Brain proteomic analysis implicates actin filament processes and injury response in resilience to Alzheimer’s disease

Zhi Huang, Gennifer E. Merrihew, Eric B. Larson, Jea Park, Deanna L. Plubell, Edward J. Fox, Kathleen S. Montine, Caitlin S. Latimer, C. Dirk Keene, James Zou, Michael J. MacCoss, Thomas J. MontineNature Communications 2023 · DOI

TL;DR

Why do some people remain cognitively sharp despite having brains riddled with the plaques and tangles characteristic of Alzheimer’s disease? This study uses deep proteomic profiling to reveal that "resilient" individuals maintain lower levels of soluble Amyloid-beta (Aβ) and show significantly less activation of cellular "injury response" pathways—specifically those involving actin filaments and detoxification—compared to those who succumb to dementia.

What Problem Does This Paper Address?

Alzheimer’s Disease (AD) is typically viewed as an inevitable slide from brain pathology to cognitive decline. However, a small subset of the population exhibits Resilience to Alzheimer’s Disease (RAD): they possess a high burden of AD pathology at autopsy but never showed symptoms of dementia during their lives.

The challenge in studying RAD has been "latent disease" and "comorbidity." Many previous studies grouped "preclinical" cases (people with low levels of disease) with truly resilient cases (people with high levels of disease but no symptoms). Furthermore, other conditions like vascular injury or Lewy body disease often cloud the data. This paper aims to isolate the specific molecular signature of pure resilience by using stringent clinical criteria and multi-region brain proteomics.

Methods

The researchers analyzed brain tissue from 43 participants categorized into three groups: Healthy Controls (HC), Resilient (RAD), and AD Dementia (ADD). They examined four distinct brain regions: the hippocampus and three cortical areas (including the caudate nucleus as a control for non-degenerating regions).

  • Proteomics: They employed Data-Independent Acquisition (DIA) mass spectrometry, a systematic method that provides more reproducible protein quantification than traditional "shotgun" proteomics.
  • Network Analysis: Beyond looking at individual proteins, they used Weighted Gene Co-expression Network Analysis (WGCNA) to identify "modules"—groups of proteins that work together.
  • Validation: To ensure their findings weren't a fluke of their small sample size, they validated their results against four massive external datasets (Banner, ROS/MAP, UPP, and BLSA).

Datasets

  • Primary Cohort: 155 samples from 43 donors (Stanford University and University of Washington).
  • Validation Cohorts:
    • ROS/MAP: Religious Orders Study and Rush Memory and Aging Project.
  • Banner: Banner Sun Health Research Institute.
  • UPP: University of Pennsylvania Proteomics study.
  • BLSA: Baltimore Longitudinal Study of Aging.

Key Insights and Ideas

  • Soluble Aβ is the Smoking Gun: While both RAD and ADD groups have high levels of insoluble plaques (the kind seen on PET scans), the RAD group had significantly lower levels of soluble Aβ. This suggests that the soluble, "liquid" form of the protein may be the primary driver of neurotoxicity and cognitive failure.
  • The "M5" Injury Module: The researchers identified a specific module of 181 proteins (labeled M5) that was highly upregulated in the dementia group but remained at near-normal levels in the resilient group.
  • Actin as a Structural Shield: The M5 module is heavily enriched for actin filament-based processes. Actin is the "skeleton" of synapses (the connections between neurons). In dementia, this system appears to be in a state of hyper-active "wound healing" or "detoxification," suggesting the brain is struggling to repair massive structural damage. Resilient individuals seem to avoid triggering this massive injury response altogether.
  • Regional Specificity: The molecular differences were most pronounced in the isocortex and hippocampus—areas critical for memory—while the caudate nucleus (often spared in early AD) showed fewer changes, acting as an internal control.

Key Results

  • Protein Count: The study quantified 7,115 proteins across the four brain regions.
  • Differential Expression: 33 unique proteins were identified as specifically associated with resilience (RAD DEPs).
  • The Aβ Threshold: Lower soluble Aβ in the SMTG (superior and middle temporal gyrus) was a defining feature of RAD compared to ADD.
  • Validation Success: 70% of the resilience-associated proteins identified in the primary study were validated in at least one external dataset, a high hit rate for proteomic studies.
  • PCA Separation: Principal Component Analysis showed that while HC and RAD groups overlap significantly in their protein profiles, the ADD group is a distinct molecular outlier, despite the RAD group having similar levels of "plaques and tangles" to the ADD group.

Limitations and Future Work

  • Bulk vs. Single-Cell: The study used "bulk" tissue, meaning they ground up the samples. This loses "cell-type specificity"—we don't know if these changes are happening in neurons, astrocytes, or microglia.
  • Sample Size: While the 43 donors were extremely well-characterized, the sample size is relatively small for high-dimensional "omics" data.
  • The Aβ Measurement Gap: Standard proteomics (using trypsin digestion) has difficulty distinguishing between Aβ fragments and its parent protein, APP. While the correlations were strong, more specialized assays are needed to confirm the exact Aβ species involved.
  • Causality: It remains unclear if lower soluble Aβ causes resilience or is a result of other protective factors. Future work using animal models or iPSCs (induced pluripotent stem cells) will be needed to move from correlation to causation.

Summary updated

Abstract

Resilience to Alzheimer's disease is an uncommon combination of high disease burden without dementia that offers valuable insights into limiting clinical impact. Here we assessed 43 research participants meeting stringent criteria, 11 healthy controls, 12 resilience to Alzheimer's disease and 20 Alzheimer's disease with dementia and analyzed matched isocortical regions, hippocampus, and caudate nucleus by mass spectrometry-based proteomics. Of 7115 differentially expressed soluble proteins, lower isocortical and hippocampal soluble Aβ levels is a significant feature of resilience when compared to healthy control and Alzheimer's disease dementia groups. Protein co-expression analysis reveals 181 densely-interacting proteins significantly associated with resilience that were enriched for actin filament-based processes, cellular detoxification, and wound healing in isocortex and hippocampus, further supported by four validation cohorts. Our results suggest that lowering soluble Aβ concentration may suppress severe cognitive impairment along the Alzheimer's disease continuum. The molecular basis of resilience likely holds important therapeutic insights.
BibTeX
@inproceedings{huang2023brain,
  author = {Zhi Huang and Gennifer E. Merrihew and Eric B. Larson and Jea Park and Deanna L. Plubell and Edward J. Fox and Kathleen S. Montine and Caitlin S. Latimer and C. Dirk Keene and James Zou and Michael J. MacCoss and Thomas J. Montine},
  title = {{Brain proteomic analysis implicates actin filament processes and injury response in resilience to Alzheimer’s disease}},
  booktitle = {Nature Communications},
  year = {2023},
  doi = {10.1038/s41467-023-38376-x},
}
Figure