Researchers have identified potential Alzheimer’s mutations by focusing their attention on people whose biomarkers reach extreme ends of the spectrum. As reported in the August 22 PLoS Genetics, scientists from Washington University in St. Louis, Missouri, found variants both known and novel when they sequenced major AD genes in people with very high or very low amounts of tau and amyloid β in their cerebrospinal fluid (CSF). Led by senior author Carlos Cruchaga, the researchers were surprised to find one of the risk variants was a polymorphism in presenilin 1 (PS1) that had previously been deemed non-pathogenic. When they considered this variant in the context of ApoE4, they found it conferred as much risk as a second copy of that allele. Someone carrying a single copy of both the PS1 variant and ApoE4 has 10 times the risk of someone with wild-type presenilin and no ApoE4, the researchers calculated. As scientists hunt for rare variants, it is important to consider that mutations may still be risk factors even if they exist in people without AD, Cruchaga said.

First author Bruno Benitez and co-authors combined two previously successful approaches. First, they focused on known AD genes, where disease-linked variants were likely: ApoE, amyloid precursor protein (APP), presenilins 1 and 2 (PS1, PS2), progranulin (GRN), and tau (MAPT; see ARF related news story). Second, to raise their chances of finding disease-linked variants with a minimum sample size, the team picked people who were at the extremes of AD biomarkers, those in the top or bottom 15 percent for Aβ, tau or phospho-tau in CSF. The team had previously found a known PSEN1 variant by focusing on people at the highs and lows of amyloid-beta levels (see ARF related news story on Kauwe et al., 2007).

In the current work, the researchers included data and DNA from subjects collected by the Alzheimer’s Disease Research Center (ADRC) at Washington University and the Alzheimer’s Disease Neuroimaging Initiative (ADNI). They selected subjects based only on the CSF biomarkers, without considering case or control status, or how far cases were in their disease progression when the CSF samples were taken. People with low Aβ and high tau levels tended to be cases, Cruchaga said, while those with high amyloid and low tau were more likely to be controls.

Eventually, the researchers plan to sequence DNA from all their subjects, Cruchaga said. But picking out the CSF extremes allowed them to take a shortcut by identifying the people most likely to have interesting genetic variants. “The approach is novel,” commented Rita Guerreiro of University College London in the U.K., who was not involved with the study. She added that few research groups possess both the DNA samples and the CSF data to perform such a project.

Other researchers are similarly selecting subjects to boost their chances of finding AD genes. For example, Guerreiro and colleagues are focusing their efforts on cases and controls who have already reached autopsy, so their disease status is certain. And in a recent paper, geneticists used a Caribbean Hispanic population who were highly inter-related to identify the loci of potential AD genes (ARF related news story).

Cruchaga and colleagues started with DNA samples from 212 volunteers. They amplified and sequenced the six known AD genes and identified variants in each. Then they returned to the individual DNA samples, plus an additional 522 from the same ADRC and ADNI sources, and genotyped all for those specific variants. This pooled approach saved time and money compared to sequencing all six genes in each sample individually, Cruchaga said.

The researchers identified 27 variants that associated with biomarker extremes: two in ApoE, four in GRN, five in MAPT, three in PSEN1 and four in PSEN2. The list included PSEN1 alanine-426-proline, which is known to cause AD (Poorkaj et al., 1998). Novel variants, whose role in disease remains uncertain, included two in APP, three in PSEN2, and one each in GRN, MAPT, and PSEN1

Cruchaga and colleagues were most interested in the previously known variant PSEN1 glutamate-318-glycine. While some studies have reported it segregates with AD in families (Albani et al., 2007, Taddei et al., 2002), others have found it in elderly controls, as well (Dermaut et al., 1999, Mattila et al., 1998), leading the field to consider it non-pathogenic. As Cruchaga and colleagues found, the role of PSEN1-E318G only became clear when they considered it in conjunction with ApoE. In the presence of ApoE4, the PSEN1 mutation turned out to be a major risk factor.

In the study dataset, PSEN1-E318G was associated with high tau levels, a sign of AD. However, it also appeared in controls, confirming it does not necessarily cause the disease. The researchers also examined the mutation’s effects in two prospective cohorts, the Rush Memory and Aging Project and the Religious Orders Study, both being conducted at Rush University, Chicago, Illinois, with people with and without AD. In addition to having ten times higher risk for AD, people with a single ApoE4 allele plus PSEN1-E318G declined faster cognitively than non-carriers who had AD. When these carriers reached autopsy they also had more amyloid-beta plaques in their brains. Cruchaga said that this is the only PSEN1 variant he knows of to interact with ApoE4 in this fashion, but many of the other AD variants are quite rare and have not been studied in this kind of detail.

Cruchaga and colleagues tested their results in an independent Washington University series of 1,855 cases and controls, and in a cohort of 4,085 people from the Genetic and Environmental Risk for Alzheimer’s Disease (GERAD) Consortium. The risk of AD due to ApoE4 plus PSEN1 E318G remained ten times that of normal. “It is very convincing,” said Ekaterina Rogaeva of the University of Toronto in Canada, who was not involved in the study. Guerreiro commented that she always likes to see results replicated by an independent group.

Now, the Washington University team has turned its attention to a possible mechanism for PS1-E318G’s effect on AD risk. Rogaeva speculated that E318G might work similarly to fully pathogenic PS1 mutations, but have a milder effect. Guerreiro, in contrast, suggested it might have a unique mechanism, because some studies have shown E318G raise Aβ42 levels (Batelli et al., 2008); if anything, it decreases them (Albani et al., 2007). The mutation sits near an RNA splice site, Cruchaga noted, so the group is checking if it affects PS1 transcript levels.

Scientists may not follow up on mutations like PS1-E318G because they do not definitively cause disease, but that may mean missing out on the role of important risk factors, Cruchaga suggested. Though this particular variant is rare and confers risk in conjunction with ApoE4, its effect is stronger than that of common variants coming out of genome-wide association studies, he said. He suggests geneticists should not discount a variant completely if it shows up in a few controls. Whether variants are pathogenic often comes up for debate. Guerreiro has proposed a rating scale for pathogenic versus non-pathogenic mutations (see Alzforum webinar on Guerreiro et al., 2010). In that scale, she and her colleagues were careful to note that they were only distinguishing causative mutations from those that do not always cause AD, and did not address risk factors. A non-pathogenic mutation may still be a risk factor.

The work of Cruchaga and colleagues is “an attractive approach to identify novel and high-risk gene variants,” commented Kaj Blennow of the University of Gothenburg in Mölndal, Sweden, in an email to Alzforum. Rogaeva suggested it would be interesting to sequence other AD genes, such as TREM2 and SORL1, in the same cohort. She speculated the cohort may turn up variants that reduce risk, and others that amplify it.—Amber Dance


No Available Comments

Make a Comment

To make a comment you must login or register.


News Citations

  1. Late- and Early Onset AD—Not So Distant at Genetic Level
  2. The Value of Biomarkers—Diagnosis and Genetic Screens
  3. Racing Toward New Alzheimer’s Mutations with Runs of Homozygosity

Webinar Citations

  1. Weeding Mendel’s Garden: Can We Hoe Dubious Genetic Associations?

Paper Citations

  1. . Extreme cerebrospinal fluid amyloid beta levels identify family with late-onset Alzheimer's disease presenilin 1 mutation. Ann Neurol. 2007 May;61(5):446-53. PubMed.
  2. . Missense mutations in the chromosome 14 familial Alzheimer's disease presenilin 1 gene. Hum Mutat. 1998;11(3):216-21. PubMed.
  3. . Presenilin-1 mutation E318G and familial Alzheimer's disease in the Italian population. Neurobiol Aging. 2007 Nov;28(11):1682-8. PubMed.
  4. . Association between presenilin-1 Glu318Gly mutation and familial Alzheimer's disease in the Australian population. Mol Psychiatry. 2002;7(7):776-81. PubMed.
  5. . The Glu318Gly substitution in presenilin 1 is not causally related to Alzheimer disease. Am J Hum Genet. 1999 Jan;64(1):290-2. PubMed.
  6. . The Glu318Gly mutation of the presenilin-1 gene does not necessarily cause Alzheimer's disease. Ann Neurol. 1998 Dec;44(6):965-7. PubMed.
  7. . Early-onset Alzheimer disease in an Italian family with presenilin-1 double mutation E318G and G394V. Alzheimer Dis Assoc Disord. 2008 Apr-Jun;22(2):184-7. PubMed.
  8. . Genetic screening of Alzheimer's disease genes in Iberian and African samples yields novel mutations in presenilins and APP. Neurobiol Aging. 2010 May;31(5):725-31. Epub 2008 Jul 30 PubMed.

External Citations

  1. ApoE
  2. APP
  3. PS1
  4. PS2
  5. progranulin
  6. tau
  7. TREM2
  8. SORL1

Further Reading


  1. . Fine mapping of genetic variants in BIN1, CLU, CR1 and PICALM for association with cerebrospinal fluid biomarkers for Alzheimer's disease. PLoS One. 2011;6(2):e15918. PubMed.
  2. . Cerebrospinal fluid APOE levels: an endophenotype for genetic studies for Alzheimer's disease. Hum Mol Genet. 2012 Oct 15;21(20):4558-71. Epub 2012 Jul 20 PubMed.
  3. . Locus-specific mutation databases for neurodegenerative brain diseases. Hum Mutat. 2012 May 11; PubMed.
  4. . PSEN1 Mutation Carriers Present Lower Cerebrospinal Fluid Amyoid-β42 Levels than Sporadic Early-Onset Alzheimer's Disease Patients but no Differences in Neuronal Injury Biomarkers. J Alzheimers Dis. 2012 Jan 1;30(3):605-16. PubMed.
  5. . Alzheimer's disease risk variants show association with cerebrospinal fluid amyloid beta. Neurogenetics. 2009 Feb;10(1):13-7. Epub 2008 Sep 24 PubMed.
  6. . Mechanism of gamma-secretase cleavage activation: is gamma-secretase regulated through autoinhibition involving the presenilin-1 exon 9 loop?. Biochemistry. 2004 May 25;43(20):6208-18. PubMed.
  7. . Is the presenilin-1 E318G missense mutation a risk factor for Alzheimer's disease?. Neurosci Lett. 2000 Jan 7;278(1-2):65-8. PubMed.

Primary Papers

  1. . The PSEN1, p.E318G Variant Increases the Risk of Alzheimer's Disease in APOE-ε4 Carriers. PLoS Genet. 2013 Aug;9(8):e1003685. PubMed.