Over 6 million Americans suffer from Alzheimer’s disease, and there aren’t many FDA-approved therapies that can halt the disease’s progression.
The most comprehensive analysis of the genomic, epigenomic, and transcriptomic alterations that take place in all cell types in Alzheimer’s patients’ brains has been carried out by MIT researchers in the hopes of finding new targets for possible Alzheimer’s treatments.
In order to examine how gene expression is altered as Alzheimer’s disease advances, the researchers used over 2 million cells from over 400 postmortem brain samples. Additionally, they monitored alterations in the epigenomic modifications of cells, which aid in identifying the genes that are active or inactive in a given cell. When combined, these methods provide the most comprehensive image of Alzheimer’s disease’s genetic and molecular causes to date.
In four papers that are published in Cell today, the researchers present their findings. The directors of MIT’s Picower Institute for Learning and Memory, Li-Huei Tsai, and Manolis Kellis, a computer science professor in the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and a member of the Broad Institute of MIT and Harvard, oversaw the studies.
The objective of the project, according to Kellis, was to combine biological and computational expertise to conduct an objective study of Alzheimer’s disease on a never-before-seen scale involving hundreds of people.
According to the research, the disease’s pathological manifestations are caused by a combination of epigenetic and genetic alterations that interact with one another.
Tsai states, “It’s a multifactorial process.” “By combining various methods, these studies suggest a convergent picture of Alzheimer’s disease in which many of the observed disease phenotypes are caused by defects in the 3D genome of the affected neurons.”
A complex interplay
The amyloid plaques that form in the brains of Alzheimer’s patients have been the focus of numerous attempts to create medications for the condition. By examining the disease’s molecular causes, the cell types that are most susceptible, and the underlying biological pathways that underpin neurodegeneration, the MIT team aimed to identify additional potential strategies in their latest series of investigations.
Towards that aim, 427 brain samples from the Religious Orders Study/Memory and Aging Project (ROSMAP), a longitudinal study that has monitored memory, motor, and other age-related changes in older adults since 1994, were subjected to transcriptome and epigenomic analyses by the researchers. 146 persons without cognitive impairment, 102 with mild cognitive impairment, and 144 with dementia linked to Alzheimer’s disease were included in these samples.
Using single-cell RNA-sequencing to examine the gene expression patterns of 54 different types of brain cells from these samples, the researchers in the first Cell paper focused on changes in gene expression and identified the cellular functions most impacted in Alzheimer’s patients. Among the most notable were deficiencies in the expression of genes related to protein complexes required to preserve the structural integrity of the genome, synaptic signaling, and mitochondrial function.
The genetic pathways linked to lipid metabolism were also found to be severely disrupted in this gene expression study, which was headed by former MIT postdoc Hansruedi Mathys, graduate student Zhuyu (Verna) Peng, and former graduate student Carles Boix. The Tsai and Kellis labs demonstrated in work published in Nature last year that the strongest genetic risk for Alzheimer’s, known as APOE4, disrupts normal lipid metabolism, which can subsequently result in abnormalities in numerous other cell processes.
In the Mathys-led study, the researchers also examined the differences in gene expression patterns between individuals with cognitive impairments and those without, including some who maintained their cognitive abilities even in the face of a certain amount of amyloid buildup in their brains—a condition known as cognitive resilience. According to that analysis, the prefrontal cortex of cognitively resilient individuals contained larger populations of two subsets of inhibitory neurons. These cells seem to be more susceptible to neurodegeneration and cell death in individuals with dementia associated with Alzheimer’s disease.
According to Mathys, “this finding implies that particular inhibitory neuron populations may be critical for preserving cognitive function even in the face of Alzheimer’s pathology.” “Our study identifies these particular subtypes of inhibitory neurons as a critical area for further investigation and may help in the creation of therapeutic interventions targeted at maintaining cognitive function in older adults.”
Epigenomics
The researchers in the second Cell paper, which included 48 healthy individuals and 44 patients with early or late-stage Alzheimer’s disease, looked at some of the epigenomic changes that occurred in 92 people. They were led by former MIT postdoc Xushen Xiong, graduate student Benjamin James, and former graduate student Carles Boix PhD ’22. Changes in the chemical modifications or packaging of DNA that impact how a specific gene is used in a given cell are known as epigenomic changes.
The researchers employed a method known as ATAC-Seq, which assesses the accessibility of sites throughout the genome at single-cell resolution, to quantify those alterations. Through the integration of this data with single-cell RNA-sequencing data, the researchers established a connection between gene expression levels and gene accessibility. Additionally, they might begin to classify genes into regulatory circuits that govern particular cell functions, like synaptic communication, which is the main means by which neurons communicate with one another and with the brain as a whole.
By employing this method, the scientists monitored alterations in epigenomic accessibility and gene expression that transpire in genes that have been connected to Alzheimer’s disease in the past. Along with identifying the cell types most likely to express these disease-linked genes, they also discovered that many of them are mostly expressed by microglia, the immune cells in charge of removing debris from the brain.
This study also showed that as Alzheimer’s disease advances, all types of brain cells experience a phenomenon called epigenomic erosion, which results in the loss of the cells’ typical pattern of accessible genomic sites and ultimately leads to cell identity loss.
The role of microglia
In a third Cell paper, research scientist Matheus Victor and graduate student Na Sun of MIT led the team that concentrated mostly on microglia, which account for 5–10% of brain cells. These immune cells not only remove debris from the brain but also react to injury or infection and facilitate neuronal communication.
This study expands upon a 2015 paper by Tsai and Kellis, who discovered that a greater number of the GWAS variants linked to Alzheimer’s disease are primarily active in immune cells, such as microglia, than in neurons or other brain cells.
Using RNA sequencing, the researchers in the new study were able to classify microglia into 12 distinct states, each of which was determined by the different levels of expression of hundreds of genes. It was also demonstrated that an increased number of microglia become inflamed as Alzheimer’s disease advances. Additionally, prior research from the Tsai lab has demonstrated that increased brain inflammation leads to a breakdown of the blood-brain barrier and impaired neuronal communication.
Simultaneously, the Alzheimer’s brain has fewer microglia in a state that supports homeostasis and aids in regular brain function. In an effort to treat Alzheimer’s disease by teaching inflammation-inducing microglia to return to a homeostatic state, the Tsai lab is currently investigating methods to activate transcription factors that the researchers identified as turning on the genes that maintain microglia in that homeostatic state.
DNA damage
In the fourth Cell study, which was directed by Boix and Vishnu Dileep, research scientists at MIT, the scientists looked at the role that DNA damage plays in the onset of Alzheimer’s disease. Previous research from Tsai’s lab has demonstrated that neuronal DNA damage can manifest long before symptoms of Alzheimer’s disease do. A contributing factor in this damage is the numerous double-stranded DNA breaks that neurons produce during the formation of memories. These breaks are quickly fixed, but as neurons deteriorate, the repair mechanism may malfunction.
This fourth study discovered that neurons find it more difficult to repair DNA damage as it builds up, which results in 3D folding defects and genome rearrangements.
According to Dileep, “when neurons sustain a lot of DNA damage, the cells make mistakes that cause rearrangements in their attempt to piece the genome back together.” “I like to compare it to a picture; if there is one crack in it, it is easy to put back together, but if the image breaks and you try to piece it back together, you will make mistakes.”
These errors in repair also give rise to a phenomenon called gene fusion, which is the result of gene rearrangements that cause dysregulation of the genes. These alterations seem to primarily affect genes linked to synaptic activity in addition to defects in genome folding, which may be a factor in the cognitive decline associated with Alzheimer’s disease.
According to the researchers, the results suggest that one strategy to slow down the progression of Alzheimer’s disease might be to find ways to improve neurons’ capacity for DNA repair.
Additionally, Kellis’ lab is now hoping to find drugs that might target some of the important genes that the researchers identified in these studies by using artificial intelligence algorithms like large language models, graph neural networks, and protein language models.
Additionally, the researchers hope that their genomic and epigenomic data will be useful to other scientists. Kellis declares, “We want the world to use this data.” “We have established digital repositories that enable users to engage with the data, retrieve it, view it, and perform real-time analysis.”
The National Institutes of Health and the Cure Alzheimer’s Foundation CIRCUITS consortium provided some funding for the study.