The goal of the present experiment was to minimize confounders introduced by artificial antidromic stimulation of dorsal root or introduction of substances to the periphery by interrupting the communication of the stimulated dorsal root with periphery. In the past, electrical stimulation of one dorsal root elicited DRRs in the neighboring roots and spread along (up to 16 spinal segment in both directions from the stimulated site) and across the spinal cord . This process is believed to operate in all-or-none manner once activated . Since a rat's paw is innervated by L4-L6 originating nerves, we assumed that stimulating the central portion of one cut dorsal root would evoke DRRs to the ipsilateral paw through the remaining two dorsal roots. In fact, in the current experiment electrical stimulation of the central stump of the dorsal root elicited a significant increase in DRR activity in the recorded fibers of the neighboring dorsal roots.
Neuropeptides (particularly, substance P and CGRP) found in peripheral terminals of nociceptive fibers contribute to neurogenic inflammation and are released in response to antidromic stimulation [11, 13, 40, 54]. Therefore, we expected that electrically evoked DRRs in the nerves innervating a rat's paw would produce both plasma extravasation and vasodilation. However, bilateral vasodilation but not plasma extravasation was observed in response to central stump stimulation.
As previously mentioned, SP acting on tachykinin receptors increases microvascular permeability and edema formation [10, 13]. CGRP, on the other hand, acting on its receptors produces arteriolar vasodilation [13, 15]. Interestingly, C-fibers contain both SP and CGRP, whereas Aδ-fibers predominantly have CGRP in their peripheral terminals [17, 55, 56]. In addition, antidromic stimulation of the saphenous nerve at C-fiber intensity produces both vasodilation and plasma extravasation, whereas stimulation at Aδ-fiber intensity produces only vasodilation [24, 54]. It has been previously shown that 1-2 pulses to lumbosacral dorsal roots are enough to cause a change in cutaneous microcirculation, and 4-16 pulses at 2 Hz evokes vasodilatation lasting for several minutes . Similar results have been shown with spinal cord stimulation . In our study, electrical stimulation of the intact dorsal root or the peripheral stump of the dorsal root produced both vasodilatation and plasma extravasation in the skin. However, electrical stimulation of the central stump with the same parameters did not elicit plasma extravasation on either side, but did produce vasodilatation bilaterally. This finding suggests that the stimulation parameters selected were sufficient to excite both myelinated and unmyelinated fibers in both the distal and central stumps of the dorsal root. However, stimulation of the central stump of the dorsal root triggers more DRRs in myelinated than unmyelinated fibers in the neighboring roots, and leads mostly to CGRP release, and in turn vasodilation.
The differential release of co-localized neurotransmitters from the same terminal depending on the firing rate is another possible explanation of the obtained results. The stimulation frequency needed to induce plasma extravasation is higher than that to produce vasodilation . Electrical stimulation should to some extent mimic peripherally evoked orthodromic action potentials. It is true that DRRs evoked by stimulating the central stump are much weaker than direct stimulation of the distal stump, due to the nature of multisynaptic connectivity inside the spinal cord. It may help to explain the differences in plasma extravasation resulting from stimulation of the central versus peripheral stump. In addition, co-packaged in the same granule, catecholamines and neuropeptides have been shown to be differentially released from adrenal medulla depending on the firing rate through a regulated activity-dependent dilation of the granule fusion pore and size-exclusion mechanism [58, 59].
In both of the proposed mechanisms, there should be a higher probability of DRR generation in Aδ-fibers compared to C-fibers in response to central stump orthodromic stimulation. First, there may be a differential effect of GABA on GABAA receptors on the central terminals of primary afferents. C-fibers have been shown to have a lower density of GABAA receptors compared to both Aδ-fibers and Aβ-fibers . Second, the threshold for generation of DRRs by PADs may be higher in C-fibers compared to Aδ- fibers.
In addition, the proportion of CGRP-containing afferents is much higher compared to SP-containing afferents in the skin. CGRP is present in both myelinated and unmyelinated nociceptive fibers, whereas SP is only found in small diameter unmyelinated fibers. CGRP is also found in larger number of unmyelinated fibers compared to SP .
Finally, the role of sympathetic nervous system needs to be addressed, since the stimulation of the central stump may increase sympathetic activity. On one hand, sympathetic activity can decrease neuropeptide release from afferent fibers by its action on prejunctional α2-adrenoreceptors , and counteract dorsal reflex-mediated neurogenic inflammation . On the other hand, sympathetic presence is important for the development of DRR-mediated neurogenic inflammation through the actions of neuropeptide Y (NPY) and norepinephrine on NPY Y2 and alpha1 receptors, respectively [63, 64].
In this study, the contribution of the sympathetic nervous system during central stump stimulation was challenged by two experiments: blood perfusion change in stimulation of the central stump of the cut dorsal root (Figure 4) and plasma extravasation in stimulation of intact dorsal root (Figure 3I). Stimulation of the central stump produced a significant bilateral increase in blood perfusion suggesting that DRRs in primary afferents surpass sympathetic vasoconstriction, if present. Stimulation of intact dorsal root on the other hand (action potentials can travel orthodromically and antidromically) produced plasma extravasation in the ipsilateral hindpaw, suggesting that even if sympathetic system is activated by orthodromic input, its subsequent effects are not strong enough to counteract plasma extravasation induced by the antidromic spikes that reached the periphery.