Functional regeneration of sensory afferents in a model of brachial plexus injury

Regeneration of damaged axons in the adult central nervous system (CNS) is limited in part by the presence of myelin-associated inhibitory proteins and a lack of adequate trophic support. Here we have tested the effect of artemin (ART), a member of the glial-derived neurotrophic factor family, and a soluble Nogo receptor (sNgR), an inhibitor of myelin-associated inhibition, following crush lesions of dorsal roots. Regeneration and sprouting were assessed using neuroanatomical, electrophysiological and behavioral analyses with a focus on determining the topographical specificity of regeneration, an important consideration in identifying therapeutic agents that promote optimal growth and functional recovery.;Cervical dorsal roots (DRs) in adult rats that receive input from the sensory component of the brachial plexus were crushed or cut. Either subcutaneous ART (2mg, Biogen Idec), intraventricular sNgR (1mg, Biogen Idec), or PBS was continuously infused over 14 or 28 days via osmotic pump. The neurotracer cholera toxin b (CTB) was used to identify the projections of medium and large diameter myelinated axonal projections in the spinal cord. Regeneration of specific classes of axons was evaluated using selective injection of CTB and immunohistochemical identification of protein markers. Extracellular recordings were made from the dorsal and ventral horns of the spinal cord in response to stimulation of peripheral nerves in the forelimb. These recordings monitor the synaptic currents elicited by activity of low-threshold, large diameter sensory afferents projecting into the spinal cord. Behavioral tests were performed to determine the functional relevance of axonal growth and synaptic recovery.;Treatment with ART promoted the regeneration of myelinated (CTB-labeled) axons in the dorsal horn of the lesioned side, where myelinated cutaneous and muscle sensory afferents normally project. An absence of label was observed in laminae I and IIo where small unmyelinated axons normally project, in a pattern similar to the unlesioned side. Injection of CTB into specific muscle nerves to visualize muscle sensory afferents and into the layers of the skin to visualize cutaneous sensory afferents revealed similar specificity of axonal projections. Treatment with sNgR, in contrast to ART treatment, promoted regeneration of sensory afferents throughout the dorsal spinal cord in the gray and white matter including in the dorsal laminae. With ART treatment, expansion of small diameter nociceptive axons was identified in laminae I and IIo; administration of sNgR did not produce an effect on these axons. ART therefore promoted axonal regeneration of several classes of sensory afferents into topographically correct regions of the spinal cord, an effect not previously described with any other treatment.;Both sNgR- and ART-treatment promoted the formation of functional synapses between three to six weeks post-lesion. Average peak synaptic responses with both treatments reached 40% of the normal amplitude by six weeks. These responses had the characteristic latency and initial time course of monosynaptic responses evoked by stimulation of unlesioned sensory axons. No synaptic responses were recorded in treated animals sacrificed at one or two weeks post-lesion or at any time point in untreated animals.;Behavioral recovery was also significantly improved after dorsal root crush lesion with sNgR and ART treatment using two behavioral tests that assess general use of the limb and contact-mediated grasping. These observations correlated with the electrophysiological results such that all of the sNgR- and ART-treated rats with behavioral recovery had significant recovery of synaptic function in the denervated region of the spinal cord. In contrast, all 14 of the PBS-treated rats had poor behavioral recovery.;To assess whether sNgR and ART promote sprouting of undamaged axons into denervated segments of the spinal cord, two cervical dorsal roots (C6 and C7) in adult rats that receive input from the sensory component of the brachial plexus were unilaterally cut and resected to prevent regeneration of the root. Extracellular recordings were made from the dorsal and ventral horns of the spinal cord in response to stimulation of peripheral nerves in the forelimb five weeks post-lesion. Dorsal roots C6 and C7 on the unlesioned side were cut acutely and synaptic responses in the denervated region of the spinal cord were measured and compared to the chronically denervated side. In three of four ART-treated rats, the synaptic responses in the denervated region on the chronically injured side were 3.5-fold larger than on the acutely denervated side. In two of three sNgR-treated rats, the synaptic responses on the chronically injured side were about three-fold larger. In untreated rats, the responses on the two sides were approximately equal, indicating that minimal functional sprouting had occurred.;These results demonstrate that administration of either sNgR or ART promotes regeneration and sprouting of sensory afferents. Regenerated afferents with ART treatment project to a restricted region of the spinal cord that is topographically correct, a finding not yet described for regenerated axons in the adult spinal cord using any other treatment. The regenerated axons form functional synapses with targets in the spinal cord that produce behavioral recovery. These data indicate that sNgR and ART might be useful as therapeutic agents in treating human brachial plexus injuries.