Experiments in a new animal model of amyotrophic lateral sclerosis suggest that the pathogenesis of genetic and sporadic forms of the disease may be fundamentally similar.
DR. SERGE PRZEDBORSKI: “I think this is a great step forward. The main question addressed here, which was acutely missing from our study, was whether a toxic factor was operating in humans, rather than just rodents, and in sporadic, rather than just familial, ALS. We never addressed that, and it is what Dr. Kaspar brings to the field.”
Most patients with amyotrophic lateral sclerosis (ALS) have the sporadic form of the disease, but all disease models to date have mimicked the much rarer genetic forms, raising questions about their relevance for most ALS patients. Now, a new model of the disease, reported in the Aug. 10 online edition of Nature Biotechnology, indicates that astrocytes in both sporadic and genetic forms secrete a soluble factor that is toxic to motor neurons, suggesting the pathogenesis of the two forms of the disease may be fundamentally similar.
Created from neural precursors in the spinal cord of ALS patients, the model is also a proof of principle that should open the way to modeling sporadic forms of other neurodegenerative diseases, including Parkinson disease and Alzheimer disease. Modeling sporadic diseases has been notoriously hard, because, by definition, there is no known genetic trigger that can be introduced into animals or cells.
A growing body of research has begun to implicate astrocytes as a prime cause of motor neuron death in ALS, and astrocytes appear to secrete a still-unknown toxic factor in a genetic cell culture model of the disease. To try to determine whether astrocytes in sporadic ALS were also doing so, Brian Kaspar, PhD, principal investigator at the Center for Gene Therapy at Nationwide Children's Hospital in Columbus, OH, turned to neural progenitor cells (NPCs) isolated from patient spinal cords at autopsy.
He chose NPCs for two reasons. First, their normal role in development is to produce astrocytes, along with other cell types of the CNS, and thus NPCs are a renewable source of cells. While induced pluripotent stem cells, derived from patient skin fibroblasts, are easier to obtain and work with, they have not yet shown any disease-related phenotype when prompted to make either astrocytes or motor neurons. Therefore, the second reason for choosing NPCs was the hope they might display an ALS phenotype.
Dr. Kaspar first induced the NPCs to become astrocytes, then co-cultured them with normal mouse motor neurons carrying a fluorescent reporter gene. He found that after 96 hours, neurons in contact with the astrocytes from sporadic ALS patients began to degenerate, just as those derived from patients with mutant superoxide dismutase 1 (SOD1) did. The interaction was specific, as there was no effect on GABAergic neurons, and no effect from fibroblast-derived astrocytes.
As previously observed in rodent models of familial ALS, motor neurons exposed to media from astrocytes, with no direct contact between the cells, also died, indicating astrocytes secreted a toxic soluble factor into the environment. The nature of the factor is unknown. Gene expression profiling indicated that both sporadic and familial astrocytes upregulated a suite of inflammatory cytokine genes, a change previously seen in rodent models. “Astrocytes really upregulate their toxic patterns when they see motor neurons,” Dr. Kaspar said. “The million dollar question is why are these astrocytes so angry?”
DR. JEFFREY ROTHSTEIN expressed some caution about the model. While the cells have been differentiated toward astrocytes, “they are not quite the same” as true spinal cord astrocytes, he said. Nonetheless, “The key thing is, they saw a phenotype in sporadic ALS cells. Whether that is a true observation or somehow derived from how the cells were cultured is unclear,” and will require further work.
Remarkably, whatever the ultimate trigger for their response, it is latent in the NPCs themselves, something that is not the case for skin fibroblasts. “I think what this is telling us is that we're deriving cells from the worst possible situation,” he said. “The patient has died from this disease, and we are taking the cells at a time when the environment is most hostile. Is the toxicity programmed in during the disease, or is it genetically present already? That we do not know. It could be either, but it is certainly very intriguing that the cells have this memory of toxicity.”
Perhaps equally remarkable, Dr. Kaspar found that suppressing expression of SOD1 in the astrocytes rescued the motor neurons, despite the fact that the gene for the protein was entirely normal and was expressed at normal levels. The results, he said, confirm that wild-type SOD1 should be considered a therapeutic target in sporadic ALS. Previous work has suggested that the normal protein may be misfolded in the disease, and that this misfolding may be part of the disease cascade.
“I think this is a great step forward,” said Serge Przedborski, MD, PhD, professor of neurology and pathology at Columbia University in New York, one the investigators who first described a toxic soluble factor from astrocytes in ALS. “The main question addressed here, which was acutely missing from our study, was whether a toxic factor was operating in humans, rather than just rodents, and in sporadic, rather than just familial, ALS. We never addressed that, and it is what Dr. Kaspar brings to the field.”
Jeffrey Rothstein, MD, PhD, professor of neurology and neuroscience at Johns Hopkins School of Medicine in Baltimore, expressed some caution about the model. While the cells have been differentiated toward astrocytes, “they are not quite the same” as true spinal cord astrocytes, he said. Nonetheless, “the key thing is, they saw a phenotype in sporadic ALS cells. Whether that is a true observation or somehow derived from how the cells were cultured is unclear,” and will require further work. He noted that it may be that the ability of NPCs to harm motor neurons has more to do with their long exposure to the toxic environment in the ALS spinal cord, rather than anything intrinsic to their derivation from ALS patients per se. He suggested it might be appropriate to see whether astrocytes from other chronic diseases, such as Alzheimer disease or even cancer, are also toxic.
DR. WALTER KOROSHETZ: “I think the most exciting thing is that they were able to get neuroprogenitor cells out of autopsy tissue. The fact that it worked is some of the biggest news. That could have implications for a lot of different diseases, where we don't know what the genetic mutation is. If we can get the cells, and they have a phenotype related to the disease, that makes the human tissue the model, rather than the mouse.”
Dr. Przedborski had fewer reservations. “The bottom line is, they are human, they look like astrocytes, and they clearly reproduced the toxicity seen in rodents,” he said.
Dr. Przedborski noted that the molecular mechanism of that toxicity is still unknown, and is an active area of investigation in his lab, which is now likely to collaborate with Dr. Kaspar's lab to isolate and identify it. The work is proceeding “very well,” he said, but it is nonetheless “incredibly and unacceptably slow for patients.”
But, he said, “even if we fail to identify the ‘X factor’ fully, as long as we know the pathway engaged by it, we win nonetheless,” because drugs could be developed to interrupt the factor's effects at several nodes along the pathway. And while Dr. Przedborski doubts that this “X factor” is the only one involved in ALS pathogenesis, targeting its effects may nevertheless be sufficient to have a therapeutic effect on the disease.
For Walter Koroshetz, MD, deputy director of the NINDS, which helped fund the study, the greatest benefit of the paper may be in its proof of principle that sporadic diseases can be modeled using patient-derived cells. “I think the most exciting thing is that they were able to get neuroprogenitor cells out of autopsy tissue. The fact that it worked is some of the biggest news. That could have implications for a lot of different diseases, where we don't know what the genetic mutation is. If we can get the cells, and they have a phenotype related to the disease, that makes the human tissue the model, rather than the mouse.”
Dr. Kaspar noted: “This is the first opportunity to study patient-specific cells in a culture dish. Now the detective work comes into play for the basic biology. We've got the tools at hand to be able to really address some important mechanistic questions about what's going wrong with these astrocytes.”
Haidet-Phillips AM, Hester ME, Kaspar PK, et al. Astrocytes from familial and sporadic ALS patients are toxic to motor neurons. 2011; E-pub 2011; 2011 Aug. 10.