Summary: Researchers have completed the most comprehensive analysis to date of genomic, epigenomic, and transcriptomic alterations in the brains of people with Alzheimer’s disease. By examining over 2 million cells from roughly 400 postmortem samples, the team mapped detailed molecular changes that underlie the disorder.
These studies reveal a complex interplay between genetic and epigenetic processes that together drive Alzheimer’s pathology, and they point to new molecular and cellular targets that could expand therapeutic strategies beyond current approaches.
Key Facts:
- The analysis covers more than 2 million cells from about 400 postmortem brain samples, offering the broadest genomic survey of Alzheimer’s patients to date.
- Results highlight an intricate interaction of genetic risk factors and epigenomic alterations that reinforce one another in disease progression.
- Specific inhibitory neuron subtypes in the prefrontal cortex were linked with cognitive resilience and may represent promising targets for preserving cognitive function.
Source: MIT
Alzheimer’s disease affects more than 6 million people in the United States, and few FDA-approved treatments slow its progression.
Aiming to identify new therapeutic targets, teams at MIT have carried out the most extensive cell-by-cell investigation so far of how gene expression, chromatin accessibility, and 3D genome organization change across cell types during Alzheimer’s disease.
Using more than 2 million cells from over 400 postmortem brain samples, the investigators tracked how gene expression becomes dysregulated as Alzheimer’s advances and how epigenomic marks that control gene activity are altered. Combining these datasets produced an unprecedented molecular atlas of the disease across cell types and stages.
The results are published today in a set of four papers in Cell, led by Li-Huei Tsai, director of MIT’s Picower Institute for Learning and Memory, and Manolis Kellis, professor of computer science at MIT CSAIL and a Broad Institute member.
“We combined computational and experimental expertise to take an unbiased, large-scale look at Alzheimer’s across hundreds of individuals — something that hadn’t been attempted at this scale,” Kellis says.
Together, the studies indicate that genetic risk factors and epigenomic defects feed into one another, creating cascades that produce many of the disease’s cellular and molecular phenotypes.
“Alzheimer’s is a multifactorial disease,” Tsai adds. “These papers point to a converging picture in which affected neurons exhibit defects in their three-dimensional genome organization, and those defects drive many downstream disease features.”
A complex interplay
Most therapeutic efforts have focused on amyloid plaques, but the MIT team explored alternative mechanisms by determining which cell types are most vulnerable and which molecular pathways lead to neurodegeneration. To do this, they performed transcriptomic and epigenomic profiling on 427 brain samples from the Religious Orders Study/Memory and Aging Project (ROSMAP), a long-running cohort that has monitored cognitive and motor changes since 1994. The cohort included 146 individuals without cognitive impairment, 102 with mild cognitive impairment, and 144 with Alzheimer’s-linked dementia.
The first Cell paper used single-cell RNA sequencing to profile 54 distinct brain cell types and identify the cellular functions most disrupted in Alzheimer’s. Prominent deficits emerged in genes tied to mitochondrial function, synaptic signaling, and protein complexes that preserve genome structure. Lipid metabolism pathways were also highly perturbed — a finding consistent with prior work showing that the APOE4 genetic risk allele disturbs lipid processing and triggers wider cellular dysfunction.
By comparing people who retained cognitive abilities despite some amyloid buildup (so-called cognitive resilience) with those who developed dementia, the researchers found that resilient individuals had larger populations of two specific inhibitory neuron subtypes in the prefrontal cortex. In Alzheimer’s patients, those neuron subsets appear particularly vulnerable to degeneration. These results nominate defined inhibitory neuron populations as potential targets for therapies aimed at maintaining cognition.
Epigenomics
A second Cell paper examined epigenomic alterations in a subset of 92 individuals (48 healthy and 44 with early or late-stage Alzheimer’s). Using single-cell ATAC-seq to measure chromatin accessibility, and integrating those data with single-cell RNA profiles, the team linked changes in gene expression with alterations in regulatory DNA accessibility. This approach allowed the researchers to map gene regulatory circuits that control cell functions such as synaptic communication.
They found that many Alzheimer’s-associated genes are most highly expressed in microglia, the brain’s resident immune cells. The study also showed that virtually every brain cell type undergoes epigenomic erosion during disease progression — a loss of the normal patterns of accessible chromatin that help define cell identity.
The role of microglia
A third Cell paper focused on microglia, which comprise about 5–10 percent of brain cells and play roles in debris clearance, immune responses, and neuronal support. Building on earlier work showing that many Alzheimer’s GWAS variants are most active in immune cells, the team classified microglia into 12 transcriptional states based on RNA sequencing of hundreds of genes.
As Alzheimer’s progresses, microglia shift toward more inflammatory states and away from homeostatic states that support normal brain function. The researchers identified transcription factors that maintain the homeostatic state and are now testing whether activating those factors can reprogram inflammatory microglia back to supportive roles — a potential therapeutic strategy to curb neuroinflammation and protect neuronal function.
DNA damage
The fourth Cell study explored how accumulated DNA damage contributes to Alzheimer’s. Prior work from the Tsai lab showed that neurons generate double-stranded DNA breaks during memory formation and normally repair them, but the repair process becomes error-prone with age. This study demonstrated that accumulating DNA damage impairs repair fidelity, producing genome rearrangements, defects in 3D genome folding, and gene fusions that dysregulate synaptic genes.
“If a genome is repeatedly broken and imperfectly repaired, the cell will make mistakes and create rearrangements,” says researcher Vishnu Dileep. Those structural errors and fusion events preferentially disrupt genes involved in synaptic activity, providing a mechanistic link to cognitive decline.
The authors suggest that enhancing neurons’ DNA repair capacity could be a promising angle for slowing disease progression. Kellis’ lab also plans to apply artificial intelligence tools — including protein language models, graph neural networks, and large language models — to search for drugs that target key genes identified in these studies.
The teams have made their genomic and epigenomic datasets publicly accessible so other researchers can explore, visualize, and analyze them. “We want the research community to use this resource,” Kellis says.
Funding: The research received support from the National Institutes of Health and the Cure Alzheimer’s Foundation CIRCUITS consortium.
About this genetics and Alzheimer’s disease research news
Author: Sarah McDonnell
Source: MIT
Contact: Sarah McDonnell – MIT
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Human Microglial State Dynamics in Alzheimer’s Disease Progression” by Li-Huei Tsai et al. Cell
Abstract
Human Microglial State Dynamics in Alzheimer’s Disease Progression
Highlights
- Single-nucleus transcriptomes and epigenomes of human microglia at scale
- Microglial states mapped with disease-stage specificity in Alzheimer’s progression
- Chromatin accessibility alone does not fully capture microglial transcriptional diversity
- Transcription factor networks drive microglial states and transitions
Summary
Altered microglial states influence neuroinflammation and neurodegeneration, yet their dynamics have been incompletely characterized.
This work reports 194,000 single-nucleus microglial transcriptomes and epigenomes from 443 human subjects representing diverse Alzheimer’s pathological phenotypes. The authors define 12 microglial transcriptional states, including homeostatic, inflammatory, and lipid-processing states that are dysregulated in AD.
They identify 1,542 AD-differentially expressed genes, covering both microglia-state-specific and disease-stage-specific alterations. By integrating epigenomic, transcriptomic, and motif datasets, the study infers upstream regulators, gene-regulatory networks, enhancer-gene links, and transcription-factor-driven state transitions.
The researchers show that forced expression of predicted homeostatic-state activators induces homeostatic features in human iPSC-derived microglia-like cells, while blocking inflammatory activators can inhibit progression to inflammatory states.
Finally, the analysis maps expression of AD-risk genes across microglial states and documents differential regulation of those genes during disease progression, providing an unprecedented resource for understanding microglial contributions to Alzheimer’s disease.