What follows are some of Greg Winget's humble transliterations from the Raven Press book "Spinal Cord Reconstruction". This book was edited by Paul J. Reier, Richard P. Bunge and Carl C. Kao several years ago. I'm not sure whether it is still available, but when I bought it, the price was about $90. It was well worth the price, as it contains some of the best information available about the status of spinal cord research at that time. You will undoubtedly recognize the names of many of the researchers, but I'm going to leave it up to you to click on the subjects that suit your fancy and see who wrote them. If you want to inquire as to the current availability, the latest address I have for Raven Press is 1140 Avenue of the Americas, New York, New York 10036. ISBN 0-89004-538-0. Although the research performed up to the beginning of this decade was professionally conducted, hopefully we have all increased our SCI knowledge base over the past few years since this was written. A couple notes of introduction. As is readily evident, emphasis herein is placed on brevity, thereby risking the obvious possibility of losing content, syntax and increasing the possibility of unintentional misunderstandings. For those of you who can suggest improvements in the techniques, methodology, theory, or lay presentation of this material, your suggestions would be very much appreciated and will likely be presented in future updates.
The hopeful outcome of presenting this work in abbreviated format is to encourage a basic understanding, or review, of some of the more interesting work on this subject, optimistically using it perhaps as a base upon which more substantial progress can be made in the near future.
There are connective tissue, peripheral nerve fiber, and Schwann cells within the normal mammalian spinal cord. However, the truly ectodermal CNS territory of the spinal cord doesn't mix with the mesodermal connective tissue elements.
One reaction to spinal cord transection is the formation of a new glial basal lamina to cover the exposed CNS surface at the stump. The original spinal cord tissue bordering the cut end becomes necrotic, detaches from the spinal cord, and is replaced by cavities that develop quite irregularly. The morphological, and probably biochemical, stability of the CNS tissue is restored despite the remaining separation of the transected spinal cord, both anatomically and electrophysiologically.
Occasional Schwann cells can be found lodging at the basal lamina and it is only through the cytoplasm of these lodged Schwann cells that a few axons bridge the CNS-connective tissue junction. On the CNS side, the axon is ensheathed by either astrocytes or oligodendrocytes, but on the connective tissue side, the same axon is ensheathed by Schwann cells.
In almost every pathological condition in which astrocytes and/or glial basal lamina is affected, CNS invasion of Schwann cell occurs: compression, transection, MS, etc.
Schwann cells have three possible origins:
1) Spinal cord perivascular space
2) Spinal root
3) Mesenchymal elements of CNS
Schwann cells invade the stumps of transected cords and lodge at the stump glial basal lamina thereby providing gates for possible passage of axons across the CNS connective tissue border. This is the only mechanism CNS axons enter alien territory and vice versa, as well as the mechanism used to form the root entry zone during embryogenesis.
Nerve grafting with Schwann cell implantation experiments are shown to have favorable results when cooling is implemented and when performed a week after injury. In some dogs, locomotor recovery of hind limbs was observed. Retransection has shown a stump glial basal lamina pierced by many axons. At the precise point of axonal crossing, the stump glial basal lamina becomes continuous with the neurilemmal basal lamina that covers the Schwann cells just like at the root entry zone in normal animals. The CNS type axons are converted to PNS-type axons in order to transverse the length of the gap.
The size of the gap is significant. A three-inch gap in humans would be too long for axons and Schwann cells, unless perhaps a vertebral body were removed.
Vertebral body resection and direct anastomosis ( the end-to-end connection of CNS neurons) of spinal cord stumps has been done in humans without success.
Renewed interest in grafting embryonic brain tissue into transected spinal cords. No graft rejection has been evidenced and the implanted neocortex not only survived but continued to undergo differentiation within the host spinal cord.
For successful CNS transplantation in the spinal cord there must be total suppression of formation of the stump glial basal lamina to ensure an undisrupted, homotypical, all CNS healing without any kind of connective tissue being interposed between the graft and the host spinal cord.
Shimizu transplanted a few folia of cerebellum into the transected spinal cords of five dogs. Three months later two of them were able to walk spontaneously and one was able to stand.
Spinal cord regeneration, in terms of continued axonal elongation, can be greatly enhanced by a variety of surgical methods which deal only with the problem at the site of the injury, not with the intrinsic regenerative capability of the nerve cell body or terminal synapsing after axons cross the lesion. Therefore use of pharmacological agents and active physical therapy should be considered.
Axons are dependent on nerve cell bodies for growth and maintenance. There are two broad classes of nerve cells:
Intrinsic neuron (confined to CNS)
Extrinsic or peripheral neuron: located in whole or in part outside CNS.
Axon reaction - axonal response - axonal retrograde - cytological & metabolic changes induced in the soma of origin by injury to an efferent process.
Severance of the axon close to the cell body is likely to have very severe consequences for the soma. The more proximal the lesion, the greater the amount of axoplasm - i.e., neuroplasm- amputated and, especially with central neurons, the greater the number of collateral axons separated from the cell body. Both the volume of axoplasm amputated and the number of collaterals removed may be important in determining the severity of the impact of the lesion.
Volume of axoplasm = cell substance
Some collaterals may be fed and sustained by retrograde
axoplasmic transport from terminal stations.
The closer the lesion to the soma, the earlier are the onset and the decline of the nucleolar response. When regeneration is prevented, nucleolar and cytoplasmic RNA content ultimately fall to less than normal values, and the ultimate RNA reduction is greatest in animals with the most proximal lesions, i.e., in those that lose the greatest amount of axoplasm.
The critical factor accounting for the usually more injurious effect of lesions proximate to the cell body is probably due to the greater amount of axoplasm removed.
LaVelle proposed that the increasing capacity of the neuron to survive axotomy with increasing age correlates with maturation of the cell's protein synthesizing machinery as manifested by the acquisition of a cytologically mature nucleolus and Nissl substance.
As a concomitant of deafferentation, both micro-glial cells and astrocytes may occupy vacated plasma membrane after bouton withdrawal. Unlike the boutons withdrawn from extrinsic neurons, which remain cytologically normal, those removed from chromatolyzed central nerve cells show various degenerative changes. These may be explained on a retrograde transneuronal basis.
A detailed reaction of axotomy to each part of the cell: cytoplasm, nucleus, nucleoli, endoplasmic reticulum, Golgi apparatus, mitochondria, neurofilaments, microtubules, lysosomes, and autophagic vacuoles.
Histometabolic observations are given for Axon reactions.
Lysosomal Hydrolyses: increased acid, phosphatase activity.
Acetylcholinesterase and other transmitter-related enzymes: loss of it.
Miscellaneous hydrolytic enzymes: butyrylcholinesterase is depleted.
Oxidative enzymes.
Synthesis of nucleic acids and proteins.
It is probable that generation of collateral sprouts is dependent on nerve cell bodies for synthesis of new axoplasm.
Other biologic phenomena of potential importance to CNS regeneration include:
a) Glial scarring at the injury site
b) Transfer of metabolites from glia to axons
c) Mechanisms of axonal elongation and target recognition.
Note: neuromas composed of intrinsic axons are not found in
mammalian CNS after injury.
Cell death is more common in central axons. They exhibit a rapid onset of cytoplasmic nuclear and nucleolar atrophy after axotomy and may persist in a shrunken state for long periods. However, in peripheral neurons cytoplasmic enlargement is common, nuclear atrophy is of lesser degree when it occurs, and nucleolar size is unchanged or increased.
The breakup of free, clustered ribosomes into single units appears to be a consistent feature of axon reaction in mammalian central neurons, but is sometimes rapidly reversible.
Enzymes associated with neurotransmitter metabolism are depleted early in the course of axonal response in both classes of nerve cells.
It would be worthwhile to attempt modification of regressive response of the cell body of origin by administration of biosynthetic agents.
Although axoplasmic transport occurs in the CNS, CNS axons fail to regenerate after transection.
Other factors influenced by the transection of axons are found in the myelin sheath, supporting glial cells, and in the microvasculature.
The interaction of factors affecting axonal regeneration can be observed at the axonal tip.
The very first event after disruption of an axon (whether by spinal cord transection or contusion) is the instantaneous escape of axoplasm from both the proximal and distal ends of the axon.
The axoplasmic leakage creates an almost immediate gap in the axoplasmic column within the otherwise intact myelin sheath tube. Within a few hours of transection the axonal tips of large
fibers are set back from the injury site leaving smaller fibers at the cut end.
The leakage of axoplasm stops within a few hours of transection as the axon tip is lined by axolemma within one hour and layers of collapsed myelin form a septum in front of the axonal tip.
There is a pictorial diagram showing stages of axonal reaction after spinal cord transection.
The process of "axonal autotomy" begins approximately one day after transection and continues for about a week. It is a process whereby the tips of axons degenerate by a means of
terminal club rupture and retrograde as much as 1 cm from the point of original transection.
The terminal club rupture is significant. Among the axoplasmic contents that buildup and escape, are lysosomes (which contain more than 50 enzymes, all hydrolytic and with acid pH
optima). The escaping lysosomes could be activated and lead to autolysis of the surrounding spinal cord tissue resulting in the destruction of the heretofore smaller intact fibers passing near
the ruptured terminal clubs.
After the one week period the final terminal club is formed at a distance of 1 to 2 mm or more from the site of transection and does not rupture again.
As there are antagonistic forces at work between the force of the axonal transport and the encasing myelin barrier, the only mechanism that could result in axonal regeneration would be the removal of the myelin encasement without rupture of the terminal club. This is precisely what occurs in the CNS of lower vertebrates that exhibit axonal elongation after transection.
The crucial difference in sheath structure is the presence of the neurilemmal basal lamina in the PNS and its absence in the CNS. In the PNS the basal lamina tube covers the myelin and the node, thus providing a continuous channel for the terminal club to pass through. It may be possible that the expanding force of the terminal club could be converted by the restraint of the basal lamina into a forward movement. Axonal regeneration could then begin.
In discussion, Dr. Blakemore suggests that trypsin may be used as a demyelinating agent to clear the path for axonal regeneration.
Axoplasm accounts for over 95% of cell volume in some neurons.
Neurofilaments are composed of three different proteins arranged in protofilamentous subunits. There is a process of continuous replenishment of the major cytoskeletal component as
evidenced by their synthesis, assembly, and transport in slow axonal flow. The disintegration of neurofilaments during pathological conditions is accounted for by the presence of calcium-activated protease in the axoplasm. The breakdown and turnover of neurofilaments in uninjured nerves could be accounted for by the activation of this protease in nerve terminals. It is possible that the recycling of neurofilament degradation products represent an important feedback mechanism regulating neurofilament production. The success or failure of axonal regeneration may be determined by the intrinsic neuronal factors that control the production of axonal cytoskeleton.
Anterograde transport is the intra-axonal movement of materials away from the cell body.
Retrograde transport is the transport of material toward the cell body. Retrograde transport carries endogenous axonal materials back to the cell body for disposal or recycling.
Areas that merit investigation:
1. Do central neurons depend on retrogradely transported neuronotrophic factors so that atrophy or death follows when the supply is cut off by axotomy? Do peripheral neurons grow and survive due to an interim supply of growth factors from Schwann cells or other sources which may replace the factors coming from target tissues? If so are such glial factors lacking in the CNS?
2. Are there any proteins transported in peripheral and developing neurons that are absent in the adult CNS which may be important for regeneration? Hoffman and Lasek noted that there is a protein in SCb in peripheral axons that increases during regeneration and has properties like tubulin. This protein is not found in SCb of central axons. Skene and Willard have found a few proteins carried by fast transport to be correlated with axon growth in toad and rabbit nerves. These proteins are transported in axons that can regenerate, and their transport may increase dramatically during regeneration. Are any of the changes in fast transport which follow axotomy lacking in the CNS?
3. At the site of axotomy do central axons fail to establish the "turnaround" of fast anterograde transport that may have a role in signaling to the cell body or in the reorganization of the cut end for out-growth? Is there a failure to reorganize the tip of the severed central axon so that cytoskeletal elements are not reassembled into a growing sprout? If so, is this failure intrinsic to the CNS axon or is it caused by influences in its local environment?
Calcium and the Mechanism of Axoplasmic Transport
in Nerve Fibers
by Sidney Ochs
Dept. of Physiology and Biophysics, Indianapolis
( p.95, Fig 4) Transport filament model with Ca2 mechanisms.
The transport filament hypothesis asserts that there is a polarity of side arms attached to microtubules. One set of microtubules is oriented for anterograde transport and the other for retrograde transport.
A growth cone is a thickened central core surrounded by a thin veil of cytoplasm. The microtubules and other organelles are concentrated in the cone interior.
One of the morphological hallmarks of a growth cone is the presence of vesicles situated in the cone interior and sometimes clustered within a mounded protrusion.
Growth cones may vary in the concentration of organelles present.
Microtubules provide support for the axon and actin-containing microfilaments provide the structural basis for mobility.
Microtubules do not extend into the motile microfilamentous zones of the growth cone.
One possible reason for failure of the adult spinal cord nerve fibers to elongate satisfactorily after damage is that the terrain or immediate environment is inadequate to provide a suitable adhesive substratum for growth.
It is likely that filopodial protrusion and attachment to a favorable substratum are more important than axonal transport of substances from the perikaryon in providing the primary force for neurite elongation.
Conclusions:
1. Nerve fiber growth rates vary depending on the age of the animal from which the ganglionic explants are taken. The most rapid growth rate occurs from neurons taken from perinatal rates. From birth on, the older the animal, the slower the nerve fibers grow out from the explant up to about a month of postnatal age.
2. The time of onset of outgrowth is increasingly delayed from ganglionic explant obtained from animals of increasing postnatal age.
All regions of the nerve cell possess the capacity for growth cone formation and motility. The growth cone is a distinctive region of the nerve cell marked by motility, enhanced endocytotic activity, increased adhesivity to the underlying surface, and paucity of plasmalemmal intramembranous particles.
Forward progression of growth cones may be accomplished primarily by filopodial extension and adhesivity to a favorable substratum to "pull" the cone ahead.
Neurites may grow profusely from embryonic neurons on one type of substratum or not at all on another.
However, ablation poses a more stringent test. Excision of a 4 mm segment may result in no regeneration whatsoever.
Decline of regenerative capacity in adults may result from the restricted ability of ependyma to reconstruct the morphogenetic sequence occurring during development.
Singer et al, "In the case of regeneration of the newt spinal cord regrowth .... repeats embryonic events. There is the appearance of a primitive ependymal tube which then gives rise to nuclear centers, glia, roots, ganglia, and all that originates from the embryonic neural tube. Adult ependymal cells therefore retain within themselves the capacity to repeat almost exactly
their embryonic history, requiring only the circumstances of injury to evoke the developmental responses.
After six months post injury, the terrain at the lesion is not conducive to axonal growth, however regenerating axons in the amphibian optic nerve grow well through similar terrain without being retarded by glial scars.
Retention of a functional microvascular system following injury may allow for the development of reversible changes in central nervous tissue bordering the lesion.
In both the optic nerve and the cord, astrocytes play the principal role in the removal of degenerating axons, and they are the primary substrate used for growth by the regenerating fibers.
Regenerative axons first appear within the lesion beneath the leptomeningeal sheath in association with the basal lamina of the glia limitans. This extracellular layer is the product of astrocyte processes that cover the spinal cord. Regeneration proceeds vigorously along the surface provided by the basal lamina except where discontinuity caused by trauma allows some of the axons to escape.
There is a conspicuous affinity of axons for the surface provided by the basal lamina especially if it is resting on a thin and well-vascularized layer of connective tissue.
Axons are observed to extend only from portions of the stump near the pia.
Even with signs of relatively complete recovery of motor function 6 months after injury neither white nor gray matter is fully constituted and fiber bundles rather than central nervous
tissue with ependyma generally bridge central portions of a large deficit.
The ependymal layer generally retains its viability and physical integrity to a much greater degree than other cellular constituents of the transitional zone. Proximity to cerebrospinal fluid in the lumen of the central canal favors survival.
New gray matter may form in the transitional zone through mitotic activity at the inner surface of the ependyma bordering the terminal vesicle and immediately outside the ependyma.
White matter in the transitional zone of the stumps is the principal source of both large and small fiber regenerating axons.
Mitotic activity of astrocytic cells has only been observed within the ependymal layer.
All parts of a fiber bridge might be populated by a process of translocation with elongated ependymal and astrocyte cells.
The existence of nerve cells in the regenerate seems to indicate that the conditions necessary for both neurogenesis and synaptogenesis can occur, however slow and sparse, in the course of spinal cord regeneration.
The source of the myelin sheath on long term spinal cord regenerates appears to be from oligodendrocytes.
Well-differentiated gray matter and neuropile doesn't develop in the spinal cord regenerate.
Summary:
The area surrounding the ependema or central cell mass consists primarily of glial elements, a few neurons and neuropile, but there is no simple extension of gray matter into the central portion of long-term regenerate. Synapotogenesis occurs only after invasion and repopulation of the larger fiber bundles by mitotic extraependymal cells, and neuropile develops only in those areas of the axonal zone colonized by neurons. We therefore find an incomplete histogenetic process unable to faithfully recapitulate the sequence of embryonic development.
The mature regenerated spinal cord of the Anolis consists of an ependymal tube, the lumen of which is the central canal of the old spinal cord. However, outside the basal lamina, which encloses the ependymal tube, occasional bundles of axons course through the loose connective tissue of the regenerated meninges. These nerve bundles were often associated with typical Schwann cells and represent the few fibers that escaped fasciculation by the ependymal cells. They may be homologous to the fiber bridges reported to span ablation gaps during spinal and regeneration in salamanders.
During regeneration of the tail of amphibians, the new spinal cord grows out into a mesenchymal blastema, whereas spinal cord regeneration at higher levels occurs without benefit of an organized blastema.
Following ablation of the spinal cord, the outgoing ependymal epithelial tube fasciculates and guides a large number of regenerating central axons. The radial processes of the ependymal cells define longitudinal tunnels that form in advance of the ingrowing central axons. Fiber bridges resulted from the regeneration of fascicles of central axons emanating from the white matter of the original amputation surface of the spinal cord.
It would appear that in terms of functional recovery, ependymal guidance is more favorable than fiber bridges.
The absence of functional regeneration in the spinal cords of higher vertebrates is likely to be related to a failure of the ependymal mechanism.
The highly complex glial-neuronal matrix of the central nervous system and the formation of inappropriate local synaptic connections are barriers that could be obviated by the
fasciculation and guidance of regenerating central axons by the ependyma.
The failure of spinal cord regeneration correlates with the absence of an organized ependymal growth and as one progresses to phylogenetically more advanced forms astrocytes are increasingly favored and ependymal tanycytes generally decline in number.
The Anolis presents an ideal experimental model for studying spinal cord regeneration as its tail will regenerate but not the cord at higher levels even though the two levels have a similar if not identical cellular organization. It could be speculated that failure of spinal cord regeneration at midtrunk levels in Anolis results from the absence of requisite epithelial-mesenchymal interactions.
In spinal cord regeneration at midtrunk levels of goldfish and newts the ablation gap is infiltrated by a mesenchymal connective tissue bed into which the organized ependymal tube grows. The absence of a mesenchymal tissue bed correlates with the absence of spinal cord regeneration at midtrunk levels in both lizard and mammals.
Attempts to reduce connective tissue infiltration into the wound area may have been counterproductive. The use of peripheral nerve sheaths and implants of cultured Schwann cells may encourage central axon regeneration across gaps in the spinal cords of mammals.
Reactive astrocytes also encapsulate exposed areas of CNS tissue by reconstituting a glial limiting membrane (glia limitans). The glia limitans separates parenchyma of the brain, spinal cord, and optic nerve, from meningeal and vascular tissue. It also separates the PNS from CNS except at nerve root entry zones where the basal laminae of Schwann cells and astrocytes are continuous.
Glial responses appear to constitute a spectrum of reparative reactions following injury, but the net result is the formation of a dense scar in the injured mammalian CNS that is considered to be a major physical impediment to successful axonal regeneration.
However, gliosis frequently coincides with the formation of a dense connective tissue scar and may contribute to the inhibitory role of reactive astrocytes.
Studies by Windle and associates demonstrated the regenerative potential of intrinsic CNS neurons in adult cats and dogs by preventing the formation of excessive glial membranes and dense collagenous scar tissue with Piromen a pyrogenic bacterial polysaccharide. The action of the Piromen also converted the dense connective tissue of the lesion to a highly vascularized, loose areolar type. So it lessened the glial response and loosened the density of the connective tissue.
Animal studies suggest that glial membranes could inhibit axonal elongation. Piromen, deoxycorticosterone, and ACTH promoted regeneration in several animals with peripheral nerve stumps inserted into cerebral cortex. Fibers grew out of the implanted peripheral nerve tissue and into the CNS so that it became difficult to distinguish the transition zone.
The beneficial effect of adrenal steroids and ACTH to elongate axons is of interest because there is little effect on glial cells. Dense connective tissue presents a more substantial
physical impediment to neuritic elongation than does gliosis.
There is unanimity with regard to the ability of axons to regenerate within the peripheral nerve portion of the dorsal root but little agreement as to their ability to penetrate or grow within the cord. At the root entry zone central and peripheral nervous tissues interdigitate. (Note: page 168 has chart summary of dorsal root regeneration experiments).
Astrocytes of the cat spinal cord impede the growth of axons having a high regenerative capacity. The lack of penetration of the uninjured glia limitans by growing nerve fibers even in the absence of a scar suggest the existence of a potent inhibitory process which must be overcome for regeneration to succeed in mammals.
The elongation of motor neurons does not appear to be compromised by reactive astrocytes but the entry of regenerating sensory fibers into the spinal cord is impeded. Also the
regeneration of motor axons is more vigorous than that of sensory fibers and the rate of regeneration of the central process of dorsal root ganglion cells is slower than that of their peripheral branches.
Kao suggests that the basal lamina of the astrocyte represents the barrier to regeneration.
The density of the glial scar near the lesion is variable and the interlacing web of astrocytic processes does not seem to constitute an impenetrable barrier to the ingrowth of axons. Limited extracellular space in the injured cord may not in itself be responsible for abortive regeneration.
In the goldfish glia do not represent a formidable obstacle to axonal regeneration although they are similar in appearance to the reactive astrocytes of mammals. Neuritic elongation is not
inhibited even though a neurological scar forms at the site of transection. Glial-ependymal bridges, which form between the stumps of the cord, fasciculate the regenerating neurites and seem to orient their elongation parallel to the axis of the spinal cord.
Experiments demonstrate a) compact astrocytic scars do not inhibit the outgrowth of axons from neurons that are capable of regeneration. b) the formation of axonal compartments, and penetrability of glial scars is not an age or developmentally dependent feature, and c) axons appear capable of actively modifying the architecture of glial cells.
Reactive mammalian astroyctes in striking contrast with their amphibian counterparts, appear to represent a major barrier to regeneration. Reactive gliosis in mammals might entail physiological changes (metabolic activity, differences in function) that adversely affect regenerating fibers.
More study needed such as optic nerve grafts in vivo and enriched astrocytic populations in vitro.
The nervous system undergoes constant renewal of most of its molecular structure, supplied by axoplasmic transport.
Possible agents that could influence the axon to grow include changes in the postsynaptic cell or the substances released by its degeneration products from an adjacent injury, a release from growth inhibition by substances released by adjacent axons, or a combination of these and as yet undefined influences. By blocking axoplasmic transport (and not nerve impulses) in peripheral nerves and CNS tracts with colchicine, evidence was found for the sprouting of intact axons in the terminal fields of the blocked fibers, in the absence of degeneration. However the etiology of induced growth is unclear.
Small diffusable peptides, among other substances, appear to stimulate axonal sprouting, enhance regenerative growth, and produce directional neuritic growth in culture.
It is essential that the growing axon be able to recognize the appropriate postsynaptic cell and form a synapse with it.
In removing an entire vertebral body the two adjacent disks must also be removed and the spinal roots preserved.
It is presently unknown whether delayed anastomosis (peripheral to CNS tissue) of the spinal cord will allow central axons to regenerate through a complete transverse lesion.
A substantial amount of the connective tissue scar probably originates from the spinal cord itself.
The direct application of the omentum to the normal spinal cord resulted in the reported development of vascular connections.
Injury has been shown to result in:
1. Decrease in spinal cord blood flow at site of injury.
2. Reduction of partial pressure of oxygen (P02).
3. Elevation of lactic acid.
4. Possible elevation of norepenephrine at the site that would have the potential to cause
vasoconstriction of blood vessels which could result in progressive ischemia and necrosis.
It was believed that the omentum could deliver the needed blood and absorb injurious biochemical products liberated at the site.
It was learned that the direction of revascularization into the cord depended on omental placement. When the omentum was accurately located over the transection site revascularization
took place equally on both sides of the spinal cord.
The pathobiology of spinal cord trauma has been well documented. In this experiment the omentum was applied immediately after injury. More study is needed to determine whether the omentum will have some effect in lessening the disastrous functional changes that result from spinal cord trauma.
Windle and associates pursued the reduction of scar formation with positive results using Piromen, a bacterially derived polysaccharide. One investigator reported return of hindlimb function in approximately 10% of his immature experimental rats.
Another hypothesis for the lack of regeneration suggested by Feringa and associates was due to an autoimmune reaction to CNS antigens released following injury. Spinal rats receiving
injections of the immuno suppressant cyclophosphamide (cytoxan) exhibited neural regeneration which was substantiated by morphological, electrophsiological, and autoradiographic evidence, but no return of hindlimb function.
In an extensive series of enzyme treatments on acute spinal rats in the Soviet Union some authors reported return of function in over 90% of their experimental animals, but the results could not be corroborated by other investigators.
Recent administration of enzymes trypsin and hyaluronidase, an autoimmune suppressant (cyclophosphamide), tissue glue (isobutyl-2-cyanoacrilate), ACTH, and Piromen have provided
various results but no permanent return of function. Two obvious major anatomical phenomena occurred at site of lesion:
1. The formation of a connective scar tissue barrier.
2. The formation and expansion of cavitations rostral
and caudal to lesion site.
Kao has attempted to eliminate "lysosomal spinal cord autotomy" (cavitation) by delayed grafting of autologous peripheral nerve implants into the lesion site. Other investigators have successfully transplanted embryonic nervous tissue into the spinal cords of adult animals and report full differentiation and anatomical integration.
Hypothermia consists of perfusing the injured area of the spinal cord with cooled saline (3 - 15(inf) C) for sustained periods. Enhanced return of function is reportedly due to reduced metabolic and oxygen requirements of nerve cells, thus preventing their destruction.
Reduction of microcirculation leads to decreased oxygen and cellular ATP levels and how massive platelet aggregation within the microvasculature compromises the blood supply to adjacent neural tissue and causes eventual cell death.
Hyperbaric oxygen therapy (HBO) on acute groups of rats all showed improved return of function. There seems to be little difference in the effectiveness of the various HBO treatments
employed. Therefore the factor determining efficiency may be the immediacy of treatment.
DMSO
It has been shown that Dimethyl Sulfoxide (DMSO) has protected axons and their myelin sheaths, reduced edema, increased blood flow, and accelerated return of function. Spinal rats showed reduced cavitation and scar formation and enhanced return of function.
As free radicals are produced by all aerobic tissue and highly toxic to living cells they are rapidly eliminated by normal enzymatic activity. Following SCI, altered metabolic activity may increase free radical production resulting in cell death and release of lysosomal enzymes. As DMSO has been shown to stabilize lysosomal membranes and scavenge free radicals, it may function in these capacities to reduce neural tissue destruction following injury.
Studies indicate a positive synergistic effect in combined HBO/DMSO treatments. De La Torre's results were not as favorable as he used different procedures. But Gelderd's showed a 60% effectiveness of acute transected spinal dogs.
Summary:
In order for axons to grow through an injured area they must be provided with a compatible biological environment such as that found in viable neuropil. Cavitation formation seems to be a result of the destruction of microcirculation which results in ischemia, hypoxia, and eventual cell death with concomitant release or production of substances that accelerate the demise of surrounding neural tissue in the spinal cord.
Factors to consider for return of function:
1. Spinal cord cavitation
2. Vigorous growth of CNS axons
3. Re-establishment of functional synaptic contacts
4. Knowledge of which ascending and descending fiber tracts are necessary to
establish meaningful volitional return of function.
5. Retention of neuropil adjacent to the lesion.
It is well established that Schwann cells are able to myelinate central axons in a variety of pathological conditions. Sensory and motor axons may each receive myelin segments from both Schwann cells and oligodendrocytes.
However, some questions remain:
1. Will Schwann cell numbers be correctly controlled within central tissue to provide optimal repair?
2. Will the trophic support of Schwann cells be adequate for some, many, or all species of neurons within the CNS?
3. Will the long term presence of Schwann cells (and/or fibroblasts) be advantageous within a tissue that normally contains few extracellular components?
4. Will the interface established between CNS neuroglial cells and Schwann cells provide
a barrier for axonal regrowth?
Schwann cells proliferate vigorously during the late fetal and early postnatal period in mammals.
The control of Schwann cell numbers may be provided by a mitogen present on the surface of the axon. P.M. Wood has shown that several types of CNS axons (e.g. those from retinal ganglion cells) have a similar capacity to stimulate Schwann cell proliferation. As axonal contact stimulates proliferation, Schwann cell numbers are expanded to provide ensheathment appropriate for the size of the axonal field present.
In the process of myelination in the CNS, the oligodendrocyte often extends laterally over several axons rather than becoming "polarized" to align along a single axon as does the Schwann cell. Also the oligodendrocyte has no extracellular connective tissue framework to hold cells in alignment after axonal degeneration, in contrast to the situation in the PNS.
Examples of portions of the PNS where axons are not directly surrounded by Schwann cell processes:
The free nerve endings within the epidermis, as the nerve fibers penetrate the basal lamina of the epidermis.
Portions of the enteric autonomic nerve plexus where nerve cell bodies and processes resemble a type of "neuroglial" cell which provide a type of ensheathment similar to that provided in the CNS by astrocytes. The neuropil of these regions resemble that of the CNS.
When a collagen-coated plastic strip is applied over suspended axons, Schwann cells will elongate along the axons, ensheathe them and within several days, begin the process of myelinating the larger axons. Serum and embryo extract are present in the culture medium. Some product of fibroblast secretion is also involved in fostering full Schwann cell development.
It is suggested by tissue culture observations that unless the histology of the CNS tissue is altered by injury or disease, the Schwann cell does not find there the simultaneous contact with both axon and extracellular matrix needed for an important intermediate step in its normal differentiation.
The Schwann cell is clearly a secretory cell. The axon-related Schwann cells produce extracellular matrix. They generate basal lamina and a relatively small population of thin collagen fibrils, as well as several types of collagen. How the extracellular matrix components guide the cell in progressing along its normal differentiative course is not known.
After implantation of whole segments of peripheral nerve into demyelinated regions of the dorsal spine, Schwann cell activity may have been limited to the region of the implant because it is only in this area that the requisite connective tissue elements are present in adequate amounts to facilitate Schwann cell function.
All PNS neurons appear to have regenerative capacity. CNS neurons that send axons into the PNS (such as cranial and spinal motor neurons) exhibit capacities for regeneration not seen in other CNS neurons. The interaction of the neuron with connective tissue components of mesodermal origin may cause an inductive signal (denied to neurons confined to the neural tube) that imparts an enhanced capability to regenerate.
Retrograde axonal support seems to provide trophic support for neuronal growth. The best known case is the dependence of adrenergic autonomic neurons on the well-characterized protein, nerve growth factor. Axonal growth and neuronal vigor depend on retrograde axonal transport of trophic material from the target as well as circulating factors, all in the PNS.
There is direct evidence that material is exchanged between Schwann cells and axons in invertebrates. It would be important to determine directly which CNS neurons are sustained by co-culture with Schwann cells.
There is direct evidence that Schwann cells release a large number of peptides into culture medium, presumably by secretion.
In general, central and peripheral glia divide during development, are relatively quiescent in the mature nervous system, and may divide after local injury.
Two methods have been shown to be effective in stimulating
growth of Schwann cells in vitro:
1. Medium containing 10% fetal calf serum.
2. A protein factor present in extracts of bovine
brain and pituitary.
A variety of purified growth factors, mitogens, and pituitary hormones did not affect Schwann cell proliferation.
It is hoped that the study of Schwann cells may be useful for analyzing the nature of the neuronal signal that induces synthesis of the peripheral myelin proteins.
Glial growth factor has been defined as a basic protein present in pituitary and brain that is a potent mitogen for Schwann cells and astrocytes. Various possibilities for the role of such a molecule:
1. The pituitary could release it into circulation as a growth factor/hormone required for maintenance or for triggering cell division.
2. It might be important in the brain as a diffusable signal that stimulates cell division.
Schwann cell invasion of the CNS follows destruction of glial cells in the area over which astrocytes are destroyed. There seems to be a balance between the destruction of astrocytes and Schwann cell invasion.
Demyelinated axons have the ability to recruit (by signal) new myelinogenic cells and ensheathing cells.
Schwann cell remyelination is a more rapid process and produces thicker myelin sheaths than oligodendrocyte remyelination.
In summary, the destruction of the glial limiting membrane appears to be the event that allows entry of Schwann cells. However, before using Schwann cells to repair demyelinated lesions in the CNS we need a better understanding of how this astrocytic barrier (glial limiting membrane) functions. We know nothing of the effects of replacing astrocytes with Schwann cells in the gray matter.
The domain of the CNS and PNS can be defined by the nature of the nonneuronal cells. The sheath cells differ. In injury, the location of the central or peripheral nerve cell body is not so important as the type of neuroglia at the site of the injured axon.
After injury, the signal for Schwann cell multiplication may come from myelin debris, regenerating axons, or elsewhere.
It is believed that growth cones adhere to Schwann cells and are guided by them in the regeneration process. Schwann cells are involved in all phases of regeneration.
In the CNS, branches from intact axons can extend at least a short distance through the spinal gray matter and also into transplanted iris tissue and embryonic neural grafts.
Injured dorsal root axons regenerate well as far as the dorsal root entry zone, but usually do not succeed in growing in the central nervous tissue.
Theories that attempt to explain the different effects on axonal regrowth of Schwann cells and astrocytes can be divided into two broad groups:
1. Those that emphasize the inhibitory action of astrocytes.
2. Those that focus on the enhancement provided by Schwann cells.
Perhaps Schwann cells have a positive effect on axonal outgrowth which central neuroglia are unable to emulate.
Activity of neuroglial cells in regeneration:
1. Proliferation
2. Migration
3. Adhesivity
4. Diffusible products
Axons of some intrinsic CNS neurons can grow from the transected spinal cord into PNS grafts.
The molecular basis of interactions between sheath cells and regenerating axons, at the time of this study, remained to be defined.
In both PNS and CNS, remyelination commences 6 to 7 days after focal injections of lysolecithin.
Using the technique of combining Schwann cells grown in vitro with in vivo axons of the PNS and CNS in the immunosuppressed mouse, demonstrated the success of ensheathment and myelination of transplanted rat cells.
Short term immunosuppression may allow enough time for regenerating axons to be guided to the target organ before the grafted sheath cells are rejected. Rejection of the foreign
transplanted cells could eventually be followed by migration of endogenous Schwann cells along axons in the grafted segment.
It has been successfully demonstrated that the peripheral nerve graft in the spinal cord will reinnervate, facilitating axonal growth from the ends of the spinal cord to the nerve graft.
However, even with very careful tailoring of the nerve graft to fit the gap in the spinal cord, there are microscopic spaces between the Bungner's bands of the nerve graft and the terminals of axons within the cord that must be bridged by outgrowth of nonneuronal cells from the graft before spinal cord axons are seen to advance.
The additional use of cultured nonneuronal cells in reconstructive surgery enhances the speed of innervation, allowing axons to reach the graft as early as one week post surgery.
The use of autologous cultured nonneuronal cells produces results consistent with results when homologous cells are used.
Therefore the use of cultured cells of peripheral nervous tissue apparently stimulates the elongation of some CNS axons. This is an extension of the previous nerve graft technique.
Note (p. 327): Schwann cells from oligodendroglia form basal lamina along their plasma membranes which ensheath axons in the CNS.
A proposed model would place an implant at the site of transection and the implant would act as a relay for motor function. It would require:
Axons from the proximal cord to make functional synapses with the neurons in the relay.
Neurons in the implant to be able to extend axons into the distal cord.
Axons from the relay neurons to make useful functional contacts with neurons in the distal cord.
The neonatal SCG neuron, when placed in culture under certain conditions, acquires cholinergic characteristics.
Influences on neurotransmitter function in sympathetic neurons:
Nonneuronal Cells
Conditional Medium
Human Placental Serum
Age
Depolarization
Potassium K+1
Elect Stim
Gluco corticosteroids
Embryo Extract
A case is presented for injections of NGF into the implant area. An explant containing adrenergic autonomic neurons would be used as a bridging implant and fibers from the implanted into the distal gray matter by direct NGF injections. Those nerve fibers entering the distal cord might adapt new transmitter mechanisms in response to the neuronal targets in the distal cord.
This approach would not require long tracts such as the corticospinal tract to traverse the lesion site.
Since then it has been shown that embryonic tissue placed in adult hosts often survived well and tissue survival is consistently demonstrable with little tissue rejection encountered.
The propensity of mammalian ependyma to provide some growth response to injury may suggest that implants of additional ependymal tissue might be considered in the hope of establishing a coherent bridge within the tissues of the gap.
Relatively few fibers may be required to cross a transection site in order to obtain useful functional return.
In summary:
(a) Implants may survive and be identified.
(b) Maturation (e.g. myelination) may occur within the implant after insertion.
(c) Implants become vascularized.
(d) Large neurons and numerous synapses are demonstrable within the implant.
In future experiments it is recommended that grafting should be delayed until after acute spinal cord edema, and that use of immobilization, hypothermia and other techniques be employed. Embryonic spinal cord may provide a more suitable terrain for axonal growth.
The dogs were transected twice, each time a segment of the entire tissue was removed and a photo taken. They began walking at four months and reached maximum recovery at eight months and the autonomous bladder regained voiding ability.
Neural transplants from the embryonic spinal cord or brain stem have a lower growth potential than from the neocortical region. They generally do not show laminar cytoarchitectural
pattern. The neurons seem to form clusters rather than layers.
In conclusion, it is possible to some extent to achieve successful transplantation when portions of the spinal cord parenchyma are exposed, there is room for growth of transplants, and neocortical tissues with a high growth potential are used as transplants.
The brainstem monomine systems apparently exert tonic regulatory influences on central neuronal networks affecting the state of activity. There is a relatively nonspecific character in monoaminergic control functions illustrated by the manner in which dopamine reverses functional defects.
It is therefore believed that the monoaminergic systems in the CNS, rather than conveying specific input and output signals, are acting on neuronal machinery, the activity levels of which are set by the activity at the monoaminergic synapses.
The transplantation model rests on two ideas:
Transplants of brainstem regulatory systems may serve to reactivate neuronal networks.
The monoaminergic and cholinergic neuron systems survive excellently provided they are taken from embryonic donors.
It has been demonstrated that the connections established between the grafted neurons and the host CNS can be highly specific and possibly functional.
The functional properties of local neuronal networks that form "central programs" and "pattern generators" are believed to cause neural control of movement. Many of the basic motor
programs generated by local neuronal networks are virtually independent of afferent feedback. Brainstem command systems can activate these motor pattern generators, tonically drive them,
and maintain them at different levels of activity.
In the spinal cord locomotion is generated by intraspinal neuronal networks (locomotion generators) controlled by supraspinal command driving systems (nonadrenergic and other brainstem command systems).
Recovery of some aspects of locomotion may not require any extensive reconnection with supraspinal systems.
Locomotion can be initiated and maintained by electrical stimulation of the so-called mesencephalic locomotor region (MLR).
**The most important factor for the recovery of locomotion in the isolated spinal cord would be the restoration of a disinhibiting regulatory input that could release the intraspinal locomotion generator and allow its operation.
Transplants of embryonic brainstem may be able to substitute for the loss of activation from the brainstem MLR, thus establishing a new "reticulospinal" command system for the
inhibited locomotion generator of the isolated spinal cord.
In the spinal cat it has been shown that adrenergic activation by systemic injections of L-Dopa or clonidine elicit treadmill walking.
Adrenergic receptor activation seems to be sufficient to imitate and maintain locomotion in the transected spinal cord. A viable experimental approach would be to reinervate the transected spinal cord by transplants of the locus coeruleus, or the entire MLR.
He concludes that in paraplegics we should consider methods for electrically bypassing the lesion site as well as surgically permitting growth across it. Dr. Bjorkland's chapter suggesting
the implantation of embryonic tissue to activate a locomotor generator within the spinal cord, is given as an example. However, he asks the question, without providing a clear answer, of how to activate the implanted tissue from higher centers.
The midbrain locomotor region is considered as one of the inputs to the strip.
Is the locomotor strip simply a tract consisting of axons originating in the hypothalamus, red nucleus, or locus coeruleus?
Excitation of the locomotor region of the brainstem activates and disinhibits stepping generators in the spinal cord.
The state of the isolated spinal cord changes dramatically and becomes similar to what can be expected during stepping after intravenous injection of L-Dopa. The injection of L-Dopa imitates excitation of noradrenergic neurons.
Although noradrenergic and fast-conducting direct reticulospinal pathways take part in controlling locomotion, stepping is possible even when one of them is eliminated.
Is it possible that activity can propagate along the spinal cord - as along the locomotor strip - by means of polysnaptic chain of activation?
Could activity propagate in a rostrocaudal direction polysnaptically in the spinal cord as in the pontomedullary locomotor strip? This columnar pathway would be the third - but now indirect - pathway for activating action of the brainstem locomotor region on generators of stepping.
Some cervical axons travel in funiculi.:
Muscle tone is one condition for evoking locomotion. Therefore destroying funiculi or disrupting spinocerebellar tracts make locomotion difficult.
In the spinal cord there are neurons giving synaptic responses to stimulation of the medullary locomotor strip.
The propriospinal polysnaptic system and direct descending pathways play essential roles in locomotor action of the brainstem on the spinal cord.
Although two or more direct descending pathways take part in control of locomotion, stepping likely persists after each one of them is eliminated. An explanation would be the transition of the spinal cord to the locomotor state being dependent to a substantial degree on propriospinal polysnaptic systems.
In summary, the two direct descending tracts:
Noradrenergic
One containing fast-conducting reticulospinal axons originating in the medial reticular formation.
In all known instances of successful spontaneous walking after low thoracic transection the dogs or cats were less than two weeks old.
Exercise was found to have a more beneficial effect on the young adult kittens than on the two week olds.
Cordotomy was found to have a greater atrophic effect on the whole soleus (SOL) muscles than on the medial gastrocnemius (MG).
The cats exercised on the treadmill demonstrated the greatest change in contraction time (CT) and histochemical profiles.
The quantity and quality of proteins of muscles regulated independently. (nutritional component?)
All of the treadmill exercised animals were able to perform treadmill walking by the end of the experiment. None of the unexercised 12-week (young adult) cats were able to perform treadmill walking. None of the 12-week cats were seen to walk spontaneously.
Treadmill exercise has a beneficial effect in maintaining muscle weight and fiber cross-sectional area in 12-week transected cats, but not in maintaining a normal contraction time of the muscle.
In the spinal cord as elsewhere in the CNS, much of the neural activity is mediated by interneurons, which is one of the three converging systems. The other two are descending and
dorsal root. The descending and dorsal root systems can readily be manipulated surgically.
A unilateral deafferentiation of the spinal cord by lumbosacral rhizotomy could be expected to elicit sprouting by some or all of the remaining systems.
Another source of sprouting is dorsal roots themselves.
Sprouting in the spinal gray matter, by the descending systems can block the sprouting of a spared root. Therefore, potential sources of sprouting may compete.
Sprouting does not appear to be random, but subject to some regulation, as well as being subject to competition between systems. Thus, axonal growth in one system may inhibit formation of persistent sprouts by another system.
Cellular mechanisms that may contribute to recovery are: a) unmasking of silent synapses, b) denervation supersensitivity, c) return of terminals displaced by desynapsis, and d) axonal
sprouting.
Postural information mediated in part, by dorsal roots of the trunk does indeed contribute to recovery of accurate limb placement when hindlimb afferents have been eliminated.
Recovery from hindlimb deafferentiation depends entirely on the ipsilateral descending systems, as demonstrated by the permanent paralysis following deafferentiation, plus ipsilateral
hemisection.
There appears to be some competitive interaction between the spared root and descending systems in mediating recovery and reflex changes after partial deafferentiation. The "hierarchical control" seems to permit the "choice" of the system to mediate recovery that can do it most efficiently.
It would be beneficial to use DOPA or clonidine to stimulate the spinal locomotor generator in monkeys as their reflex capacity is not as good as that of cats or dogs.
Residual brain influence on segmental reflexes in paralyzed muscles is a frequent neurophysiological finding. Paralysis of the limbs does not always correspond to the degree of spinal cord tissue destruction. Absence of some functions can be due to a dysfunction resulting from disequilibrium between noninjured elements. Paralysis is not the only criterion to define the
degree of impaired supraspinal influence.
Although function is still possible with much of the cord damaged, it is essential to have both anterior and posterior neuronal elements present for neural control of gait. However, patients with diffuse lesions of the spinal cord and partial survival of the ascending, descending and propriospinal elements will usually regain motor activity and locomotion but will have spasticity and impaired sensation.
Eidelberg and associates report that a residual fiber count in excess of 5,000 are associated with useful hindlimb locomotion in animals.
Any patient with complete absence of all qualities of sensation was unable to walk after SCI. Only when tonic supraspinal control is present does some residual control for standing and stepping remain.
All patients who demonstrated some volitional control and were able to walk also exhibited partial preservation of function of the long ascending tracts and definitive somatosensory responses. Thus partial preservation of primary sensory neuron functions together with descending tracts seem to be essential for establishing voluntary control of gait.
Spinal cord stimulation can activate sensory, sensory-motor, and motor spinal cord mechanisms, depending on the patterns and site of stimulation.
A permanent system for spinal cord stimulation was implanted in seven SCI patients. There were varying degrees of modest improvement in five of them. Potential candidates must
demonstrate signs of partially preserved descending motor control of voluntary or even reflex motor functions as well as a partially functioning ascending system.
Despite both legs being paralyzed a conscious attempt to move one leg may elicit EMG activity in the opposite leg.
Restoration of function in patients with complete
transection.
.
The initial phase of treatment consists of massage and passive motion.
Training of voluntarily contractible muscles in trunk.
Sitting straight, turn trunk while sitting, and have free movement of arms.
Lifting the entire leg while shifting the pelvis toward the shoulder girdle.
Note: All seven (7) of their patients were able to walk with assistance of leg braces and canes.
An illustration (p.478, fig.2) described on preceding page explains how the patient lies on one side and uses the shoulder girdle and long trunk muscles to lift and extend the leg as if to
thrust the pelvis toward the shoulder girdle.
Somehow, the skin displacement must act as a stimulus for the contraction of the lower portions of the muscles. The gluteus medius muscle eventually begins to contract and pulls the
thigh upward.
In summary, this method of physiotherapy aims to activate spinal reflexes, which eventually cause useful contraction of the muscles innervated by the isolated spinal cord.
With intensive treatment and training, spasticity is usually reduced considerably.
Degeneration is a frustrated attempt at regeneration. He suggests more research in manipulating the axonal swelling. (Axonal swelling is what causes degeneration).
Dr. Bjorklund:
Spinal locomotion may be activated without axonal reconnection to the brain. As Dr. Goldberger emphasized, it is a matter of having the right type of axon regenerate across the transection.
Dr. Grillner:
Since only a certain number of fibers need regenerate across the gap from different descending systems, these fibers must make synaptic connections with the isolated spinal cord. What is needed is to find a very simple model with relatively few neurons with readily identifiable connections.
Dr. Bunge:
Cellular implants must continually provide trophic substances for the growth and guidance of the proliferating nerve fibers. As yet, we do not know if the Schwann cell offers trophic support for central axons. We may want to attempt the use of ependymal cells for guidance and trophic support. He looks to living (implants) prosthesis to help solve the regeneration problems.
Dr. Gelderd:
Surgically resection vertebral bodies, retransection of the cord, and juxtaposition of the cut ends of the spinal cord through the use of living prosthesis may be best, as described by Dr. Bunge. However, one important aspect of this procedure is to include hyperbaric oxygenation therapy at the point of retransection.
Dr. Schimizu:
Suggests that a three-pronged approach be taken:
1. Surgery
2. Medicine
3. Physiotherapy
It will take a joint effort similar to a research group organized in 1974 at the University of Tokyo, under the ministry of Science and Technology.
Dr. Dimitirjevic:
Physiotherapy
Dr. Derlon:
Interaction of clinicians with basic researchers is essential. The major problem is prevention of spinal cord self-destruction at time of retransection.
Dr. Kao:
Four areas of importance have emerged:
Drugs or tropic substances must be found that will enhance axonal elongation and synaptogenesis.
Surgical instrumentation that will remove one or two vertebral bodies to approximate the ends of the cord without causing tension in the cord.
A technique that produces perfect wound healing in the cord may be attained through the combined use of modern microsurgical techniques, neural grafts, physical agents and/or drugs.
Physiotherapeutic method of retraining the paralyzed locomotor apparatus to maintain the neuromuscular junctions and later activate the isolated spinal cord into function.
Kao emphasizes that the patient choose the best surgeon as 40 years were lost because Sugars surgical results were not duplicated until Richardson and Aguayo in 1980.
Perhaps 5 million axons will need to cross the gap little by little in order to regain function.
Dr. Tator:
Grafting with the use of embryonic brain grafts containing "autonomic" nerve cells is an important future direction.
Other areas deserving more study are promoting factors such as thyroid hormones, as well as axoplasmic transport studies, neuronal reaction to injury, collateral sprouting and synaptogenesis. The studies should be done both "in vivo" and in tissue culture, using electron microscopy and histochemical techniques. Growth inhibiting factors also deserve intense future study. Studies in amphibians and goldfish are also important.
Internet Services Provided by
