ARTICLE IN BRIEF
✓ Two separate research groups conducted whole-genome scans of patients with ALS and healthy controls, identifying genes for cytoskeleton and neural adhesion as major determinants of risk for sporadic ALS.
Two new whole-genome scans of patients with amyotrophic lateral sclerosis (ALS) point to genes for cytoskeleton and neural adhesion as major determinants of risk for sporadic ALS. The two studies were presented back to back at the 17th International Symposium on Amyotrophic Lateral Sclerosis/Motor Neuron Disease, held in Yokohama, Japan, in December.
Two separate research groups conducted the studies. The first was led by Bryan Traynor, MD, at the National Institute of Mental Health (NIMH), and John Hardy, PhD, of the National Institute on Aging (NIA). The second was led by Dietrich Stephan, PhD, director of neurogenomics at the Translational Genomics Institute (TGen), a nonprofit research institute in Phoenix, AZ.
Although the details of the methods used differed, the two studies were similar in both purpose and broad outline. The specific genes they obtained also differed, but both found that variations in actin cytoskeleton genes are likely to be a major source of risk leading to development of ALS in those without a family history.
“Sporadic ALS is a complex genetic disorder,” Dr. Stephan told Neurology Today, “with multiple subtly predisposing genes that have to come together, along with environmental exposures, to begin the disease process.” Because of this complexity, only a whole-genome association study is likely to turn up good candidate genes for increasing the risk of sporadic ALS, he said.
Both groups used silicon chips carrying thousands of indexed gene fragments, with known variations called SNPs, or single nucleotide polymorphisms (see sidebar “Chips, SNPs, and HapMaps”).
Chips are exposed to patient or control DNA, and any differences in binding patterns between the two, which result from or indicate differences in DNA sequence, appear as different colors on the chips. The precise location of these differences is read by a computer, and correlated with the related gene.
While “gene chips” have been available for almost a decade, it has only been within the last two years that enough information could be packed onto a single chip to allow the entire genome to be scanned. Three years ago, only 400 genes could be assessed. Two years ago, that number rose to 10,000. Today, it stands at 500,000, which provides sufficient density to give an independent report on every gene in the human genome.
“Only at this density can we tease apart these complex diseases,” Dr. Stephan said. Publication of the human haplotype map (“HapMap”) was also critical, because it provides a fine-scale outline of normal human variation across the entire genome.
ACTIN CYTOSKELETON GENES STAND OUT
While still short of independently sequencing the entire genome of each person, this kind of study gives an extraordinarily detailed picture of the pattern of gene variations in patients with sporadic ALS. Unlike typical Mendelian disorders like Huntington disease, in which a single gene is at fault, complex diseases like ALS may involve many genes.
In Dr. Stephan's study, ALS patients as a group had approximately 10 percent more of certain gene variants than did controls. No one variant is likely to be found in all cases; instead, each is “subtly predisposing,” he said, and a constellation of the variants adds up to an increased risk for the disease.
Dr. Stephan's study, which was funded in part by the Muscular Dystrophy Association, examined DNA from 1,251 ALS patients from research centers across the country, along with 1,000 neurologically normal controls. The 400 SNPs identified as most promising in the first round were re-examined, to give a set of 42 highly significant genes associated with sporadic ALS.
The study by Dr. Traynor and Dr. Hardy, funded by the NIMH and NIA, the Packard Center at Johns Hopkins University, and the ALS Association, examined 276 patients and 276 controls. The group identified about a dozen loci of interest, most of which were associated with known genes.
Both groups found that genes responsible for building and maintaining the actin cytoskeleton, the framework that gives shape to the cell, were strongly represented among the mix. Given the pathogenesis of ALS, this makes sense, Dr. Traynor said in an interview. “Motor neurons have to produce huge axons, so scaffolding becomes important to their function.”
While the theoretical importance of the actin cytoskeleton to ALS has long been recognized, this is the first study to implicate specific genes in patients with sporadic ALS. Curiously, the exact genes found by the two groups differed, probably because they used different methods and different patient groups, and also because so many actin-related genes may be involved.
One gene identified by Dr. Traynor's group was formin2, which also has been implicated with infertility in women. More work will need to be done to determine how this and the other genes found alter risk for the disease, and the magnitude of the increased risk conveyed by each.
No other major themes have emerged from the NIMH/NIA study, Dr. Traynor said, but that is likely to change as they and others expand on this work with more patients. “One important aspect of our study is that all of our raw genotype data are being made available publicly. Other researchers can mine this data, and we hope others will add patients to it. It doesn't end with us.” As more patients are added, he said, so grows the power to find new genes.
A PRIMARY FAILURE OF NEURONAL ADHESION?
From the TGen study, another set of genes appears to be involved in neural cell adhesion. “This leads us to believe there may be a problem with the motor neuron sticking to the neuromuscular junction,” Dr. Stephan said. “We know neurons disconnect and retract in ALS. This is usually thought to be a primary defect in the cell body, resulting from a sick neuron not doing its job. We hypothesize that instead the neuron is fine.” Rather, he suggested, the back-and-forth signaling that keeps the muscle and nerve connected is lost, leading to loss of cell adhesion and retraction of the axon.
There is a precedent for this in embryogenesis, Dr. Stephan pointed out, where many more spinal cord neurons extend axons to muscle than are finally connected. “Whether the cell bodies die at that point is debatable,” he said. If they don't, and if instead they are merely dormant, it may be possible to reawaken them, and coax them to reconnect with the target muscle.
Dr. Stephan hopes to use the results from this study to begin to identify drug candidates that can strengthen the neuron-muscle bond, and perhaps also trigger an emergence from dormancy of quiescent neurons.
“I don't claim to fully understand the disease yet,” Dr. Stephan said, but he suggested that other proposed disease-related mechanisms, such as immune activation and free radical formation, “will all be downstream of this primary defect” in neuromuscular connection.
“This isn't a cure, but it is the first step toward knowledge-based therapeutics. I think we have finally gleaned a major insight into the causes of ALS. This isn't about false hope. We have more information that we did nine months ago,” he said. “I firmly believe we can stop the progression of the disease.”
‘AN ENORMOUS STEP FORWARD’
According to Stephan Pulst, MD, this kind of study, which analyzes a very large number of DNA variants, “is an enormous step forward. One can hope that this is the first step towards understanding the genetic contributions to sporadic ALS.” Dr. Pulst is professor of medicine and neurobiology at the David Geffen School of Medicine at the University of California-Los Angeles.
Dr. Pulst pointed out that neither study examines another source of genetic variation, namely variation in gene copy number, but both groups have the resources to examine this potentially very important type of genetic variation. Risk of Parkinson disease, for instance, is increased by an increase in number of alpha-synuclein genes.
Finally, he said, “It is curious that none of the DNA variants appears to be shared between the two study groups. This may point to considerable genetic complexity of sporadic ALS, and the need for additional studies.” Those studies may be aided by the public availability of the NIMH/NIA data.
CHIPS, SNPS, AND HAPMAPS
In recent years, DNA microarrays (“gene chips”) have become a key tool in analyzing the vast amounts of data available in the human genome. In the investigation of disease, microarrays are typically used to ask two major types of questions: how does gene expression differ between patients and controls; and how does the genotype differ between patients and controls? The current studies examine genotype differences between ALS patients and controls.
Expression differences are detected by exposing the microarray to RNA products from patients or controls, thereby revealing which genes differ most in their activity – either more or less – between the two. Genotype differences are detected by exposing the microarray to DNA from patients or controls. The results indicate where and how the genetic sequence of the two differs.
The microarray for a genotyping experiment is embedded with short DNA fragments containing SNPs – single nucleotide polymorphisms – from across the entire human genome. SNPs (pronounced “snips”) have become the most important kind of landmark in the genome map, because they represent a very fine-scale set of “signposts,” and because they are polymorphic (they differ in sequence) to some extent between individuals.
SNP microarrays detect variations within both genes and non-gene sequences. Such variations may be associated with differences in gene activity, but genotyping studies, including the current ones, cannot determine if that is the case.
More than 12 million SNPs have been identified, far more than can currently be embedded in one microarray. But the Human HapMap shows that, because SNPs are clustered and are often inherited as a block, or haplotype, a far smaller set captures almost all the variation within the human genome. In the current studies, chips with 500,000 SNPs were used, which is believed to encompass more than 90 percent of the variation within the genome. (For more information about the International HapMap Project, see Neurology Today, “International HapMap Project Utilizes a ‘Fast-Track’ Approach in Genomics Research,” Dec. 5; pages 26–27).