14 Jun 2019

Does the head honcho of genetic Alzheimer’s risk occlude contributions of smaller genes? In the June 10 JAMA Neurology, researchers led by Lindsay Farrer at Boston University exposed some hidden variants that have languished in the shadow of ApoE. The researchers stratified, by ApoE genotype, the whole exome sequencing data from the Alzheimer’s Disease Sequencing Project and replication cohorts. This data had been vaguely underwhelming when it was first analyzed, but stratification by family history identified some new risk variants. Now, the scientists uncovered some other new variants whose sway over a person’s risk of developing late-onset AD was either dependent on, or masked by, ApoE4. The new findings include variants in ISYNA1 that may protect ApoE4 carriers, and one in GPAA1 that doubled risk in noncarriers.

Last year, Farrer and colleagues unveiled findings from the largest whole exome sequencing (WES) study to date. It used some 11,000 whole exome sequences in the ADSP data set and found two new AD risk variants (Aug 2018 news). Since then, the researchers have been analyzing the ADSP data in different ways to tease out more information. At the AD/PD meeting in Lisbon, Farrer reported five risk genes that emerged when he limited analysis to people with a strong AD family history. A separate study looked for variants predicted to nix protein function; it unearthed NOTCH3 (Apr 2019 conference news). These eight rare variants contribute little to overall AD risk in the general population but, all the same, they offer clues about disease pathways. So far, Farrer noted, genetic studies point to six broad pathways in AD: inflammation, lipid metabolism, vesicular trafficking, neuronal signaling, Aβ, and tau.

In this study, first author Yiyi Ma and colleagues divvied the data by ApoE genotype. They sought variants that modulate the AD risk attributable to the E4 allele, as well as variants that affect risk independently of E4. Their discovery data set included 7,358 ApoE4 noncarriers, of whom 3,145 had been diagnosed with AD and 4,213 were healthy controls, plus 2,377 AD patients and 706 controls who carried at least one copy of ApoE4.

Initial WES analysis pulled out 22 variants with suggestive associations with AD risk, including 12 in ApoE4 carriers, and 10 in noncarriers. Among the latter, two loci reached genome-wide significance, one in the gene for ACO99552, a long noncoding RNA, and one in GPAA1, which encodes glycosylphosphatidylinositol (GPI) anchor attachment 1 protein. In ApoE4 carriers, multiple hits within single genes combined to indicate gene-level association with AD for OR8G5, SLC24A3, and IGHV3-7; however, these associations did not repeat in subsequent analysis of other cohorts because the primary SNPs were not found. Even so, Farrer thinks IGHV3-7, which encodes immunoglobulin heavy variable chain, is notable, because variants in other members of this gene family have been tied to increased AD risk. A protective variant in the ISYNA1 gene encoding inositol-3-phosphate synthetase-1 did associate with AD in the discovery and two of three replication data sets.

For replication, the researchers ran a meta-analysis of their discovery cohort and more than 10,000 AD cases and 12,000 controls from the Alzheimer’s Disease Exome Sequencing-France Project, the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium, and the Alzheimer’s Disease Genetics Consortium (ADGC).

Among ApoE4 carriers, the top SNP in ISYNA1, rs2303697, protected in all cohorts except CHARGE. Several other protective polymorphisms in this gene nearly reached statistical significance in the meta-analysis. The protein, inositol-3-phosphate synthetase-1, converts glucose-6-phosphate to myoinositol-1-phosphate. Brain myoinositol concentration has been correlated to tau pathology and cognitive impairment, and ApoE4 carriers reportedly have higher myoinositol levels than noncarriers (Mullins et al., 2018; Waragai et al., 2017; Voevodskaya et al., 2016). Scyllo-inositol, aka ELND005, which purportedly reduces myoinositol levels in the brain, did not pass muster in AD clinical trials.

In ApoE4 noncarriers, rs138412600, the GPAA1 variant found in the discovery data set, also reached significance in the replication cohorts. GPI anchor attachment 1 protein constitutes part of an enzyme complex that supports translational modification of GPI. In the meta-analysis, the GPAA1 minor A allele nearly doubled AD risk. The variant resides within the second exon of the gene, within a domain that binds FOXG1, a repressive transcription factor previously linked to Rett syndrome (Mitter et al., 2018). 

To examine the link between GPAA1 and AD risk, the researchers turned to data from the Religious Orders Study/Rush Memory and Aging Project. In ROSMAP, rs138412600 associated with poorer global cognition. ROSMAP transcriptome data indicated that the variant also associated with higher expression of FOXG1, and lower expression of GPAA1. Expression varied the most among ApoE4 noncarriers, suggesting ApoE genotype somehow modulates functional effects of this SNP. How the GPAA1 variant doubles AD risk and how ApoE genotype contributes remain to be investigated, Farrer said.

In ApoE4 noncarriers, the researchers also pulled out known variants in TREM2 and MAPT that were only nominally associated with AD risk in E4 carriers. They also discovered novel candidates in NSF, a gene encoding N-ethylmaleimide-sensitive factor. Because the NSF gene lies near MAPT, the researchers could not nail it down as a bona fide independent risk gene. While the ACO99552 lncRNA variant appeared to increase AD risk by 88-fold in the discovery analysis, Farrer cautioned that it was too rare—occurring in just 10 people in the discovery cohort and none in the replication cohorts—to draw conclusions about its relationship to disease risk.

“APOE so dominates the risk for AD that this is undoubtedly an approach which needs to be attempted,” commented John Hardy of University College London. “It is, however, very difficult to do, both because of the correspondingly small (n) in Apoe4-AD cases and in elderly Apo4+ controls,” he noted. He added that the error rate of AD diagnoses, upon which the AD risk associations rely, further complicates the findings. 

Farrer said that all variants identified in ADSP will be put to the test in larger, whole genome sequencing cohorts. The Alzheimer’s Disease Genomic Consortium aims to sequence the whole genomes of at least quadruple the number of people included in the current ADSP WES analysis, Farrer said. “We are in the middle of an ongoing saga of filling in the genetic architecture of AD, which clearly is not accounted for by common variants.”—Jessica Shugart

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Largest AD Whole-Exome Study to Date Finds Two New Risk Genes 17 Aug 2018
2. At AD/PD Conference, New Alzheimer’s Genes Reinforce Known Pathways 19 Apr 2019

Therapeutics Citations

1. ELND005

Paper Citations

1. Mullins R, Reiter D, Kapogiannis D. Magnetic resonance spectroscopy reveals abnormalities of glucose metabolism in the Alzheimer's brain. Ann Clin Transl Neurol. 2018 Mar;5(3):262-272. Epub 2018 Jan 29 PubMed.
2. Waragai M, Moriya M, Nojo T. Decreased N-Acetyl Aspartate/Myo-Inositol Ratio in the Posterior Cingulate Cortex Shown by Magnetic Resonance Spectroscopy May Be One of the Risk Markers of Preclinical Alzheimer's Disease: A 7-Year Follow-Up Study. J Alzheimers Dis. 2017;60(4):1411-1427. PubMed.
3. Voevodskaya O, Sundgren PC, Strandberg O, Zetterberg H, Minthon L, Blennow K, Wahlund LO, Westman E, Hansson O, Swedish BioFINDER study group. Myo-inositol changes precede amyloid pathology and relate to APOE genotype in Alzheimer disease. Neurology. 2016 May 10;86(19):1754-61. Epub 2016 Apr 15 PubMed.
4. Mitter D, Pringsheim M, Kaulisch M, Plümacher KS, Schröder S, Warthemann R, Abou Jamra R, Baethmann M, Bast T, Büttel HM, Cohen JS, Conover E, Courage C, Eger A, Fatemi A, Grebe TA, Hauser NS, Heinritz W, Helbig KL, Heruth M, Huhle D, Höft K, Karch S, Kluger G, Korenke GC, Lemke JR, Lutz RE, Patzer S, Prehl I, Hoertnagel K, Ramsey K, Rating T, Rieß A, Rohena L, Schimmel M, Westman R, Zech FM, Zoll B, Malzahn D, Zirn B, Brockmann K. FOXG1 syndrome: genotype-phenotype association in 83 patients with FOXG1 variants. Genet Med. 2018 Jan;20(1):98-108. Epub 2017 Jun 29 PubMed.

Further Reading

No Available Further Reading