A Molecular Survey of Astrocytes in Alzheimer’s Disease
A comprehensive molecular survey was conducted to analyze astrocytes across five brain regions commonly affected by Alzheimer’s Disease (AD). This meticulous research utilized single-nucleus RNA sequencing (snRNA-seq) to delve into the differences in astrocyte behavior as influenced by pathology.
Figure 1 presents an experimental overview of our snRNA-seq study, developed using BioRender.com, highlighting the structured approach undertaken for this investigation. Panel B depicts UMAP visualization, showcasing the clustering patterns of nuclei that are negative for NEUN and OLIG2.
Panel C features violin plots that elucidate the expression levels of cell type-specific marker genes within NEUN−/OLIG2− nuclei, illustrating variation across the five examined brain regions. The subsequent results indicate a measurable Aβ plaque load, expressed through the percentage of immunoreactive area fraction, as well as the pTau/tau ratio determined via ELISA, taken from adjacent samples utilized in the snRNA-seq analysis across varying brain regions and stages of pathology.
To better understand the changes in astrocyte morphology in relation to Aβ plaques and pTau neurofibrillary tangles (NFTs), we quantified the local burden of Alzheimer’s neuropathological changes (ADNC) in nearby tissue samples. The assessment of Aβ-immunoreactive area fractions was performed through immunohistochemistry, alongside the measurement of the pTau/total tau ratio via ELISA.
The 32 donors were classified into four distinct pathology stages, informed by global quantitative measures of neuritic plaques, utilizing the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) neuritic plaque score and Braak NFT staging. This stratification resulted in: (1) no signs of AD/low ADNC burden, (2) moderate ADNC burden with sparse NPs, (3) high ADNC burden indicated by moderate or frequent NPs, and (4) severe ADNC burden with frequent NPs.
The analysis of Aβ plaque load revealed relative consistency across brain regions; however, elevated levels were observed within the prefrontal cortex (PFC) at advanced disease stages. Conversely, the pTau/tau ratio was significantly highest in the entorhinal cortex (EC) and followed a patterned hierarchy across stages: EC > ITG > PFC > V2 > V1, corroborating the stereotypical neuroanatomical progression of NFTs.
Table 1 summarizes the demographic and neuropathological characteristics of the study participants, including detailed quantitative measures of Aβ and pTau in Supplementary Data 1.
While numerous transcriptomics studies have characterized astrocytes in mice, the human aging brain’s astrocyte transcriptomic diversity remains underexplored. Eight control donors provided a rare opportunity to assess regional differences in astrocytic gene expression, with these individuals displaying minimal Aβ plaques and low pTau/tau ratios.
We integrated data from 246,464 nuclei, conducting a differential gene expression analysis across the brain regions. Notably, the EC and V1 exhibited the highest counts of differentially expressed genes (DEGs), emphasizing their unique transcriptomic profiles in comparison to other brain areas. Subsequent analyses of genes such as APOE, APP, and GJA1 revealed distinct patterns of upregulation and downregulation in these regions, further confirmed through immunohistochemistry.
The next phase involved examining whether astrocyte expression changes align with the stereotypical progression of AD neuropathology. This study was pivotal in asserting the hypothesis that the extent of astrocyte transcriptomic changes correlates with both spatial vulnerability of neural networks and the accumulation of ADNC in specific brain regions.
We rank-ordered the brain regions based on their vulnerability to neurofibrillary degeneration, performing a differential expression analysis that revealed 504 DEGs across adjacent regions. The findings were then categorized into six spatial gene sets that demonstrated distinct expression trajectories across the AD network.
To establish whether regional variations in expression patterns were influenced by the amounts of Aβ and pTau, we scrutinized overlaps between EC-specific and V1-specific signatures from normal controls and the created spatial trajectory gene sets. Remarkably, significant enrichments were identified, suggesting that intrinsic regional properties along with local AD pathology influence astrocyte gene expression.
Following the spatial analysis, we turned our attention to the temporal dynamics of astrocyte transcriptomics over the course of AD. We executed differential expression analysis comparing the four pathology stages. This approach yielded a list of 798 DEGs indicative of temporal associations with AD pathology, further categorized into six disparate temporal gene sets.
From the analysis, it was evident that some gene sets exhibited trends correlating with early stage homeostatic signatures, while others peaked at stages signifying moderate ADNC. Additionally, distinct functional characteristics associated with several of these sets highlighted the varying roles astrocytes play through disease progression.
Furthermore, our clustering analysis made significant contributions to understanding astrocyte diversity. From an initial sample of 500 nuclei per donor, we delineated nine distinct subclusters of astrocytes, some mirroring the characteristics of other cell types like neurons and microglia.
Investigation into the dynamic states of astrocyte subclusters revealed variability in frequency and density along AD’s spatial and temporal axes. Specifically, the proportion of homeostatic astH0 astrocytes decreased in the entorhinal cortex (EC) compared to all other regions, while the increased presence of astR1 and astR2 reactive astrocytes was noted.
Moreover, we employed pseudotime analysis, confirming a sequential pattern, illustrating the transition from homeostatic astrocytes to reactive states. The findings suggest that astMet and astTinf are terminal subclusters influenced by cumulative toxic exposure throughout AD progression.
For broader accessibility to our research, we have developed the AD progression atlas, a web-based tool designed for querying and visualizing gene expression across various parameters.
**Interview with Dr. Sarah Mitchell, Neuroscientist and Lead Author of the Recent Study on Astrocytes in Alzheimer’s Disease**
**Editor:** Thank you for joining us today, Dr. Mitchell. Your recent study provides some fascinating insights into astrocytes’ role in Alzheimer’s Disease. Can you explain the significance of focusing on astrocytes in this research?
**Dr. Mitchell:** Absolutely. Astrocytes, a type of glial cell in the brain, play a critical role in maintaining homeostasis and supporting neuronal function. In Alzheimer’s Disease, they significantly alter their behavior, which can contribute to disease progression. By studying astrocytes, we can better understand the molecular changes that occur in different brain regions and how they relate to the pathology of Alzheimer’s.
**Editor:** You utilized single-nucleus RNA sequencing (snRNA-seq) in your study. Could you elaborate on how this technique enhanced your research findings?
**Dr. Mitchell:** SnRNA-seq allows us to analyze the gene expression of individual nuclei from specific brain regions. This high-resolution technique enables us to capture the complexity of astrocytic responses to Alzheimer’s pathology, revealing regional differences and specific changes in gene expression that were previously underexplored in human tissues.
**Editor:** In your study, you categorized the 32 donors according to four distinct pathology stages. What did these stages reveal about the relationship between astrocytic behavior and Alzheimer’s progression?
**Dr. Mitchell:** The stratification provided invaluable insights. We found that as the pathology deepened—from no signs of AD to severe AD burden—there were significant alterations in astrocyte gene expression. This correlation suggests that astrocyte changes might not only reflect the state of the disease but could also play a role in its progression.
**Editor:** Your findings highlighted significant differences in Aβ plaque load and pTau ratios across different brain regions. Why is this information critical for understanding Alzheimer’s?
**Dr. Mitchell:** This data is essential because it matches the neuroanatomical progression pattern of Alzheimer’s. For example, we noted heightened pTau levels in the entorhinal cortex compared to other regions, which aligns with the known trajectory of neuronal degeneration in AD. Understanding these patterns can help in identifying potential therapeutic targets.
**Editor:** You mentioned integrating data from a surprising number of nuclei. What did your differential gene expression analysis reveal, particularly regarding genes like APOE and APP?
**Dr. Mitchell:** Our analysis showed distinct patterns of regulation for these key genes across different brain regions. For instance, there was significant upregulation in certain areas, highlighting their potential roles in astrocytic responses to AD pathology. This could help in developing targeted therapies that address specific regional vulnerabilities.
**Editor:** What do you see as the next steps in this line of research following your findings?
**Dr. Mitchell:** The next phase involves investigating the temporal changes in astrocyte gene expression as the disease progresses. We want to link these changes more closely with functional outcomes in neurodegeneration and explore how they might influence treatment strategies and patient outcomes.
**Editor:** Thank you, Dr. Mitchell, for sharing your insights. Your work is sure to advance our understanding of Alzheimer’s Disease significantly. We look forward to seeing where this research leads next!
**Dr. Mitchell:** Thank you for having me! I’m excited to share our findings and their implications for the future of Alzheimer’s research.
