Paraventriculospinal tract as a model for axon injury: spinal cord.

The response of immunocytochemically identified hypothalamic axons innervating the rat spinal cord was examined at varying times after cord hemisection in a model of axonal injury using the paraventriculospinal projection. The purpose was to determine whether these long descending peptidergic axons would show signs of regrowth after injury. From 1 to 180 days after hemisection, horizontal sections of the spinal cord were stained with peroxidase immunocytochemistry. Antiserum against neurophysin was used to identify axons projecting from the hypothalamic paraventricular nucleus to the spinal cord. The paraventricular nucleus innervates all rostrocaudal segments of the cord, yet the projection is not massive, allowing the trauma response of individual axons to be studied. Immediately caudal to a T4 hemisection, axons began decreasing in number by 2 days after surgery. Ten days postoperatively, only a few axons could be found caudal to the cut; these remaining axons arose from the contralateral cord. A substantial increase in the number of stained axons was found rostral to the hemisection 3-12 weeks after surgery. In that an increase in axon number could be due to both increasing staining efficacy and sprouting, the orientation of axons in control and hemisected rats was studied. Three millimeters rostral to the hemisection, axons had a greater variance in orientation and were more likely to project medially out of the dorsolateral white matter compared with the contralateral control side. Rostral to the hemisection, a statistically significant two- to fourfold increase in the number of branches per axons was found in comparison to the contralateral control side. Axons were found in the dorsal white matter 4 months after surgery; in controls, immunostained axons were not found here. At all intervals after surgery, structures suggestive of growth were found, including terminal growth cones and lateral filopodia and lamellipodia extending from axons whose distal ends had been severed by hemisection. Similar structures were not found in control spinal cord. Together, these data suggest that after cord hemisection, axons from the paraventricular nucleus sprout rostral to the injury.