15 Dec 2023

The Alzheimer's disease pathological cascade unfolds in the brain over decades, amidst a menagerie of cell types reacting to change. How, and when, does each cell type contribute to the process? A study published November 30 in Nature Communications used a genetic approach to chip away at this question.

Researchers led by Hyun-Sik Yang of Brigham and Women’s Hospital in Boston deployed a cell-type-specific version of polygenic risk scores, by tallying the burden of AD risk variants near genes active in specific cell types. In an autopsy cohort spanning the disease spectrum, astrocyte-specific risk variants associated with amyloid plaques, while the burden of microglia-specific variants linked to plaques, tau tangles, and even cognitive decline. Among living, cognitively normal people in another cohort, astrocytic risk variants associated with early signs of cognitive impairment, while people with a higher microglial PRS had more Aβ plaques and tau tangles according to PET scans. In all, the findings point to when and how astrocytes and microglia play their parts in the AD cascade. However, because the study is limited to only the most cell type-specific genes, it would have missed associations between genetic variants that are active in several different cell types and other aspects of AD.

Alzforum reported initial findings from this work at AAIC 2021 (Aug 2021 conference news). Since then, the study's scope has expanded, both in terms of sample size and the number of AD endophenotypes examined. The published study also draws from a larger, newer GWAS for AD risk variants, and sets more stringent cutoffs to designate variants as cell-type specific.

The researchers tapped two well-characterized cohorts. One was the ROSMAP neuropathology cohort, the other the Anti-Amyloid Treatment in Asymptomatic AD (A4) trial, which boasts neuroimaging data from cognitively unimpaired volunteers. For each of the 1,457 ROSMAP and 2,291 A4 participants, the scientists calculated their overall AD PRSs by tallying the combined effects of risk variants they carried. These variants were derived from a GWAS, and included those that fell below the threshold for genome-wide significance (Apr 2022 news).

Next, the scientists referenced single-nuclei RNA sequencing data from a separate, neurologically normal subset of 24 ROSMAP participants to pick out the top 10 percent—or 1,343—most cell-type-specific genes for each of six major brain cell types. By tallying the AD risk imparted by variants within the neighborhood of these specific genes, they derived cell-type-specific PRSs for excitatory neurons, inhibitory neurons, astrocytes, microglia, oligodendrocytes, and oligodendrocyte precursor cells.

What came out of this analysis? In ROSMAP, the scientists explored connections between cell-type-specific PRS and multiple endophenotypes, such as autopsy-confirmed AD dementia, several different measures of Aβ accumulation, neurofibrillary tangle burden, and cognitive decline. A person's overall PRS, excluding ApoE, associated with his or her status on all of these endophenotypes. Their astrocytic PRS was tied strongly to both diffuse and neuritic Aβ plaque burden, while higher microglial or oligodendrocyte PRSs were linked to neuritic but not diffuse plaques. The authors interpreted this finding to mean that risk variation in astrocytes may influence amyloid deposition early on, while microglia and oligodendrocytes play more of a role in the development of mature plaques.

Microglia, oligodendrocyte, and astrocyte PRSs associated with tau tangles, with microglia doing so most strongly. Risk variants expressed in microglia and astrocytes contributed to cognitive decline. Perhaps unsurprisingly, a person's microglial PRS associated with a higher proportion of activated microglia in their postmortem brain, as gauged by assessments of microglial morphology in subset of ROSMAP participants.

What Each Cell Does. Statistical modeling hints at causal relationships between astrocyte- and microglia-specific polygenic risk (ADPRS) and Alzheimer's endophenotypes. Black arrows indicate phenotype-phenotype associations; red arrows indicate genotype-phenotype associations. DP=diffuse plaque, NP=neuritic plaque, NFT=neurofibrillary tangle. [Courtesy of Yang et al., Nature Communications, 2023.]

Focusing on the astrocyte and microglial PRS, the researchers ran a series of mediation analyses to estimate how each cell type might influence different aspects of the complex AD cascade (image above). The most likely model to emerge from this analysis suggests that risk variants in astrocytes affect AD pathology primarily through amyloid deposition, whereas microglia exert a broader impact, influencing Aβ and tau accumulation as well as cognitive decline. Notably, a person's microglial PRS influenced his or her cognitive decline above and beyond AD pathology, suggesting microglia could play a role in cognitive resilience to a given amount of pathology, Yang told Alzforum.

The researchers also studied this in living people. They hunted for associations between the various polygenic risk scores and brain imaging in the A4 study, using its baseline screening data. Here, only the microglial PRS significantly associated with amyloid-PET; risk variants in astrocytes and excitatory neurons just missed the statistical cutoff. Likewise, for tau-PET only the microglial PRS associated. This is at odds with findings from ROSMAP, where microglia, oligodendrocytes, and astrocytes associated with tangle burden. However, the researchers noted that tau deposition is likely at an early stage in the A4 cohort, most of whom were largely cognitively unimpaired when they enrolled. Interestingly, among cognitively normal people in ROSMAP, only microglial PRS associated with tau tangle burden.

Thus, despite important differences in the way tau tangles were measured, results from ROSMAP and A4 suggest a coherent biology, the authors wrote. “Microglia may exacerbate tau pathology starting from the preclinical stage of AD, while other cell types may contribute to tau pathology later in the symptomatic disease stages.”

None of the cell-type-specific PRS linked to hippocampal volume in A4, although oligodendrocyte PRS came close. The authors noted that the A4 study is likely underpowered to detect associations between neurodegeneration or cognitive impairment, neither of which were extensive among participants.

Curiously, only the astrocytic PRS was tied to poor scores on the preclinical Alzheimer’s cognitive composite (PACC), a test battery that picks up early inklings of cognitive impairment. This held regardless of a person's Aβ or tau-PET measures, suggesting astrocytes might directly influence early changes in cognition.

This conflicts with what Yang had presented at AAIC back in 2021, when he reported that the excitatory neuron PRS tied to poor PACC scores, suggesting a role for neuronal genes in cognition. Why the different result now? Yang said that it comes down to changes in the way cell-type-specific PRSs are calculated. Previously, the scientists had calculated PRSs based on the top 10 percent of transcripts found in a given cell type. By this measure, even transcripts that are expressed in several cell types could have made the cut. In the new analysis, the researchers limited their analysis to genes that were most specific to each cell type. Therefore, Yang said that genes highly expressed in neurons as well as some other cell types would have been excluded from the neuronal PRS. The findings don’t rule out a role for genetic variation in neurons in cognitive decline.

Yang believes the findings help elucidate what different cell types contribute to AD pathophysiology, and point to mechanisms involved. For example, the astrocyte contribution to early amyloid deposition may relate to these cell’s role in Aβ clearance at the blood-brain barrier. However, he does not think the scores should be used for diagnostic or trial screening purposes. This is because each cell-type-specific PRS explained no more than 3 percent of the variance in each AD endophenotype. This may seem shockingly low, but Yang said it was to be expected, particularly when keeping in mind that these PRS calculations do not include ApoE, and are considered independently of age.

Yang and colleagues plan to continue using these cell-type specific PRSs to understand how different cells contribute to levels of other endophenotypes, including fluid biomarkers such as p-tau and GFAP, he said.—Jessica Shugart

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Polygenic Scores Paint Microglia as Culprits in Alzheimer's 27 Aug 2021
2. Paper Alert: Massive GWAS Meta-Analysis Published 6 Apr 2022

Further Reading

No Available Further Reading