained with the pSer antibody in the WT mice compared with the saline-treated WT mice (Fig. 6, A to C). The increase in pSer-GluN1 amounts induced by morphine treatment was largely blocked in Gα_q-cKD mice. The total GluN1 amounts in the spinal cords did not differ significantly between morphine-treated WT mice and morphine-treated Gα_q-cKD mice (Fig. 6, A and B). These findings suggest that opioid treatment potentiates NMDAR phosphorylation in the spinal cord through Gα_q in DRG neurons.

Fig. 6. Gα_q knockdown in DRG neurons reverses morphine treatment–induced serine phosphorylation of GluN1 and synaptic trafficking of α2δ-1–GluN1 complexes in the spinal cord.

(A to C) Representative blotting images (A) and quantification of GluN1 (B) and pSer-GluN1 (C) in total proteins (Input) and pSer immunoprecipitates from the spinal cords of WT and Gα_q cKD mice treated with saline (S) or morphine (M). n = 6 mice per group. (D to F) Representative blotting images (D) and quantification of GluN1 and α2δ-1 in total synaptosomal proteins (E) and GluN1 immunoprecipitates (F) from the spinal cords of WT and Gα_q cKD mice treated with saline or morphine. n = 6 mice per group. WT and Gα_q-cKD mice were intraperitoneally injected with morphine (10 mg/kg, twice per day) or saline for 7 consecutive days. pSer or GluN1 immunoprecipitates were blotted for GluN1, pSer, and α2δ-1. The precipitated protein was normalized to the amount in the input. β-Tubulin and PSD95 were used as loading controls. Data are means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001 (one-way ANOVA followed by Tukey’s post hoc test).

Morphine treatment promotes α2δ-1–NMDAR interaction and synaptic trafficking of α2δ-1–bound NMDARs in the spinal cord through Gα_q in DRG neurons

Given our finding that conditional Gα_q knockdown in DRG neurons attenuated morphine-induced NMDAR phosphorylation, we further investigated whether Gα_q in DRG neurons was involved in synaptic trafficking of α2δ-1–bound NMDARs in the spinal cord potentiated by morphine treatment. The synaptosomal proteins were extracted from the dorsal spinal cord tissues of Gα_q-cKD and WT mice treated with morphine for 7 days. The total synaptosomal proteins and proteins precipitated using a GluN1-specific antibody were analyzed by immunoblotting for GluN1 and α2δ-1. Similar to the results in rats, morphine treatment in WT mice significantly increased the amounts of GluN1 and α2δ-1 proteins in the spinal cord synaptosomes. The potentiating effect of morphine on synaptic GluN1 and α2δ-1 proteins was largely prevented in Gαq-cKD mice (Fig. 6, D to F).

In WT mice, morphine treatment also caused a large increase in the amount of α2δ-1 in the GluN1 immunoprecipitates (Fig. 6, D to F). Furthermore, the increase in the α2δ-1–GluN1 protein complex in spinal cord synaptosomes induced by morphine treatment was substantially reduced in Gα_q-cKD mice (Fig. 6, D and F). These data suggest that Gα_q expressed in DRG neurons plays a key role in the opioid-induced synaptic trafficking of α2δ-1–bound NMDARs in the spinal cord.

Gα_q at the central terminals of DRG neurons is essential for opioid-elicited rebound LTP in the spinal dorsal horn

Acute opioid exposure can rapidly elicit rebound long-term potentiation (LTP), which is mediated by stimulation of presynaptic NMDARs at primary afferent central terminals (3, 13, 14). To investigate whether Gα_q in the central terminals of DRG neurons mediated rapid activation of presynaptic NMDARs triggered by opioid exposure, we used [D-Ala², N-MePhe⁴, Gly-ol⁵]-enkephalin (DAMGO), a highly specific and short-acting MOR agonist (13, 14), to induce rebound LTP in spinal cord slices devoid of DRG neuronal somas. In tissues from WT mice, bath application of 1 μM DAMGO for 3 min induced a rapid reduction of the baseline amplitude of evoked EPSCs in all 22 lamina II neurons tested (Fig. 7). After DAMGO washout, the amplitude of evoked EPSCs in 11 neurons (50%) returned to the baseline within 6 to 9 min. The other 11 lamina II neurons showed a rebound and persistent increase in the EPSC amplitude (LTP) after DAMGO washout, consistent with our previous reports (13, 14). In spinal cord slices from Gα_q-cKD mice, bath application of DAMGO still inhibited the amplitude of evoked EPSCs in all 16 lamina II neurons examined (Fig. 7). The amplitude of evoked EPSCs gradually returned to the baseline after DAMGO washout. None of the 16 neurons showed a rebound LTP after DAMGO washout in spinal cord slices from Gα_q-cKD mice (Fig. 7). These findings suggest that Gα_q at primary afferent central terminals is required for opioid-triggered NMDAR hyperactivity in the spinal dorsal horn by shifting the opioid effects on spinal nociceptive transmission from inhibitory to excitatory.

Fig. 7. Gα_q knockdown in DRG neurons blocks rebound LTP in the spinal dorsal horn elicited by MOR stimulation.

(A and B) Representative recording traces (A) and quantification (B) show the time course of changes in the amplitude of monosynaptic EPSCs evoked by dorsal root stimulation during baseline, during bath application of 1 μM DAMGO, and after DAMGO washout in lamina II neurons. Data are presented for WT mice (n = 22 neurons from six mice, including 11 neurons with rebound LTP and 11 neurons with no rebound LTP) and mice with conditional Gα_q knockdown (Gα_q-cKD) (n = 16 neurons from six mice). Current traces 1 to 4 correspond to the time points indicated on the time-course plot. Data are shown as means ± SEM. *P < 0.05 compared with the respective baseline (−3 min) in the same group immediately before DAMGO application (repeated measures ANOVA followed by Dunnett’s post hoc test).

Gα_q in DRG neurons mediates morphine treatment–induced hyperalgesia and analgesic tolerance

Last, we investigated the role of Gα_q in DRG neurons in hyperalgesia and tolerance induced by morphine treatment. To this end, Gα_q-cKD and WT mice were treated with morphine twice per day for 7 consecutive days. Hindpaw withdrawal thresholds were examined before (baseline) and 30 min after the first daily morphine injection. In addition, the time course of the acute analgesic effect of the first morphine injection was assessed on the first and seventh days. There were no significant differences in baseline mechanical or thermal withdrawal thresholds between Gα_q-cKD mice and WT mice before morphine treatment (Fig. 8, A to C).

Fig. 8. Gα_q signaling in DRG neurons contributes to hyperalgesia and analgesic tolerance induced by morphine treatment.

(A to C) Time course of changes in baseline paw withdrawal thresholds and the morphine analgesic effect in response to repeated treatment with morphine in WT and Gα_q-cKD mice. n = 10 mice per group. Mice were intraperitoneally injected with morphine (10 mg/kg) twice per day for 7 consecutive days. Tactile (A), pressure (B), and heat (C) thresholds were tested before (baseline) and 30 min after the first morphine injection every day. (D to F) Time course of the acute analgesic effect of morphine in WT and Gα_q-cKD mice. Tactile (D), pressure (E), and heat (F) nociceptive thresholds were tested every 30 min after the first morphine injection on day 1 and day 7. n = 10 mice per group. Data are means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001 versus day 1 or time 0. #P < 0.05, ##P < 0.01, and ###P < 0.001 versus respective WT groups at the same time (two-way ANOVA followed by Tukey’s post hoc test). (G) Schematic representation of the cross-talk between MORs and Gα_q in opioid-induced phosphorylation and synaptic trafficking of NMDARs at the central terminals of primary sensory neurons. Activation of MORs at primary afferent central terminals typically inhibits VGCCs and reduces nociceptive transmission through Gβγ release. Under normal conditions, NMDARs are minimally phosphorylated and mainly located within primary afferent terminals. Opioid treatment activates MORs and dimeric mGluR5, leading to the release of Gβγ and Gα_q, respectively. Gα_q and Gβγ cooperratively activate PLCβ3, which in turn stimulates PKC, resulting in NMDAR phosphorylation. α2δ-1 physically interacts with phosphorylated NMDARs, enhancing their synaptic expression and activity, thereby augmenting glutamatergic input to spinal dorsal horn neurons that express AMPA receptors.

In WT mice, daily morphine treatment led to a gradual reduction in baseline mechanical and thermal withdrawal thresholds, measured before each daily injection, indicating the development of hyperalgesia. A gradual decrease in the acute analgesic effect of morphine was also observed over the course of treatment (Fig. 8, A to C), suggesting the development of analgesic tolerance. However, in Gα_q-cKD mice, although there was a gradual reduction in both baseline withdrawal thresholds and acute analgesic effect of morphine, these values remained significantly higher compared with those in WT mice on the same days (Fig. 8, A to C).

Furthermore, morphine injection transiently increased mechanical and thermal withdrawal thresholds in WT and Gα_q-cKD mice when tested on both the first and seventh days (Fig. 8, D to F). The acute analgesic effects of morphine were more pronounced in Gα_q-cKD mice compared with WT mice (Fig. 8, D to F). These findings suggest that Gα_q in DRG neurons counteracts the opioid analgesic effect and promotes the development of opioid-induced hyperalgesia and tolerance.

DISCUSSION

Our study reveals that Gα_q in primary sensory neurons plays a critical role in opioid-induced hyperalgesia and tolerance. Opioids typically activate Gα_i/_o-coupled MORs to produce potent analgesia by inhibiting VGCCs involved in nociceptive transmission from primary sensory neurons to spinal dorsal horn neurons (3, 8–10). However, opioids can paradoxically cause hyperalgesia and analgesic tolerance through signaling mechanisms that are not yet fully understood. Although conditional knockout of MORs in primary sensory neurons eliminates both the analgesic and hyperalgesic effects of opioids (3), the signaling pathways underlying these divergent effects remain unclear. In TRPV1-expressing DRG neurons, which are predominantly involved in opioid-induced hyperalgesia and tolerance (39), Gα_i/_o-coupled MORs and Gα_q-coupled mGluR5 are colocalized (40, 41). In this study, we found that inhibiting Gα_q activity with FR900359 at the spinal cord level not only potentiated the acute analgesic effect of morphine but also attenuated hyperalgesia and tolerance induced by prolonged treatment with morphine. Furthermore, using CRISPR-Cas9 to conditionally knock down Gα_q in DRG neurons, we showed that Gα_q knockdown not only diminished Gα_q amounts in the DRG but also markedly reduced Gα_q in spinal synaptosomes. This suggests that a large portion of Gα_q proteins in the dorsal spinal cord originates from DRG neurons and is anterogradely transported to their central terminals. Moreover, Gα_q removal in DRG neurons enhanced the analgesic effect of morphine and reduced morphine treatment–induced hyperalgesia and tolerance. These findings provide new evidence that Gα_q in primary sensory neurons mediates opioid-induced hyperalgesia and tolerance.

A key finding of our study is that Gα_q signaling is essential for opioid-induced hyperactivity of NMDARs at the central terminals of primary sensory neurons. NMDAR antagonists reduce opioid-induced hyperalgesia and tolerance in both animal models and humans (11, 12, 42, 43). NMDARs at central terminals of DRG neurons are normally inactive but become tonically activated in opioid-induced hyperalgesia and tolerance (12, 15, 32). Consistent with this, the genetic ablation of GluN1 in primary sensory neurons has no effect on normal nociception but reduces opioid-induced hyperalgesia and tolerance (13, 44). The potentiated NMDAR activity at primary afferent central terminals amplifies excitatory nociceptive input to spinal dorsal horn neurons, resulting in hyperalgesia. Furthermore, because MOR agonists do not inhibit NMDAR-mediated nociceptive hypersensitivity, this presynaptic NMDAR hyperactivity diminishes opioid analgesic efficacy. In the present study, we used the spinal cord slice preparation that expresses native MORs and preserves the synaptic contact between primary afferent nerves and dorsal horn neurons. We showed that inhibiting Gα_q with FR900359 in spinal cord slices from morphine-treated animals diminished NMDAR-mediated increases in the mEPSC frequency and the amplitude of EPSCs monosynaptically evoked from the dorsal root, highlighting the important role of Gα_q activity in opioid-induced presynaptic NMDAR hyperactivity.

Acute opioid exposure can rapidly induce hyperalgesia in both animals and humans (45–47) and trigger presynaptic NMDAR-dependent rebound LTP in spinal glutamatergic excitatory neurons (3, 13, 14). In this study, we used DAMGO-elicited rebound LTP as a real-time readout of synaptic NMDAR activity in spinal cord slices. We found that conditional Gα_q knockdown in DRG neurons did not affect the initial inhibitory effect of DAMGO on glutamatergic input to spinal dorsal horn neurons involved in the acute analgesic effect of opioids. Ablating Gα_q in DRG neurons diminished DAMGO withdrawal–elicited rebound LTP in spinal dorsal horn neurons. Therefore, Gα_q at the central terminals of DRG neurons probably acts as an upstream signal for opioid-induced presynaptic NMDAR hyperactivity in the spinal cord. This finding highlights the crucial role of Gα_q in switching the opioid effects from inhibitory to excitatory at the spinal cord level.

Another salient finding of our study is that Gα_q activity in DRG neurons was crucially involved in opioid-induced serine phosphorylation of NMDARs in the spinal cord. Our current understanding of signaling mechanisms underlying opioid-induced hyperalgesia and tolerance is still fragmented. The activation of PLCβ-PKC signaling is typically a downstream effect of Gα_q-coupled receptors (48). The phosphorylation regulation of NMDARs by PKC is an important process increasing NMDAR activity (49, 50). Also, treatment with opioids increases PLCβ3 and PKC activity in DRG neurons (51). NMDAR phosphorylation by PKC is crucial for morphine treatment–induced NMDAR hyperactivity and synaptic trafficking in the spinal cord (12, 13), which contributes to opioid-induced hyperalgesia and tolerance (11, 52). In the present study, Gα_q inhibition at the spinal cord level or Gαq ablation in DRG neurons largely diminished the serine phosphorylation of GluN1 in the spinal cord that was augmented by morphine treatment, suggesting that Gα_q in DRG neurons plays a pivotal role in opioid-induced PKC activation and NMDAR phosphorylation. In addition, the activation of Gα_q-coupled receptors stimulates mitogen-activated protein kinase (MAPK) through PKC, and Gβγ is responsible for MAPK activation by Gα_i-coupled receptors (53, 54). Because MAPK in the spinal cord contributes to opioid-induced presynaptic NMDAR hyperactivity and analgesic tolerance (37), MAPK may also mediate NMDAR phosphorylation caused by opioids.

Under normal conditions, both NMDARs and mGluR5 are localized in the central terminals of primary afferents but are not present on their synaptic surface, rendering presynaptic NMDARs and mGluR5 functionally inactive in the spinal cord. α2δ-1, commonly known as a VGCC subunit, is also a phospho-binding protein that preferentially interacts with phosphorylated NMDARs (23, 36). The recruitment of NMDARs to spinal nociceptive synapses by α2δ-1 is the fundamental mechanism underlying synaptic NMDAR hyperactivity in opioid-induced hyperalgesia and neuropathic pain conditions (29, 32, 36, 55). Moreover, α2δ-1 is essential for PKC activation–induced NMDAR synaptic trafficking in the spinal dorsal horn (23). Prolonged morphine administration increases the α2δ-1–NMDAR interaction in the spinal cord (32). Inhibition of α2δ-1 with gabapentin or disruption of the α2δ-1–NMDAR interaction using an α2δ-1 C terminus peptide reduces opioid-induced presynaptic NMDAR hyperactivity, hyperalgesia, and tolerance (13, 32). Opioid exposure promotes synaptic trafficking and expression of NMDARs through α2δ-1 and mGluR5 at primary afferent central terminals (29, 32). Consequently, synaptically released glutamate can activate presynaptic NMDARs and mGluR5, strengthening glutamatergic input from primary afferents to spinal dorsal horn neurons, thereby contributing to opioid-induced hyperalgesia and tolerance. Our current study showed that Gα_q inhibition or knockdown in DRG neurons effectively reversed the morphine-induced increase in synaptic accumulation of α2δ-1–bound NMDARs in the spinal cord. These findings support the notion that Gα_q functions as an upstream regulator of opioid-induced synaptic trafficking of α2δ-1–bound NMDARs at the central terminals of DRG neurons.

In the present study, we demonstrated that morphine treatment markedly increased the association of Gα_q with mGluR5 in the spinal cord. However, the mGluR5 monomer was the only form present in the DRG and did not interact with Gα_q in the DRG. Therefore, only mGluR5 dimers at the central terminals of DRG neurons are likely functional and capable of forming a large signaling complex with NMDARs and Gα_q. We found no evidence of a direct interaction between MORs and Gα_q proteins in either the DRG or spinal cord. In addition, MORs and mGluR5 do not directly interact in either the DRG or spinal cord (29). Because MORs have no physical interactions with mGluR5 or Gα_q, the coincident activation of the PLCβ3-PKC-NMDAR signaling pathway by MOR and mGluR5 stimulation is likely mediated by their respective G proteins. In this regard, Gβγ released downstream of MOR and Gα_q activation by mGluR5 may act together to activate PLCβ3 and PKC, resulting in NMDAR phosphorylation and synaptic trafficking. In cell lines, Gβγ induces PLCβ3 activation only in the presence of Gα_q (21, 30, 31). In the hypothalamus, activation of mGluR5 switches the effect of Gα_i/_o-coupled group III mGluRs from inhibitory to excitatory (56), also suggesting cross-talk between Gα_q-coupled and Gα_i/_o-coupled GPCRs. mGluR5 antagonists increase the analgesic efficacy of morphine and reduce opioid analgesic tolerance (29, 57). Morphine treatment promotes mGluR5 trafficking from DRG neurons to their central terminals and enhances mGluR5-NMDAR interactions in the spinal cord (29). In addition, ablating mGluR5 in DRG neurons attenuates opioid-induced presynaptic NMDAR hyperactivity and associated hyperalgesia and tolerance (29). Therefore, mGluR5 and MORs at the central terminals of DRG neurons likely play a mutually important role in opioid-induced coincident stimulation of the PLCβ3-PKC-NMDAR signaling cascade. Nonetheless, although mGluR5 is the most likely Gα_q-coupled receptor involved in opioid-induced hyperalgesia and tolerance, Gα_q may also be activated by other Gα_q-coupled receptors, such as neurokinin 1 receptors, GPR139, and odd-numbered muscarinic acetylcholine receptors. These GPCRs are expressed in the DRG and spinal cord and could contribute to the attenuation of opioid analgesic actions (58–61).

In summary, our findings provide new in vivo evidence linking Gα_q signaling in DRG neurons to opioid-induced presynaptic NMDAR hyperactivity, hyperalgesia, and tolerance. We propose that Gβγ from activated MORs and Gα_q from mGluR5 cooperate to activate PLCβ3-PKC, leading to NMDAR phosphorylation and the synaptic trafficking of α2δ-1–bound NMDARs to the central terminals of primary sensory neurons (Fig. 8G). This mechanism underscores the critical role of Gα_q in transitioning opioid effects from antinociceptive to pronociceptive. This information not only extends our mechanistic understanding of how Gα_q and MORs are reciprocally involved in opioid-induced hyperalgesia and tolerance but also suggests new strategies for treating this condition. Inhibiting Gα_q activity or disrupting the MOR-Gα_q cross-talk could enhance opioids’ analgesic efficacy and reduce the risk of opioid use disorder.

MATERIALS AND METHODS

All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee of University of Texas MD Anderson Cancer Center and conformed to the NIH guidelines on the ethical use of animals. The sample sizes used in the study were estimated with an α value of 0.05 and a statistical power of 85%, aligning with those commonly used in the field (12, 29, 32, 62). Animals were assigned to the control and treatment groups with 1:1 allocation based on availability. Investigators performing behavioral tests and electrophysiological recordings were blinded to treatment groups. Data from male and female animals were pooled, given that no sex differences were observed in the degree or time course of morphine-induced hyperalgesia and tolerance and spinal NMDAR hyperactivity, consistent with findings from previous reports (3, 13, 29, 32).

Animals

Adult male and female Sprague-Dawley rats (180 to 250 g) used in this study were obtained from Envigo and housed with no more than three rats per cage. Cas9^Flox+/+ mice were purchased from the Jackson Laboratory. Cas9^Flox+/+ mice have a floxed-STOP cassette preventing Cas9 expression. Avil^Cre+ mice (63) were provided by F. Wang (Massachusetts Institute of Technology). To induce Cas9 knockin in primary sensory neurons, Avil^Cre+::Cas9^Flox+ mice were obtained by breeding the male Avil^Cre+ mice with female Cas9^Flox+/+ mice. Mouse genotypes were confirmed by using ear biopsies. All mice had a C57BL/6 genetic background and were housed with no more than five mice per cage. Adult male and female mice (8 to 12 weeks old) were used for final experiments.

For induction of opioid-induced hyperalgesia and tolerance, morphine was injected intraperitoneally in mice (10 mg/kg, twice per day) or rats (5 mg/kg, twice per day) for 7 consecutive days (29). To inhibit Gα_q activity at the spinal cord level, the Gα_q inhibitor FR900359 (35) was intrathecally injected at the L4/L5 level via lumbar puncture.

Construction of lentiviral vectors expressing Gnaq-specific gRNA

We constructed a lentivirus expressing a gRNA targeting mouse Gnaq using previously established protocols (29, 64, 65). Briefly, we first cloned the Gnaq-specific gRNA into a lentiGuide-Puro plasmid (Addgene). To assess the effectiveness of the gRNA-containing lentiGuide-Puro plasmids on the knockdown of Gα_q, we cloned the full-length mouse Gnaq coding sequence (gene ID: 14682) into a pcDNA6 plasmid. The gRNA-containing lentiGuide-Puro plasmids were cotransfected with pcDNA6- Gnaq and lentiCas9 plasmids (Addgene) into HEK293FT cells by using Polyjet transfection reagent (SignaGen Laboratories). After a 3-day coincubation, the cells were harvested for immunoblotting analysis to confirm the Gα_q knockdown. The lentiGuide-Puro plasmid without the gRNA insertion served as a negative control. Subsequently, the lentiGuide-Puro plasmid with the most effective Gα_q knockdown was packaged into the lentivirus by cotransfecting the packaging plasmids pCMV-VSV-G (Addgene) and psPAX2 (Addgene) into HEK293FT cells. The virus-containing culture medium was harvested 72 hours postculture. The virus was then concentrated 100-fold using the Lenti-X Concentrator reagent (Takara Bio), and its titer was determined using the QuickTiter Lentivirus Quantitation Kit (Cell Biolabs). The concentrated virus, with a titer of 2 × 10¹² virus particles/ml, was stored at −80°C until use. A negative control virus was generated using the empty lentiGuide-Puro plasmid with the same packaging method.

Tissue collection, spinal synaptosome preparation, and immunoprecipitation

The animals were euthanized via decapitation after being deeply anesthetized with 5% isoflurane. The DRGs and the dorsal spinal cords at the L3 to L5 level were collected for protein analysis. Total proteins from the DRG and dorsal spinal cord tissues were extracted using radioimmunoprecipitation assay (RIPA) lysis buffer (MilliporeSigma) containing a protease inhibitor cocktail. To isolate synaptosomal proteins, the spinal cord tissues were homogenized in ice-cold homogenization buffer consisting of 0.32 M sucrose, 10 mM Hepes, 2 mM EDTA (pH 7.4), and a protease inhibitor cocktail, as previously described (55, 66). The homogenate was then centrifuged at 800g for 10 min at 4°C to remove insoluble cell debris. The resulting supernatant was further centrifuged at 13,000g for 20 min to obtain crude synaptosomes. The synaptosome pellets for immunoblotting were solubilized in RIPA lysis buffer with a protease inhibitor cocktail for 1 hour on ice, followed by centrifugation at 10,000g for 15 min at 4°C.

For immunoprecipitation, the proteins from the DRG and spinal cord were solubilized in immunoprecipitation lysis buffer containing a proteinase inhibitor cocktail. Proteins were incubated at 4°C overnight with protein G beads (MilliporeSigma) along with either a rabbit anti-pSer antibody (1:100; catalog no. AB1603, MilliporeSigma), a rabbit anti-GluN1 antibody (1:500; catalog no. G8913, MilliporeSigma), a rabbit anti-Gα_q antibody (1:100; catalog no. 14373, Cell Signaling Technology), or normal rabbit immunoglobulin G (IgG) (catalog no. 12-371, MilliporeSigma). After incubation overnight, all samples were washed three times with immunoprecipitation lysis buffer. Subsequently, the beads were incubated with 1× NuPage loading buffer supplemented with an additional 100 mM dithiothreitol and 1% SDS for 10 min and then boiled for 5 min. The eluted proteins were then used for further immunoblotting analysis.

Immunoblotting

Immunoblotting was conducted as described previously (29, 67). In brief, proteins were separated on 4 to 12% SDS NuPage bis-tris gels (Thermo Fisher Scientific) and subsequently transferred to polyvinylidene fluoride membranes (MilliporeSigma). The membranes were then incubated with primary antibodies overnight at 4°C. This was followed by incubation with horseradish peroxidase (HRP)–conjugated secondary antibodies (1:5000, anti-rabbit IgG, catalog no. 7074; 1:5000, anti-mouse IgG, catalog no. 7076; Cell Signaling Technology) for 1 hour at 22°C. To detect protein expression from immunoprecipitation samples, HRP-conjugated TrueBlot secondary antibodies were used (1:5000, anti-rabbit IgG, catalog no. 18–8816-31; 1:5000, anti-mouse IgG, catalog no. 18-8817-31, Rockland Immunochemicals). The membranes were thoroughly washed after each antibody incubation. Immunoblots were developed using a chemiluminescence kit (Thermo Fisher Scientific) or an enhanced chemiluminescence kit (Thermo Fisher Scientific). The protein bands were visualized using an Odyssey Fc Imager (LI-COR Biosciences) and quantified using ImageJ software. The primary antibodies included rabbit anti-Gα_q (1:2000; catalog no. 14373, Cell Signaling Technology), rabbit anti-GluN1 (1:2000; catalog no. G8913, MilliporeSigma), rabbit anti-MOR (1:2000; catalog no. AB1580-I, MilliporeSigma), mouse anti–β-tubulin (1:10,000; catalog no. 32-2600, Thermo Fisher Scientific), mouse anti-PSD95 (1:5000; catalog no. MABN1190, MilliporeSigma), mouse anti–α2δ-1 (1:2000; catalog no. NB120-2864, Novus Biologics), rabbit anti-mGluR5 (1:1000; catalog no. AB5675, MilliporeSigma), and mouse anti-Cas9 (1:2000; catalog no. 14697, Cell Signaling Technology). The specificity of primary antibodies used in this study has been validated using genetic knockout or knockdown (3, 4, 29, 36). For protein quantification, the intensity of the protein bands was normalized to that of PSD95 or β-tubulin on the same gels.

Nociceptive behavioral tests

To measure the tactile withdrawal threshold, animals were placed in an individual plastic box on a mesh floor. A series of calibrated von Frey filaments was applied perpendicularly to the plantar surface of the hindpaw with sufficient force to bend the filaments for 6 s. A brisk paw withdrawal or flinch was considered a positive response. In the absence of a response, the next filament with greater force was applied. If a response occurred, the next filament with lower force was used. Six consecutive responses after the first change were used to calculate the withdrawal threshold using the “up-down” method (68, 69).

The pressure withdrawal threshold was assessed using a digital Randall-Selitto paw pressure device (catalog no. 2500, IITC Life Science). The device gently held the animal’s hindpaw while a steadily increasing force was applied via a pointed end to the midplantar glabrous surface. The force was immediately stopped upon a withdrawal response, and the threshold was recorded (29, 70).

The thermal withdrawal latency was measured with a thermal testing apparatus (catalog no. 390G, IITC Life Science). Animals were placed on a glass surface maintained at 30°C and allowed to acclimate to the device. A mobile radiant heat stimulus was then applied to the plantar surface of the hindpaw until the animal lifted or licked the hindpaw (10, 67). The time taken for hindpaw withdrawal was recorded as the thermal withdrawal latency.

Electrophysiological recordings in spinal cord slices

The animals were anesthetized with 3% isoflurane, and the lumbar spinal cords at the L4 to L6 levels were quickly removed via laminectomy. The animals were then killed via inhalation of 5% isoflurane, followed by rapid decapitation. The spinal cords were immediately put into 95% O₂ and 5% CO₂ presaturated ice-cold artificial cerebrospinal fluid containing the following reagents: 25 mM glucose, 234 mM sucrose, 3.6 mM KCl, 26 mM NaHCO₃, 1.2 mM NaH₂PO₄, 2.5 mM CaCl₂,and and 1.2 mM MgCl₂. The tissue was then glued onto the stage of a vibratome and cut into 400-μm-thick transverse slices. The slices were incubated in Krebs solution containing 11 mM glucose, 117 mM NaCl, 3.6 mM KCl, 25 mM NaHCO₃, 1.2 mM NaH₂PO₄, 2.5 mM CaCl₂, and 1.2 mM MgCl₂ (gassed with 95% O₂ and 5% CO₂) at 34°C for at least 1 hour before recordings. We then placed the spinal cord slices in the recording chamber with continuous perfusion of oxygenated Krebs solution (3 ml/min) at 34°C.

We identified neurons in the lamina II outer layer using infrared illumination and differential interference contrast under a microscope. EPSCs from these neurons were recorded using the whole-cell voltage-clamp mode at the holding potential of −60 mV, as we described previously (13, 69). A glass pipette electrode (4 to 7 megohm) was filled with the internal solution containing 135 mM K-gluconate, 5 mM KCl, 2 mM MgCl₂, 0.5 mM CaCl₂, 5 mM EGTA, 5 mM Hepes, 0.5 mM Na₂–guanosine 5′-triphosphate, 5 mM Mg-adenosine 5′-triphosphate, and 10 mM lidocaine N-ethyl bromide (QX314; 280 to 300 mOsm, pH 7.3). QX314 was included in the pipette recording solution to suppress postsynaptic neuronal firing. EPSCs were evoked by electrical stimulation (0.6 mA, 0.5 ms, and 0.1 Hz) of the dorsal root to elicit glutamate release from primary afferent nerves. Monosynaptic EPSCs in lamina II neurons were identified on the basis of the constant latency and absence of conduction failure of evoked EPSCs in response to 20-Hz electrical stimulation (14, 71). Neurons with rebound LTP were defined as those showing the increased amplitude of EPSCs at least 20% above the baseline and lasting 15 min after DAMGO washout (3, 71).

To determine the PPR, we generated a pair of stimuli at 50-ms intervals to evoke EPSCs. The PPR was expressed as the ratio of the amplitude of the second synaptic response to the amplitude of the first synaptic response (67, 72). Also, mEPSCs were recorded in the presence of 0.5 μM tetrodotoxin. The recording was discontinued if the input resistance changed by more than 15% during the session. All signals were recorded using an amplifier (MultiClamp700B, Molecular Devices), filtered at 1 to 2 kHz, and digitized at 10 kHz. Only one neuron was recorded from each spinal cord slice, and at least four mice or rats were used for each recording protocol.

AP5 and tetrodotoxin were purchased from Hello Bio Inc., and FR900359 was obtained from Cayman Chemical Company. DAMGO was obtained from MilliporeSigma. Agents were diluted in artificial cerebrospinal fluid before slice recordings. All drugs were delivered at their final concentrations via syringe pumps.

Statistical analysis

All data are expressed as means ± SEM. The amplitude and frequency of mEPSCs were analyzed using the MiniAnalysis peak detection program (Synaptosoft). The cumulative probability of the amplitude and interevent interval of mEPSCs was compared using the Komogorov-Smirnov test, which evaluates the probability that two cumulative distributions are similar. The evoked EPSCs and PPR were analyzed using Clampfit 10.0 software (Molecular Devices), and the amplitude of EPSCs was quantified by averaging six consecutive currents. Two-tailed Student’s t tests were used to determine the differences between two groups. One-way or two-way analysis of variance (ANOVA) followed by Tukey’s or Dunnett’s post hoc tests was used to compare more than two groups. All statistical analyses were performed using Prism software (GraphPad Software Inc). P values of less than 0.05 were considered statistically significant.

Data and materials availability:

All data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials. All materials generated in this study, such as viral vectors, are available from the authors upon reasonable request.