A multi-omic atlas in the African turquoise killifish reveals increased glucocorticoid signaling as a hallmark of brain aging

This study presents a multi-omic atlas of the African turquoise killifish brain revealing that aging is characterized by microglial expansion and conserved glucocorticoid signaling activation, which can be mitigated by pharmacological inhibition to rescue key aging phenotypes.

The Gist

Imagine the brain as a bustling, high-tech city. As the city ages, the roads get potholed, the power grid flickers, and the sanitation crew (the immune system) starts overreacting, causing traffic jams and chaos. For a long time, scientists have been trying to figure out exactly why this happens and if there's a way to fix it, but studying human brains takes decades.

Enter the African turquoise killifish. Think of this tiny fish as a "fast-forward button" for aging. While a human might live 80 years, this fish lives only about 4 to 6 months. It goes from a baby to a grandparent in the time it takes a human to finish a college degree. This makes it the perfect "time-lapse camera" for studying how brains age.

Here is the story of what this team of scientists discovered, broken down into simple concepts:

1. The Great Map (The Atlas)

The researchers created the first-ever "Google Maps" of the killifish brain as it ages. They didn't just look at the whole brain; they zoomed in to see every single type of cell (neurons, support cells, and the brain's immune cells) in both male and female fish, and in two different "breeds" of fish (one that lives a short life and one that lives a bit longer).

The Big Discovery: As the fish got old, one specific group of cells exploded in number: Microglia.

• The Analogy: Imagine the brain's immune cells (microglia) are like the city's sanitation workers. In a young brain, they are efficient, picking up trash and fixing small potholes. In an old brain, they panic. They multiply uncontrollably, start shouting (inflammation), and accidentally damage the roads (neurons) while trying to clean up. The study found this "panic mode" happens in both short-lived and long-lived fish, regardless of whether they are male or female.

2. The Stress Hormone Culprit

The researchers wanted to know what was causing this panic. They looked at the genetic "instruction manuals" inside the cells and found a common theme: Glucocorticoid Signaling.

• The Analogy: Think of glucocorticoids (like cortisol in humans) as the body's "Stress Alarm." In a young brain, the alarm rings briefly when there's a real emergency, then turns off. In the aging brain, the alarm gets stuck in the "ON" position. It's like a smoke detector that won't stop beeping, even when there's no fire. This constant screaming stresses out the cells, makes the sanitation workers (microglia) go crazy, and accelerates the city's decay.

They even checked the fish's kidneys (where fish make stress hormones) and found that old fish had much higher levels of this "stress hormone" circulating in their bodies, confirming the alarm was indeed stuck on.

3. The "Off Switch" Experiment

The most exciting part of the study was the intervention. The scientists asked: If we turn off this stuck alarm, can we slow down the aging of the brain?

They used a drug called Mifepristone (which is already FDA-approved for other conditions) to block the "stress alarm" receptors. They started giving this drug to middle-aged fish (the equivalent of a human in their 40s).

The Results:

• The Reset: The brains of the treated fish looked much younger. The "panic" in the genetic instructions calmed down.
• The Cleanup Crew: The number of overactive microglia (the panicked sanitation workers) dropped back down to normal levels.
• The Repair: Pathways related to protein building and cell repair, which usually break down with age, started working again.

The Takeaway

This study is like finding a master key for the aging brain. It suggests that a major driver of brain aging isn't just random wear and tear, but a chronic stress response that gets stuck in the "on" position.

By using the killifish as a fast-forward model, the scientists proved that if you can calm down this stress signal in the middle of life, you can actually reverse some of the damage and keep the brain's "city" running smoothly for longer. While this was done in fish, the fact that these stress pathways are very similar in mice and humans suggests this could be a powerful new strategy for keeping human brains healthy as we age.

Technical Summary

1. Problem Statement

Aging is the primary risk factor for cognitive decline and neurodegeneration, yet the specific molecular and cellular mechanisms driving brain aging across different sexes and genetic backgrounds remain poorly understood. Traditional long-lived vertebrate models (e.g., mice, humans) make longitudinal interventional studies difficult due to their extended lifespans. While the African turquoise killifish (Nothobranchius furzeri) is a short-lived vertebrate model (~4–6 months) that spontaneously develops age-related pathologies, previous studies have been limited by:

• Focus on single sexes or single genetic strains.
• Lack of cell-type resolution (bulk tissue analysis).
• Limited integration of transcriptomic and epigenomic data.
• A lack of actionable therapeutic targets identified through mechanistic studies in this model.

2. Methodology

The authors generated a comprehensive, multi-omic atlas of the aging brain using two independent genetic strains of turquoise killifish: the short-lived, inbred GRZ strain and the longer-lived, wild-derived ZMZ-1001 strain.

Experimental Design:

• Cohorts: Both male and female fish were profiled at four time points: Young (6 weeks), Middle-aged (10 weeks), Old (16 weeks), and Geriatric (26 weeks, for ZMZ-1001 only).
• Multi-omic Profiling:
  • snRNA-seq: Single-nucleus RNA sequencing (10x Genomics) for both strains to capture transcriptomes (44 libraries, ~210k nuclei).
  • snATAC-seq: Single-nucleus ATAC sequencing (10x Genomics) for the GRZ strain to map chromatin accessibility (8 libraries, ~27k nuclei).
  • Bulk ATAC-seq: Omni-ATAC protocol on whole brain tissue for both strains (48 libraries).
  • Bulk RNA-seq: Used for validation and intervention studies.
• Validation:
  • RNAscope (RNA-ISH): Multiplex fluorescent in-situ hybridization to validate cell proportions and gene expression in intact tissue.
  • 3D Reconstruction: Imaris software used to quantify microglial soma volume.
• Computational Analysis:
  • Annotation: Semi-supervised approach combining unsupervised clustering, cross-species marker mapping (mouse/zebrafish), and reference-based transfer (scMAGIC) to identify 16 distinct cell types.
  • Differential Analysis: Pseudobulking strategies (muscat) followed by DESeq2 to identify age-related changes while controlling for batch effects and sex.
  • Regulon Activity: Used decoupleR to infer Transcription Factor (TF) activity from snRNA-seq and TOBIAS/HOMER for TF motif accessibility in ATAC-seq.
  • Machine Learning: Built aging clocks (ElasticNet and Random Forest) using microglia transcriptomes to predict biological age.
• Intervention:
  • Drug: Mifepristone (a glucocorticoid receptor antagonist).
  • Paradigm: Treatment initiated at middle-age (10 weeks) via agar-coated food, continuing until old age (16 weeks).

3. Key Contributions

1. First Multi-Omic Atlas: Created the largest and most comprehensive single-cell atlas of the turquoise killifish brain, encompassing both sexes, two distinct genetic strains, and three 'omic' layers (snRNA-seq, snATAC-seq, bulk ATAC-seq).
2. Cell-Type Resolution: Defined 16 major brain cell types (8 neuronal, 8 non-neuronal) and established a searchable interactive resource (R Shiny app) for the community.
3. Conserved Aging Signatures: Demonstrated that despite genetic and lifespan differences, core molecular aging signatures are conserved across strains and sexes, suggesting a universal "rate of aging" mechanism.
4. Identification of a Master Regulator: Identified Glucocorticoid Receptor (NR3C1) signaling as a conserved, upregulated driver of brain aging across vertebrates (fish, mouse, human).
5. Therapeutic Rescue: Provided proof-of-concept that mid-life pharmacological inhibition of glucocorticoid signaling can reverse key molecular and cellular aging phenotypes.

4. Key Results

A. Cellular Composition and Microglia Expansion

• Microglia Expansion: Aging leads to a robust, conserved expansion of the microglial population in the brain, regardless of sex or strain. This was validated by both scRNA-seq proportion analysis and RNAscope quantification.
• Morphological Changes: Microglia in aged fish exhibited increased soma volume (reactive phenotype), consistent with activation.
• Other Cell Types: While some changes were sex-specific (e.g., NSPC decline in females), the microglial expansion was the most robust cross-strain feature.

B. Transcriptional and Epigenomic Remodeling

• Gene Expression: Aging caused widespread upregulation of inflammatory pathways (e.g., "Immune response," "Innate immune system") and downregulation of cell proliferation pathways.
• Conserved Signatures: Genes upregulated in aged mouse and human brains showed consistent upregulation in aged killifish.
• Chromatin Accessibility: Granule excitatory neurons showed the most dramatic changes in chromatin accessibility (increased heterochromatin domain accessibility), while microglia showed the most significant transcriptional changes.

C. Glucocorticoid Signaling as a Hallmark

• TF Activity: Multi-omic integration revealed a significant increase in the activity of the Glucocorticoid Receptor (nr3c1) regulon across all cell types and strains.
• Conservation: This upregulation of NR3C1 activity was also observed in re-analyzed public datasets of aging mouse and human brains.
• Mechanism: LC-MS/MS analysis of kidney tissue (the site of cortisol synthesis in teleosts) confirmed a significant age-related increase in circulating cortisol levels in both strains, suggesting that increased ligand availability drives the receptor activation.

D. Microglia Aging Clock

• Machine learning models trained on GRZ microglia transcriptomes could accurately distinguish young from old fish (AUROC ~0.85).
• When applied to the longer-lived ZMZ-1001 strain, the models predicted that microglia remained "younger" for longer, reaching "old" status only at the geriatric stage, correlating with the strain's extended lifespan.

E. Pharmacological Rescue (Mifepristone)

• Transcriptomic Reversal: Mid-life initiation of mifepristone treatment shifted the brain transcriptome of old fish toward a "younger" state. Age-regulated genes were reversed in direction.
• Pathway Rescue: Treatment specifically reversed age-related dysregulation in proteostasis (ribosome/translation), membrane biology, and RNA metabolism.
• Cellular Rescue: Mifepristone treatment significantly blunted the age-related expansion of microglia, reducing their abundance to levels comparable to young controls.

5. Significance

• Mechanistic Insight: The study identifies chronic glucocorticoid signaling as a central, conserved driver of vertebrate brain aging, linking systemic stress hormone production (cortisol) to local neuroinflammation and microglial activation.
• Therapeutic Potential: It demonstrates that targeting the glucocorticoid receptor in mid-life is a viable strategy to mitigate brain aging phenotypes, offering a potential repurposing avenue for existing drugs (like mifepristone) for age-related neurodegeneration.
• Resource Utility: The provided atlas and interactive tools significantly lower the barrier for researchers to study brain aging in a genetically tractable, short-lived vertebrate, facilitating rapid screening of interventions.
• Generalizability: The findings bridge the gap between fish models and mammals, suggesting that the mechanisms of brain aging identified in killifish are highly relevant to human neurobiology.