A Recipe for Recovery: Transplanted Embryonic Stem Cells Used to Create Neuromuscular Junctions and Restore Function in Paralyzed Rats
Douglas A. Kerr, MD, PhD

Dr. Douglas A. Kerr and a team of researchers recently published the results of a study that documents their use of embryonic stem cells to replace damaged motor neurons and restore function in paralyzed adult rats.  Dr. Kerr has graciously accepted our invitation to write an article for the TMA newsletter, which provides us with a description of this groundbreaking study.  A citation to the originally published article is provided below. 

The significance of this work for the TMA community cannot be overstated.  Dr. Kerr’s studies are focused on animals that have experienced spinal cord impairment from a virus that simulates neuronal damage typical of transverse myelitis.  It has been Dr. Kerr’s contention that transverse myelitis provides an excellent model for understanding and working through the restoration of myelin and nerve function.  The damage caused by an inflammatory attack from TM can be fairly small and might offer more controlled conditions for restorative research.  In these cases, it is possible that small amounts of myelin and nerve restoration could bring about significant recovery of function for a person.  Additionally, for the majority of people with monophasic TM, once the repair has been administered, there would likely not be risk of future damage, as is the case currently with multiple sclerosis. 

What is learned from the transverse myelitis model will be an important first step in developing wider treatments for more complex conditions, such as multiple sclerosis, ALS, Parkinson’s disease or traumatic injuries to the spinal cord. 

Original Publication: Deepa M. Deshpande, MS, MBIOT, Yun-Sook Kim, PhD, Tara Martinez, BS, Jessica Carmen, PhD, Sonny Dike, MD, Irina Shats, MS, Lee L. Rubin, PhD, Jennifer Drummond, BA, Chitra Krishnan, MHS, Ahmet Hoke, MD, PhD, Nicholas Maragakis, MD, Jeremy Shefner, MD, PhD, Jeffrey D. Rothstein, MD, PhD,and Douglas A. Kerr, MD PhD,  “Recovery from Paralysis in Adult Rats Using Embryonic Stem Cells”  Annals of Neurology 2006; 60:32–44

The Johns Hopkins study represents the first published account of the repair of a motor neuron circuit in an adult animal.  Our research provides compelling evidence of embryonic stem cells’ potential for healing.  We engineered new, completed, fully-working motor neuron circuits -- neurons stretching from spinal cord to target muscles -- in paralyzed adult animals.  We demonstrated that the motor neuron circuit was anatomically repaired (embryonic stem cell-derived motor neurons replaced motor neurons that were destroyed) and that the animals had restored motor function.  This work builds on previous studies, which are cited in our article recently published in the Annals of Neurology (2006). 

During the process of early development in humans and in all animals, motor neurons grow in the spinal cord.  These motor neurons (axons) leave the spinal cord and grow out into the body’s periphery and connect to muscles.  When the brain tells a set of muscles to move in your hand, this activity involves a motor neuron circuit that goes from the brain down through the spinal cord.  It exits the spinal cord and travels up the arm to your hand.  This is a very long way for a motor neuron to grow.  The nerves are connected to muscle and they stimulate chemical-electrical activity, which moves the hand. 

During this process of development, motor neuron growth is completed, and axons no longer grow outside of the spinal cord.  As adults, our cells no longer respond to the early developmental cues, because those cues are usually gone. That’s why we don’t recover well from severe injuries. 

The great challenge of our work was to show that cells can be made to re-trace complex pathways of nerve development, which are long shut off in adult mammals.  Our research is the proof of principle that we can recapture and replicate what happens in early stages of motor neuron development and use what we have learned about that process to repair damaged nervous systems.  We asked what was there when motor neurons were born, and specifically what let motor neurons extend outward. Then we tried to bring that environment back, in the presence of adaptable, receptive stem cells.  

The approach we used is analogous to an electrical repair.  Paralysis is like turning on a light switch and the light doesn’t go on. The connectivity is messed up because there was damage within the pathway (spinal cord) between the switch (brain) and the light bulb (muscle).  We’ve asked stem cells to go where needed to fix the circuit.

For a brief period after a nerve dies, it leaves behind what’s essentially an empty shell, with some scaffolding and non-nerve substances remaining.  But with embryonic stem (ES) cell injections at the right time and place, and by adding the right cues, we’ve learned to restore the biological ‘memory’ for growing neurons, which is clearly still in place.  The motor circuit engineering combines recent discoveries on stem cell differentiation, a growing understanding of early development of the nervous system, and insights into behavior of the nervous system in traumatic injury.  Our research has created a protocol or a cookbook recipe to restore lost nerve function.

The rats in our study were paralyzed; they had virus-damaged spinal cords, which modeled nerve disease.   The rats lost motor neurons to an aggressive infection with Sindbis virus -- one that, in rodents, specifically targets motor neurons and kills them. 

The repair protocol begins in our study with the differentiation of mouse (ES) cells into spinal motor neurons.  This is accomplished by exposing ES cells to retinoic acid (RA) and sonic hedgehog (Shh) protein.  We’ve learned that this exposure in the first weeks of life directs ES cells into motor neurons.  The differentiating ES cells were placed in a substance 3.5 days later that contained dbcAMP, a growth factor that increases the survival of the cells and also allows the new axons to extend into the peripheral nervous system. 

Sixty thousand mouse (ES) cells were injected into the ventral gray matter of the lumbar spinal cords of the paralyzed adult rats 28 days after they were inoculated with the motor neuron-killing virus.  At the time of transplantation, approximately 12,000 of the cells can be identified from their characteristics as early motor neurons. 

We understood from previous research that extending new motor neurons in an adult nervous system meant overcoming a number of hurdles. One involved myelin, the fatty material that insulates mature motor neurons. Like the coating on electrical wire, myelin prevents weakening of the traveling electrical impulse and lets it continue long distances. In humans, the myelinated sciatic nerve, for example, exits the spinal cord and extends to the leg muscles it activates, carrying impulses several feet.  Once laid down, however, myelin inhibits further nerve growth.  This is nature’s way of discouraging excessive wiring in the nervous system. We had to overcome inhibition from myelin lingering in the dead nerve pathways.

Two recently-developed agents, rolipram and dbcAMP, were used to overcome myelin-mediated axon repulsion.  As noted earlier, dbcAMP was used in the transplant solution to stimulate axonal growth and survival.  Rolipram was administered to the rats subcutaneously and served to neutralize the inhibitory effects of myelin on axonal outgrowth; Rolipram allowed the axons to grow through the spinal cord and to extend slightly into the outlying nervous system.
 
The next hurdle involved the stimulation of axon growth out into the periphery, attracting them to skeletal muscle targets, and providing cues to stimulate the formation of neuromuscular junctions between the newly transplanted motor axons and the existing paralyzed muscles.  Based on the earlier work of our research team member, Ahmet Hoke, we knew that GDNF (glial cell–derived neurotrophic factor) is a powerful stimulator of neuron growth.  At the time we transplanted ES cells into the spinal cord, we also transplanted glial cells into the remains of the newly-dead (from the virus) sciatic nerve at a point near its former leg muscle contacts.  The GDNF attracted the extending motor neurons, “luring” them to the muscle targets. To ensure a continuous supply of GDNF, we relied on injected fetal mouse neural stem cells, a known source of the molecule.  Our study demonstrated that GDNF did attract the ES cell-derived motor axons and stimulated the creation of neuromuscular junctions.  Finally, we administered CsA to inhibit the rejection of the transplanted ES cells.

The ES cell-derived motor neurons were distinguishable from host (existing) motor neurons as the transplanted neurons were “programmed” to express GFP (green fluorescent protein).  We were thus able to both observe and count these new motor neurons within the rat spinal cords and sciatic nerves.     

At three months from the time of transplantation some 4,100 new motor neurons were created in the spinal cord.  Roughly 200 exited the cord and 120 reached skeletal muscle, forming typical nerve-muscle junctions, with appropriate, typical chemical markers. Microscopically, the neurons and their muscle associations appear identical to natural ones in healthy animals.  Through electrophysiological analysis, we were able to demonstrate that the neuromuscular junctions between the transplanted motor neurons and the existing skeletal muscles were functionally active.  Fifty of the new neurons were found to carry electrical impulses. Because such testing is time and labor intensive, only a small area of leg muscle was analyzed.  The improved ability of treated rats, however, suggests more functional neurons are likely. 

Finally, we performed blinded evaluations of functional recovery from a behavioral assessment based on hind-limb grip strength.  This assessment progressed for six months from the time of transplantation.  Recovery was defined as the ability to flex the leg under the animal and to push off with the foot.  Eleven of the 15 treated rats gained significant, though partial, recovery from paralysis. The animals recovered enough muscle strength to bear weight and step with the previously paralyzed hind leg.  The rats gained weight, were more mobile in their cages and measures of muscle strength increased.

One of the more important aspects of the study entailed experiments that were designed to determine whether all of the ingredients in the recipe were required to achieve the desired result.  We created eight different groups and each was composed of 15 rats.  Only one group received all of the treatments in the protocol: intraspinal dbcAMP, subcutaneous rolipram, GDNF cells into the sciatic nerves, and CsA to inhibit rejection of transplanted cells.  The remaining seven groups of animals received different combinations of these treatments to determine whether any of these factors were not necessary for the complete restoration of nerve function in the rats.  For instance, one group did not receive dbcAMP, another group did not receive rolipram, and so forth.  In one of the groups, we tested the importance of using motor neurons in the protocol.  This group received ES cells that were differentiated to neural cells rather than motor neurons, and received the full compliment of other treatments. Each of the groups was then tested to assess evidence that implanted motor neurons were forming functional connections to skeletal muscles, that these motor units were functional (electrically active) and that there was functional recovery from hind-limb paralysis (behavioral observations). 

The tests demonstrated that of the seven groups that were treated without even one component of the “cocktail,” the animals experienced no recovery.  Animals that were transplanted with only neurons and not motor neurons did not experience functional recovery.  Animals that were not treated with the myelin inhibitors (rolipram and dbcAMP) or with GDNF (to attract axon growth in the peripheral nervous system) did not experience functional recovery.  For the group of animals that received all treatments in the protocol, we observed 125 new connections with host skeletal muscle, and we found evidence of 50 of these connections in the hind limb, which were active electrically.   We also found that only the animals that experienced the new and active neural muscular junctions experienced observable behavioral recovery.

Our study has developed and defined a protocol for the restoration of motor function in paralyzed rats using mouse embryonic stem cells.  The first step in the process involved the transplantation of ES cells that had been differentiated to motor neurons.  The ES cells were treated to increase their survival and also to inhibit their rejection by the host animal.  The next step entailed the administration of dbcAMP and rolipram to turn off the characteristics of myelin, which prevent the axons from growing outside of the spinal cord and into the periphery where muscles are located in the body.  GDNF was then used in the periphery, within the muscle “targets” to attract and stimulate the growth of the axons.  The result of this process was the connection of the newly transplanted motor neurons (axons) with the animals’ existing muscles to form active neural-muscular junctions.  The animals had been paralyzed, because the motor neurons that had been connected to these hind limb muscles had been destroyed.  Our transplanted precursor motor neurons, coaxed by our protocol, were able to form functional motor units (active nerve and muscle connections).  From the results we have observed in the restoration of function in paralyzed adult rats, we conclude that ES cells represent a potential therapeutic intervention for humans with paralysis.

We are in the process of beginning research to see how well the technique applies to human nerve recovery, using federally-approved human ES cells in larger mammals like pigs. Each of six academic institutions in a new collaboration will tackle a different major question of safety and effectiveness. Questions of tumor-formation, often a concern with ES cells, of the safety of surgery and of the ES cells’ ability to form healthy motor circuits are major questions to answer. Several years of testing and thorough data evaluation would occur before applying to the FDA to approve human clinical trials.

The study was supported by Families of SMA, Andrew’s Buddies/Fight SMA, the ALS Association and The Robert Packard Center for ALS Research at Johns Hopkins, the Muscular Dystrophy Association, Wings Over Wall Street, and a grant from the NIH.

Dr. Kerr is a grantee of The Packard Center for ALS Research at Johns Hopkins. He also directs Project RESTORE, a Hopkins-based undertaking to advance therapies for transverse myelitis and multiple sclerosis.

Others on the research team from the Department of Neurology at the Johns Hopkins University School of Medicine include Jeffrey Rothstein, M.D., Ph.D.; Ahmet Hoke, M.D., Ph.D.; Nicholas Maragakis, M.D.; Yun Sook Kim, Ph.D.; Sonny Dike, M.D.; Deepa Deshpande, M.S.; Chitra Krishnan, M.S., and Jennifer Drummond. Researchers from the Department of Molecular Microbiology and Immunology at the Johns Hopkins University Bloomberg School of Public Health include Jessica Carmen, Tara Martinez and Irina Shats. Jeremy Shefner, M.D., Ph.D., of the Department of Neurology at the State University of New York Upstate Medical University, Syracuse, N.Y., also contributed to the study.