Cell by cell
Researchers release a single-cell atlas examining changes in gene expression in Alzheimer’s disease beyond neurons
SINGAPORE—An international collaboration between Duke-NUS Medical School and Monash University has resulted in a comprehensive online atlas of gene expression changes in the different types of brain cells associated with Alzheimer’s disease. Typically, the focus in this disease has been on neurons, but this analysis has shown that more questions could be answered by looking into other types of brain cells as well.
“Limited information has been available about how individual cell types in the brain contribute to Alzheimer’s disease,” said study co-senior author Assistant Professor Owen Rackham, from Duke-NUS’ Cardiovascular and Metabolic Disorders (CVMD) program. “Although various genes have been implicated in Alzheimer’s disease, we do not know which cell types harbor these differences in gene expression.”
For this work, single-nucleus RNA sequencing known as DroNCSeq was applied to samples from both healthy brains and those with Alzheimer’s disease, providing a total of 13,214 nuclei. Cells were taken from the entorhinal cortex, which is part of the medial temporal lobe and hippocampal memory system in the brain, connecting the hippocampal and neocortex sections of the brain. These structures are involved with processing and storing memories, as well as our ability to perceive the passing of time and anticipate events in the future. A paper titled “On areas of transition between entorhinal allocortex and temporal isocortex in the human brain. Normal morphology and lamina-specific pathology in Alzheimer’s disease,” had suggested that the first pathological changes of Alzheimer’s disease take place in the second layer of the entorhinal cortex. This section of the brain has also been associated with temporal lobe epilepsy and potentially schizophrenia.
In this latest work, the researchers’ screening identified subpopulations of cells that are only present in the brains of Alzheimer’s disease patients, in addition to common and distinct networks of genes and functions that are co-regulated across a variety of cell types.
The cell types examined included microglia, astrocytes, neurons, oligodendrocytes, oligodendrocyte progenitor cells (also known as oligodendrocyte precursor cells) and endothelial cells. As for the genes, the team looked in-depth at the expression of more than a dozen.
Enrico Petretto, associate professor at Duke-NUS’ CVMD program and co-senior author of the study, explained, “Our research sought to explore whether the answer to treating Alzheimer’s lies in understanding how non-neuronal cells are affected during the disease. Using DroNCSeq, we were able to study differences in gene expressions at single-cell resolution, which is key to understanding how genes identified by genome-wide association studies in specific cell subpopulations are associated with Alzheimer’s disease.”
In a previous paper published on this research—“A single cell brain atlas in human Alzheimer’s disease,” which can be found on Cold Spring Harbor Laboratory’s bioRxiv server—the authors reported that “a recent study of regional cell density in the human brain revealed that approximately 40 percent of cortical grey matter is composed of neurons, whereas neurons only comprised ~15 percent of the hippocampal formation, which included entorhinal cortex, and this proportion was further reduced to approximately 5 percent in AD.” These numbers highlight the importance of exploring Alzheimer’s disease’s impact on cell types beyond neurons.
The researchers separated Alzheimer’s patient cells from the controls, and identified marker genes for each of the cell types mentioned above: “HLA-DRA, CX3CR1, C1QB, and CSF1R for microglia; AQP4, SLC1A2 for astrocytes; SYT1, glutamate receptors (i.e., GRIK2, GRIA1 and GRIN2B), RBFOX1 for neurons; MOBP, MBP and PLP1 for oligodendrocytes; PCDH15 and MEGF11 for OPCs; and FLT1 and CLDN5 for endothelial cells.”
Interestingly, when looking at clusters of genes with coordinated gene expression differences, they found that some had high populations of genes linked to cell stress response.
“For instance, both DEG5 and DEG7 clusters of highly and co-ordinately upregulated genes in AD brains were enriched for genes involved in responses to topologically incorrect protein and cell stress responses, including a number of mitochondrial, heat shock and chaperone genes (i.e., MT-ND1-4, MT-CO2, MT-CO3, MT-ATP6, HSPA1A, HSP90A1, DNAJA1 Fig. 1f, g), consistent with previous reports in SH-SY5Y cells and primary mouse neurons,” the authors reported. “It is possible that the cell-independent responses to topologically incorrect protein may be responses to extracellular amyloid deposition, as tau is only intraneuronal and there is evidence of plaque and oligomer interactions with multiple brain cell types.
“Our analyses highlighted that, in addition to the well-studied cell-type-specific responses (i.e., inflammatory responses of astrocytes and microglia in AD, oxidative stress in neurons in AD 22–24), there is a further coordinated response which may act to boost molecular chaperone levels to protect cells against protein misfolding and which may provide additional therapeutic avenues aimed at enhancing endogenous cell-type independent responses.”
The team intends to further explore this work by looking into identified genes for druggable targets. In addition to the Nature Neuroscience paper, the researchers also created an online resource so that interested parties can examine their dataset. (This interactive website can be found at http://adsn.ddnetbio.com.)