Protective Aggregation in Amygdala CRF Neurons as a Feedback Buffer Against Stress-Induced Hyperexcitability

I propose that chronic stress triggers protective protein aggregation in amygdala CRF neurons, serving as a negative feedback mechanism to contain stress-induced proteotoxicity and suppress circuit hyperexcitability. These aggregates function as a proteostatic sink, buffering toxic oligomeric intermediates. The age-related upregulation of this pathway may explain why aged rodents actually show decreased anxiety-like behaviors under chronic stress—a seemingly paradoxical finding that makes sense if aggregates are protective rather than pathological.

CRF receptor signaling directly modulates the HSP90 co-chaperone FK506BP, connecting stress pathways to the proteostasis machinery 1. Under chronic stress, sustained CRF receptor activation appears to shift proteostatic priorities: stress-responsive chaperone synthesis increases, metastable proteins get sequestered into ordered aggregates instead of being refolded, and monomeric proteins essential for synaptic function and CRF signaling itself are selectively preserved. This strategy converts toxic oligomeric intermediates into thermodynamically stable, inert assemblies that no longer disrupt neural excitability.

This model predicts that CRF neurons employ a triage approach under proteostatic stress. Rather than attempting energy-intensive disaggregation—a youthful priority—neurons under chronic CRF overload redirect misfolded proteins into structured aggregates. Aging neurons appear to follow this same strategy. This explains why aged rats show decreased anxiety despite elevated aggregate loads: the aggregates aren't causing dysfunction; they're preventing it.

Testable Predictions:

1. Aggregate Prevention Worsens Phenotype: If we pharmacologically inhibit aggregation (using repurposed anti-amyloid compounds) in amygdala CRF neurons of chronically stressed young mice, we should see increased anxiety-like behaviors and impaired fear extinction compared to vehicle-treated stressed controls. Blocking the protective sink would release toxic oligomers that worsen circuit dysfunction.

2. Aggregate Induction Improves Resilience: Expressing aggregation-prone proteins specifically in amygdala CRF neurons of unstressed mice should induce modest protective aggregates and produce stress-resilience phenotypes—reduced fear generalization, enhanced extinction—without requiring actual stress exposure. This would show that aggregates can substitute for stress-adaptive downregulation.

3. Age x Stress Interaction: Young mice lacking key aggregation-promoting factors (like Hsp104 or other proteostasis modulators) won't exhibit the protective shift and should show exacerbated stress-induced proteotoxicity, phenocopying aged animals that can't mount this response. Conversely, aged wild-type mice will show resilience that's abolished by aggregate inhibition.

4. CRF-Specific Proteostasis Markers: Chronic stress will increase ubiquitin-positive aggregates co-localized with CRF neurons in the amygdala of aged (but not young) mice, paralleling the mRNA decreases in CRF and CRF-BP observed previously 3. This indicates proteostatic sequestration has replaced the need for continued CRF signaling.

Falsifiability: If aggregates in CRF neurons are merely pathological rather than protective, preventing their formation should improve fear-related phenotypes in chronically stressed animals. If aggregation has no functional role in CRF circuits, genetic or pharmacological manipulation should produce no behavioral effects. The hypothesis is also falsified if aged animals with high aggregate loads show worse anxiety under chronic stress, contrary to existing evidence.

Novel Contribution: This model bridges the proteostasis-aging literature with stress-neurocircuit research by proposing that protein aggregates function as activity-dependent feedback regulators in specific neuronal populations. It challenges the therapeutic assumption that aggregate dissolution necessarily benefits neurological function, instead predicting context-dependent harms when the proteome's final organizational strategy is disrupted.