Senescent cells produce more clearance proteins yet clear less debris: stage-resolved proteomics traces the collapse of the cell's entire protein quality-control system
Senescent cells produce more clearance proteins yet clear less debris: stage-resolved proteomics traces the collapse of the cell's entire protein quality-control system
Scientists at the Institute of Molecular Biology in Mainz used mass spectrometry to track how the protein composition of a human cell changes across four stages of senescence. Each step in the breakdown of protein quality-control systems has a precise, experimentally verified cause.
A cell enters senescence when it exhausts its division limit. Telomeres, the protective caps at the ends of chromosomes, shorten after dozens of doublings until division stops permanently. This limit was discovered in 1961, but how the cell's proteins changed along the way was unknown: individual markers had been observed only in isolation. The authors of a study published on September 16 in Nature Communications drove the human lung fibroblast line IMR90 to replicative senescence and, rather than taking a single "young cell versus old cell" snapshot, measured the protein composition at four sequential stages of that trajectory, quantifying 5,923 proteins at each.
The breakdown inside the cell follows a strict order, step by step. Nuclear and chromatin proteins, the molecular scaffold on which DNA is packaged, are lost earliest. DNA replication proteins and one class of histones (the spool proteins that wind DNA) persist the longest. Nearly a tenth of the changing proteins behaved in opposition to their own mRNA: proteins that promote gene transcription accumulated at the mRNA level but disappeared at the protein level. Part of the breakdown, in other words, occurs after the gene is read, at a stage invisible to standard RNA profiling methods. The same discordance was found in the brain of the killifish, a model organism for aging research.
The most unexpected finding concerns the cell's clearance systems: autophagy (the digestion of damaged cellular components) and the proteasome (the molecular shredder that degrades unwanted proteins). More protein subunits were present, yet the machinery stalled: autophagic flux and proteasome activity both declined. Energy supply was not the explanation. ATP levels in senescent cells were already 4.4-fold lower, but when ATP was halved in young cells, proteasome activity was unaffected. The cause lay elsewhere: the fraction of immunoproteasome subunits, those normally switched on during infection, had increased.
The chemistry of degradation tags changed as well. Ubiquitin formed atypical chain linkages more frequently, linkages less efficient at routing proteins to the proteasome. Among the enzymes that attach ubiquitin, 19 declined and only three increased. The second modifier, SUMO, followed the same pattern: the free pool was depleted while the SUMO load on proteins that aggregate into insoluble clumps grew. This mirrors the way toxic aggregates accumulate in neurodegenerative diseases.
550 proteins lost solubility entirely. When ATP was added back to the cell extract, three quarters of them re-dissolved. Aggregation turned out to be an active, energy-dependent process rather than passive wear. The exception was mitochondrial proteins: their aggregation could not be reversed by restoring energy, making them the hardest target for repair. The same ordering was predicted in March by a yeast aging atlas: proteostasis failed before mitochondria, not the other way around.
To rule out the explanation that the observed changes were simply a stress response, the authors subjected young cells to six different challenges, including proteasome inhibition, autophagy blockade, oxidative stress, and the senescence mimetic Inflachromene. None reproduced the full pattern. Replicative senescence has its own proteomic signature.
The search for senolytic targets, vulnerable points whose disruption could selectively kill senescent cells, is one of the goals the authors name for this work. Yet the map they have built opens another path: repairing the proteasome, the ubiquitin tagging system, SUMO conjugation, or mitochondria within the senescent cell itself.