MIT researchers reveal why certain brain cells are vulnerable in Alzheimer’s disease (AD) and how some individuals retain cognition. Analyzing over 1.3 million cells from six brain regions, they link loss of Reelin-positive neurons to decline, and astrocyte programs to resilience. Open-access data provide a resource for understanding AD and developing therapies.
A recent study from researchers at the Massachusetts Institute of Technology (MIT) provides new evidence on how specific cells and neural circuits become vulnerable in Alzheimer’s disease (AD), while also exploring factors that may help some individuals maintain cognitive abilities despite clear pathological signs of the disease.
Published online in Nature on July 24, 2024, the study entitled “Single-cell multiregion dissection of Alzheimer’s disease” compares gene expression across multiple brain regions in individuals with or without AD and validates key findings through laboratory experiments.
By analyzing over 1.3 million cells from six brain regions of 48 donors, the study reveals detailed differences in cellular activity associated with cell type, brain region, disease pathology, and cognitive performance, offering potential targets for interventions to preserve cognition and memory.
Cellular Vulnerability and the Role of Reelin
To highlight potential targets for maintaining cognition and memory, researchers compared gene expression across multiple brain regions of individuals with or without Alzheimer’s disease and conducted laboratory experiments to test and validate their main findings.
Although all brain cells share the same DNA, their identities and activities differ based on how these genes are expressed. This new study measured gene expression differences in more than 1.3 million cells across over 70 cell types in six brain regions from 48 tissue donors, including 26 individuals who died from Alzheimer’s disease and 22 who died from non-AD causes. As a result, this study provides unique, extensive, and detailed data revealing differences in brain cell activity in AD patients across cell types, brain regions, disease pathology, and cognitive assessments performed before death.
Li-Huei Tsai, the study’s co-corresponding author and Picower Professor of Neuroscience at MIT, said, “Specific brain regions in Alzheimer’s patients are vulnerable, so we need to understand how these regions or particular cell types become susceptible. The brain contains not only neurons but many other cell types. How these cells respond differently based on location is a fascinating area that we are only beginning to study.”
Manolis Kellis, co-corresponding author and head of MIT’s Computational Biology Group, likened the single-cell RNA analysis used to measure gene expression comparisons to a more advanced “microscope.”
Kellis explained, “Under a conventional microscope, Alzheimer’s appears as amyloid plaques and phosphorylated tau tangles, but our single-cell ‘microscope’ can tell us, cell by cell and gene by gene, the thousands of subtle yet important biological changes induced by pathology. Linking this information to patients’ cognitive states can reveal relationships between cellular responses and cognitive decline or maintenance, guiding new approaches to treating cognitive loss.”
Neuronal Vulnerability and Reelin Protein
The research team analyzed brain regions including the prefrontal cortex, entorhinal cortex, hippocampus, anterior thalamus, angular gyrus, and middle temporal cortex.
They found that one type of excitatory neuron in the hippocampus and four types of excitatory neurons in the entorhinal cortex were markedly reduced in AD patients. Patients with these cell losses performed significantly worse on cognitive assessments. Many vulnerable neurons are interconnected within a common neural circuit and either directly express a protein called Reelin or are influenced by Reelin signaling.
These findings highlight particularly vulnerable neurons that share a neural circuit and a molecular pathway, whose loss is associated with cognitive decline.
Professor Li-Huei Tsai emphasized the significance of Reelin in Alzheimer’s research, noting a recent study of a Colombian man with a rare Reelin gene mutation that enhanced Reelin protein activity. Despite a strong family history of early-onset AD, he maintained healthy cognition into old age. This new study further indicates that loss of Reelin-producing neurons is associated with cognitive decline.
To further validate their findings, researchers directly examined human brain tissue samples and two AD mouse models. As expected, results showed that Reelin-positive neurons in the entorhinal cortex were reduced in both humans and mice.
Astrocytes, Cholinergic Metabolism, and Cognitive Resilience
To identify factors allowing cognition to persist despite pathology, researchers investigated which genes, cells, and brain regions were most closely associated with cognitive resilience—defined as residual cognitive function exceeding the typical expected decline given observed pathology.
The analysis revealed a surprising result: across multiple brain regions, astrocytes expressing genes related to antioxidant activity, cholinergic metabolism, and polyamine biosynthesis were significantly associated with sustained cognitive ability, even in the presence of high levels of tau and amyloid proteins.
These results further support prior findings by Tsai and Susan Lundqvist, suggesting that dietary cholinergic supplementation can help astrocytes cope with lipid dysregulation induced by the APOE4 variant, the most significant AD risk gene.
Additionally, antioxidant studies identified a molecule, spermidine, potentially suitable as a dietary supplement with anti-inflammatory properties. However, this association requires further research to confirm causality.
Advanced Analytical Methods and Open Data
To analyze the massive single-cell dataset, researchers developed a robust new approach based on co-expressed gene clusters, or “gene modules,” which leverages the correlated expression patterns of functionally related genes within the same module. Kellis explained, “In theory, 1.3 million cells expressing 20,000 genes could form astronomical combinations. In practice, we observe only a small subset of coordinated changes. By identifying these patterns, we infer more reliable changes because they are based on multiple genes within the same functional module.”
He likened this to the human body: although the body has many joints capable of complex movements, coordinated actions like walking, running, or dancing involve far fewer combinations. This method allows scientists to identify coordinated gene expression programs as functional gene clusters.
While the Kellis and Tsai labs have already reported several notable findings from this dataset, they anticipate more discoveries await exploration. To facilitate further research, they have released user-friendly analysis and visualization tools along with the dataset on Kellis’ lab website.
Looking ahead, researchers are investigating regulatory circuits associated with differentially expressed genes to understand genetic variations, regulatory factors, and other drivers. These elements could potentially be modulated to reverse pathogenic circuits in specific brain regions, cell types, and stages of disease progression.