Potentiation of Amitriptyline Anti-Hyperalgesic-Like Action By Astroglial Connexin 43 Inhibition in Neuropathic Rats

Abstract

Antidepressants, prescribed as first line treatment of neuropathic pain, have a limited efficacy and poorly tolerated side effects. Because recent studies pointed out the implication of astroglial connexins (Cx) in both neuropathic pain and antidepressive treatment, we investigated whether their blockade by mefloquine could modulate the action of the tricyclic antidepressant amitriptyline. Using primary cultures, we found that both mefloquine and amitriptyline inhibited Cx43-containing gap junctions, and that the drug combination acted synergically. We then investigated whether mefloquine could enhance amitriptyline efficacy in a preclinical model of neuropathic pain. Sprague-Dawley rats that underwent chronic unilateral constriction injury (CCI) to the sciatic nerve (SN) were treated with either amitriptyline, mefloquine or the combination of both drugs. Whereas acute treatments were ineffective, chronic administration of amitriptyline reduced CCI-SN-induced hyperalgesia-like behavior, and this effect was markedly enhanced by co-administration of mefloquine, which was inactive on its own. No pharmacokinetic interactions between both drugs were observed and CCI-SN-induced neuroinflammatory and glial activation markers remained unaffected by these treatments in dorsal root ganglia and spinal cord. Mechanisms downstream of CCI-SN-induced neuroinflammation and glial activation might therefore be targeted. Connexin inhibition in astroglia could represent a promising approach towards improving neuropathic pain therapy by antidepressants.

Figure 1. Synergic inhibitory effects of mefloquine and amitriptyline on astroglial Cx43 coupling.

Cultured astrocytes were treated during 24 h before imaging. (A) Typical fluorescence photomicrographs of intercellular Lucifer yellow spreading after a 10-min scrape loading under the following conditions: no treatment (control, CTRL, black bar); mefloquine (MEFLO 0.5 μM); amitriptyline (AMIT 10 μM); AMIT 10 μM + MEFLO 0.5 μM. Calibration scale: 20 μm. (B) Fluorescence area of Lucifer yellow spreading after 24 h treatment with AMIT and/or MEFLO at the indicated concentrations (μM). Values are expressed with respect to CTRL fluorescence area. Each bar is the mean ± S.E.M. of 3–6 independent determinations. ***P < 0.001 compared with CTRL group, one-way ANOVA, Newman-Keuls; ^≠≠≠P < 0.001 comparison between AMIT-treated groups, one-way ANOVA, Newman-Keuls test.

Figure 2. Unchanged inhibitory effect of amitriptyline on astroglial Cx43 expression and hemichannels upon concomitant treatment with mefloquine.

(A) Photomicrographs of fixed cells examined at 40× with a confocal laser-scanning microscope one day after EtBr uptake in cultured astrocytes treated with: none (control, CTRL, black bar); LPS (1 μg/mL) alone; LPS + MEFLO 0.5 μM; LPS + AMIT 10 μM; LPS + AMIT 10 μM + MEFLO 0.5 μM. Stacks of 10 consecutive confocal images were taken at 0.49 μm intervals. Calibration scale: 20 μm. (B) Effects of AMIT and/or MEFLO on LPS-induced HeC activity measured by EtBr uptake (fluorescence intensity). Values are expressed with respect to CTRL fluorescence intensity. Each bar is the mean ± S.E.M. of 4–6 independent determinations. ***P < 0.001 compared with treatment with LPS alone, Kruskal-Wallis, Dunn’s test. (C) Western blots of Cx43 and Gapdh in extracts from cultured astrocytes after 24-hour treatment by amitriptyline 10 μM and/or mefloquine 0.5 μM. (D) Quantification of Cx43 immunolabeling under these respective treatment conditions. Each bar is the mean ± S.E.M. of 5 independent determinations. *P < 0.05, as compared with control (no treatment, black bar) (Kruskal-Wallis, Dunn’s test).

Figure 3. Unaltered CCI-SN-induced mechanical hyperalgesia after acute treatment with amitriptyline and/or mefloquine.

Amitriptyline (AMIT, 12 mg/kg i.p.) or its vehicle (saline) was co-administered with mefloquine (MEFLO, 1 mg/kg i.p.) or its vehicle (saline) in rats that had undergone unilateral CCI-SN two weeks before. Mechanical hyperalgesia was assessed at various times after treatment (abscissa) by determining pressure threshold value (in g) to trigger vocalization in the Randall-Selitto test. Each point is the mean ± S.E.M. of independent determinations in n rats: CTRL (saline + saline), n = 6; AMIT, n = 9; MEFLO, n = 4; AMIT + MEFLO, n = 10). C on abscissa: control naïve rats before CCI-SN; 0 on abscissa: two weeks after CCI-SN just prior to treatments. None of the pressure threshold values determined after treatment significantly differed from those at time 0; two-way ANOVA, Bonferroni test.

Figure 4. Potentiation by mefloquine of the anti-hyperalgesic effect of chronic treatment with amitriptyline in CCI-SN rats.

Rats underwent unilateral CCI-SN and two-week-treatments with amitriptyline (12 mg/kg s.c. daily, via osmotic mini-pump) or its vehicle (saline s.c. via osmotic mini-pump), and mefloquine (0.5 mg/kg i.p., twice daily) or its vehicle (saline) started on day 15 post-surgery. Mechanical hyperalgesia was assessed by determining pressure threshold values to trigger hindpaw withdrawal (A,B) and vocalization (C,D) in the Randall-Selitto test performed at various times during treatment (abscissa, in days). (A,C) Time-course changes in pressure threshold values. Each point is the mean ± S.E.M. of independent determinations in n rats (“AMIT + MEFLO”, n = 14; “AMIT”, n = 12; “MEFLO”, n = 10; “CTRL”, n = 8). C on abscissa: control rats before CCI-SN; 0 on abscissa: two weeks after CCI-SN just prior to treatments. (B,D) AUC values (g × day) calculated from respective time course-curves illustrated in A, C. Each bar is the mean ± S.E.M. of n rats. *P < 0.05, **P < 0.01, ***P < 0.001 compared with CTRL group, two-way ANOVA, Bonferroni test; ^≠P < 0.05, ^≠≠P < 0.01, one-way ANOVA, Newman-Keuls test.

Figure 5. Unchanged serum and brain levels of amitriptyline by co-administration of mefloquine in CCI-SN rats after a two-week treatment with amitriptyline.

Amitriptyline (12 mg/kg s.c. daily, via osmotic mini-pump) was co-administered with mefloquine (0.5 mg/kg i.p., twice daily, AMIT + MEFLO) or its vehicle (0.9% NaCl, AMIT) for 14 days in rats which had undergone CCI-SN two weeks before. At the end of treatment, amitriptyline levels were measured in serum (ng/ml, A1) and in brain (μg/g, A2). The ratio of amitriptyline levels in serum over those in brain (A₃) was also calculated for each rat. Each bar is the mean ± S.E.M. of independent determinations in 8 rats. No significant difference was noted between both treatment groups (unpaired t-test).

Figure 6. Unchanged CCI-SN-induced expression of mRNAs encoding glial and neuro-inflammatory markers in dorsal root ganglia and spinal cord in rats treated with amitriptyline and/or mefloquine or their vehicles for two weeks.

Amitriptyline (AMIT, 12 mg/kg s.c. daily, via osmotic mini-pump) or its vehicle (CTRL, 0.9% NaCl) was co-administered with mefloquine (MEFLO, 0.5 mg/kg i.p., twice daily) or its vehicle (0.9% NaCl) for 14 days in rats which had undergone unilateral CCI-SN two weeks before. At the end of treatments, mRNA levels of ATF3 (A) IL-6 (B) IL-1β (C) OX-42 (D) GFAP (E) and Cx43 (F) encoding mRNAs were measured by RT-qPCR in ipsilateral dorsal root ganglia (L4-L6) and dorsal quadrant of the L4-L6 segment of the spinal cord. Determinations were also made in naïve rats (D-1, black bars) for comparison. Each bar is the mean ± S.E.M. of independent determinations in 5–10 rats. *P < 0.05, **P < 0.01, ***P < 0.001 compared with naïve rats (D-1, black bars), Kruskal-Wallis and one-way ANOVA, followed respectively by Dunn and Newman-Keuls tests. None of the treatments significantly changed CCI-SN-induced up regulation of any of those markers in both dorsal root ganglia and lumbar cord (comparison with CTRL CCI-SN rats, empty bars).