Exercise-induced motor improvement after complete spinal cord transection and its relation to expression of brain-derived neurotrophic factor and presynaptic markers

Background: It has been postulated that exercise-induced activation of brain-derived neurotrophic factor (BDNF) may account for improvement of stepping ability in animals after complete spinal cord transection. As we have shown previously, treadmill locomotor exercise leads to up-regulation of BDNF protein and mRNA in the entire neuronal network of intact spinal cord. The questions arise: (i) how the treadmill locomotor training, supplemented with tail stimulation, affects the expression of molecular correlates of synaptic plasticity in spinal rats, and (ii) if a response is related to BDNF protein level and distribution. We investigated the effect of training in rats spinalized at low thoracic segments on the level and distribution of BDNF immunoreactivity (IR) in ventral quadrants of the lumbar segments, in conjunction with markers of presynaptic terminals, synaptophysin and synaptic zinc. Results: Training improved hindlimb stepping in spinal animals evaluated with modified Basso-Beattie-Bresnahan scale. Grades of spinal trained animals ranged between 5 and 11, whereas those of spinal were between 2 and 4. Functional improvement was associated with changes in presynaptic markers and BDNF distribution. Six weeks after transection, synaptophysin IR was reduced by 18% around the large neurons of lamina IX and training elevated its expression by over 30%. The level of synaptic zinc staining in the ventral horn was unaltered, whereas in ventral funiculi it was decreased by 26% postlesion and tended to normalize after the training. Overall BDNF IR levels in the ventral horn, which were higher by 22% postlesion, were unchanged after the training. However, training modified distribution of BDNF in the processes with its predominance in the longer and thicker ones. It also caused selective up-regulation of BDNF in two classes of cells (soma ranging between 100-400 mu m(2) and over 1000 mu m(2)) of the ventrolateral and laterodorsal motor nuclei. Conclusion: Our results show that it is not BDNF deficit that determines lack of functional improvement in spinal animals. They indicate selectivity of up-regulation of BDNF in distinct subpopulations of cells in the motor nuclei which leads to changes of innervation targeting motoneurons, tuned up by locomotor activity as indicated by a region-specific increase of presynaptic markers.