The following is a one page synopsis of a recent one-hour seminar.
The impressions are my own and do not reflect accuracy of the facts
contained in the presentation. Corrections and discussion are welcomed.
The considerations of grafts in spinal cord injury repair need
not be restricted to the regeneration of original circuits, but
could include partial function restoration through new circuits
created between host and graft, i.e. with the graft playing a relay
role between host neurons. Besides forming alternate circuits, a
graft might also promote sparing of remaining neurons at the
periphery of a lesion, and/or may provide a physiological milieu
conducive to regeneration of original host circuitry.
To examine these issues, a hemisection of spinal cord was
scooped out of a segment of several rats' spinal cords. Embryonal
(E14) spinal cord tissue (T2-T9, meninges and DRG removed) was then
grafted into the area lesioned. Although visible glial scarring
occurred between graft and host, PHAL tracer injected into graft
neurons showed some processes of graft neurons invading host cord
tissue, usually through areas of least scarring. (Most processes,
however, ended at the graft/host interface.) In addition, some
corticospinal neurons were demonstrated to have grown from host
into graft. CGRP staining also showed DRG process ingrowth into the
graft, ending in synaptic connections. Lastly, there was apparently
lots of host/graft interaction at the interface.
When DRG neurons that had invaded a graft were stimulated at
14 uA, an extracellular discharge in the graft could occassionally
be found. A 6 msec delay, typical of spinal cord transmission
rates, was consistently seen, supporting the implication that a
functional connection was being maintained. (Interestingly, the
graft also has spontaneous intrinsic activity typical of inhibitory
Similar results were obtained when a delay of 7 weeks to
months following the S.C. injury elapsed before the graft was
inserted. The grafts took well, CGRP neuron processes grew in, and
functional connections were made.
If a suspension of dissociated fetal brain 5-HT cells were
injected into the injured host S.C., such cells tended to migrate
to areas where 5-HT cells would be expected to localize in S.C.
If the S.C. is contused, which would cause petechial
hemorrhages releasing proteases to bring about cystic necrosis of
the S.C., dissociated cells injected at one spot through the dura
(to minimize further trauma) were found to weave into damaged
areas. CGRP and robust 5-HT containing processes grew from host
into areas where these graft cells had localized. The H-reflex in
rats, a functional test, showed a decrease in hyper-reflexia as
time went on following the graft, indicating that the inhibitory
functions of the spinal cord which were lost with lesioning were
Similar experiments were carried out in cats. Cyclosporin A
needed to be given, since inbred strains were not available, to
prevent graft rejection. In these studies, graft acceptance,
ingrowth of host fibers (as seen by CGRP), and some functional
improvement were also seen. However, in one rejected graft,
function was not restored, even though neurons surrounding the
rejection appeared to not be affected by the rejection. When rat
fetal tissue was used as the graft in cat lesions, the normal grey-
grey matter integration occurred, but, strangely, some integration
of white-grey tissue also occurred, which was not seen in
homogenous species grafts, since white tissue is usually walled off
from a graft by vigourous glial scarring.
Lastly, MRI was being used to image graft viability in the cat
S.C., with a 2 Tesla magnet for gross imaging. A 7 Tesla magnet was
required for slice imaging of histological quality.