Carbon monoxide reduces neuropathic pain and spinal microglial activation by inhibiting nitric oxide synthesis in mice

Abstract

Background: Carbon monoxide (CO) synthesized by heme oxygenase 1 (HO-1) exerts antinociceptive effects during inflammation but its role during neuropathic pain remains unknown. Our objective is to investigate the exact contribution of CO derived from HO-1 in the modulation of neuropathic pain and the mechanisms implicated.

Methodology/principal findings: We evaluated the antiallodynic and antihyperalgesic effects of CO following sciatic nerve injury in wild type (WT) or inducible nitric oxide synthase knockout (NOS2-KO) mice using two carbon monoxide-releasing molecules (CORM-2 and CORM-3) and an HO-1 inducer (cobalt protoporphyrin IX, CoPP) daily administered from days 10 to 20 after injury. The effects of CORM-2 and CoPP on the expression of HO-1, heme oxygenase 2 (HO-2), neuronal nitric oxide synthase (NOS1) and NOS2 as well as a microglial marker (CD11b/c) were also assessed at day 20 after surgery in WT and NOS2-KO mice. In WT mice, the main neuropathic pain symptoms induced by nerve injury were significantly reduced in a time-dependent manner by treatment with CO-RMs or CoPP. Both CORM-2 and CoPP treatments increased HO-1 expression in WT mice, but only CoPP stimulated HO-1 in NOS2-KO animals. The increased expression of HO-2 induced by nerve injury in WT, but not in NOS2-KO mice, remains unaltered by CORM-2 or CoPP treatments. In contrast, the over-expression of CD11b/c, NOS1 and NOS2 induced by nerve injury in WT, but not in NOS2-KO mice, were significantly decreased by both CORM-2 and CoPP treatments. These data indicate that CO alleviates neuropathic pain through the reduction of spinal microglial activation and NOS1/NOS2 over-expression.

Conclusions/significance: This study reports that an interaction between the CO and nitric oxide (NO) systems is taking place following sciatic nerve injury and reveals that increasing the exogenous (CO-RMs) or endogenous (CoPP) production of CO may represent a novel strategy for the treatment of neuropathic pain.

Figure 1. Effect of CORM-2, CORM-3 and CoPP on sciatic nerve-injured WT and NOS2-KO mice.

The development of the mechanical allodynia (A and D), thermal hyperalgesia (B and E) and thermal allodynia (C and F) in sciatic nerve-injured (continuous lines) and sham-operated (discontinuous lines) WT or NOS2-KO mice treated during 11 consecutive days with vehicle, CORM-2, CORM-3 or CoPP at 10, 14 and 20 days after surgery is shown. For each genotype, test, day and drug evaluated, *indicates significant differences when compared vs. vehicle sham-operated group (p<0.001, one-way ANOVA followed by the Student Newman Keuls test) and +indicates significant differences when compared vs. vehicle nerve-injured group (p<0.001, one-way ANOVA followed by the Student Newman Keuls test). Results are shown as mean values ± SEM; n = 7 animals per experimental group.

Figure 2. Effect of CORM-2 and CoPP on HO-1 protein expression from sciatic nerve-injured WT and NOS2-KO mice.

The protein expression in the ipsilateral site of the dorsal root ganglia (A) and the lumbar section of the spinal cord (B) from sciatic nerve-injured (CCI) WT and NOS2-KO mice treated with vehicle, CORM-2 or CoPP at 20 days after surgery is represented. The expression of HO-1 in the dorsal root ganglia and spinal cord from sham-operated WT and NOS2-KO mice treated with vehicle has been also represented as controls (sham-vehicle). In both figures and genotypes, *indicates significant differences when compared vs. their respective sham-operated vehicle treated mice (*p<0.05, one-way ANOVA followed by the Student Newman Keuls test), +indicates significant differences when compared vs. their respective sciatic nerve-injured vehicle treated mice (+p<0.05, one-way ANOVA followed by the Student Newman Keuls test), #indicates significant differences when compared vs. their respective sciatic nerve-injured CORM-2 treated mice (#p<0.05, one-way ANOVA followed by the Student Newman Keuls test). Representative examples of western blots for HO-1 protein (32 kDa) in which β-actin (45 kDa) was used as a loading control are also shown. Data are expressed as mean values ± SEM; n = 5 samples per group.

Figure 3. Effect of CORM-2 and CoPP on HO-2 protein expression from sciatic nerve-injured WT and NOS2-KO mice.

The protein expression in the ipsilateral site of the dorsal root ganglia (A) and the lumbar section of the spinal cord (B) from sciatic nerve-injured (CCI) WT and NOS2-KO mice treated with vehicle, CORM-2 or CoPP at 20 days after surgery is represented. The expression of HO-2 in the dorsal root ganglia and spinal cord from sham-operated WT and NOS2-KO mice treated with vehicle has been also represented as controls (sham-vehicle). In both figures and genotypes, *indicates significant differences when compared vs. their respective sham-operated vehicle treated mice (*p<0.05, one-way ANOVA followed by the Student Newman Keuls test). Representative examples of western blots for HO-2 protein (37 kDa) in which β-actin (45 kDa) was used as a loading control are also shown. Data are expressed as mean values ± SEM; n = 5 samples per group.

Figure 4. Effect of CORM-2 and CoPP on CD11b/c protein expression from sciatic nerve-injured WT and NOS2-KO mice.

The protein expression in the ipsilateral site of the lumbar section of the spinal cord from sciatic nerve-injured (CCI) WT and NOS2-KO mice treated with vehicle, CORM-2 or CoPP at 20 days after surgery is represented. The expression of CD11b/c in the spinal cord from sham-operated WT and NOS2-KO mice treated with vehicle has been also represented as controls (sham-vehicle). In this figure *indicates significant differences when compared vs. sham-operated vehicle treated WT mice (*p<0.05, one-way ANOVA followed by the Student Newman Keuls test). Representative examples of western blots for CD11b/c protein (97 kDa) in which β-actin (45 kDa) was used as a loading control are also shown. Data are expressed as mean values ± SEM; n = 5 samples per group.

Figure 5. Effect of CORM-2 and CoPP on NOS1 and NOS2 protein expression from sciatic nerve-injured WT and NOS2-KO mice.

The protein expression in the ipsilateral site of the lumbar section of the spinal cord of NOS1 (A) and NOS2 (B) from sciatic nerve-injured (CCI) WT and NOS2-KO mice treated with vehicle, CORM-2 or CoPP at 20 days after surgery is represented. The expression of NOS1 and NOS2 in the spinal cord from sham-operated WT and NOS2-KO mice treated with vehicle has been also represented as controls (sham-vehicle). In both figures and genotypes, *indicates significant differences when compared vs. their respective sham-operated vehicle treated mice (*p<0.05, one-way ANOVA followed by the Student Newman Keuls test). Representative examples of western blot for NOS1 (155 kDa) and NOS2 (130 kDa) proteins in which β-actin (45 kDa) was used as a loading control are also shown. Data are expressed as mean values ± SEM; n = 5 samples per group.