Constitutive μ-opioid receptor activity leads to long-term endogenous analgesia and dependence

Abstract

Opioid receptor antagonists increase hyperalgesia in humans and animals, which indicates that endogenous activation of opioid receptors provides relief from acute pain; however, the mechanisms of long-term opioid inhibition of pathological pain have remained elusive. We found that tissue injury produced μ-opioid receptor (MOR) constitutive activity (MOR(CA)) that repressed spinal nociceptive signaling for months. Pharmacological blockade during the posthyperalgesia state with MOR inverse agonists reinstated central pain sensitization and precipitated hallmarks of opioid withdrawal (including adenosine 3',5'-monophosphate overshoot and hyperalgesia) that required N-methyl-D-aspartate receptor activation of adenylyl cyclase type 1. Thus, MOR(CA) initiates both analgesic signaling and a compensatory opponent process that generates endogenous opioid dependence. Tonic MOR(CA) suppression of withdrawal hyperalgesia may prevent the transition from acute to chronic pain.

Fig. 1. Injury-induced pain sensitization is tonically opposed by spinal MOR-G-protein signaling

(A) Progression of mechanical hyperalgesia following intraplantar CFA (5 μl) (n = 10). (B) Resolution of hyperalgesia during and 14d after infusion of NTX (10 mg/kg/d, s.c.) in Sham and CFA mice (n = 5–6). ★ P < 0.05 compared to CFA+saline, ◇ P < 0.05 compared to Sham+NTX. (C) Time course of reinstatement of hyperalgesia following subcutaneous NTX (3 mg/kg) in CFA-21d mice (n = 6–13). (D) Dose-response effects of NTX on hyperalgesia (n = 6 per dose). MPE: maximal possible effect. (E–F) Effect on hyperalgesia of (E) subcutaneous or (F) intrathecal NTX (3 mg/kg or 1 μg) or NMB (3 mg/kg or 0.3 μg) (n = 5–10). (G–J) Effect of NTX (1 μg, i.t.) on reinstatement of (G) mechanical hyperalgesia in Sham and CFA mice (n = 5–8), (H) heat hyperalgesia (n = 5–10), (I) spontaneous pain (n = 4–8), and (J) post-operative pain (n = 6–11). (K) Effect of pertussis toxin (0.5 μg, i.t.) on hyperalgesia (n = 6). (L) Representative radiograms and (M) dose-response effects of DAMGO-stimulated GTPγS³⁵ binding in lumbar spinal cord; inset: binding E_max (n = 7–9). (N) Effect of DAMGO (i.t.) on hotplate latency (n = 8). (O) Effect of CTOP (100 ng, i.t.) on hyperalgesia (n = 6–7). (P–R) Representative images and (S) dorsal horn laminar quantification (I–II and III–V) of light touch-evoked pERK after NTX (1μg, i.t.) (n = 5–7). (T) Confocal image of pERK+ cells. (U–W) From boxed region in panel T: Co-localization of pERK with NeuN. All scale bars = 200 μm. ★ P < 0.05 for all panels. All data shown as mean±s.e.m. See fig. S1 for full time course data of panels E–J, O.

Fig. 2. Pain reinstatement requires potentiated N-methyl-D-aspartate receptor activation of calcium-sensitive adenylyl cyclase type 1

(A) Time course of glutamate-evoked (0.3 mM) [Ca²⁺]_i in spinal cord slices from sham and CFA mice (n = 4–7 mice). (B) Effect of CTOP (1 μM), NTX (10 μM) or NTX+MK-801 (100 μM) on [Ca²⁺]_i. Values are relative to pre-drug control responses (n = 3–5 mice). (C) Representative F₃₈₀ nm image of dorsal horn neurons from a CFA-21d slice responding to glutamate before (top) and after NTX (10 μM, bottom) (yellow arrows). Decrease in fluorescence intensity corresponds to increase in [Ca²⁺]_i. The red traces illustrate the rise in [Ca²⁺]_i for the indicated cell (red arrow). Insets depict area in white box. Scale bars: 0.02 ΔF/F (vertical) and 3 min (horizontal), and 100 μm and 10 μm (inset). Effect of MK-801 (1 μg, i.t.) on NTX-precipitated (1 μg, i.t.) (D) hyperalgesia and (E) touch-evoked dorsal horn pERK expression (n = 5–10). (F) Spinal cord cAMP levels after intrathecal vehicle (n = 14–18), CTOP (100 ng; n = 6) or NTX (1 μg; n = 6–10). (G) Effect of MK-801 (1 μg, i.t.) on NTX-precipitated spinal cAMP overshoot (n = 5). (H–K) Effect of intrathecal NMDA (3 pmol; n = 5–7) or forskolin (1.5 μg; n = 6) on (H and J) spontaneous nocifensive behaviors and (I and K) spinal cAMP levels. (L) Progression of mechanical hyperalgesia and (M) effect of intrathecal NTX in AC1^−/− and wild-type mice 21d after CFA (n = 5–9). Effect of NB001 (1.5 μg, i.t.) on NTX-precipitated (N) hyperalgesia (n = 4–7), (O) spinal cAMP levels (n = 6–9), and (P) affective pain (n = 6–12). (Q) Schematic of cellular pathways involved in endogenous opioid withdrawal and pain reinstatement. ★ P < 0.05.

Fig. 3. Spinal MORs acquire constitutive activity after injury

(A,B,C) Effects of NTX (1μM or 1 μg), 6β-naltrexol (1μM or 10 μg) or co-administration of 6β-naltrexol+NTX in sham and CFA-21d mice on (A) [Ca²⁺]_i in spinal cord slices (n = 5–6), (B) spinal cAMP levels (n = 6–11) and (C) hyperalgesia (n = 6–7). (D) Effect of intrathecal 6β-naltrexol and/or NTX on hyperalgesia in Paw Incision-21d mice (n = 6–7). (E) Effect of intrathecal β-funaltrexamine (β-FNA; 2.5 μg) on hyperalgesia (n = 6–7). (F) Representative radiograms and (G) dose-response effects of β-FNA on basal GTPγS³⁵ binding in lumbar ipsilateral dorsal horn; inset: binding E_max (n = 7–9). ★ P < 0.05. All data shown as mean ± s.e.m. See fig. S9 for full time course data of panels C–E.

Fig. 4. Prolonged endogenous MOR signaling generates psychological and physical dependence

(A) Behavioral signs of psychological withdrawal (aversion associated with spontaneous pain), reflected by place preference for intrathecal lidocaine upon naloxone administration. Left: Intragroup chamber analysis for intrathecal saline (5 μl) or lidocaine (0.04%) in sham and CFA-21d mice treated with intraperitoneal saline or naloxone (3 mg/kg). Right: Intergroup difference score analysis illustrating time spent in intrathecal lidocaine-paired chambers. (n = 6 per group). (B,C) Behavioral signs of physical withdrawal, recorded for 60 min after injection of NTX (3 mg/kg), NMB (3 mg/kg) or vehicle. (n = 6 – 7). (D) Progression paw edema and (E) effects of repeated subcutaneous vehicle or NTX (3 mg/kg) on hyperalgesia over 105 days after CFA. (n = 7 per group) (F) Effect of repeated NTX (3 mg/kg) on the number of precipitated escape jumps over 77 days after CFA. (n = 8). ★ P < 0.05. All data shown as mean ± s.e.m.