Cold temperature extends longevity and prevents disease-related protein aggregation through PA28γ-induced proteasomes

Moderate decreases in body temperature can be beneficial for lifespan across species, whereas extreme cold is detrimental. Lifespan inversely correlates with temperature in poikilotherms (e.g., C. elegans), and mild body temperature reductions increase longevity in homeotherms (rodents), with correlations also reported in humans. Human body temperature varies diurnally and can reach 36 °C during sleep, and has declined slightly over the past 160 years, coinciding with increased longevity. While historically attributed to slowed metabolism (“pace of living”), evidence in C. elegans shows cold-induced longevity is an actively regulated process, not solely thermodynamic: the TRPA-1 cold-sensitive channel senses low temperature in neurons and intestine to extend lifespan, and chaperone systems (TRiC/CCT, DAF-41/p23) are induced to preserve proteostasis. Given that aging promotes aggregation-prone proteome changes underlying neurodegeneration (Alzheimer’s, Parkinson’s, Huntington’s, ALS) and that proteasomes degrade damaged/misfolded proteins, the authors hypothesized that cold-induced mechanisms might enhance proteasome function to mitigate pathological aggregation. The proteasome’s 20S core exhibits three activities (caspase-like/β1, trypsin-like/β2, chymotrypsin-like/β5) and is activated by regulators including 19S (forming ubiquitin-dependent 26S) and PA28 complexes. PA28γ/PSME3 forms homo-heptamers, acts ubiquitin-independently, and preferentially enhances trypsin-like activity. The study investigates whether cold modulates proteasome activity via PA28γ/PSME-3 to extend lifespan and reduce disease-related aggregation in C. elegans and human cells.

Prior work established temperature as a determinant of lifespan in multiple species, with TRPA-1 required for cold-induced longevity in C. elegans and induction of proteostasis factors (TRiC/CCT, DAF-41/p23) at low temperature. Age-related decline in the ubiquitin-proteasome system leads to accumulation of aggregation-prone proteins; 26S proteasomes require ubiquitination, whereas PA28γ-activated 20S complexes can degrade unstructured proteins ubiquitin-independently. In vitro, PA28γ promotes degradation of unfolded proteins and polyglutamine-containing peptides. These findings suggest PA28γ could complement 26S function under conditions (such as aging) where ubiquitination is impaired, motivating investigation of PA28γ’s role in cold-induced longevity and proteostasis.

C. elegans: Worms were maintained at 20 °C and shifted after development to 15 °C, 20 °C, or 25 °C. Sterile fer-15(b26);fem-1(hc17) and wild-type N2 strains were used alongside mutants (trpa-1(ok999)) and transgenics expressing aggregation-prone proteins (neuronal polyQ19/67, muscle polyQ40, neuronal FUSP525L, neuronal TDP-43M337V). psme-3 overexpression lines were generated under the sur-5 promoter; endogenous PSME-3 was GFP-tagged by CRISPR. RNAi targeting psme-3 and transcription factors (daf-12, nhr-49, mdt-15, daf-16) or rpn-6.1 was delivered by feeding, including tissue-specific RNAi using sid-1 or rde-1 rescue strains (germline, intestine, muscle, neurons). Lifespan assays (n=96/condition) started at day 1 adulthood; cold-shock survival was tested by 12 h at 4 °C. Motility (thrashing) was measured at day 3 adulthood. Proteasome activity assays measured trypsin-like (Ac-RLR-AMC), chymotrypsin-like (Z-GGL-AMC), and caspase-like (Z-LLE-AMC) activities in lysates by fluorescence kinetics. Protein levels and aggregation were quantified by SDS-PAGE/Western blot and filter trap assays (SDS-resistant aggregates) with antibodies to GFP (polyQ), FUS, TDP-43, IFB-2, proteasome subunits. qPCR measured mRNA levels (normalized to cdc-42 and Y45F10D.4); Statistics used paired/unpaired two-tailed t-tests with FDR correction. Human cells: HEK293T/17 cells were cultured at 37 °C and shifted to 36 °C (24 h) or 35 °C. TRPA1 function was modulated via shRNA or HC-030031 (25 µM), PSME3 via shRNA or lentiviral overexpression. Models expressed Q23-HTT-GFP or Q100-HTT-GFP, FUS(WT)-HA or FUS(P525L)-HA. Proteasome activities were assayed as in worms; native blue gel immunoblotting assessed PA28γ complex assembly. iPSCs (FUSWT/WT and FUSP525L/P525L) were differentiated to motor neurons using a monolayer protocol with stage-specific small molecules; apoptosis was assessed by cleaved caspase-3 immunostaining. Subcellular localization studies used immunocytochemistry; leptomycin B (20 nM, 6 h) inhibited nuclear export to test compartment-specific degradation. Statistical analyses used paired/unpaired t-tests with FDR correction.

• Cold selectively induces proteasome trypsin-like activity via PA28γ/PSME-3 in C. elegans:
  • Shifting adults to 15 °C increased trypsin-like activity, with no increase in caspase-like or chymotrypsin-like activities; effect seen in sterile and wild-type worms and as early as day 1–3 adults.
  • Proteomics and Western blots showed increased PSME-3 protein and mRNA at 15 °C; 19S/20S core subunits largely unchanged.
  • psme-3 RNAi specifically blunted the 15 °C-induced trypsin-like activity, with minimal effect at 20 °C; psme-3 overexpression further increased trypsin-like activity at 15 °C and 25 °C, but not at 20 °C.
• TRPA-1–NHR-49 axis drives psme-3 induction at cold temperature:
  • trpa-1 mutants showed reduced trypsin-like activity and reduced PSME-3 protein/mRNA induction at 15 °C; knockdown of nhr-49 (and co-regulator mdt-15) suppressed psme-3 induction and trypsin-like activity, whereas daf-16 was dispensable.
  • trpa-1 loss did not further reduce trypsin-like activity when psme-3 was knocked down, placing TRPA-1 upstream of PSME-3.
• PSME-3 is required for cold-induced longevity and acts across tissues:
  • psme-3 RNAi shortened lifespan at 15 °C but not at 20 °C or 25 °C.
    • 15 °C: Vector RNAi 28.26 ± 0.69 d vs psme-3 RNAi 24.19 ± 0.67 d.
    • 20 °C: 18.92 ± 0.34 d vs 18.71 ± 0.42 d (NS).
    • 25 °C: 13.03 ± 0.30 d vs 13.02 ± 0.30 d (NS).
  • rpn-6.1 RNAi reduced lifespan at all temperatures (15 °C: 21.13 ± 0.62 d vs 13.73 ± 0.36 d; 20 °C: 16.60 ± 0.55 d vs 10.72 ± 0.23 d; 25 °C: 12.75 ± 0.44 d vs 7.98 ± 0.15 d).
  • Tissue-specific psme-3 RNAi at 15 °C reduced lifespan when targeted to germline, neurons, intestine, or muscle (largest effect in germline).
  • psme-3 overexpression in somatic tissues: decreased lifespan at 20 °C (17.19 ± 0.41 d vs 15.48/16.31 d) and 25 °C (13.83 ± 0.31 d vs 10.60/10.67 d), but extended lifespan at 15 °C (23.26 ± 0.43 d vs 26.65/26.38 d).
  • psme-3 overexpression did not enhance survival after acute 4 °C cold shock.
• Cold-induced PSME-3 ameliorates age-related proteostasis deficits:
  • At 15 °C, aged worms had reduced intestinal IFB-2 accumulation and aggregation; psme-3 RNAi abolished these improvements.
• Cold reduces aggregation of disease-related proteins via PSME-3 in C. elegans:
  • Neuronal polyQ67: 15 °C lowered total polyQ67 levels and SDS-resistant aggregates vs 20–25 °C; psme-3 RNAi increased aggregation at 15 °C and worsened motility; neuron-specific psme-3 RNAi was sufficient to increase neuronal aggregation. psme-3 overexpression reduced polyQ67 levels/aggregation and improved motility at 25 °C.
  • Muscle polyQ40: 15 °C reduced levels and aggregation; psme-3 RNAi blocked these benefits and impaired motility improvements.
  • ALS models: 15 °C decreased mutant FUSP525L and TDP-43M337V levels and aggregation; psme-3 RNAi restored higher levels/aggregation and negated motility benefits.
• Human cell findings (HEK293 and iPSC-derived motor neurons):
  • 36 °C (24 h) selectively increased trypsin-like activity without altering chymotrypsin- or caspase-like activities; 35 °C reduced trypsin- and caspase-like activities.
  • TRPA1 knockdown or HC-030031 inhibited the 36 °C induction of trypsin-like activity.
  • 36 °C increased PSME3 transcript and protein, and assembly into PA28γ complexes; PSME3 shRNA blocked trypsin-like induction, while PSME3 overexpression increased trypsin-like activity at 36 °C and at 37 °C.
  • Mutant HTT (Q100): 36 °C reduced protein levels and SDS-resistant aggregates; TRPA1 knockdown/inhibition or PSME3 knockdown prevented these effects. PSME3 overexpression reduced mutant HTT aggregation at 37 °C.
  • Mutant FUSP525L: 36 °C reduced levels and aggregates; PSME3 knockdown or TRPA1 inhibition/knockdown abrogated benefits. PSME3 overexpression reduced FUS aggregates at 37 °C.
  • ALS iPSC-derived motor neurons (FUSP525L/P525L) showed elevated apoptosis vs isogenic controls at 37 °C; 36 °C reduced apoptosis. Motor neurons had higher basal trypsin-like activity than HEK293; 36 °C further increased trypsin-like activity without affecting other proteasome activities. PSME3 knockdown reduced trypsin-like activity (more strongly at 36 °C) and blunted the anti-apoptotic effect of 36 °C.
• Subcellular mechanism:
  • PSME3 predominantly nuclear in HEK293, present in soma and nucleus of human motor neurons; distribution unaffected by 36 °C. Leptomycin B increased nuclear retention of FUSP525L and further enhanced its cold-induced degradation (but not at 37 °C), suggesting preferential nuclear degradation for FUS. HTT remained cytoplasmic and its degradation at 36 °C was unaffected by leptomycin B, indicating cytoplasmic PA28γ-proteasomes can act on HTT.

The study demonstrates that moderate cold activates a regulated proteostasis program rather than merely slowing metabolism. Sensing of cold via TRPA-1 induces the transcription factor NHR-49 and upregulates PA28γ/PSME-3, selectively elevating proteasomal trypsin-like activity. This mechanism is required for lifespan extension at 15 °C, acting across germline and somatic tissues, and counteracts age-associated declines in ubiquitination-dependent protein clearance by enabling ubiquitin-independent degradation of aggregation-prone proteins. PA28γ/PSME-3 reduces aggregation and toxicity of disease-related proteins (polyQ, FUS, TDP-43) in C. elegans and promotes degradation of mutant HTT and FUS in human cells at 36 °C, with TRPA1 upstream of PSME3 in cold signaling. In human cells, PSME3 overexpression can boost trypsin-like activity even at 37 °C and mitigate aggregation, indicating translational potential independent of cooling. Compartment-specific analyses suggest PA28γ-proteasomes act in nucleus (FUS) or cytoplasm (HTT) depending on substrate localization. Collectively, the findings link environmental temperature sensing to proteasome regulation and lifespan, revealing a conserved, targetable axis for proteostasis and neuroprotection.

Cold temperature triggers a conserved TRPA-1→NHR-49→PA28γ/PSME-3 pathway that selectively increases proteasomal trypsin-like activity, extends lifespan at 15 °C in C. elegans, and reduces aggregation and toxicity of disease-associated proteins across tissues. In human cells, moderate cooling to 36 °C similarly induces PA28γ-proteasomes via TRPA1, decreasing mutant HTT and FUS aggregation and reducing apoptosis in ALS patient-derived motor neurons. Overexpressing PSME3 at 37 °C enhances trypsin-like activity and suppresses aggregation, suggesting therapeutic avenues without systemic cooling. Future work should identify cold-activated cofactors and normal-temperature inhibitors/activators of PA28γ, define direct substrates in vivo, and explore safe pharmacologic strategies to enhance PA28γ-proteasome function in neurodegenerative disease.

• Extreme cold (4 °C) remains harmful; PA28γ/psme-3 overexpression did not improve survival after cold shock.
• psme-3 overexpression shortens lifespan at standard and warm temperatures in C. elegans despite reducing aggregation, indicating a trade-off and context dependence.
• In C. elegans, PSME-3 overexpression alone does not increase trypsin-like activity at 20 °C, implying unidentified cold-induced activators or normal-temperature inhibitors; these regulators are not yet defined.
• Cellular studies may not capture organismal inter-tissue signaling; differences between species and culture models may affect PSME3 regulation.
• Exact human journal publication day is not specified in the provided text.