A brain gene-activity atlas spanning 1494 donors shows that Alzheimer's disease, Lewy body dementia, vascular dementia, and Parkinson's disease are molecularly closer to one another than to schizophrenia or bipolar disorder
A brain gene-activity atlas spanning 1494 donors shows that Alzheimer's disease, Lewy body dementia, vascular dementia, and Parkinson's disease are molecularly closer to one another than to schizophrenia or bipolar disorder
The PsychAD consortium built a single-cell atlas of gene activity in the prefrontal cortex: 6.3 million cell nuclei from tissues of 1494 donors, ranging from newborns to 108-year-olds, including healthy individuals and patients with eight brain diseases. The results were published on 23 September in Nature.
The dorsolateral prefrontal cortex supports working memory, planning, and the ability to hold several tasks in mind simultaneously, and it is affected by many of the eight diseases in the atlas. Panos Roussos, who led the project at the Icahn School of Medicine at Mount Sinai, explains the rationale: comparing different diseases in the same brain region makes it easier to relate the results both to one another and to existing genetic data.
Earlier single-cell studies examined one disease at a time in small cohorts and could not separate the background shared across many diseases from the molecular signature specific to a given diagnosis. The PsychAD authors took a different approach: they identified genes that change in the same way across all eight diseases (these genes govern basic processes of protein assembly and transport within cells) and subtracted this shared signal from the comparison. After that subtraction, Alzheimer's disease, Lewy body dementia, vascular dementia, and Parkinson's disease clustered together at the molecular level and stood apart from schizophrenia and bipolar disorder. For the same disease pairs, the authors independently calculated the fraction of shared genetic risk from genome-wide data and obtained a consistent result: the more genetic risk two diseases share, the more similar the cellular changes in the brains of their patients.
"Single-cell studies of the brain used to be limited to small numbers of individuals. Scaling to a population level changes the very type of questions you can ask," says Zhichao Miao, a computational biologist at the Guangzhou National Laboratory who was not involved in the project.
The atlas also draws a line between disease and normal aging. Vascular smooth muscle cells normally decline in number with age, yet in Alzheimer's disease their numbers increase. A similar pattern holds for layer II/III cortical neurons: Alzheimer's disease drives their loss, while ordinary aging does not. The authors confirmed this loss with a separate tissue-section experiment. The same consortium drew a comparable boundary in an aging atlas of the same cortex: after age 60, gene activity in glia (cells that support neurons) shifts sharply, and many of these genes are known Alzheimer's risk genes.
The authors constructed a causal chain: genetic risk promotes the accumulation of amyloid plaques (aggregates of protein deposited between neurons), plaques accelerate tau pathology (damage to the tau protein, which normally maintains the shape of neuronal processes but in disease aggregates into tangles inside the cell), and tau pathology leads to dementia. Somatostatin-expressing inhibitory neurons are statistically associated with reduced plaque accumulation. Expansion of microglia, the brain's resident immune cells, is statistically associated with suppression of precisely these protective neurons.
The brain's immune response changes over the course of the disease. In early stages it is protective: the innate immune response, including monocyte maturation, is statistically associated with resilience to dementia. In later stages, microglia accumulate lipid droplets, and this is statistically associated with disease acceleration. One exception is the response of vascular cells to the inflammatory signal IL-17: it is harmful already at early stages, and blocking IL-17 has already improved cognitive function in mice. This finding gives an existing class of drugs a basis for testing in Alzheimer's disease.