Afferent information modulates spinal network activity in vitro and in preclinical animal models

Primary afferents are responsible for the transmission of peripheral sensory information to the spinal cord. Spinal circuits involved in sensory processing and in motor activity are directly modulated by incoming input conveyed by afferent fibres. Current neurorehabilitation exploits primary afferent information to induce plastic changes within lesioned spinal circuitries. Plasticity and neuromodulation promoted by activity-based interventions are suggested to support both the functional recovery of locomotion and pain relief in subjects with sensorimotor disorders. The present study was aimed at assessing spinal modifications mediated by afferent information. At the beginning of my PhD project, I adopted a simplified in vitro model of isolated spinal cord from the newborn rat. In this preparation, dorsal root (DR) fibres were repetitively activated by delivering trains of electrical stimuli. Responses of dorsal sensory-related and ventral motor-related circuits were assessed by extracellular recordings. I demonstrated that electrostimulation protocols able to activate the spinal CPG for locomotion, induced primary afferent hyperexcitability, as well. Thus, evidence of incoming signals in modulating spinal circuits was provided. Furthermore, a robust sensorimotor interplay was reported to take place within the spinal cord. I further investigated hyperexcitability conditions in a new in vivo model of peripheral neuropathic pain. Adult rats underwent a surgical procedure where the common peroneal nerve was crushed using a calibrated nerve clamp (modified spared nerve injury, mSNI). Thus, primary afferents of the common peroneal nerve were activated through the application of a noxious compression, which presumably elicited ectopic activity constitutively generated in the periphery. One week after surgery, animals were classified into two groups, with (mSNI+) and without (mSNI-) tactile hypersensitivity, based on behavioral tests assessing paw withdrawal threshold. Interestingly, the efficiency of the mSNI in inducing tactile hypersensitivity was halved with respect to the classical SNI model. Moreover, mSNI animals with tactile hypersensitivity (mSNI+) showed an extensive neuroinflammation within the dorsal horn, with activated microglia and astrocytes being significantly increased with respect to mSNI animals without tactile hypersensitivity (mSNI-) and to sham-operated animals. Lastly, RGS4 (regulator of G protein signaling 4) was reported to be enhanced in lumbar dorsal root ganglia (DRGs) and dorsal horn ipsilaterally to the lesion in mSNI+ animals. Thus, a new molecular marker was demonstrated to be involved in tactile hypersensitivity in our preclinical model of mSNI. Lastly, we developed a novel in vitro model of newborn rat, where hindlimbs were functionally connected to a partially dissected spinal cord and passively-driven by a robotic device (Bipedal Induced Kinetic Exercise, BIKE). I aimed at studying whether spinal activity was influenced by afferent signals evoked during passive cycling. I first demonstrated that BIKE could actually evoke an afferent feedback from the periphery. Then, I determined that spinal circuitries were differentially affected by training sessions of different duration. On one side, a short exercise session could not directly activate the locomotor CPG, but was able to transiently facilitate an electrically-induced locomotor-like activity. Moreover, no changes in reflex or spontaneous activity of dorsal and ventral networks were promoted by a short training. On the other side, a long BIKE session caused a loss in facilitation of spinal locomotor networks and a depression in the area of motor reflexes. Furthermore, activity in dorsal circuits was long-term enhanced, with a significant increase in both electrically-evoked and spontaneous antidromic discharges. Thus, the persistence of training-mediated effects was different, with spinal locomotor circuits being only transiently modulated, whereas dorsal activity being strongly and stably enhanced. Motoneurons were also affected by a prolonged training, showing a reduction in membrane resistance and an increase in the frequency of post-synaptic currents (PSCs), with both fast- and slow-decaying synaptic inputs being augmented. Changes in synaptic transmission onto the motoneuron were suggested to be responsible for network effects mediated by passive training. In conclusion, I demonstrated that afferent information might induce changes within the spinal cord, involving both neuronal and glial cells. In particular, spinal networks are affected by incoming peripheral signals, which mediate synaptic, cellular and molecular modifications. Moreover, a strong interplay between dorsal and ventral spinal circuits was also reported. A full comprehension of basic mechanisms underlying sensory-mediated spinal plasticity and bidirectional interactions between functionally different spinal networks might lead to the development of neurorehabilitation strategies which simultaneously promote locomotor recovery and pain relief