4,068 posters, 219 videos, 101 topics, 69 sessions, 14,365 authors, 6,128 institutions

Alzheimer's Association International Conference 2026

July 12-15, 2026 | London, United Kingdom, and Online

3130

Disentangling Brain-region Specific From Global Brain-wide Microglial Responses To Pathologies Of Alzheimer’s Disease

View Poster Contact Presenter

Poster Presenter

Kevin Lin

Authors

Amy Watt, David Fardo, Suman Jayadev, Katherine Prater, Shubhabrata Mukherjee, Kevin Lin

Affiliations

University of Kentucky, University of Washington

Part of Topic

Disentangling brain-region specific from global brain-wide microglial responses to pathologies of Alzheimer’s disease

Amy Watt1, David Fardo2, Suman Jayadev3, Katherine E. Prater3, Shubhabrata Mukherjee4, Kevin Z. Lin1

Motivation:

Most transcriptomic studies of Alzheimer’s disease (AD) sample tissue from one brain region, and conclude cellular mechanisms of the entire brain.
However, these cellular mechanisms could strongly differ from region-to-region, for instance, due to how co-occurring neurodegenerative diseases spread throughout the brain.
Which cellular mechanism are region-specific vs. brain-wide?
We focus on microglia, the resident immune system of the brain.
After correcting for baseline snRNA-seq gene expression, correlation b/w brain regions is only modestly positive – this suggests possible region-specific mechanisms.

Fig. 1: A) Co-pathologies of AD are correlated with increased severity of AD (NACC data). B) Different pathologies spread along different routes, suggest the importance of distinguishing local-vs-global cellular mechanisms.

Fig. 2: Correlation of microglial transcriptomic profiles b/w brain regions (SEA-AD, 32 donors, spanning AD spectrum)

Goal and dataset:

Fig. 3: A) Goal – distinguish microglial brain-wide mechanisms those localized to the MTG. B) Stainings used to derive the HALO quantitative neuropathology (% area in gray matter), GFAP not shown.

Analysis workflow:

Step 1: Confirmatory factor analysis to derive MTG-specific burden
Step 2: Associate MTG-specific burden and global burden (Sinha, et al., 2023) to gene expression using NEBULA (He et al., 2021).

Fig. 4: A) Schematic of association b/w MTG burden and gene expression, adjusting for covariates. B) Flowchart of how HALO neuropathology and snRNA-seq is used.

Latent MTG burden:

Fig. 5: Confirmatory factor analysis including 2 co-localization variables to improve model fit (CFI = 0.959, RMSEA = 0.058).

Fig. 6: A) SEA-AD cohort, showing last CASI score (Nested F-test p-value: 0.0581) or Braak. MTG is staged at Braak 4. The MTG-specific burden captures variability overlooked by the global BPS. B) Corroboration of the local-vs-global burden using ROS-MAP cohort. *Not exact replication is the quantified neuropathology is only amyloid and tau from the medial temporal cortex (MTC).

Associated genes:

Key genes (CHI3L1, apoptosis; SORL1, phagocytosis; CTSD, lysosome) implicated as responding locally in MTG burden.

Fig. 7: A) Schematic of how we compare genes associated w/ (global) BPS or (local) MTG. B) Comparison of the gene associations. Dashed line denotes the signed FDR multiple-testing thresholding.

Fig. 8: The microglial sig. genes associated with MTG are corroborated in the ROSMAP cohort

Microglial composition:

What shifts in microglial states (composition) are associated w/ MTG burden?
We transfer microglial state labels from Sun et al., 2023 (ROS-MAP) to SEA-AD.

Fig 9: A) Transferred microglial states in SEA-AD.. B) Compositional shifts along BPS (global) or MTG burden (local), w/ Pearson corr. shown.

Comments

Email is required but is not shown publicly.

Sort comments

Loading comments...