Involvement of reactive oxygen species in long-term potentiation in the spinal cord dorsal horn

Abstract

Recent studies suggest that reactive oxygen species (ROS) are functional messenger molecules in central sensitization, an underlying mechanism of persistent pain. Because spinal cord long-term potentiation (LTP) is the electrophysiological basis of central sensitization, this study investigates the effects of the increased or decreased spinal ROS levels on spinal cord LTP. Spinal cord LTP is induced by either brief, high-frequency stimulation (HFS) of a dorsal root at C-fiber intensity or superfusion of a ROS donor, tert-butyl hydroperoxide (t-BOOH), onto rat spinal cord slice preparations. Field excitatory postsynaptic potentials (fEPSPs) evoked by dorsal root stimulations with either Abeta- or C-fiber intensity are recorded from the superficial dorsal horn. HFS significantly increases the slope of both Abeta- and C-fiber evoked fEPSPs, thus suggesting LTP development. The induction, not the maintenance, of HFS-induced LTP is blocked by a N-methyl-D-aspartate (NMDA) receptor antagonist, D-2-amino-5-phosphonopentanoic acid (D-AP5). Both the induction and maintenance of LTP of Abeta-fiber-evoked fEPSPs are inhibited by a ROS scavenger, either N-tert-butyl-alpha-phenylnitrone or 4-hydroxy-2,2,6,6-tetramethylpiperidine-N-oxyl. A ROS donor, t-BOOH-induced LTP is inhibited by N-tert-butyl-alpha-phenylnitrone but not by D-AP5. Furthermore, HFS-induced LTP and t-BOOH-induced LTP occlude each other. The data suggest that elevated ROS is a downstream event of NMDA receptor activation and an essential step for potentiation of synaptic excitability in the spinal dorsal horn.

Fig. 1.

A: compound action potentials were recorded with a hook electrode from the isolated dorsal rootlet while stimulating the cut proximal end of the dorsal root by a suction electrode. B and C: examples of extracellular compound action potential recordings evoked at 2 different stimulus intensities (B, 30–60 μA; C, 1 mA). The threshold intensities for Aβ-, Aδ- and C-fiber activation were 40, 60, and 700 μA, respectively. The stimulus duration was 0.5 ms. Calculated conduction velocities were 17.6, 6.1, and 1.5 m/s for Aβ-, Aδ- and C-fibers, respectively. D: field excitatory postsynaptic potential (fEPSP) recording setup and examples of Aβ-fiber-evoked fEPSPs. fEPSPs were recorded from the superficial medial dorsal horn while stimuli were applied to the cut proximal end of the attached dorsal root with a suction electrode. The fEPSPs evoked by Aβ-strength stimulus was completely blocked by bath application of 20 μM 6-cyano-7-nitroquinoxalene-2,3-dione (CNQX), suggesting that fEPSP is mediated by glutamate AMPA/kinate receptors. ▲, stimulus artifact. E: an outline of the dorsal quadrant of a transverse spinal cord slice with recorded fEPSPs at 4 locations: superficial medial (SM), superficial lateral (SL), deep medial (DM), and deep lateral (DL) dorsal horn. The probabilities of recording fEPSPs with a clear 1st peak were signified by symbols (, 6–8 of 10 samples; ○, 1–3 of 10 samples; ×, 0 of 10 samples). Typical examples of raw traces of fEPSP from each location (SM, SL, DM, and DL) are shown. Calibration bars are 3 ms and 0.1 mV. Note that most of fEPSPs (6 of 10) from DM have multiple peaks.

Fig. 2.

Recordings of Aβ- and C-fiber-evoked fEPSPs before and after LTP induction in spinal cord slices. A: each fEPSP slope value is the average of 4 recordings made every 2 min in each specimen and then averaged again from the recordings of 6 different specimens. Slope values were plotted as percent of control against time (n = 6). The intensities of test stimuli to elicit Aβ- and C-fiber-evoked fEPSPs were 30–50 μA (0.5-ms duration) and 1–1.2 mA (0.5-ms duration), respectively. Baseline fEPSPs in response to the test stimuli were recorded for 20 min. The conditioning high-frequency stimuli (HFS), which consisted of 5 1-s trains of 100-Hz pulses (1.2 mA, 0.5 ms) given at 10-s intervals, were delivered at 20 min (↑). After HFS, recording was paused for 10 min for stabilization of preparation. Responses to test stimuli were then recorded for an additional 40 min. The slopes of fEPSPs were significantly increased after HFS, indicating the induction of LTP. B and C: examples of Aβ- and C-fiber-evoked fEPSP recordings before (a) and after (b) HFS. ▲, stimulus artifacts.

Fig. 3.

LTP of Aβ-fiber-evoked fEPSPs and the effect of a N-methyl-d-aspartate (NMDA) receptor antagonist, d-AP5. A: LTP-inducing HFS (indicated by ↑) was delivered twice in spinal cord slices (n = 5). The first stimulation was delivered during the superfusion with 50 μM of d-2-amino-5-phosphonopentanoic acid (d-AP5, indicated by the horizontal bar). The 2nd HFS was delivered 30 min after washing out the d-AP5 (2nd ↑ at 80 min). HFS failed to induce LTP of Aβ-fiber-evoked fEPSPs in the presence of d-AP5, suggesting that NMDA receptor activation is essential for LTP induction by HFS. B: to test NMDA receptor involvement in maintenance phase of LTP, d-AP5 was applied after spinal cord LTP is fully established (n = 6) by HFS (↑). The results show that d-AP5 had no effect on the maintenance of LTP of Aβ-fiber-evoked fEPSPs. The data suggest that NMDA receptor activation is necessary for the induction but not the maintenance of LTP of Aβ-fiber-evoked fEPSPs.

Fig. 4.

The effect of a ROS scavenger [1 mM of N-tert-butyl-α-phenylnitrone (PBN), n = 6, A] and superoxide dismutase (SOD) mimetics [5 mM of 4-hydroxy-2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPOL), n = 3, B] on the induction of LTP of Aβ-fiber-evoked fEPSPs. After 20 min of baseline recordings of fEPSPs, drug was superfused for 40 min as indicated by the horizontal bar (from the 20th to 60th min). HFS was delivered twice, 10 min before the end of drug superfusion and 30 min after washing out drug as indicated (↑). The data show that HFS-induced spinal cord LTP is blocked in the presence of PBN and TEMPOL but restored when those were washed out.

Fig. 5.

The effect of ROS scavengers, PBN and TEMPOL, on the maintenance of LTP of Aβ-fiber-evoked fEPSPs. A: Aβ-fiber-evoked fEPSPs before and after HFS (↑) and during PBN application (n = 6). Fifty minutes after HFS, the spinal cord slice was superfused with 1 mM PBN for 30 min (indicated by a horizontal bar). fEPSPs were recorded for an additional 30 min after washing out the PBN. B: the same experiment as A was repeated with 5 mM TEMPOL for 20 min instead of PBN for 30 min (n = 5). The data suggest that ROS are involved in the maintenance of LTP of Aβ-fiber-evoked fEPSPs.

Fig. 6.

Induction of LTP of Aβ-fiber-evoked fEPSPs by a ROS donor, tert-butyl hydroperoxide (t-BOOH), and its relationship to NMDA receptor activation. A: changes in the averaged relative slope values of Aβ-fiber-evoked fEPSPs with superfusion of t-BOOH (5 mM, 30 min) and then PBN (1 mM, 30 min). The fEPSP slopes were gradually increased during t-BOOH superfusion and further thus reached a significantly potentiated fEPSP, LTP state, by 60–80 min after initiation of t-BOOH (n = 6). PBN (1 mM) superfuion (80–110 min) after washing out t-BOOH (30 min) significantly reduced fEPSPs that were potentiated by t-BOOH. B: the effect of a NMDA receptor antagonist d-AP5 on t-BOOH induced spinal cord LTP. Application of d-AP5 (50 μM, 20 min) did not change the baseline Aβ-fiber-evoked fEPSPs. When t-BOOH (5 mM, 30 min) is also added to the bath, fEPSPs gradually increased and reached to greatly potentiated fEPSPs. These potentiated fEPSPs were sustained another 40 min after removal of both d-AP5 and t-BOOH. The data show that t-BOOH induce spinal cord LTP is independent from NMDA receptor activation.

Fig. 7.

LTP induced by a reactive oxygen species (ROS) donor attenuated HFS-induced LTP and vice versa. A: after 20 min of baseline recordings of fEPSPs, 5 mM of t-BOOH was superfused for 30 min as indicated by the horizontal bar (n = 6). After confirming the development of potentiated fEPSPs, t-BOOH was washed out for 20 min and then HFS (↑) was delivered. Enhanced fEPSPs by t-BOOH were moderately attenuated after HFS (n = 6, P < 0.05). B: the sequence of LTP induction methods were reversed in this experiment (n = 6). Thus HFS (↑) was delivered after 20 min baseline recordings to induce spinal cord LTP. When LTP was established 30 min after HFS, 5 mM of t-BOOH was superfused into the bath (20 min). fEPSPs were greatly attenuated (n = 6, P < 0.001) during t-BOOH application. These data indicate that LTP induction by t-BOOH is occluded by that of HFS and vice versa. The significant reduction of fEPSPs after the 2nd LTP induction stimulus indicates that too high levels of ROS (beyond that necessary to maintain LTP) due to added stimulus, seem to interfere with LTP.