Joint first authors John Mitchell and Praveen Paul led the study with principal investigator Jackie de Belleroche. In a whole-genome screen, they discovered the mutation in a three-generation family affected by ALS. D-amino oxidase (DAO) oxidizes D-serine, a co-factor in activation of NMDA-type glutamate receptors. The researchers found that mutant DAO has reduced activity. Expressing the mutant in cultured cells caused cell death, although the exact mechanism is not yet known. This study, implicating synaptic NMDA receptors, fits in with the hypothesis that excitotoxicity from too much glutamate mediates cell damage in ALS (reviewed in van Damme et al., 2005).

De Belleroche has been working with families affected by ALS for more than two decades, collecting samples to screen for genes as well as offering information to patients about the disease. Even before the Internet, de Belleroche kept families apprised of new research with newsletter updates. DNA samples from these families were part of the analyses that identified TDP-43 and FUS as ALS genes (see ARF related news story on Sreedharan et al., 2008 and ARF related news story on Vance et al., 2009).

Mutations in superoxide dismutase-1 (SOD1), TDP-43, and FUS account for perhaps one-quarter of inherited ALS cases, but scientists have also identified rarer mutations in ALS gene candidates such as VAPB, alsin, senataxin, FIG4, and angiogenin, as well as disease-associated loci of as-yet-unknown genes (reviewed in Pasinelli and Brown, 2006). Only one-tenth of people with ALS inherit the condition; the majority of cases are sporadic.

Genomewide association screens often use single nucleotide polymorphisms, which require a large cohort to find a significant marker. De Belleroche, working with 200 families, instead looked for microsatellites, short tandem repeats of a sequence such as “CA.” Since the number of repeats is variable, microsatellites come not just in the four nucleotide flavors that can fill a single spot, but a dozen or more varieties depending on their length. This makes microsatellites a poor choice for large, automated sequencing studies, but useful for smaller studies, or when SNPs may not be present near the gene of interest.

Having located a disease-associated microsatellite in the q arm of chromosome 12 in a single pedigree, the researchers sequenced nearby genes to look for genetic variants. The DAO mutation, R199W, showed up in three affected family members, one obligate carrier and three members at risk to inherit ALS, but not in three unaffected members.

The authors note that the results do not unequivocally prove that DAO is the right gene; another mutation co-inherited with the microsatellite could conceivably be the real problem. However, the mutation was not present in 1,000 other samples of people with sporadic or familial ALS and unaffected controls. In addition, de Belleroche and colleagues are examining non-coding DAO variants, which might affect splicing, from other people with ALS.

D-serine, as a co-agonist for the synaptic NMDA receptor, is involved in activating neurons (reviewed in Wolosker et al., 2002). The enzyme serine racemase, found primarily in glia, synthesizes D-serine from L-serine. DAO, mainly an astrocyte protein, breaks the amino acid down by converting it from an amino to a keto acid. DAO is highly expressed in the cerebellum and spinal cord, just where you might expect to find a gene involved in motor neuron disease.

The R199W mutation is smack in the middle of DAO’s active site. To test DAO activity in ALS, the researchers obtained spinal cord samples from one family member who had died. Compared to DAO from control and sporadic ALS cases, this person’s DAO evinced lower enzymatic activity.

To further analyze the effects of the R199W mutation, de Belleroche and colleagues expressed mutant and wild-type DAO in a motor neuron cell line (NSC-32). They transduced each form of the gene into primary rat motor neuron cultures, and analyzed cell survival with the TUNEL assay. Cells that received mutant DAO had twice as many TUNEL-positive, dying cells as those with wild-type DAO. Since DAO is an astrocytic protein, the researchers also expressed it in astrocytes co-cultured with primary motor neuron cultures. Again, the cultures with the mutant gene had more TUNEL-positive cells than those with wild-type DAO.

De Belleroche proposes the following hypothesis for DAO and ALS: Young people carrying only one functional DAO gene have enough of the enzyme to mop up D-serine, keeping the NMDA receptors from going overboard. But some change—caused by aging, cells stress, or inflammation—increases serine racemase activity. If there is not enough DAO to get rid of the excess D-serine, it causes overactivity at the NMDA receptor, leading to neuron excitation that can become toxic.

The new results sync up nicely with a 2007 paper on D-serine in ALS (Sasabe et al., 2007). The authors of that study found that spinal cord D-serine and serine racemase levels are elevated in ALS model mice overexpressing mutant human SOD1, as well as in a few people with sporadic or familial ALS. Either increased serine racemase, as in the 2007 study, or decreased DAO activity, as in the current work, would boost D-serine. Both studies linked this extra D-serine to ALS.

Given that serine racemase levels normally increase with age, it is conceivable that D-serine is involved in many forms of ALS, not just the few cases where it is mutated. “Normally you would expect the wild-type DAO will have the capacity to deal with any sort of toxic insults,” de Belleroche speculated. “The accumulation of cell stress, aging, and so on, might just be the threshold that gets exceeded.”

How much science can be built off of a single family’s rare mutation? “It is one family, but it is a novel mechanism, and that is what is important,” said John Fink of the University of Michigan, who was not involved in the study. “This not only provides insight into a new molecular mechanism that can cause ALS, but also it highlights the importance of these D-amino acids in maintenance of the nervous system.”

The D-serine pathway has therapeutic potential, Fink said. The current data suggest it would be beneficial to somehow reduce D-serine levels or prevent the amino acid binding to the receptor. De Belleroche also pointed out that a DAO mutant mouse might be a useful model for ALS research, which sorely needs new models. She is already working on transgenic mice.—Amber Dance.

Reference:
Mitchell J, Paul P, Chen HJ, Morris A, Payling M, Faichi M, Habgood J, Panoutsou S, Winkler S, Tisato V, Hajitou A, Smith B, Vance C, Shaw C, Mazarakis ND, de Belleroche J. Familial amyotrophic lateral sclerosis is associated with a mutation in D-amino acid oxidase. Proc Natl Acad Sci U S A. 2010 Apr 5. Abstract