Either brain-derived neurotrophic factor or neurotrophin-3 only neurotrophin-producing grafts promote locomotor recovery in untrained spinalized cats

Abstract

Background. Transplants of cellular grafts expressing a combination of 2 neurotrophic factors, brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3) have been shown to promote and enhance locomotor recovery in untrained spinalized cats. Based on the time course of recovery and the absence of axonal growth through the transplants, we hypothesized that recovery was due to neurotrophin-mediated plasticity within the existing locomotor circuitry of the lumbar cord. Since BDNF and NT-3 have different effects on axonal sprouting and synaptic connectivity/strengthening, it becomes important to ascertain the contribution of each individual neurotrophins to recovery. Objective. We studied whether BDNF or NT-3 only producing cellular grafts would be equally effective at restoring locomotion in untrained spinal cats. Methods. Rat fibroblasts secreting one of the 2 neurotrophins were grafted into the T12 spinal transection site of adult cats. Four cats in each group (BDNF alone or NT-3 alone) were evaluated. Locomotor recovery was tested on a treadmill at 3 and 5 weeks post-transection/grafting. Results. Animals in both groups were capable of plantar weight-bearing stepping at speed up to 0.8 m/s as early as 3 weeks and locomotor capabilities were similar at 3 and 5 weeks for both types of graft. Conclusions. Even without locomotor training, either BDNF or NT-3 only producing grafts promote locomotor recovery in complete spinal animals. More clinically applicable delivery methods need to be developed.

Keywords: kinematics; locomotion; neurotrophin; plasticity; spinal cord injury.

Figure 1

(Top) Experimental timeline for the cats in the study. Animals were acclimated and locomotor kinematics on the treadmill at speed from 0.3–0.8 m/s were measured pre-transection. The animals received a transplant of NT-3 or BDNF producing fibroblasts at the time of transection and locomotor performance was measured at 3 and 5 weeks post-transection. At the completion of the experiments the animals were perfused and the spinal tissue retrieved for histological analysis. The locomotor recovery obtained in cats receiving no locomotor training and a transplant of either unmodified fibroblasts or a mix of fibroblasts modified to produce either BDNF or NT-3 served as comparison for the locomotor recovery observed in this study. (Bottom) Experimental apparatus showing the marker locations and kinematics measurements used to evaluate locomotor recovery.

Figure 2

“Stick-diagram” representations of the pre- and post-transection (5 weeks) hindlimb kinematics for one step of a representative cat from both groups (NT-3 and BDNF) at 2 different treadmill speeds: 0.4 and 0.8 m/s. The position of the hindlimb is presented as stick figures throughout the step cycle at 30 ms intervals. Both cats showed similar stepping characteristics with good plantar weight-bearing stepping at all the speed tested but with reduced swing length and height. Solid lines: Stance portion of the step; dashed lines: Swing portion of the step.

Figure 3

Post-transection (3 and 5 weeks) swing length and swing height index for all cats at the 6 treadmill speeds tested (0.3 to 0.8 m/s). A: swing length index for the BDNF cats. B: swing length index for the NT-3 cats. C: swing height index for the BDNF cats. D: swing height index for the NT-3 cats. Data were normalized with pre-transection values for each cat. In both groups, most cats were able to locomote with plantar weight bearing steps at speeds up to 0.8 m/s by 5 weeks post-transection. However, one cat in each group did not recover. Number of steps n = 20 for each situation. Data are mean±SD of the 20 steps for each cat. Recovery in swing length index obtained with body-weight supported locomotor training is indicated by a dashed line in panel B (Belanger et al/De Leon et al), while the recovery obtained with transplant producing both neurotrophins but no locomotor training is indicated by a dotted line (Boyce et al).

Figure 4

Average post-transection swing length (A) and height (B) indexes for both groups (BDNF and NT-3). Averages of 3 cats per group were calculated for 2 different treadmill speeds (0.4 and 0.8 m/s) at 3 and 5 weeks post-transection. Number of steps n = 60 for each situation, and data represent mean±Sd. The 95% CI of the estimated marginal means of the linear mixed model of swing length and height overlap in most conditions except where indicated (i.e. at 0.8 m/s treadmill speed) where recovery of swing length and height is shown to be slower for the NT-3 treated animals.

Figure 5

Swing length (A) and swing height (B) indexes for each group as a function of speed and time post-transection. Swing length index increased slightly in the range of speed studied while swing height index was essentially constant. The only index that was significantly lower than the other was the swing length or height for the NT-3 treated group at 3 weeks post-transection and at the highest speed tested (95% CI of the estimated marginal mean for that point did not overlap with the intervals of the others, linear mixed model analysis of swing length or height indexes).

Figure 6

Kinematics of gait for each individual cat at 5 weeks post-transection for a treadmill speed of 0.4 m/s. A) Da (anterior foot placement), Dp (posterior foot placement) and hip height indices. B) step cycle duration index. C) Percent of step cycle at which swing onset occurs (cycle define as stance onset to stance onset). D) Variability in left hindlimb stance onset with respect to right hindlimb stance onset as percentage of the step cycle. E) maxima, minima and ranges for the hip, knee and angle angles. Data are mean±SD of 20 steps for each cat, except for D where 10 steps were used.