Gene activity in the mouse gut changes in abrupt steps, with the largest shift in midlife coinciding with shutdown of the cellular protein quality control system
Gene activity in the mouse gut changes in abrupt steps, with the largest shift in midlife coinciding with shutdown of the cellular protein quality control system
On October 8, the Fritz Lipmann Institute on Aging in Jena, Germany, published a preprint examining gene activity in the colon of 83 male mice across five age groups, from three to twenty-eight months. The largest shift occurred between midlife and old age, when almost three thousand genes changed their activity, with clear signs that the system responsible for monitoring protein folding had shut down.
In 2024, the same group, Steve Hoffmann, Maja Olecka and colleagues, identified two abrupt shifts in DNA methylation in the same colon samples, at 3→9 and 15→24 months. The new gene activity analysis identified almost the same windows: 1618 genes changed at 3→9 months, just 116 during the relatively stable 9→15 interval, and 2722, the largest number, at 15→24. Two molecular layers in the same mice undergo abrupt changes at the same points in life.
The same window also saw the shutdown of UPR, a cellular protein quality control system: the proportion of its key protein XBP1 in the active form fell, even though misfolded proteins generally become more abundant with age. The authors' findings suggest that the cellular stress sensor itself fails:
“Our data suggest a collapse of the protein quality control system itself, rather than a simple depletion of its reserve capacity.”
Other studies have already shown that restoring this kind of sensing activity in old animals can have rejuvenating effects, making the inactive sensor a specific target for future intervention.
These nonlinear changes also complicate the interpretation of p16, a well-known marker of cellular senescence: expression of the Cdkn2a gene rises sharply early in life, then increases only gradually in old age. A similar nonlinear pattern appears in the gut's circadian clock: circadian genes act in coordination in young mice, but that coordination breaks down in some old mice by twenty-eight months.
A similar pattern appears in transcriptomic entropy, a measure of how evenly activity is distributed across genes in the gut. It falls in young mice, then reverses after midlife and rises nonlinearly as the tissue shifts toward a gene expression profile resembling that of stem cells. Increases in this form of entropy across different organs have already been linked to poorer cancer survival. Genes whose expression increases are active in enteric neurons, which make up the gut's nervous system; genes whose expression decreases protect the intestinal barrier and are associated with survival in colorectal cancer. The gut is losing its distinct cellular identity, supporting a prediction Olecka made in 2024:
“Genes that undergo abrupt changes in both DNA methylation and expression have important roles in the intestinal barrier, colorectal cancer and the enteric nervous system.”
Across the full dataset, only nine genes changed significantly at each of the four transitions between age groups. The researchers selected the three with the clearest patterns, Map10, Dmrta2 and Vsig2, and used their relative expression levels to build a model that predicted age to within 0,8 months. After calibration using external data and qPCR (a rapid assay of gene activity), its accuracy was around two months. Such clocks typically rely on hundreds of genes and transfer less reliably between laboratories; this model transferred successfully despite using just three genes.
Some mice die before reaching twenty-eight months, so some signals in the oldest age window may reflect the greater resilience of the survivors. However, if protein quality control, a senescence marker, circadian rhythms and cellular identity all lose coordination within the same window, the timing provides a stronger rationale for intervention in midlife than for treatment over decades.