Spinal Cord Regeneration

Published: January 12, 2004

For successful Regeneration to occur following spinal cord injury, several things must happen. First, damaged nerve cells and supporting cells must survive or be replaced, despite the acute effects of trauma and the conspiracy of processes that cause secondary damage. Replacement of lost cells in the CNS is unlikely without intervention because adult nerve cells in the brain and spinal cord cannot divide. Nerve cells that survive the injury often must regrow axons, despite tissue changes such as cavity formation that obstruct growth. Axons also must navigate among the myriad possibilities to find appropriate targets. Once the axons locate their targets, they must construct specialized structures to release neurotransmitters at synapses, while target cells must assemble and precisely locate the structures needed to respond to neurotransmitters. Finally, the neural circuitry may have to compensate for changes that have occurred in the spinal cord circuitry following injury.

Until recently, most scientists believed that nerve cells in the CNS of adult mammals could not regenerate. Dramatic findings, some presented at this workshop, are now changing that pessimistic outlook. For example, some studies have demonstrated that nerve cells in the brain and spinal cord make unsuccessful attempts to regenerate and can regrow under some conditions. New findings also demonstrate that the spinal cord has more active repair mechanisms than previously suspected. Although researchers recognize the many obstacles to obtaining regeneration in the human spinal cord, they believe successful regeneration of even a small percentage of nerve fibers will produce significant recovery of function.

Implications of Development

Scientists favor the spinal cord for studying the CNS because it is simpler than the brain. The long tradition of anatomical and physiological research on the cord provides a solid framework for studying development. Developing nerve cells perform the same steps needed for regeneration — they grow, navigate, and make appropriate connections. Regenerating nerve fibers face problems that are quite different from those faced by developing nerve cells, however. For example, the tissue through which axons move is more loosely connected during development, and an injured spinal cord may become quite disorganized near the injury site. Also, distances in the adult CNS are much greater than in the embryo, and chemical signposts for navigating axons may have changed in the adult. While the extent to which regeneration resembles development is uncertain, research about nervous system development is a source of crucial insights about how to promote regeneration following spinal cord injury.

Nerve Cell Differentiation

The adult spinal cord is an intricate assembly of cells and nerve fibers arrayed in specific locations with very precise interconnections. Nerve cells in the spinal cord include several types of Motor neurons, sensory neurons, and interneurons, each of which varies in shape, electrical activity, Neurotransmitter release, and many aspects. Glial Cells also include several specialized types of cells in the mature CNS, and the major nerve pathways of the spinal cord white matter are highly organized anatomically. How all of this comes about has been a subject of speculation and experiments for more than a century. The mystery is finally giving way to traditional neuroscience research methods, augmented by new technologies such as Molecular genetics.

The factors causing cell types in the spinal cord to become distinct from one another are cell lineage (which cells arise from which by cell division) and cues from within the developing embryo. Research is now identifying these chemical cues and discovering how cells respond to them. Two major signaling systems control the fate of embryonic brain and spinal cord cells. One system controls the specialization of the nervous system along the long axis from the brain down through the spinal cord. The other system controls specialization along the dorsoventral plane, that is, in a cross-section of the spinal cord (“dorsal” refers to the back portion and “ventral” denotes the abdominal direction). So far, the general operating principles of the two systems appear to be the same.

The control of cell identity along the dorsoventral axis of the spinal cord illustrates how these developmental systems operate. Among the essential tools scientists developed to study this process are chemical markers that stain specific cell types before they fully specialize in the embryonic spinal cord. Three cell types form in the ventral part of the early embryonic spinal cord. Glial cells form in the most ventral part, called the floor plate; motor neurons and interneurons form more dorsally. Experiments have shown that the key signal that determines the fate of all three cell types is a protein called sonic hedgehog. (The name arises because this mammalian molecule was identified by its resemblance to the “hedgehog” protein of the fruitfly. Flies with a mutation in the hedgehog gene have a peculiar prickly appearance.) To simulate the situation in the developing embryo, scientists placed pieces of ventral spinal cord in cell culture and exposed them to different concentrations of sonic hedgehog protein. These pieces produced motor neurons, glia, or interneurons depending on the concentration of protein to which they were exposed. In the embryo, a structure called the notochord releases the sonic hedgehog protein signal. Spinal cord cells that lie closest to the notochord are exposed to the highest concentration of the signal and become glial cells. Those in more dorsal positions are exposed to lesser concentrations and become, respectively, motor neurons and interneurons.

Although scientists are rapidly identifying the signals that drive the generation of cell types in the developing spinal cord, many basic questions remain. Many signals have yet to be discovered, and it is not yet clear how cells sense small differences in concentrations and respond to become specialized cell types. Interactions among the various signaling systems are likewise obscure.

Answers to these questions may have implications for spinal cord regeneration. In the last few years, scientists have discovered that even the mature CNS may harbor latent progenitor cells that can divide and specialize to form new nerve and glial cells. In a rat model of spinal cord trauma, the single layer of cells lining the central canal of the spinal cord expands to multiple layers of cells about 48 hours after a contusion Lesion. The central canal is continuous with the brain ventricles, large fluid-filled spaces inside the brain. During development, new nerve cells arise from cells lining these structures. Cells from the expanded central canal of injured animals appear to stream out into the spinal cord; these may be neural progenitor cells attempting to repair damaged tissue. It is important to know whether developmental signals that might guide Neuron growth persist in the adult. Another reason studies of cell specialization may be relevant to spinal cord injury is that the molecules involved in this developmental process may have other important functions in the adult. Understanding the signals that control cell specialization in development may be critical for learning how to help them repair damaged spinal cords.

Many new findings presented at the workshop reflected the ways researchers now study the molecular machinery by which cells operate. Knowing the genetic code for proteins allows scientists to detect similarities among proteins. By comparing genes among different species, researchers can rapidly apply insights from lower organisms to mammals. Comparing newly identified genes and proteins to known ones within the same animal can also help scientists understand what a newly discovered protein does. Identifying one protein often helps reveal other members of the same protein family that have related functions, as in families of growth factors, cell adhesion molecules, and neurotransmitter receptors. Scientists are also learning to recognize Functional regions that many proteins share in different combinations. Gene sequences predict many aspects of a protein’s function, such as whether it will respond to certain regulator molecules. Thus, the chemical language that orchestrates development provides crucial clues about regeneration, even if the processes differ.

Axon Pathfinding

Developing nerve cells of the brain and spinal cord grow axons over long distances, along specific routes, and to precise targets. The tip of a growing axon forms a specialized structure called a growth cone. These growth cones sense cues, integrate that information, and make choices that steer the axon in one direction or another. Scientists have identified attractants and repellents that diffuse over long distances, as well as chemical attractants and repellents with fixed locations. Together, these cues allow axons to navigate through the developing brain and spinal cord. The identification of one family of long-distance attractants, the netrins, illustrates this area of research and its potential relevance to spinal cord regeneration.

A century ago, the Spanish neuroanatomist Ramón y Cajal speculated that diffusible chemical signals might guide growing axons. The first such signals, called netrins, were discovered just a few years ago in the chick spinal cord. “Commissural” neurons in the dorsal part of the spinal cord send axons from the front of the cord around toward the back. When the growth cones of these axons approach the midline of the developing spinal cord, they make a beeline for the floor plate, a specialized region of the embryonic spinal cord at the ventral edge of the midline. When scientists placed pieces of developing spinal cords in various arrangements in culture, they found that something in the ventral floorplate attracted growing commissural axons from a distance. They isolated the attractants and named the identified proteins netrins. When scientists further examined the effects of netrins, they found that these molecules also repelled growing axons from other types of developing nerve cells. This finding was predicted by studies in worms of molecules that closely resemble netrins. Experiments in normal and mutant mice confirmed that these molecules guide developing axons in living mammals.

Many guidance molecules were only recently identified, and certainly more remain to be found. Similarities between guidance molecules in mammals and those in simple organisms like worms and fruitflies are speeding progress in this area of neurobiology. Whether regenerating axons respond to guidance signals in the same manner as developing axons and whether these cues are still present in the adult spinal cord are particularly important questions for spinal cord regeneration. Ultimately, scientists hope to find ways to manipulate these signaling mechanisms to enhance regeneration.

Synapse Formation

For regeneration to be successful, axons must not only grow but also find and connect to appropriate targets. Axons must construct the highly specialized structures that release neurotransmitters from nerve terminals. Cells that receive signals across synapses also must participate in forming new synapses at a time when they would not normally do so. They must assemble in precisely the right places the specialized structures necessary to respond to neurotransmitters. Finally, the developing spinal cord must insure that synapses of the correct type form only on proper cells and on the appropriate parts of these cells so that the neural circuits will work.

Although scientists know very little about how new synapses form in the adult mammalian spinal cord, they are learning how synapses develop in the skeletal neuromuscular junctions (NMJs), the synapses by which motor neurons activate muscle cells. NMJs are much more accessible for study than synapses in the spinal cord, and scientists have therefore used them to learn about the basic principles of synapse development and function. These nerve-muscle synapses also regenerate, which allows comparison of development and regeneration.

Although axons and muscle cells can each synthesize the specialized components they need to form synapses, development of synapses requires back-and-forth signaling between the two cell types. At the turn of the century, scientists demonstrated that regenerating motor nerves form synapses at the exact sites of former synapses, even though synapses cover only a tiny percentage of the available muscle surface. This means muscle cells must have “stop signals” that axons can recognize. Muscles also regulate their receptivity to synapse formation according to whether they already have a nerve connection. An implanted nerve will not form a synapse with a muscle unless the original nerve to that muscle has been removed. Muscles that have lost their nerve connections may also release molecules that entice axons to make new synapses.

In the modern era, scientists have added the tools of genetics to traditional methods of developmental neuroscience research. They can now test the role of particular molecules in synapse formation by creating mutant mice, such as “knockout mice,” that lack a particular protein. Studies with knockout mice have shown how a protein called agrin helps the developing muscle aggregate molecules called acetylcholine receptors at the synapse where they are needed. Acetylcholine receptors enable muscle cells to respond to the neurotransmitter acetylcholine, which is released from the nerve terminal. Agrin knockout mice died before birth or were stillborn because of defective NMJs. Interestingly, inactivating the agrin gene not only affected muscle, but also the perturbed axons. This reflects the complex interactive nature of the signaling process between axons and their targets, which scientists are just beginning to understand. Scientists are now creating genetically altered mice to study other molecules that control the development of the NMJ.

In many ways, synapses within the brain and spinal cord resemble the NMJ, but not completely. Some, but not all, of the molecules that control development of the NMJ are present in the developing CNS. Each muscle cell receives synapses from only one axon, and all of these use the same neurotransmitter (acetylcholine). A single spinal cord nerve cell, on the other hand, may receive thousands of synapses from nerve cells of the brain and spinal cord and from sensory nerves of the body. Each spinal cord neuron may also respond to several different neurotransmitters. So, while spinal cord synapses and NMJ probably share the same general principles, the spinal cord must need additional signals to form synapses.

Understanding synapse formation is becoming increasingly important as the prospect improves for obtaining survival and growth of spinal cord cells after injury. So far, nerve fibers regrowing in experimental animal models of spinal cord regeneration have developed few new synapses, and this may be the limiting factor in recovery of movement. Understanding how synapses develop may reveal whether spinal regeneration stops because regenerating axons lack the ability to form synapses or whether nerve cells below the lesion are unreceptive to synapse formation. This may lead to ways of encouraging the formation of new synapses by regenerating fibers.

Basic Regeneration Studies

Scientists have long known that nerve cells outside the brain and spinal cord can regenerate, but they believed that nerve cells in the CNS of adults could not regrow. In the early 1980s, experiments in the spinal cords of animals showed that CNS neurons can regrow under certain conditions. These experiments were inspired by the idea that adult CNS cells might be able to grow if given a permissive growth Environment. Scientists grafted segments of sciatic nerve — a Peripheral nerve that can regenerate — to the spinal cord, circumventing the lesion site. Some nerve cells, usually from near the site of the lesion, grew axons through the nerve bridges as far as 3 or 4 centimeters and reached the other end of the bridge. Some nerve cells in the lower parts of the brain also grew into the graft. Because of the complexity of the spinal cord, researchers could not accurately assess whether regrowing nerve cells made functional synapses or exactly what about the sciatic nerve bridges was “permissive.” For this reason, some scientists turned to another model system to study CNS regeneration—the retina.

The retina of the eye is an outpost of the brain. Like the spinal cord, it is part of the CNS. Retinal neurons called ganglion cells carry signals from the retina to the brain. Their axons, together with supporting cells, form the optic nerve. Cutting or crushing the optic nerve, and thus the axons of the retinal ganglion cells, has become an important model for injury and regeneration in the CNS.

Retinal ganglion cells normally do not regenerate after the optic nerve has been transected. In early experiments, scientists placed peripheral nerve bridges from the site of damage in optic nerves (usually near the retina) to appropriate targets in the brain. The nerve grafts bypassed the problem of pathfinding by funneling growing axons directly to the correct region of the brain. Some ganglion cell axons grew distances equivalent to nearly twice their normal length. However, at best only a small percentage of axons regrew in these experiments since most cells died soon after transection. Some axons did penetrate the brain and make synapses, restoring the simple Reflex response of pupils to light and the animals’ light-avoidance behavior. Axons that reached the brain found the appropriate layers and parts of cells in the brain, but failed to recreate the proper, orderly representation of the visual world on the brain.

These retinal regeneration experiments raised many questions. What makes Peripheral Nervous System tissue supportive for growth? Are there growth factors in the nerve grafts that are not available in the adult CNS, or are growth inhibitors normally present in the adult CNS absent from the grafts? Why do so many ganglion cells die and so few regenerate? Do intrinsic genetic programs of these cells affect success and failure? How does regeneration resemble and differ from development? Experiments are beginning to answer these kinds of questions.

Trophic Factors

For nerve cells to regenerate axons, they must first survive the injury. Trophic factors are signals that promote the survival and growth of nerve cells. The classical studies of trophic factors in development showed that nerve cells become dependent on these substances during the period when they specialize and begin to connect with their targets. The developing nervous system produces many more nerve cells than the adult nervous system needs. Cells compete with one another to obtain trophic factors supplied by appropriate target cells. Those neurons that do not succeed in competing for appropriate connections die through apoptosis.

The first trophic factor isolated was NGF (Nerve Growth Factor). NGF is essential for the survival of some types of nerve cells in the PNS. Withdrawing NGF from peripheral neurons in culture is an important way of studying apoptosis in nerve cells. In the last several years, scientists have found that trophic factors are important in the development of the CNS as well. At the workshop, participants reported experiments suggesting that there are important differences in the trophic factor requirements of central and peripheral nerve cells. Those differences may help explain why CNS nerve cells do not regenerate.

Here again, retinal ganglion cells are favorable subjects for CNS research. Scientists developed methods to isolate these cells with 99 percent purity, allowing precise studies of the factors these cells need to survive and grow. The scientists then attempted to sustain these cells in culture using trophic factors that the cells might encounter in their normal course from the retina to the brain. None of these factors alone was sufficient for more than 1 percent of retinal ganglion cells to survive for even 3 days.

Studies in peripheral nerve cells have shown that activating the cyclic AMP “second messenger” system augments the effects of trophic factors. Cyclic AMP is a small molecule that carries messages from cell surface receptors activated by “first messengers” (hormones, neurotransmitters, or other signals) to sites within the cell. Like other second messenger systems, this biochemical pathway allows a single first messenger to control several cellular processes and helps in regulating and integrating the many signals cells receive. Although activating the cyclic AMP second messenger pathway with the drug forskolin did not sustain retinal ganglion cells in culture, and trophic factors alone were insufficient, the two combined saved more than a third of the cultured cells. Combining multiple trophic factors with forskolin allowed survival of more than half of the cultured cells for more than a month. Adding other, as yet unpurified, factors boosted survival to more than 80 percent.

These experiments suggest that combinations of trophic factors may be essential for survival of CNS neurons. Another important insight is that the responsiveness of CNS cells to trophic factors is not static, but can change depending on the level of second messengers. Electrical activity and signals from other cells stimulate second messenger systems, and these influences change dramatically for cells below a spinal cord injury. Administering trophic factors and controlling responsiveness to these factors may promote nerve cell survival in the damaged spinal cord. However, these powerful and poorly understood substances can also have serious side effects.

Scientists are not yet certain whether adult spinal cord nerve cells need combinations of trophic factors, which trophic factors affect which cell types, or what controls the cells’ sensitivity to these factors. In one important finding from the retina culture experiments, scientists learned that survival and axon growth were always coupled; that is, any interventions that allowed cells to survive also prompted them to extend their axons.

Intrinsic Growth Programs

Nerve cells’ intrinsic capacity to grow is another factor that may contribute to the success or failure of regeneration. Scientists have studied intrinsic growth capacity by comparing cells from young animals to those from older ones. Very young animals generally recover more completely from CNS damage than do adult animals. By placing the retina and the tectum (the brain area to which retinal ganglion cells connect) from animals together in culture, scientists demonstrated that regeneration in culture is also age-dependent. They then independently varied the age of the tectal and retinal pieces placed together in culture to determine to what extent each contributed to the failure to grow in older animals. The results showed two major reductions in the ability of ganglion cells to grow as the animal ages. The first, larger reduction was due to changes in the growth capacity of the retinal cells themselves. The later, smaller reduction, was more gradual and appeared to be due mostly to changes in the target tissue. Providing growth factors partially increased survival and growth but could not overcome the early large decline in growth ability.

Biologists believe changes in growth capacity probably reflect changes in the specific genes that are active in each cell. Several genes were inactivated at about the time that regeneration declined, but one gene was turned off just as the capacity to regenerate was lost. Surprisingly, that gene was bcl-2, which is well-known because its product is an important regulator of apoptosis. Retinal ganglion cells taken from mice with an inactivated bcl-2 gene (bcl-2 knockout mice) did not show the normal sharp decline in growth ability. Even cells from the adult retina of these knockout mice could grow if they were given embryonic tissue as a target. Experiments with drugs directed at enzymes in the apoptosis pathway showed that the bcl-2 gene’s effects on growth were separate from its effects on apoptosis. This gene apparently acts as a “switch” that controls axon growth in the CNS. Finding ways to control this switch may yield a new approach to therapy for spinal cord injury that may complement other therapies such as trophic factors. While this treatment approach appears beyond genetic technology at the moment, understanding the role played by these intrinsic programs in regulating the neuron growth will provide important insights intoregeneration.

Barriers to Growth

Scientists have now identified a long list of molecules in the adult CNS that actively inhibit growth. For example, oligodendrocytes produce a Myelin-associated growth inhibitor that may be one of the most important inhibitors of growth in the adult spinal cord. One way these inhibitors act is by making growth cones collapse. Growth inhibition may be quite specific for each nerve cell type; that is, different cells may be most sensitive to different inhibitors. Another way inhibitors act is by modifying the extracellular matrix, the noncellular material surrounding cells through which axons must grow. For example, some substances act as “anti-adhesives,” preventing growing axons from sticking to surrounding tissue, which is necessary for them to grow forward. How inhibitors block axon growth and which of the many inhibitors are clinically important following spinal cord injury are essential questions that scientists are now trying to answer.

Scientists need to determine the normal physiological roles of the many substances that inhibit growth in the adult spinal cord. Similarities in how these inhibitors work might allow generic strategies for overcoming their effects. One possibility would be to find common pathways, such as second messenger systems, through which these factors operate. Experiments with a component of pertussis toxin (a toxin from the bacteria that causes whooping cough) suggest that this might be possible. This toxin, which affects second messengers, blocked growth cone collapse from three very different inhibitory factors (collapsin-1, thrombin, and the myelin-associated factor). Because the extracellular matrix that surrounds cells is a repository for many inhibitory substances, understanding the interaction of cells with the extracellular matrix is an important focus of research. Finally, signals that inhibit and stimulate growth might converge on common intracellular machinery so that sufficient stimulation might overcome some of the inhibition. Experiments with trophic factors in retinal ganglion cells support this idea.

Applied Regeneration Studies

Researchers are beginning to apply knowledge about nerve growth and inhibitory factors and other aspects of neuron regeneration by testing new therapeutic approaches in animal models of spinal cord injury. The partial success of several of these animal experiments has led to optimism by many experts that, with the right combination of strategies, regeneration will eventually become possible in humans. However, it now appears unlikely that there will be a single magic bullet for repairing the spinal cord. Instead, a combination of approaches will probably be necessary.


One approach for repairing spinal cords that is being tested in animals is to transplant cells and tissues into the damaged spinal cord. In particular, scientists are transplanting cells or pieces of peripheral nerves that produce substances that create an environment for axons to grow. This idea was first advocated about 100 years ago by the neurologist Ramón y Cajal. He suggested implanting cells from the PNS into the area of a CNS injury. Since the environment of the PNS supports axon regeneration, he believed re-creating this environment in the spinal cord might allow CNS axons to regrow after an injury. Ideally, this environment would also point growing nerves to the correct targets. Experiments with PNS transplants in rat models of spinal cord injury have led to axon elongation and cell body changes associated with regrowth. Transplants from the PNS also seem to reduce scarring around the injury that may impede regrowth. One technique tested in rats is transplanting Schwann Cells — glial cells that help myelinate axons in the PNS — into the spinal cord after injury. These transplants supported regrowth of the damaged nerves in rats with spinal cord injury. Researchers are now studying human Schwann cells to determine if this technique will work in humans.

Another way of encouraging regeneration is to implant fetal tissue. Tissue from a growing fetus contains stem cells, progenitor cells, and many substances that support growth. Such tissue also presents fewer obstacles to growing axons. Stem cells can differentiate into several cell types, depending on the signals they receive. Transplanting them into the spinal cord may, with the right chemical signals, help them develop into neurons and supporting cells in the spinal cord, re-establishing lost circuits.

Studies in rats show that fetal transplants promote survival and regrowth of some damaged nerve cells. Transplanting fetal CNS tissue into the spinal cord of both mature and newborn rats yielded axon growth that terminated within a few millimeters of the border of the transplant. Researchers still need to learn exactly how fetal tissue transplants promote nerve regrowth. The transplants appear to “rescue” axons and provide a bridge across which regenerating axons can grow. While both adult and newborn rats regrew descending nerve fibers from the brain, the growth of descending pathways into the transplants was substantially greater in the newborns. This suggests that other changes in the maturing CNS, such as the production of inhibitory factors or a loss of certain axon guidance molecules, may influence axonal regrowth after injury.

Trophic Factors

Using insights from retina and culture experiments, researchers are beginning to test whether trophic factors can enhance regrowth in the spinal cords of rats. Growth factors may be responsible for much of the nerve regeneration normally seen in the PNS and in CNS axons near transplanted PNS tissue.

Different pathways in the spinal cord may require particular combinations of growth factors for survival after injury. While nerve cells usually do not survive after axons have been severed close to the cell body, recent experiments in the rat spinal cord have shown that two trophic factors, brain-derived neurotrophic factor (BDNF) and neurotrophin 3 (NT3), can rescue nerve cells from which the axons have been recently severed. Although NT3 has short-term effects, BDNF can help nerve cells survive for 4 weeks or more after injury. When the trophic factors BDNF, NT3, and NT4 (neurotrophin 4) were combined with fetal tissue transplants, axons no longer stopped growing at the border of the transplant but instead greatly expanded the territory into which they projected.

The combination of transplants and trophic factors also led to an increase in c-jun, an important immediate early gene. Immediate early genes respond rapidly to many stimuli and regulate many cell functions. Interestingly, these experiments showed that axons from cells that use the neurotransmitter serotonin responded to trophic factors more vigorously than axons from cells that use other neurotransmitters. This illustrates the importance of finding the right combination of growth factors for each type of cell.

Anti-Inhibitory Factors

Myelin-associated neurite growth inhibitor, which is produced by oligodendrocytes, is the most important CNS growth inhibitor so far identified. When researchers blocked this growth inhibitor with an Antibody called IN-1, which binds to and masks the factor from growth cones, severed axons began extending past the oligodendrocytes and reconnecting with their targets. After this treatment, rats with severed spinal cords moved more normally and partially regained their contact-placing reflexes (in which rats move their legs to support their bodies when they are placed against a surface).

Combination Therapies

Evidence that combining some therapies may have an additive effect has prompted researchers to focus effort on finding a combination that will achieve regeneration. Some combination therapies recently tested in rats have shown exciting results. One approach used neurotrophin 3, fetal cell transplants, and IN-1, the antibody to myelin-associated neurite growth inhibitor. Rats treated with this approach showed faster and more extensive recovery after spinal cord injury than those given any single treatment alone. Their recovered reflexes disappeared after researchers destroyed the cerebral cortex, showing that the brain, rather than reorganization within the spinal cord, controlled the reflexes. Researchers still need to learn if this therapy can be a general approach or if specific nerve pathways have specific requirements for growth. They also need to carefully define the time windows for effective combination treatment.

Another approach using nerve fiber transplantation combined with growth factors showed the first functional regeneration of completely transected rat spinal cords. Researchers carefully transplanted 18 pieces of peripheral nerves (one to three pieces for each of the normal nerve tracts) taken from the rats’ chests to “bridge” 5-millimeter gaps at the T8 segment of rats’ spinal cords. To evade inhibitory proteins from oligodendrocytes, the bridges routed regenerating axons away from white matter, where they would normally grow, and into gray matter. The researchers fixed the grafts in place with a glue based on a blood-clotting factor called fibrin. The glue also contained acidic fibroblastic growth factor, or aFGF, which enhances nerve fiber development. Finally, the scientists wired the Vertebrae to keep the spine in place while the area healed.

After 3 weeks, rats that had received this type of graft began to recover function in their hind legs. Some of the treated rats regained some movement on both sides of their bodies, while others regained movement on only one side. The rats that recovered on both sides of their bodies eventually began partially supporting their weight with their hind limbs. They also displayed walking movements and contact-placing reflexes. The rats continued to improve gradually over the course of a year, though they never walked normally. Rats with bridges from white matter to other white matter, rats in which the fibrin glue had no aFGF, and rats that did not receive transplants did not recover any function over time.

Anatomical studies of spinal cords from rats that recovered function after this therapy showed that the nerve fibers grew into the gray matter on the opposite side of the gap. The fibers then grew at the interface between the gray matter and the white matter, an area that corresponds to the normal corticospinal tract in rats. The degree of recovery corresponded significantly to the degree of motor fiber regeneration.


Basic research has led to a better understanding of trophic factors, growth barriers in the CNS, and the intrinsic capacity of nerve cells to grow. These insights are being applied in animal models of spinal cord injury using transplantation, trophic factors, and anti-inhibitory molecules. The exciting results of strategies that combine these interventions suggest that such approaches will ultimately prove the most successful for regenerating spinal cord pathways in humans. Developmental studies of cell specialization, axon growth and pathfinding, and synapse formation are leading to promising new avenues for improving on these combination approaches.