Peripherally Selective Cannabinoid 1 Receptor (CB1R) Agonists for the Treatment of Neuropathic Pain

Abstract

Alleviation of neuropathic pain by cannabinoids is limited by their central nervous system (CNS) side effects. Indole and indene compounds were engineered for high hCB1R affinity, peripheral selectivity, metabolic stability, and in vivo efficacy. An epithelial cell line assay identified candidates with <1% blood-brain barrier penetration for testing in a rat neuropathy induced by unilateral sciatic nerve entrapment (SNE). The SNE-induced mechanical allodynia was reversibly suppressed, partially or completely, after intraperitoneal or oral administration of several indenes. At doses that relieve neuropathy symptoms, the indenes completely lacked, while the brain-permeant CB1R agonist HU-210 (1) exhibited strong CNS side effects, in catalepsy, hypothermia, and motor incoordination assays. Pharmacokinetic findings of ∼0.001 cerebrospinal fluid:plasma ratio further supported limited CNS penetration. Pretreatment with selective CB1R or CB2R blockers suggested mainly CB1R contribution to an indene's antiallodynic effects. Therefore, this class of CB1R agonists holds promise as a viable treatment for neuropathic pain.

Figure 1

Reversible suppression of SNE-induced mechanical allodynia by representative indenes. (A) Schematic of sciatic nerve entrapment and relevant peripheral nerve and spinal ganglia. (B) Graph of withdrawal thresholds to mechanical stimulation of hindpaws ipsilateral and contralateral to SNE at 1 h before and 3, 6, 24, and 48 h after 2-5u (0.3 mg/kg, ip) injection. At 3 h postinjection, ipsilateral thresholds are increased to levels of predrug contralateral thresholds and are indistinguishable from thresholds measured prior to neuropathy development (pre-SNE). Also note the drug-induced small, but significant increases in contralateral thresholds (mean ± SEM, n = 8 rats). (C) In the same rats, administration of 2-5j (0.3 mg/kg), but not vehicle alone, results in similar increases in ipsilateral thresholds to pre-SNE values. (D) 2-5a (0.3 mg/kg) produces considerably smaller increases in thresholds than 2-5u or 2-5j. (E) Changes in plasma [2-5u] and calculated pharmacokinetic parameters after injection (0.3 mg/kg, ip) in naïve rats (n = 3) are consistent with the time course of its effects on SNE neuropathy symptoms. Brain and CSF/plasma ratios of PrNMI obtained from samples collected from 3 other rats at ~75 min after 2-5u (0.3 mg/kg, ip) suggest minimal CNS penetration. (F) Oral administration of 2-5u (3 mg/kg) reversibly suppresses SNE neuropathy symptoms (n = 8 rats). *, p < 0.05 vs predrug (−1 h) values (one-way RM ANOVA).

Figure 2

Activity of 1, 2-5u, 2-5t, and 2-5j in the “tetrad” assays. (A–D) Rats were tested in each assay 1 h prior and up to 48 h following intraperitoneal injection of vehicle, 1 (0.01 mg/kg) or 2-5u (0.3 mg/kg). Note the profound CNS side effects of 1 vs 2-5u in the ring, core temperature, and rotarod tests. The small analgesic effect of 2-5u in the tail-flick assay is consistent with peripheral activation of CBRs. *, p < 0.05 vs predrug (−1 h) values (one-way RM ANOVA). (E–H) Dose-dependence of brain-permeant 1 and compounds 2-5t, 2-5u, or 2-5j in the tetrad assays after intraperitoneal administration. Each point represents mean peak effect ± SEM of 1 (n = 6 rats), 2-5t (n = 8 rats), 2-5u (n = 8 rats), and 2-5j (n = 8 rats), each subtracted from its vehicle control. (I–L) Dose-dependence of 2-5t, 2-5u, and 2-5j in the tetrad assays after oral administration. Each point represents mean peak effect ± SEM of 2-5t (n = 8 rats), 2-5u (n = 8 rats), and 2-5j (n = 8 rats), each subtracted from its vehicle control. Note the relative lack of side effects in the catalepsy, motor incoordination, and hypothermia assays. The small effects in the tail-flick assay are consistent with antinociceptive effects due to activation of peripheral CBRs in naïve rats.

Figure 3

CB1Rs mediate antiallodynic effects of 2-5u in the SNE neuropathy. (A) Suppression of SNE-induced mechanical allodynia by 2-5u (0.6 mg/ kg). (B) In the same rats (n = 8), pretreatment with the CB1R inhibitor 2 (rimonabant, 3 mg/kg, ip) completely blocks the response to 2-5u. (C) A peripherally restricted rimonabant analogue, 18A, also blocked the response to 2-5u. (D) By contrast, the selective CB2R inhibitor 3 (3 mg/kg, ip) produced only a small decrease in the response to 2-5u. (E) Summary of effects of selective CB1R and CB2R inhibition on antiallodynic effects of 2-5u. *, p < 0.05 vs treatment with 2-5u alone (one-way ANOVA).

Scheme 1

Scheme 2

Scheme 3