Reactive oxygen species contribute to neuropathic pain by reducing spinal GABA release

Abstract

Although both a loss of spinal inhibitory neurotransmission and the involvement of oxidative stress have been regarded as important mechanisms in the pathogenesis of pain, the relationship between these 2 mechanisms has not been studied. To determine whether reactive oxygen species (ROS) involvement in pain mechanisms is related to the diminished inhibitory transmission in the substantia gelatinosa (SG) of the spinal dorsal horn, behavioral studies and whole-cell recordings were performed in FVB/NJ mice. Neuropathic pain was induced by a tight ligation of the L5 spinal nerve (SNL). Pain behaviors in the affected foot were assessed by behavioral testing for mechanical hyperalgesia. Pain behaviors developed by 3 days and lasted more than 8 weeks. Both systemic and intrathecal administration of an ROS scavenger, phenyl-N-tert-butylnitrone (PBN), temporarily reversed mechanical hyperalgesia up to 2 hours, 1 week after SNL. In nonligated mice, an intrathecal injection of an ROS donor, tert-butyl hydroperoxide (t-BOOH), dose-dependently induced mechanical hyperalgesia for 1.5 hours. In whole-cell voltage clamp recordings of SG neurons, perfusion with t-BOOH significantly decreased the frequency of mIPSCs, and this effect was reversed by PBN. Furthermore, t-BOOH decreased the frequency of GABA(A) receptor-mediated mIPSCs without altering their amplitudes but did not affect glycine receptor-mediated mIPSCs. In SNL mice, mIPSC frequency in SG neurons was significantly reduced as compared with that of normal mice, which was restored by PBN. The antihyperalgesic effect of PBN on mechanical hyperalgesia was attenuated by intrathecal bicuculline, a GABA(A) receptor blocker. Our results indicate that the increased ROS in spinal cord may induce pain by reducing GABA inhibitory influence on SG neurons that are involved in pain transmission.

Figure 1. The time course of mechanical hyperalgesia in SNL and sham mice

Response rates to von Frey filament 3.0, corresponding to 0.1 g, were measured at 0 day for pre-surgical levels and were gradually increased following SNL. The mice exhibited signs of mechanical hyperalgesia at 1 day which peaked by 3 day and were stably maintained, lasting longer than 8 weeks in all operated mice (n = 6). Sham operation (n = 6) did not result in a significant change in response rates to the stimuli throughout the testing period. Data are presented as means ± SEM. Arrow with L5 Lig or Sham, time of L5 SNL or sham surgery; *, the value is significantly (p < 0.05) different from that of the sham control by two-way repeated-measures ANOVA.

Figure 2. The effects of a ROS scavenger on mechanical hyperalgesia in SNL mice

(A) The effect of i.p. PBN. SNL resulted in significantly increased response rates to von Frey filament 3.0 from pre-surgical levels (P) in all operated mice (n = 18). One week after surgery, a single systemic injection of 150 mg/kg PBN (n = 8) significantly alleviated mechanical hyperalgesia up to 1 hr after injection. Vehicle treatment (n = 10) resulted in little change in response rates. (B) The effect of i.t. PBN. SNL resulted in significantly increased response rates from pre-surgical levels (P) in all operated mice (n = 14). One week after surgery, a single intrathecal injection of 100 μg PBN in 5 μl saline (n = 8) significantly alleviated mechanical hyperalgesia up to 1 hr. Vehicle treatment did not affect response rates (n = 6). Data are presented as means ± SEM. Arrow with L5 Lig, time of L5 SNL, arrow with ip, intraperitoneal injection; arrow with it, intrathecal injection; *, the value is significantly (p < 0.05) different from that of the vehicle control by two-way repeated-measures ANOVA followed by Duncan's post hoc tests.

Figure 3. The effects of an ROS donor on paw withdrawal response in non-ligated mice

Response rates to von Frey filament 3.0 were measured prior to and after injection of an ROS donor. A single intrathecal injection of 0.05, 0.10, or 0.25 μg t-BOOH dissolved in 5 μl saline (n = 8, 9, 9, respectively) dose-dependently increased response rates, which lasted up to 1.5 h. Vehicle injection did not affect response rates (n = 9). Data are presented as means ± SEM. Arrow with it, intrathecal injection; *, the value is significantly (p < 0.05) different from that of the vehicle control by two-way repeated-measures ANOVA followed by Duncan's post hoc tests.

Figure 4. The effects of an ROS donor and a ROS scavenger on mEPSCs of SG neurons

This experiment was performed in ten neurons in the substantia gelatinosa (SG). The SG neurons were sequentially perfused with control solution (Control), the same solution containing 2 mM t-BOOH (t-BOOH) or 1 mM PBN (PBN). (A) Representative tracings of mEPSCs were observed before and during t-BOOH followed by PBN infusion. Cumulative probability analysis for the current amplitude (B) and inter-event interval (C) of spontaneous mEPSCs were recorded from the same SG neuron. (D, E) The summary graph of the averaged amplitudes and the frequency of the mEPSCs at various conditions: control baseline, during t-BOOH superfusion (t-BOOH), and during PBN superfusion (PBN). Overall, there were no significant differences between the control condition and the test conditions. Statistical analyses were carried out by one-way analysis of variance (ANOVA).

Figure 5. The effects of an ROS donor and ROS scavenger on mIPSCs of SG neurons

The SG neurons were sequentially perfused with control solution (Control), the same solution containing 2 mM t-BOOH (t-BOOH) or 1 mM PBN (PBN). (A) Representative tracings of mIPSCs observed before and during t-BOOH and then PBN infusion in ten neurons in the substantia gelatinosa (SG). Cumulative probability analysis for the current amplitude (B) of spontaneous mIPSCs and inter-event interval (C, p < 0.001, control vs t-BOOH; p = 0.23, t-BOOH vs PBN) were recorded from the same neuron. (D, E) Averaged data show that t-BOOH caused a significant decrease in the frequency of the mIPSCs while their amplitudes remained unaffected. Perfusion with 1 mM PBN partially recovered the frequency of mIPSCs (*p < 0.05, t-BOOH vs. control and PBN). Statistical analyses were carried out by one-way analysis of variance (ANOVA).

Figure 6. The effect of an ROS scavenger on mIPSCs of SG neurons in SNL mice

(A, B) Representative tracings of mIPSCs before (Control) and during 1 mM PBN perfusion (PBN) in the substantia gelatinosa (SG) neuron in normal (A) and SNL (B) mice. (C, D) The summary graphs of the averaged amplitudes and the frequency of the mEPSCs in normal and SNL mice (6 neurons/group). *, the value is significantly (p < 0.05) different from that of normal control or SNL control by one-way repeated-measures ANOVA followed by the Tukey's post hoc tests.

Figure 7. The effects of an ROS donor and a ROS scavenger on GABA_A- vs. glycine-mediated mIPSCs of SG neurons

0.5 μM strychnine with 1 μM TTX were used to isolate only GABA_A-receptor mediated mIPSCs (A, C, E, n = 9) and 10 μM bicuculline with 1 μM TTX were used to isolate glycine-receptor mediated mIPSCs (B, D, F, n = 7). The SG neurons were sequentially perfused with control solution (Control), the same solution containing 2 mM t-BOOH (t-BOOH), 1 mM PBN (PBN), 10 μM bicuculline (Bicuculline) or 0.5 μM strychnine (Strychnine). (A) Representative tracings of GABA_A-receptor mediated mIPSCs (GABA_A-R mIPSC) during control baseline (Control), t-BOOH superfusion (t-BOOH), PBN superfusion (PBN) and bicuculline superfusion (Bicuculline). (C, E) The summary graph of the averaged amplitudes (C) and frequencies (E) of the GABA_A-R mIPSCs at various conditions. The ROS donor caused a significant reduction in the GABA-R mIPSC frequency compared with control (**p < 0.001), with a near complete recovery by PBN (**p < 0.001, t-BOOH vs PBN). (B) Representative tracings of glycine-receptor mediated mIPSCs (glycine-R mIPSC) during control baseline (Control), t-BOOH superfusion (t-BOOH), PBN superfusion (PBN) and strychnine superfusion (Strychnine). (D, F) The summary graph of the averaged amplitudes (D) and frequencies (F) of the glycine-R mIPSCs at various conditions. The glycine-mediated mIPSC frequency was unchanged by either the ROS donor or the ROS scavenger. *, the value is significantly (p < 0.05) different from that of normal control or SNL control by one-way repeated-measures ANOVA followed by Tukey's post hoc tests.

Figure 8. The effects of a GABA_A antagonist on the PBN induced analgesic effect in SNL mice

SNL resulted in significantly increased response rates to von Frey filament 3.0 from pre-surgical levels (P) in all operated mice (n = 19). One week after surgery, a single injection of 150 mg/kg PBN i.p. given immediately after 5 μl saline i.t. (n = 5) alleviated mechanical hyperalgesia up to 2 hr. A single intrathecal injection of 0.05 or 0.10 μg bicuculline dissolved in 5 μl saline (n = 7, 7 respectively) given immediately before PBN (150 mg/kg, i.p.) dose-dependently antagonized PBN's analgesic effect. Single injection of bicuculline (0.10 μg, i.t., n=6) itself did not affect mechanical hyperalgesia in SNL mice (n=6). Data are presented as means ± SEM. Arrow with L5 Lig, time of L5 SNL, arrow with Inj, time of intrathecal followed by intraperitoneal injection. *, the value is significantly (p < 0.05) different from that of the vehicle control by two-way repeated-measures ANOVA followed by Duncan's post hoc tests.