Network Centrality Determines Whether Age-Related Protein Aggregates Act as Protective Depots or Pathogenic Sinks

Hypothesis: The functional outcome of protein aggregation during aging hinges on the topological position of the aggregating protein within the interactome. High‑betweenness hub proteins that become sequestered into amyloid‑like aggregates disrupt communication between network modules, turning a potential protective depot into a pathogenic sink. In contrast, low‑connectivity peripheral proteins that form inert aggregates serve as a sink for misfolded species, buffering proteostatic stress without compromising signaling. This dichotomy is regulated by a phosphorylation‑dependent switch that alters surface hydrophobicity and determines whether an aggregating protein is recruited to liquid‑liquid phase separated condensates (protective) or to irreversible amyloid aggregates (pathogenic).

Mechanistic rationale: Aging cells experience a gradual decline in chaperone capacity and an increase in kinase activity that preferentially targets hub proteins (e.g., insulin/IGF‑1 signaling nodes). Phosphorylation of specific serine/threonine residues on these hubs increases their propensity to adopt β‑sheet rich conformations while reducing their solubility. When phosphorylated, hub proteins nucleate amyloid aggregates that physically obstruct the flow of information across module interfaces, because their many interactions act as bottlenecks. Dephosphorylation or inhibition of the responsible kinases keeps hubs in a more soluble state, allowing misfolded proteins to be captured by peripheral, low‑betweenness proteins that form dynamic, reversible aggregates. These peripheral aggregates act as a "hold‑and‑release" system, sequestering dangerous species until proteasome or autophagy activity can clear them.

Testable predictions:

1. In C. elegans, RNAi knockdown of a hub‑specific kinase (e.g., PKC‑1) will reduce phosphorylation of high‑betweenness proteins such as DAF‑2/Insulin receptor, decrease their amyloid aggregation, and extend lifespan only when those hubs are confirmed to have high betweenness centrality (>8).
2. Conversely, overexpression of an aggregation‑prone domain fused to a low‑betweenness peripheral protein (e.g., a cytosolic metabolic enzyme) will increase its insoluble fraction, improve resistance to heat shock and oxidative stress, and extend lifespan without affecting hub protein solubility.
3. Phosphoproteomic analysis of young versus aged worms will reveal a significant increase in phosphorylation on hub proteins that correlates with their insolubility status, while peripheral proteins show unchanged or decreased phosphorylation.
4. Pharmacological inhibition of the identified kinase in mammalian cultured neurons will lower phosphorylated tau and amyloid‑β hub aggregation, restore neuronal network activity measured by calcium imaging, and reduce toxicity only when the targeted proteins have high betweenness in the human brain interactome.

Falsifiability: If reducing hub protein aggregation fails to rescue lifespan or network function despite confirmed decrease in their insoluble fraction, or if enhancing peripheral protein aggregation does not confer stress resistance, the hypothesis would be refuted. Similarly, if phosphorylation status does not differ between hub and peripheral aggregates, the proposed regulatory switch would be invalid.

This framework integrates the observations that aggregation is widespread yet only partially detrimental, that pro‑longevity genes are highly connected, and that insulin/IGF‑1 signaling modulates both aging and solubility. It shifts the focus from viewing aggregates as uniform waste to recognizing them as topology‑dependent tools that the cell employs—or misemploys—as a last resort to maintain order.