Removing sensory input disrupts spinal locomotor activity in the early postnatal period

Abstract

Motor patterns driving rhythmic movements of our lower limbs during walking are generated by groups of neurons within the spinal cord, called central pattern generators (CPGs). After suffering a spinal cord injury (SCI), many descending fibers from our brain are severed or become nonfunctional, leaving the spinal CPG network without its initiating drive. Recent studies have focused on the importance of maintaining sensory stimulation to the limbs of SCI patients as a way to initiate and control the CPG locomotor network. We began assessing the role of sensory feedback to the locomotor CPG network using a neonatal mouse spinal cord preparation where the hindlimbs are still attached. Removing sensory feedback coming from the hindlimbs by way of a lower lumbar transection or by ventral root denervation revealed a positive correlation in the ability of sensory input deprivation to disrupt ongoing locomotor activity on older versus younger animals. The differences in the motor responses as a function of age could be correlated with the loss of excitatory activity from sensory afferents. Continued studies on this field could eventually provide key information that translates into the design of novel therapeutic strategies to treat patients who have suffered a SCI.

Fig. 1

Spinal Cord lower limb-attached preparation. a Suction recording electrodes were placed to monitor motor activity from ventral root nerve before, during and after perfussion of drugs. b Actual preparation showing simultaneous extracellular recordings c Extracellular recordings showing rhythmic locomotor-like activity from lumbar nerve 2 and the left and right 5th lumbar nerves (scale bar in b panel: 3 mm)

Fig. 2

Effects of sensory deprivation on P0, P1 and P2 mouse spinal cord lower limb-attached preparations. All measurements were performed 15, 30 and 60 minutes after sensory deprivation via sacral transection. a1 Mean data on the effects of the loss of sensory input on the burst amplitude of the ventral L2 and L5 nerve recordings from P0 mouse spinal cord preparations (n = 5). a2 Circular plots showing the effects of sensory feedback to the phasing between the rL2 and rL5 ventral roots (Note: cycles of motor burst activity located at 0.5 are considered in alternation while 0 is considered synchronous activity). b Similar recording as in a panels but from a P1 mouse spinal cord lower limb-attached preparation. c Similar recording as in a panels but from a P2 mouse spinal cord lower limb-attached preparation (n = 5 for each age group)

Fig. 3

Effects of sensory deprivation on P0, P1 and P2 mouse spinal cord preparations. All measurements were performed 15, 30 and 60 minutes after sensory deprivation after cutting all lumbar ventral nerve roots. a1 Mean data on the effects of the loss of sensory input on the burst amplitude of the ventral L2 and L5 nerve recordings from P0 mouse spinal cord preparations (n = 5). a2 Circular plots showing the effects of sensory feedback to the phasing between the rL2 and rL5 ventral roots (Note: cycles of motor burst activity located at 0.5 are considered in alternation while 0 is considered synchronous activity). b Similar recording as in a panels but from a P1 mouse spinal cord lower limb-attached preparation. c Similar recording as in a panels but from a P2 mouse spinal cord lower limb-attached preparation (n = 5 for each age group)

Fig. 4

Comparing ventral nerve activity to dorsal root activity before and after sensory deprivation via ventral nerve root cuts on P2 mice. a Rectified ventral root activity from the right 2^nd lumbar nerve root and raw dorsal root activity from the right 4^th lumbar dorsal nerve root before and, 1 and 15 minutes after cutting all ventral nerve roots (n = 5). b panels: Mean data on the effects of loss of sensory input on burst amplitude of P2 mouse spinal cord lower limb-attached preparations (b1) and the firing activity as recorded from both L4 dorsal roots (b2) (n = 5)

Fig. 5

Attempting to regain alternating locomotor-like activity after sensory deprivation by increasing the concentration of 5-HT, NMDA and dopamine (locomotion-inducing drug cocktail) in spinal cord-leg attached preparations of P2 mice. After the disruption of the recorded locomotor-like rhythm was achieved via sensory deprivation (ventral root denervation) we proceeded to raise the initial drug cocktail concentrations (5-HT: 9 M / NMDA: 6 M / Dopamine: 18 M) by 1 M increments up to 5 M (from 9 M 5-HT / 6 M NMDA / 18 M DA up to 14 M 5-HT / 11 M NMDA / 23 M DA) in 20 minutes time periods and check (via extracellular ventral root recordings) if the alternating motor pattern was recovered with a higher concentration of the drugs (n = 5)

Fig. 6

Diagram of the potential mechanism by which stopping the flow of sensory input disrupts an ongoing pharmacologically-induced locomotor pattern. The included diagram is intended to suggest a potential mechanism by which depriving sensory input to a spinal CPG network which was activated pharmacologically in the presence of limb afferent feedback and no supraspinal inputs. We suggest that the developmental maturity of the sensory network could explain why depriving sensory input in a P0-early P1 animal did not produce a significant perturbation of motor output since apparently this motor pattern was mostly produced by the intrinsic circuitry of the spinal CPG network (arrows). In contrast, in late P1-P2 animals the spinal CPG network could have produced a motor pattern which depended in part on afferent feedback from the limbs (arrows). We suggest that this potential mechanism is enhanced in the absence of supraspinal inputs and the source of this afferent signal which is lost after sensory input deprivation is yet to be elucidated