ARTICLE IN BRIEF
Investigators were able to use a transcription factor to induce neurogenesis in an adult animal spinal cord, providing proof of principle that cell reprogramming could occur without transplantation.
Turning stem cells into specific cell populations and then transplanting them into the body is a promising approach to treat certain diseases. Now, scientists are experimenting with reprogramming cells directly within the area of the diseased tissue, avoiding the problems inherent in transplantation — and it seems to work.
In the latest study, scientists at the University of Texas Southwestern Medical Center were able to deliver a single transcription factor into astrocytes in the adult spinal cord and to turn these cells into neurons. Tests of the reprogrammed neurons showed that they developed synapses with neighboring cells in the spinal cord.
Ultimately, the scientists say that it may be possible to create new nerve cells that would replenish the area around the injured cord and help restore function. The science is still in its infancy but this proof of principle study, published in Nature Communications in February, could pave the way for such reprogramming techniques to treat a variety of neurological problems, including spinal cord damage, traumatic brain injury, Alzheimer's disease, and other neurodegenerative conditions.
Chun-Li Zhang, PhD, an assistant professor of molecular biology at UT Southwestern, and his colleagues have published a series of papers on their reprogramming method. They initially tested a dozen transcription factors and it turned out that they needed only one — SOX2, also known as SRY (sex determining region Y)-box 2 — to coax astrocytes into becoming neurons and forming networks. The study outlining the work was published in Nature Cell Biology last October.
In their latest paper in Nature Communications, they took astrocytes from the spinal cord and provided the same cellular signal — SOX2; the cells subsequently became neurons when transplanted into the spinal cord. Then they worked directly in normal animals and in those with an acute injury to the spinal cord.
“Our work is showing us that we may be able to reprogram cells without the need of cell transplantation,” said Dr. Zhang, the senior author of the study.
While the researchers have found that neurons make connections to healthy spinal cord cells, they do not yet know whether the reprogramming of the cells leads to any functional improvement in the paralyzed animal. Those studies are underway.
Spinal injury may not be the easiest model to prove that their technique works, Dr. Zhang explained. While astrocytes are plentiful in the brain and spinal cord, they are the cells that respond to spinal injury by sending out signals that create a scar at the site of the damage. This has been a hurdle to scientists trying to get cells above and below the scar to communicate with one another to promote functional recovery.
Still, Dr. Zhang and his colleagues believe that astrocytes “are the ideal target for in vivo reprogramming in the spinal cord and there are good reasons to use a patient's own endogenous cells.”
Cell transplantation is potentially a promising technique for repair to the damaged cord. Recent studies in the cell transplant field showed that grafted fetal- or induced pluripotent stem cell-derived neurons or neural precursors are able to grow and function in the otherwise inhibitory environment of the spinal cord. The transplanted cells look like they can repair some of the spinal cord damage. Collecting and preparing the cells for transplantation is time-consuming, and it may not be possible to infuse the cells early enough on the heels of an injury.
There is also the potential that residual undifferentiated stem cells in the injured environment could expand and become cancerous. And the transplantation itself could trigger some damage to the already injured spinal cord.
For these reasons, Dr. Zhang's team set out to develop an alternative technique to transplantation. Could they reprogram endogenous non-neuronal cells (such as the scar-forming astrocytes) into neurons in the spinal cord? The reprogramming field had advanced enough for the scientists to try out their ideas.
They began their study by screening for transcription factors that could induce neurogenesis in the adult spinal cord. After testing a dozen factors they identified one — SOX2 — that was sufficient “to convert endogenous spinal astrocytes to proliferative neuroblasts that would eventually generate mature neurons,” said Dr. Zhang. Then they found that treating the animals with valproic acid — a drug that encourages the survival of neuroblasts and their differentiation into mature neurons — doubled the number of new neurons.
Only a small fraction of the astrocytes in the region — 3 to 6 percent — turned into neurons but the scientists are now working on ways to expand the number of cells. Also, in vivo growth of the cells from the astrocyte stage to neuroblasts to mature neurons took eight weeks, which is slower than the same process in a lab dish. Dr. Zhang said that this slower course may be beneficial for the cells to integrate into their new environment, but he is not sure.
He said that the SOX2-induced mature neurons were still there 210 days after the experiment started. They have yet to go out farther to see just how long the cells will survive. Mindful of the possibility of tumor growth, they followed the animals for one year. None of the animals developed tumors in the spinal cord.
The researchers have not yet done the studies to show that these new neurons are functioning as healthy spinal cord cells.
“By themselves, the cells can't become mature or survive a long time because the environment is not good for newborn neurons,” said Dr. Zhang. “But if we change the environment and make it more permissive to new neurons, which we did when we added valproic acid, we [can] promote neuronal production.”
The investigators have done extensive electrophysiological testing and showed that the neurons seem to make connections with neighboring cells in the adult brain.”
“It's a proof of principle,” said Dr. Zhang. “We got some cells but not enough yet to repair the damaged cord or brain.”
The scientist added that for such a technique to work in humans it would be ideal to identify small molecules that would trigger the expression of SOX2 rather than using viral vectors to deliver the gene for the transcription factor.
This work follows on the heels of findings from other transplant experts like Marius Wernig, MD, PhD, an assistant professor of pathology at the Institute for Stem Cell Biology and Regenerative Medicine at Stanford University School of Medicine, who turned fibroblasts into a stem-like state and then into neurons, and used them successfully in animal models of Parkinson's disease. The same group converted mouse fibroblasts into neurons using three transcription factors. They replicated the experiment using human cells as well.
“This study is potentially exciting,” said Dr. Wernig. “But it remains to be seen whether these in vivo-generated “neurons” are actually functional. He was surprised that SOX2 alone could coax astrocytes into neurons.
Paul Tesar, PhD, an assistant professor of genetics at Case Western Reserve University, said he believes that “all these reprogramming techniques are laying the foundation for what is possible.”
There are many unanswered questions, however. Do the reprogrammed cells revert back to their former cell type? How long do they keep their fate? Is there long-term stability and functionality? More specifically for the Nature Communications study, are they even the right cell type to make a difference at the site of the damaged cord? Would creating neurons in that environment even be useful?
“We are still in a feasibility phase of the reprogramming field,” said Dr. Tesar. “The next steps will be to see whether it is effective and safe over a long period of time. All of these studies will teach us about the biology of the systems and the disease.”
Wise Young, MD, PhD, chair of neuroscience in the department of cell biology and neuroscience at the W. M. Keck Center for Collaborative Neuroscience at Rutgers, the State University of New Jersey, said that he “read this paper with interest because it suggests that upregulating a single gene (SOX2) can cause astrocytes to become neuroblasts that produce synapse forming neurons.
“The authors claim that they are spinal astrocytes because they express GFAP [glial fibrillary acidic protein, a marker of mature astrocytes]. They also claim that they are not NG2+ oligodendroglial precursor cells by using cells from a mouse that is genetically labeled so that NG2+ cells express yellow fluorescence (YFP) and showing that none of the induced cells express YFP.
“I am not necessarily convinced by these arguments, however, since neural stem cells in the subventricular express GFAP and look like astrocytes. They may well be inducing cells that are already inclined towards being precursor cells to produce neurons by transfecting them to express SOX2. On the other hand, perhaps it doesn't matter. Clearly, there are some cells in the injured spinal cord that can give rise to neurons. Whether they are mature astrocytes or not probably does not matter. The data indicate the neurogenesis occurs in the injured spinal cords if you add virus that causes SOX2 to express in cells.
“But, the authors convinced me that they are indeed getting some astrocytes to produce neurons when they transplanted induced astrocytes into the spinal cord and found that they produced neurons,” said Dr. Young, an expert on spinal cord damage and repair.
Dr. Young wonders whether or not astrocytic production of neurons is a natural occurrence in injured spinal cords, that is, whether SOX2 may be spontaneously upregulated after injury in these cells, thereby leading to spontaneous neurogenesis in injured spinal cords.
“It is interesting that the induction appears to be very slow, taking weeks,” said Dr. Young, “unlike the induction of neurons from fibroblasts shown earlier by Marius Wernig at Stanford, using three or four genes to get induction within 48 hours. Perhaps this is because the cells had to downregulate the SOX2 expression before they can start making neurons.”
The study findings are “of interest and potentially applicable, although we have never had any FDA-approved clinical trial of viral therapy of the spinal cord or brain,” he said. “It is possible that inducing SOX2 expression can be done without a virus. If so, we may be able to make neurons from blood cells of the patient and then inject them into the spinal cord.
“This finding illustrates how much progress is being made in the field of programming cells,” he said. “We are entering into a brave new world where we will soon be able to not only make stem cells but make any kind of cell that we want.”
Jerry Silver, PhD, a professor of neurosciences at Case Western, also agreed that it is an interesting and novel approach. “The question is whether it will lead to any functional improvements,” he said. “We don't know if cells are linking up in an appropriate way.” He added that the model — a hemi-section lesion — is not a relevant clinical model of human spinal cord damage. A contusive injury would be a better model to test whether the new cells are helping to repair the injured spine and facilitate walking.