A spinal cord injury physically interrupts the long distance sensory and motor
connections between cortex and the spinal cord as well as local circuits resulting in
permanently dysfunctional or paralytic body parts. The axotomised neurons that
survive the injury cannot regenerate their lost axons in the adult CNS. Therefore,
treating the spinal cord injured is a daunting challenge posed to the neuroscience
community. Our largely incomplete understanding of how the CNS responds to the
injury adds to the difficulty in designing effective treatments.
Over the last two decades researchers have focused on the question: why do
neurons in the adult CNS not regenerate? The experiments driven by this question
have led to the key concept of neurite outgrowth inhibitors, e.g. the protein Nogo-A.
The concept and its significance in the injured spinal cord is summarised in chapter
one. In the same two decades the adult cortex has been found to be surprisingly
plastic. After peripheral injuries, the denervated cortex is largely reactivated by injury
spared body parts. What little is known about spinal cord injury from a cortical
perspective is reviewed in chapter two.
A few years back, through anatomical tract tracing and intracortical
stimulation after thoracic spinal cord injury in rats, it was suggested that the injury
affected hindlimb cortex participates in forelimb movements. Axotomised hindlimb
corticospinal neurons were proposed to influence forelimbs after the injury. In
chapter three we find that indeed axotomised hindlimb corticospinal neurons can
rewire to the forelimb after spinal cord injury. This rewiring was accompanied by
representational changes in the cortex, such that the forelimb area now encroached
into the hindlimb field. The involvement of axotomised corticospinal neurons in the
plastic process was indeed surprising. The subsequent question was: does the spinal
cord injury influence the synapses in the hindlimb area? We answered this question
by taking advantage of transgenic mice that expressed a fluorescent protein in some
cortical neurons to reveal the specialised synaptic protrusions or spines. Our results
from a time course study in which we determined the synaptic spine density of both
axotomised and non-axotomised cortical output neurons in the hindlimb field are
described in chapter four. Our findings point towards a selective disconnection of
cortical output neurons from a dysfunctional hindlimb circuitry while maintaining
connectivity with the intact forelimb.
The fifth chapter reveals the cortical plasticity that occurs after a spinal cord
injury that models the Brown-Séquard syndrome in rats. A unilateral cervical spinal
cord hemisection induced sprouting of corticospinal neurons from the intact side to
the injury inflicted half of the spinal cord in both cervical and lumbar segments. We
mapped these neurons in the cortex to find that forelimb-hindlimb somatotopy
emerges over time in the newly formed corticospinal representation. The sensory
input from the injured hindlimb was able to reach this representation in the ipsilateral
cortex. In the sixth chapter we describe a simple modification of BOLD-fMRI
analysis that enabled the monitoring of representational changes in the adult cortex
after spinal cord injury.
In summary, the adult cortex undergoes significant remodelling after a spinal
cord injury. This remodelling incorporates neurons both directly and indirectly
affected by the injury. The cellular changes of corticospinal neurons appear to be
related to the larger scheme of representational changes involving the injury spared
and sensory deprived cortices. Yet the cortical perspective of spinal cord injury
remains incomplete. Now it is to be determined, if these changes are beneficial or
need to be reversed to optimally treat the spinal cord injured patient.