R-flurbiprofen reduces neuropathic pain in rodents by restoring endogenous cannabinoids

Abstract

Background: R-flurbiprofen, one of the enantiomers of flurbiprofen racemate, is inactive with respect to cyclooxygenase inhibition, but shows analgesic properties without relevant toxicity. Its mode of action is still unclear.

Methodology/principal findings: We show that R-flurbiprofen reduces glutamate release in the dorsal horn of the spinal cord evoked by sciatic nerve injury and thereby alleviates pain in sciatic nerve injury models of neuropathic pain in rats and mice. This is mediated by restoring the balance of endocannabinoids (eCB), which is disturbed following peripheral nerve injury in the DRGs, spinal cord and forebrain. The imbalance results from transcriptional adaptations of fatty acid amide hydrolase (FAAH) and NAPE-phospholipase D, i.e. the major enzymes involved in anandamide metabolism and synthesis, respectively. R-flurbiprofen inhibits FAAH activity and normalizes NAPE-PLD expression. As a consequence, R-Flurbiprofen improves endogenous cannabinoid mediated effects, indicated by the reduction of glutamate release, increased activity of the anti-inflammatory transcription factor PPARgamma and attenuation of microglia activation. Antinociceptive effects are lost by combined inhibition of CB1 and CB2 receptors and partially abolished in CB1 receptor deficient mice. R-flurbiprofen does however not cause changes of core body temperature which is a typical indicator of central effects of cannabinoid-1 receptor agonists.

Conclusion: Our results suggest that R-flurbiprofen improves the endogenous mechanisms to regain stability after axonal injury and to fend off chronic neuropathic pain by modulating the endocannabinoid system and thus constitutes an attractive, novel therapeutic agent in the treatment of chronic, intractable pain.

Figure 1. Time course of the nociceptive behavior in the Spared Nerve Injury (SNI, left panel) and the Chronic Constriction Injury (CCI, right panel) model of neuropathic pain.

Rats were treated with R-flurbiprofen, S-flurbiprofen, gabapentin or vehicle (n = 6 per group). The daily treatment (twice daily i.p.) started 10 days after nerve injury. a Mechanical hyperalgesia was assessed by recording the paw withdrawal threshold to stimulation with a Dynamic Plantar Aesthesiometer, b mechanical allodynia as paw withdrawal threshold to von Frey hairs, c, d cold allodynia as response time in the acetone test and e, f cold hyperalgesia by counting withdrawal reactions during exposure to a cold plate at 2°C. Comparison of the areas under the effect x time curves revealed statistically significant differences between R-flurbiprofen and gabapentin treated animals compared with vehicle or S-flurbiprofen. For R-flurbiprofen 4.5 or 9 mg/kg twice daily provided statistically significant antinociceptive effects. For SNI, respective P values were 0.002 for mechanical, 0.026 for cold allodynia, 0.018 for cold hyperalgesia. For CCI, P values were 0.001 mechanical, 0.041 cold allodynia, n.s. for cold hyperalgesia.

Figure 2. Time course of the nociceptive behavior in nerve-injured and naïve rats treated with R-flurbiprofen or vehicle (n = 6 per group) and glutamate release in the dorsal horn of the spinal cord.

a In SNI, the daily drug treatment started the day after nerve injury. Mechanical hyperalgesia was assessed by recording the paw withdrawal threshold to stimulation with a Dynamic Plantar Aesthesiometer, b cold allodynia as paw withdrawal latency after exposure to a 10°C cold plate. Comparison of the areas under the effect x time curves revealed statistically significant differences between R-flurbiprofen and vehicle treatment (P<0.05). c In naïve rats mechanical sensitivity was assessed with a Dynamic Plantar Aesthesiometer, and d heat sensitivity as withdrawal latency to noxious heat stimulation in the Hargreaves test. R-flurbiprofen was injected at time zero, it had no effect. e Time course of glutamate release in the dorsal horn of the lumbar spinal cord in rats treated with 9 mg/kg R-flurbiprofen i.p. or vehicle 30 min before application of capsaicin cream (0.1%) onto the exposed sciatic nerve (time “zero”). Glutamate was analyzed in microdialysates from the dorsal horn in a colorimetric enzyme assay. The peak glutamate release after capsaicin application differed significantly between the treatment groups (n = 6; P<0.05).

Figure 3. Expression and activity of fatty acid amide hydrolase (FAAH) and NAPE-PLD in DRGs and spinal cord after nerve injury.

a, b Time course of FAAH mRNA levels in the L5 DRGs and the dorsal horn of the lumbar spinal cord ipsi lateral to a sciatic nerve lesion in three different models of neuropathic pain, i.e. the spared nerve injury, SNI, the chronic constriction injury, CCI and the spinal nerve ligation, SNL analyzed by Affymetrix U34 microarray in triplicate. Pooled samples of three animals each were used. The asterisks indicate statistically significant results with P<0.05. c Results were confirmed in the SNI model in mice by quantitative RT-PCR and d protein analysis in Western Blots. e Concentration dependent in vitro inhibition of recombinant FAAH by R-flurbiprofen and the FAAH specific inhibitor URB597 f QRT-PCR and Western Blot analysis of NAPE-PLD mRNA and protein expression in DRGs 7 days after SNI and treatment with R-flurbiprofen 4.5 mg/kg twice daily or vehicle. For QRT-PCR and Western Blot analysis pooled samples of each 3 mice were used and two independent experiments were performed. Mice were treated with 4.5 mg/kg R-flurbiprofen or vehicle twice daily (representative images; n = 6 per group).

Figure 4. Modulation of endocannabinoid levels by R-flrubiprofen in microglial cells (EOC20) and tissue.

a Endocannabinoids in EOC20 at baseline and after stimulation with LPS (1 mg/ml) with or without treatment with R-flurbiprofen (10 µM); n = 6 experiments per group. b Endocannabinoids in ipsi lateral L4/5 DRGs, c ipsi lateral dorsal horn of the spinal cord d and contra lateral forebrain cortex in the SNI model 7 days after nerve injury and treatment with 4.5 mg/kg R-flurbiprofen or vehicle twice daily in mice (n = 8 per group; P<0.05). AEA, anandamide; OEA, oleylethanolamine; PEA, palmitoylethanolamine; 1AG and 2AG, 1- and 2-arachidonoylglycerol were analyzed by LC-MS/MS. P<0.05 for all tests. e PPARγ activity in nuclear protein extracts of L4/5 DRGs, spinal cord and forebrain cortex in naïve mice and in nerve injured mice 7 days after SNI. Mice were treated with vehicle or 4.5 mg/kg R-flurbiprofen twice daily. Results represent pooled extracts of 6 mice in each group. f Core body temperature in mice (mean ± s.e.m. shadowed area) treated with the CB1 agonist, WIN55,212-2, the FAAH inhibitor URB597, R-flurbiprofen and respective vehicles (n = 6 per group). Vehicles 10% or 50% DMSO did not produce changes of the core temperature as compared to phosphate buffered saline and were summarized. The decrease of core temperature in vehicle treated mice is caused by resting during the day. Hypothermia in mice is defined as core temperature <35°C.

Figure 5. Time course of the nociceptive behavior in the SNI model of neuropathic pain in mice deficient of the CB1 receptor in DRG neurons (SNS-CB1^−/−) compared with control mice carrying the floxed CB1 alleles (CB1^fl/fl).

Mice were treated with R-flurbiprofen (4.5 mg/kg twice daily) or vehicle (n = 8 per group). The daily drug treatment started the day after nerve injury. a Mechanical hyperalgesia was assessed by recording the paw withdrawal latency to stimulation with a Dynamic Plantar Aesthesiometer, b cold hyperalgesia as counts of paw withdrawal reactions after exposure to a 4°C cold plate. Comparison of the time courses by ANOVA revealed statistically significant differences between R-flurbiprofen and vehicle treatment in CB1^fl/f mice in both tests. In SNS-CB1^−/− mice R-flurbiprofen significantly reduced mechanical hyperalgesia but not cold hyperalgesia (n = 8 per group, P<0.05). The grey hatched areas indicate the difference in R-flurbiprofen and vehicle treated SNS-CB1^−/− mice. c Anandamide (AEA) levels in L4/5 DRGs and dorsal horn of the spinal cord in SNS-CB1^−/− and CB1^fl/fl mice treated with R-flurbiprofen or vehicle (n = 8, * indicates P<0.05). 5d, e, f Effects of CB1 and CB2 cannabinoid receptor antagonists, AM251 and AM630, respectively on R-flurbiprofen mediated antinociception after nerve injury in the SNI model. R-flurbiprofen or vehicle was continuously administered in the drinking water starting after SNI surgery up to the end of the observation period. Fifteen days after SNI AM251 (3 mg/kg) and AM630 (3 mg/kg) were injected i.p. in R-flurbiprofen treated mice (black arrow) and heat (d), mechanical (e) and cold (f) pain analyzed 1 and 3 hours after AM251 and AM630 injection (grey area). To assess effects of AM251 and AM630 alone the drugs were injected (each 3 mg/kg) in vehicle treated mice at day 17 (white arrow) and nociception analyzed 1 and 3 hours after drug injection.

Figure 6. Immunofluorescence analysis of microglia in the spinal cord dorsal horn seven days after nerve injury.

Microglia was identified by Iba-1 immunoreactivity (green). Mice were treated with R-flurbiprofen (4.5 mg/kg twice daily, left panel) or vehicle (right panel). Figures show the side ipsi lateral to the nerve lesion (a, c, e, f, g, h, i, j, k, l) and the contra lateral side (b, d). The inserts in i and j are z-projections. In I, J neurons were visualized by NeuN immunofluorescence (red). In k, l astrocytes were stained for glial fibrillary acidic protein (GFAP, red). Scale bars: 100 µm in a–f; 50 µm in g–l; 20 µm in the inserts of I, j.

Figure 7. Quantitative analysis of microglia marker proteins in the dorsal horn of the spinal cord.

a Western Blot analysis of Iba-1 immunoreactivity in the ipsi lateral L4/5 mouse spinal cord dorsal horn 7 days after SNI and treatment with R-flurbiprofen (4.5 mg/kg twice daily) or vehicle. Representative blot of n = 4 per group. B. The densitometric ratio of Iba-1 versus loading control β-actin was used for statistics; P<0.05. b Quantitative EIA analysis of phopho-p38 MAP kinase in the L4/5 mouse spinal cord dorsal horn ipsilateral to 7 days SNI or CCI and treatment with R-or S-flurbiprofen (4.5 mg/kg twice daily), dexamethasone (positive control) or vehicle. c Respective representative Western Blot of phospho-p38 and p38 alpha. d Immunofluorescence and e Western Blot analysis of microglia CD68 immunoreactivity in the spinal cord dorsal horn 7 days after nerve injury and treatment with R-flurbiprofen (4.5 mg/kg twice daily) or vehicle ipsi lateral to the nerve lesion (n = 3 per group). f, g CD11b immunofluorescence of dorsal horn microglia and densitometric CD11b image analysis in the ipsilateral dorsal horn 21d after SNI. Scale bar: 100 µm.