Geneticists are increasingly turning to endophenotypes, i.e., measurable disease traits, to help uncover genes that might be missed in heterogenous clinical cohorts. In the October 8 Nature Genetics, scientists led by David Fardo at the University of Kentucky, Lexington, described the largest GWAS yet to link genetic variants to neuropathological measures of dementia. The study of nearly 8,000 autopsied participants found four new risk loci: a microglial gene associated with tangles, two genes related to hardening of cerebral arteries, and an independent signal near the APOE locus that correlated with cerebral amyloid angiopathy. The study also tied 19 known Alzheimer’s genes to specific endophenotypes.

  • Neuropathology GWAS found four new genes linked to dementia endophenotypes.
  • Three associated with cerebrovascular disease, including a new locus near APOE.
  • The study also tied 19 known AD genes to specific pathologies.

Commenters praised the approach. “By linking specific genetic variants to distinct pathologies, the paper offers a more detailed understanding of AD’s genetic complexity and opens avenues for targeted therapeutic interventions,” Vivek Swarup at the University of California, Irvine, wrote to Alzforum. Michael Belloy at Washington University in St. Louis noted that the findings could help distinguish between genes that directly affect pathology, and those that instead help maintain cognition in the face of plaques and tangles. “The results will help future studies disentangle AD resistance genes from AD resilience or general dementia genes, which in turn should improve genetic risk counseling as well as genetically informed drug targeting,” he wrote.

Endophenotype GWAS are often limited by small sample sizes. First author Lincoln Shade overcame this by combining data from three different cohorts, comprising 5,940 participants from The National Alzheimer’s Coordinating Center, 1,183 participants in the Religious Orders Study and Rush Memory and Aging Project, and 681 from the Adult Changes in Thought study, an observational study in Seattle. Overall, 70 percent were clinically diagnosed with AD, 10 percent with other dementias, and 20 percent were cognitively unimpaired.

Shade correlated genetic and autopsy data for these 7,804 brains, focusing on 11 different endophenotypes. These comprised three measures of Alzheimer’s pathology (total plaques, CERAD score, tangles), five of cerebrovascular disease (CAA, microinfarcts, large infarcts, atherosclerosis, arteriolosclerosis), and three of non-Alzheimer’s pathologies (Lewy bodies, TDP-43, hippocampal sclerosis).

The GWAS turned up eight dementia genes, four known and four new. First, the known genes: APOE, BIN1, TMEM106B, and GRN. APOE was associated with all three AD pathologies, as well as CAA and TDP-43, while the AD gene BIN1 associated with CERAD and tangles, but not total plaques. The CERAD score measures dense-core neuritic plaques that also incorporate tau fibrils, leading the authors to speculate that BIN1’s associations were driven by tau pathology. Meanwhile, TMEM106B and GRN, which are linked to frontotemporal dementia as well as AD, correlated with TDP-43 and hippocampal sclerosis, but not with specific AD pathologies.

The new genes? A locus in the broader APOE region but close to the microglial apolipoprotein gene APOC2 was related to CAA, and microglial kinase PIK3R5 to tangles. Endothelial gene LZTS1 associated with arteriolosclerosis, a hardening of the small arterioles of the brain. For the endothelial gene, COL4A1, a component of collagen IV, atherosclerosis of large arteries provided the link. “The identification of loci such as COL4A1 and LZTS1 highlights the critical role of vascular health in AD,” Swarup noted.

A couple of previous studies had suggested the existence of a second risk locus near APOE, but the evidence had been inconclusive, the authors noted (Cervantes et al., 2011; Bellenguez et al., 2022). “The present study is the first to confirm that this association is independent of the known effects of APOE alleles,” the authors wrote. They could link the putative functional risk variant to hypomethylation of the APOC2 region, suggesting it would boost transcription.

In a separate analysis, the authors searched for associations between the 83 known AD genes and the 11 endophenotypes. This produced 15 more hits, with variants near ABCA7, CR1, FERMT2, INPP5D, PICALM, PTK2B, SNX1, SORL1, and ZCWPW1 associating with CERAD score, HLA-DQA1 and MME variants with tangles, PLCG2 and TPCN1 loci with microinfarcts, and IL34 and MAPT with hippocampal sclerosis. PICALM and TPCN1 were also correlated with CAA.

Julie Williams at Cardiff University, Wales, noted that tying genetic factors to neuropathology instead of clinical diagnosis could produce cleaner, less noisy relationships and offer clues to how each gene affects the brain. “There is much to follow up in identifying disease mechanisms,” she wrote.—Madolyn Bowman Rogers

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

Paper Citations

  1. . Genetic variation in APOE cluster region and Alzheimer's disease risk. Neurobiol Aging. 2011 Nov;32(11):2107.e7-17. PubMed.
  2. . New insights into the genetic etiology of Alzheimer's disease and related dementias. Nat Genet. 2022 Apr;54(4):412-436. Epub 2022 Apr 4 PubMed.

Further Reading

Primary Papers

  1. . GWAS of multiple neuropathology endophenotypes identifies new risk loci and provides insights into the genetic risk of dementia. Nat Genet. 2024 Oct 8; PubMed.