Sympathetic-sensory coupling after L5 spinal nerve lesion in the rat and its relation to changes in dorsal root ganglion blood flow

Transection of the L5 spinal nerve in rats results in allodynia- and hyperalgesia-like behavior to mechanical stimulation which are thought to be mediated by ectopic activity arising in lesioned afferent neurons mainly in the dorsal root ganglion (DRG). It has been suggested that the neuropathic pain behavior is dependent on the sympathetic nervous system. In rats 3-56 days after L5 spinal nerve lesion, we tested responses of axotomized afferent fibers recorded in the dorsal root of the lesioned segment to norepinephrine (NE, 0.5 mu g/kg) injected intravenously and to selective electrical stimulation of the lumbar sympathetic trunk (LST). In some experiments we measured blood flow in the DRG by laser Doppler flowmetry. The majority of lesioned afferent fibers with spontaneous activity responded to neither LST stimulation (82.4%) nor NE (71.4%). In those which did react to LST stimulation, responses occurred only at high stimulation frequencies (likely to be above the physiological range), and they could be mimicked by non-adrenergic vasoconstrictor drugs (angiotensin II, vasopressin). Excitatory responses to LST stimulation were closely correlated with the stimulation-induced phasic vasoconstrictions in the DRG. We therefore hypothesized that the activation of lesioned afferents might be brought about indirectly by an impaired blood supply to the DRG. To test this hypothesis we induced a strong and sustained baseline vasoconstriction in the DRG by blocking endothelial nitric oxide synthesis with N-G-nitro-L-arginine methyl ester (L-NAME) applied systemically. L-NAME enhanced baseline vascular resistance in the DRG about threefold and also increased stimulation-induced vasoconstrictions. After L-NAME, the majority of axotomized neurons with spontaneous activity were activated by LST stimulation (76%) or NE (75%). Again, activations closely followed stimulation-induced phasic vasoconstrictions in the DRG provided that a critical level of vasoconstriction was exceeded. In the present study, inhibitory responses to LST stimulation were generally rare and could be reversed to activation by prolonged stimulation or after L-NAME. These results show that sympathetic-sensory coupling occurs only in a minority of axotomized afferents after L5 spinal nerve injury. Like previous studies, they cast doubt on the notion that the L5 spinal nerve lesion is a good model for sympathetically maintained pain. Since responses of lesioned afferent neurons to LST stimulation and NE could be provoked with high reliability after inducing vasoconstriction in the DRG, and since they mirrored stimulation-induced vasoconstrictions in the DRG, it appears that in this model the association of sympathetic activity with afferent discharge occurs mainly when perfusion of the DRG is impaired. (C) 2000 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved.