Capsaicin-evoked release of immunoreactive calcitonin gene-related peptidefrom the spinal cord is mediated by nitric oxide but not by cyclic GMP

Recent data support a role for nitric oxide (NO) in pain processing at the level of the spinal cord, possibly via regulation of neuropeptide release. The goal of this study was to determine whether capsaicin, which selectivel y activates primary afferent neurons and evokes neuropeptide release, acts in an NO-dependent manner. Our results indicate that capsaicin (1 mu M)-evo ked release of immunoreactive calcitonin gene-related peptide (iCGRP) is si gnificantly reduced in the presence of the NO synthase inhibitor, L-NAME (1 0-400 nM; F-3,F-45 = 68.38; P<0.001) and, the selective nNOS inhibitor, 3-b romo-7-nitroindazole (170-680 nM; F-5,F-48 = 56.2; P < 0.01). D-NAME (200 n M) had no effect on capsaicin-evoked iCGRP release. Hemoglobin (an extracel lular scavenger of NO; 3 mg/ml) significantly reduced the effect of capsaic in on the release of iCGRP (F-1,F-8 = 9.12; P < 0.05). The NOS substrace, L -arginine, effectively reversed the inhibitory effect of 3-bromo-7-nitroind azole on capsaicin-evoked iCGRP release. To determine whether the NO-mediat ed release was NMDA-driven, we superfused spinal cord slices with competiti ve and nan-competitive NMDA antagonists in the presence and absence of caps aicin. MK-801 (0.1-10 mu M; F-4,F-33 = 8.49; P < 0.0001) and AP-5 (0.01-10 mu M; F-4,F-38 = 3.34; P < 0.05) reduced capsaicin-evoked iCGRP release. CN QX, an AMPA/kainate antagonist (10 nM-10 mu M), significantly decreased cap saicin-evoked release of iCGRP (F-6,F-42 = 8.76; P < 0.01) in a dose-depend ent fashion. Additionally, our results demonstrate that while capsaicin-evo ked release is significantly reduced in the presence of LY-83583 (10 mu M; F-2,F-18 = 3.46; P < 0.01; a cyclic GMP lowering agent), there is no effect of ODQ (a potent and selective inhibitor of guanylate cyclase). Moreover, the application of a cell permeable analog of cyclic GMP (8-bromo-cGMP; 0.0 1-1000 mu M) is without effect on both basal and evoked iCGRP release. Fina lly, we observed no colocalization of immunoreactive neuronal NOS (nNOS) wi th CGRP in the, dorsal horn. in summary, these data indicate that. capsaici n evokes the release of iCGRP, in part, via the production of NO which ente rs the extracellular space prior to having an effect. Moreover, iCGRP and n NOS are produced in distinct populations of neurons within the dorsal horn. We conclude that capsaicin-evoked release involves the activation of the N MDA receptor but is also modified by the activation of AMPA or kainate rece ptors. Finally, these data suggest that while capsaicin-evoked iCGRP releas e is modified by NO, this release does not require the activation of guanyl ate cyclase and subsequent production of cyclic GMP. (C) 2000 Elsevier Scie nce B.V. All rights reserved.