Single-Cell Spatial Proteomics Uncovers Molecular Interconnectivity among Hallmarks of Aging

By leveraging a single-cell spatial proteomics atlas of aging yeast, this study reveals that hallmark phenotypes arise from compartment-specific protein disorganization, identifies conserved molecular changes in human aging, and establishes a hierarchical sequence of cellular failures driven by early nucleolar, proteostatic, and mitochondrial dysfunction.

The Gist

Imagine your body as a bustling, high-tech city. Every cell is a neighborhood, and inside each neighborhood, there are specialized factories (organelles) like the power plant (mitochondria), the library (nucleus), and the shipping docks (ribosomes).

For a long time, scientists studying aging looked at this city from a helicopter, taking a "bulk" photo. They could see that the city was getting messy overall, but they couldn't see where the mess was happening or why. They knew the power plant was failing and the library was chaotic, but they didn't know if the power plant was failing because the library was messy, or if they were just two separate problems happening at the same time.

This new study is like sending a fleet of tiny, high-definition drones into every single cell of a yeast organism (a simple, single-celled creature that acts as a model for human aging). These drones didn't just count the workers; they watched exactly where every single protein (the workers) was standing, how much of them there were, and whether they were forming toxic piles (aggregates).

Here is what they discovered, translated into everyday terms:

1. The "Lost and Found" Problem (Spatial Chaos)

The biggest surprise wasn't that the workers stopped working; it was that they stopped staying in their assigned rooms.

• The Analogy: Imagine a library where the librarians (proteins) are supposed to stay in the reading room. As the city ages, the librarians start wandering out into the hallway, the cafeteria, and the parking lot. Some of them even trip and fall into a pile (aggregation).
• The Result: The library becomes a mess because the books aren't being organized, and the hallway becomes clogged with confused librarians. The study found that this "wandering" happens to hundreds of proteins, breaking the specific zones where life-sustaining chemistry needs to happen.

2. The Domino Effect: The "Nucleolus" is the First to Fall

The researchers found a specific order to the collapse. It didn't happen all at once.

• The Analogy: Think of the Nucleolus (a tiny structure inside the nucleus) as the Main Assembly Line that builds the city's delivery trucks (ribosomes). These trucks are needed to build everything else in the city.
• The Discovery: The study found that the Assembly Line starts breaking down very early. The workers on this line start wandering off, and the trucks they build are defective.
• The Consequence: Because the trucks are broken, the rest of the city (the mitochondria, the DNA repair crews, the nutrient sensors) starts to fail. The study suggests that fixing the Assembly Line might stop the rest of the city from crumbling.

3. The Power Plant's "Import" Failure

The mitochondria (the power plant) are famous for failing in old age. But this study showed how it happens.

• The Analogy: The power plant has a strict security gate. Only specific workers with the right badges are allowed inside to fix the generators.
• The Discovery: As the cell ages, the security gate breaks. The workers who are supposed to fix the power plant (DNA repair enzymes) get stuck outside in the rain, while the workers who belong in the library wander in.
• The Result: The power plant gets damaged because the right tools aren't there, and the library gets cluttered with the wrong tools. This explains why the power plant fails: it's not just old; it's been invaded by the wrong stuff and abandoned by the right stuff.

4. The "Reset Button" (Daughter Cells)

One of the most hopeful findings involves how yeast reproduce. When an old yeast cell divides, it creates a brand-new "daughter" cell.

• The Analogy: Imagine an old, messy house. When the family moves out and a new family moves in, they don't just inherit the mess. They get a brand new, clean house with fresh paint and organized rooms.
• The Discovery: The study showed that the "daughter" cells actually reset their protein organization. They kick out the wandering workers and clear the toxic piles. The "mother" cell keeps the mess, but the "daughter" starts fresh. This proves that the aging process is reversible at the cellular level if you can just reset the organization.

5. The Human Connection

The researchers checked if this happens in humans, too. They looked at human proteins that are similar to the yeast ones.

• The Stat: A massive 91.6% of these "aging" proteins in yeast have human counterparts that also change as humans age.
• The Takeaway: The rules of the city are the same for us. If we can figure out how to keep the workers in their rooms and the assembly lines running in yeast, we might find a way to slow down aging in humans.

The Big Picture

This study changes how we view aging. It's not just that things wear out like an old car. It's that the blueprint of the city gets scrambled. The workers lose their maps, the factories lose their security, and the assembly lines break down.

The authors suggest that aging is a loss of organization. If we can develop therapies that act like a "city planner"—helping proteins find their way back to the right room and keeping the assembly lines running—we might be able to fix the root cause of aging, rather than just treating the symptoms.

Technical Summary

1. Problem Statement

Aging is characterized by conserved "hallmarks" (e.g., genomic instability, loss of proteostasis, mitochondrial dysfunction, epigenetic alterations). However, current understanding is limited because:

• Isolation of Hallmarks: Most studies examine these hallmarks in isolation rather than investigating their mechanistic interconnectivity.
• Lack of Spatial Resolution: Bulk transcriptomics and proteomics average out cell-to-cell variability and obscure subcellular organization. Many aging phenotypes (e.g., protein aggregation, organelle morphological changes, loss of compartmentalization) are inherently spatial and cannot be captured by bulk measurements.
• Missing Atlas: There was no comprehensive, proteome-wide, single-cell dataset capturing protein expression, localization, and aggregation dynamics across the entire replicative lifespan of an organism.

2. Methodology

The authors leveraged a companion study (Yoo et al., 2026a) to construct and analyze a massive spatial proteomic atlas in Saccharomyces cerevisiae (yeast).

• Experimental Platform:
  • Strains: A library of yeast strains with >94% of the proteome endogenously tagged with mNeonGreen (mNG).
  • Aging Model: Replicative aging (mother cells separated from daughters).
  • Imaging: High-content 3D confocal microscopy capturing >90 million cells across the lifespan.
  • Automation: Automated liquid handling (Tecan Fluent) for cell wall biotinylation and magnetic capture of aged mother cells to ensure high purity.
• Computational Analysis:
  • Deep Learning: A dual-stream ensemble framework (2D CNN + 3D ResNet-50) quantified protein localization into 17 subcellular compartments, concentration, and aggregation states at single-cell resolution.
  • Metrics:
    • Localization: Probability vectors for compartment assignment.
    • Concentration: Mean intensity within organelle masks.
    • Aggregation: Puncta detection using intensity thresholds and Laplacian-of-Gaussian (LoG) filters.
  • Temporal Trajectories: Cells were binned by age (Young, 3–5, 6–8, 9–11, 12–14, 15+ generations) to map the progression of dysfunction.
  • Cross-Species Validation: Yeast findings were compared against human aging transcriptomic and proteomic datasets to assess ortholog conservation.

3. Key Contributions

• First Proteome-Wide Spatial Atlas of Aging: Provided a high-resolution map of how protein localization, concentration, and aggregation change across the entire proteome during aging.
• Redefinition of Hallmarks: Demonstrated that hallmark phenotypes often manifest as compartment-specific erosion of spatial confinement (relocalization, loss of organelle identity, and aggregation) rather than simple changes in total protein abundance.
• Discovery of Molecular Conduits: Identified specific proteins that act as "conduits" linking different hallmarks, revealing a hierarchical sequence of cellular failures.
• Identification of Early Vulnerabilities: Pinpointed nucleolar dysfunction and ribosome biogenesis defects as early, upstream drivers that precede and potentially cause other hallmark failures.

4. Key Results

A. Genomic Instability and Epigenetic Alterations

• DNA Damage: Increased ssDNA foci and upregulation of DNA repair proteins (Rad52, Rad1, etc.) in aged cells.
• Nuclear Morphology: ~30% of aged cells showed nuclear shape abnormalities and enlargement.
• Epigenetic Drift:
  • Silencing Loss: Reduced Sir3 and Cohibin components (Csm1, Lrs4) led to telomeric and rDNA silencing defects.
  • Histone Modifiers: Subunit-specific remodeling of HDAC complexes (Rpd3L/S, Set3C, Hda1) and increased Dot1 (H3K79 methyltransferase) in a subset of cells, correlating with transcriptional noise.
  • Chromatin Remodelers: Increased variability in SWI/SNF and RSC components, exacerbating transcriptional heterogeneity.

B. Loss of Proteostasis and Autophagy

• Autophagy Impairment: Accumulation of Cvt pathway intermediates (Ape1, Ams1) and stalled mitophagy (Atg33 foci), indicating a bottleneck in vacuolar clearance.
• Protein Aggregation: Aggregation began early (generations 3–5) in the nucleus and nucleolus, preceding widespread mitochondrial and cytosolic aggregation.
• Translation Stress: Formation of P-bodies (Scd6, Dcp1/2) and nuclear accumulation of mRNA export factors (Mex67), indicating global translation repression and mRNA export bottlenecks.
• UPS Failure: Mislocalization of 19S proteasome subunits (Rpn5, Rpn9) from nucleus to cytosol and reduced ubiquitin-processing enzymes.

C. Organelle-Specific Dysfunction

• Mitochondria:
  • Modular Failure: Not a uniform loss, but specific collapse of Fe-S cluster biogenesis (early), mitochondrial translation, and lipid metabolism.
  • Import Defects: Loss of mitochondrial import machinery (Tim21, Tom20/70 aggregation) led to the selective depletion of mitochondrial DNA repair enzymes (Ogg1, Apn1), linking proteostasis failure to mtDNA instability.
  • Metabolic Rewiring: Mislocalization of enzymes (e.g., Leu4, Adh3) and cofactor depletion (Thiamine, Lipoic acid).
• Peroxisomes: Destabilization of import machinery (Pex10, Pex13) leading to heterogeneous peroxisome populations within single cells.
• Rejuvenation: Asymmetric cell division successfully reset organelle proteostasis in daughter cells, removing aggregates and restoring stoichiometry (e.g., Pex13:Pex3 ratios).

D. Nucleolar Dysfunction and Ribosome Biogenesis (The "Early Driver")

• Spatial Collapse: The nucleolus lost its identity as a compartmentalized hub. Processing factors (SSU and LSU processome) escaped into the nucleoplasm and aggregated.
• Key Lesions:
  • SSU: Disorganization of the SSU processome (U3, UTP complexes) and aggregation of Kre33 (human homolog NAT10).
  • LSU: Fragility of the ITS2-centered network (Nop4, Nop7-Erb1-Ytm1) and loss of late-stage licensing factors (Tif6).
• Consequence: This spatial uncoupling of ribosome assembly steps likely produces aberrant ribosomes, driving proteotoxic stress and amplifying downstream aging phenotypes.

E. Nutrient Signaling Dysregulation

• Despite nutrient-replete conditions, aged cells showed:
  • Amino Acids: Vacuolar enrichment of Gtr1/2 (TORC1 activator) and depletion of amino acid permeases from the plasma membrane.
  • Glucose: Reduced surface glucose transporters (Hxt2/7) and Hexokinase (Hxk2) variability.
  • Phosphate/Inositol: Loss of Pho84 (phosphate transceptor) and nuclear accumulation of Opi1 (repressing lipid synthesis genes), indicating a decoupling of extracellular nutrients from intracellular sensing.

5. Significance and Conclusion

• Interconnectivity: The study reveals that aging is not a collection of independent failures but a hierarchical, interconnected network.
  • Example: Early nucleolar/ribosome defects $\rightarrow$ Proteostasis collapse $\rightarrow$ Mitochondrial import failure $\rightarrow$ mtDNA instability $\rightarrow$ Genomic instability.
• Temporal Hierarchy: The authors propose two groups of hallmarks based on temporal trajectories:
  • Group 1 (Early/Steep): Nucleolar disruption, proteostasis decline, mitochondrial dysfunction.
  • Group 2 (Gradual): Epigenetic alterations, genomic instability, disabled autophagy, deregulated nutrient sensing.
  • Implication: Group 1 failures likely drive the emergence of Group 2.
• Cross-Species Conservation: 91.6% of the yeast hallmark-linked proteins have human orthologs that show age-associated abundance changes in human proteomes, validating the yeast model for human aging mechanisms.
• Therapeutic Insight: The findings suggest that interventions targeting spatial organization (e.g., restoring nucleolar confinement or mitochondrial import) rather than just protein abundance could yield pleiotropic benefits by rebalancing the entire aging network.

In summary, this paper provides a foundational framework for understanding aging as a spatiotemporal reorganization of the proteome, where the loss of compartmental integrity serves as the primary mechanism linking diverse cellular hallmarks.