Cellular and axonal plasticity in the lesioned spinal cord of adult zebrafish
MetadataShow full item record
Zebrafish, in contrast to mammals, are capable of functional regeneration after complete transection of the spinal cord. In this system I asked: (1) Which spinal cell types regenerate in the lesioned spinal cord? (2) To what extent do the dopaminergic and 5-HT systems regenerate and (3) do dopaminergic axons from the brain influence cellular regeneration in the spinal cord? (1) Lost motor neurons are replaced by newly born motor neurons that mature and are integrated into the spinal circuitry after a spinal lesion in adult zebrafish. Using immunohistochemical and transgenic markers in combination with BrdU labeling, we showed that also 5-HT, parvalbuminergic, Pax2+ and Vsx1+ cells are newly born after lesion. Thus, my work shows that diverse cell types are newly generated in the lesioned spinal cord of adult zebrafish. (2) After spinal cord lesion, zebrafish completely recover locomotion within six weeks. Previous work suggested that axonal regeneration is crucial for functional recovery. Here I analyzed changes in the density of 5-HT and dopaminergic axon terminals in the lesioned spinal cord during recovery. Rostral to the lesion site, I observed die-back and sprouting of dopaminergic axons within two weeks post-lesion. Caudal of the lesion, axons are lost indicating Wallerian degeneration. At six weeks post-lesion I tested functional recovery with a behavioral swim test. In recovered fish, a third of the axonal density was restored just caudal of the lesion site, but not at far caudal levels. In contrast, in fish that had non-recovered, only few axons had bridged the lesion site. Thus dopaminergic axon regrowth correlates with functional recovery. Re-transection of the spinal cord in recovered animals abolished re-gained swimming capability, suggesting that behavioral recovery critically depends on axons that crossed the spinal lesion site and not on an intraspinal circuit. 5-HT axon terminals are of both intra- and supraspinal origin. The overall time course of changes in axon terminal density during recovery is similar to that of dopaminergic axon terminals and also correlates with functional recovery. Overall, the organization of the spinal dopaminergic and 5-HT systems, consisting of neuronal somata in the spinal cord and descending axons, differs significantly from their unlesioned organization. I observe sprouting rostral to the lesion site and limited innervation of the caudal spinal cord, as axons do not regrow into the far distal spinal cord. (3) We further hypothesized that signals released by descending axons are involved in cellular regeneration around the lesion site. Dopaminergic axons of supraspinal origin sprout rostral, but are almost completely absent caudal to the lesion site at two weeks post-lesion. Moreover, we observe that expression of the dopamine receptor drd4a is only increased rostral to the lesion site in the ventricular zone of progenitor cells, including olig2 expressing motor neuron progenitor cells. Correlated with these rostro-caudal differences, numbers of regenerating motor neurons are almost two-fold higher rostral than caudal of the lesion site. To functionally test whether dopamine is involved in motor neuron regeneration, we ablated tyrosine hydroxylase positive, mostly dopaminergic axons by injecting the toxin 6-hydroxydopamine. This treatment significantly reduced motor neuron numbers only rostral to the lesion site. As a gain-of-function experiment, we injected the dopamine agonist NPA after spinal lesion, which increased motor neuron numbers only rostral to the lesion site at two weeks post-lesion. These results suggest that dopamine released by descending axons, augments the generation of motor neurons in the lesioned spinal cord of adult zebrafish. In summary, during spinal cord regeneration I observe generation of various cell types and plastic changes of descending axonal projections. Dopamine released by descending axons is able to increase motor neuron regeneration, showing for the first time that signals from descending axons influence cellular regeneration in the spinal cord.