14 April 2010. A genetic variant near the gene for MTHFD1L on chromosome 6 likely doubles a person’s risk of late-onset Alzheimer disease, according to research presented today at the American Academy of Neurology meeting in Toronto, Canada. The discovery, from a large genomewide association screen, jibes well with what is known about the protein’s role in metabolizing homocysteine, a known risk factor for Alzheimer’s. First author Adam Naj and principal investigator Margaret Pericak-Vance, both of the University of Miami, Florida, outlined the finding in a poster session.
The study authors hunted for AD-linked genes in a genomewide association study encompassing nearly 500,000 SNPs and including more than 2,000 AD subjects, and over 3,000 controls in the combined discovery and confirmation datasets. They found that SNPs in and around the MTHFD1L gene were linked to AD risk, with the disease-associated variants doubling risk. In comparison, the ApoE4 allele increases AD risk by a factor of 3.68. However, the presenters noted that the odds ratio of two is likely to go down a bit with further study.
MTHFD1L encodes a protein, methylenetetrahydrofolate dehydrogenase (NADP+ dependent) 1-like, that may help convert homocysteine into methionine, so problems with this gene might lead to elevated homocysteine levels. Increased plasma homocysteine is a well-known AD risk factor (see ARF related news story on Seshadri et al., 2002). In addition, MTHFD1L variants have been linked to coronary disease (Coronary Artery Disease Consortium, 2009) and dementia can be caused by reduced blood flow to the brain, so the link among MTHFD1L, Alzheimer’s, and coronary disease is compelling, Pericak-Vance said.
Other studies have not found a significant association between SNPs near MTHFD1L AD (see AlzGene), although some GWAS have come close. Pericak-Vance and Naj suggested that their study met stringent statistical GWAS thresholds because of the particular SNPs they analyzed, and the size of the cohort.
The authors speculated that carriers of the MTHFD1L disease-linked variant are unable to properly convert homocysteine to methionine, allowing homocysteine to build up. This molecule could then cause disease by altering vasculature, increasing oxidative stress, or poisoning cortical neurons. However, Naj noted that it is too soon to be sure what happens in people with different MTHFD1L variants. “We have no evidence to go on as to what it actually does,” he told ARF. As with any GWAS, it will be necessary to look for the significance of this gene in other datasets to confirm its importance. The researchers are currently sequencing MTHFD1L and testing plasma homocysteine levels in control and AD samples.—Amber Dance.