4 March 2009. Glia, once the ignored country cousins of the neurons that scientists were really interested in, continue to show that they, too, are sophisticated players in health and disease. These supporting cells have been implicated as stealth neuron killers in amyotrophic lateral sclerosis (ALS), and a new paper, published online in Nature Neuroscience February 22, suggests they also harm nerve cells in the neurodevelopmental disorder Rett syndrome. A separate report, published online in PNAS February 27, finds that, unlike other glia that are toxic to motor neurons when they express mutant superoxide dismutase 1, Schwann cells producing mSOD1 appear to slow disease.
The role of glia in ALS has been getting a fair share of attention in ALS research for a few years now (see ARF related news story), but not so in Rett syndrome. Sporadic mutations in methyl-CpC-binding protein 2, a transcriptional repressor, cause this developmental syndrome that includes autism-like symptoms and motor problems. Neurons highly express MeCP2, and have been the focus of most research. In glia, MeCP2 was generally thought to be not expressed, or not important.
Nurit Ballas of the State University of New York in Stony Brook, first author on the Nature Neuroscience paper, suspected researchers were missing glial MeCP2; other non-neuronal cells such as fibroblasts and muscle cells do express the protein. Working with Gail Mandel, who has since moved to the Vollum Institute in Portland, Oregon, Ballas generated a new MeCP2 antibody that is more efficient than the commercially available one. Armed with this tool, she and colleagues showed that MeCP2 is indeed present, if at lower levels than in neurons, in primary cultures of rat glia and in protein isolated from rat optic nerve, which contains glia but no neuronal cell bodies.
Ballas and colleagues cultured wild-type mouse hippocampal neurons with a feeder layer of wild-type astrocytes or astrocytes from a mouse model for Rett syndrome. In a dish with Rett astrocytes, the neurons extended fewer processes than did neurons in the control cultures. Many neurons with Rett astrocytes also died; after six days of co-culture, the Rett dishes contained one-third fewer nerve cells than the control dishes. Similarly, up to four-fifths of wild-type neurons cultured in astrocyte-conditioned media had stunted dendrites after growing in media from Rett glia, suggesting a soluble factor is responsible for the neuron pathology.
The neurons could be affected by the presence of a toxic factor, or the absence of a nutrient they normally get from astrocytes. To distinguish between these two possibilities, the scientists mixed half wild-type and half Rett astrocyte-conditioned medium. This situation is likely to more closely match the conditions in the brain of people with Rett syndrome because the gene is X-linked; due to X-inactivation, wild-type protein is expressed in half of an affected girl’s cells. Again, dendrites were affected. This suggested to Ballas and Mandel that astrocytes carrying mutant MeCP2 secrete something that poisons neurons, as astrocytes may do in ALS, although the identity of such a factor remains to be discovered. If indeed MeCP2-deficient astrocytes produce a neurotoxic substance, then that pathway could be an appealing target for pharmaceutical intervention, Mandel said. Mandel discussed the research in an interview with the Rett Syndrome Research Trust.
“It certainly comes as a huge surprise to hear that [the MeCP2 mutation] may, in fact, be acting in the glia only, or in the glia in addition to the neurons,” said Ben Barres of Stanford University in Palo Alto, California, who called the results “dramatic.” Next, Ballas is making a mouse model with MeCP2 selectively downregulated in glial cell populations. “These phenotypes are so strong, it would be really surprising if she didn’t see the same phenotype in the mouse,” Barres said.
In ALS, as well, motor neuron death is non-cell autonomous. Studies have shown that wild-type neurons die at the hands of astrocytes overexpressing mSOD1, a gene that, when mutated, causes approximately one-fifth of inherited ALS cases. Using the enzyme cre under glial-specific promoters to selectively delete lox-flanked mSOD1 from astrocytes and microglia of mouse models, scientists from the University of California, San Diego, laboratory of Don Cleveland showed previously that disease progresses more slowly when the glia lack mSOD1 (Boillée et al., 2006; Yamanaka et al., 2008).
Knowing this, Christian Lobsiger, first author of the PNAS paper, and Cleveland and colleagues wondered if the same would be true of Schwann cells. These cells envelope the axons, putting them right where the action is, Lobsiger said. “The most important component, the ‘wire’ that actually links to the muscle…is insulated by Schwann cells,” he said.
Lobsiger, now at INSERM in Paris, France, and colleagues crossed lox-flanked SOD1-G37R mice with mice carrying cre controlled by the Schwann cell-specific myelin-protein-zero promoter. SOD1-G37R levels in the sciatic nerve tissue of the resulting mice were down by two-thirds. “I thought it would have a strong protective effect,” Lobsiger said. He was wrong: while the single mutant SOD1-G37R mice exhibited complete hind-limb paralysis and were sacrificed at 13.5 months, the double mutants reached this point after a year. Though he was blinded to which mice were which, Lobsiger noticed the phenotype during the experiment: “Some of these mice, they just dropped dead. They had a very, very rapid disease progression compared to other mice in the cage,” he said.
To explain these surprising results—the opposite of what the authors expected—they note that the G37R SOD1 mutant protein retains its dismutase activity, and when expressed in Schwann cells, may actually protect nerves by mopping up free radicals. The obvious next question, then, is what happens when dismutase-inactive mSOD1 is downregulated in Schwann cells? According to the author’s hypothesis, dismutase-inactive SOD1 in Schwann cells would not exert any protective effect, and deleting it should not accelerate disease. Cleveland’s lab attempted to make that mouse line, but it took too long to develop ALS-like symptoms, making it useless for experiments, Lobsiger said. Raymond Roos, of the University of Chicago in Illinois, is working on a similar experiment, but said, “It’s too early to say what’s what.” If the hypothesis were shown to be correct, it would imply that anti-SOD1 therapeutics should be specifically targeted to avoid downregulating SOD1 in the periphery where it may do some good, Lobsiger suggested.
“Until recently, glia have not been thought to be interesting cells,” Mandel said, but that is no longer the case. “The scientific community is getting more and more interested in understanding glial biology.”—Amber Dance.
Lobsiger CS, Boillée S, McAlonis-Downes M, Khan AM, Feltri ML, Yamanaka K, Cleveland DW. Schwann cells expressing dismutase active mutant SOD1 unexpectedly slow disease progression in ALS mice. PNAS Early Edition 2009 Feb 27. Abstract
Ballas N, Lioy DT, Grunseich C, Mandel G. Non-cell autonomous influence of MeCP2-deficient glia on neuronal dendritic morphology. Nat Neurosci 2009 Mar;12(3):311-317. Abstract