THE ARISING · Longevity & Healthspan
Causal Gene Network Made 55-Year-Old Cells Resemble 25-Year-Old Cells
Skin cells from 55-year-old donors, grown in a dish, have been shifted toward the profile of cells from 25-year-olds. That is the claim of a new preprint on bioRxiv. It rests on a way of separating genes that merely change with age from genes that may drive those changes.
This is early evidence. The work is a preprint, so it has not been peer reviewed. It combines a computer analysis of gene activity data from dermal fibroblasts (connective-tissue cells in skin) of 490 donors in the GTEx collection with experiments in fibroblasts from 55-year-old donors. It is lab work in cells, not a study in people or animals.
The problem the authors set out to solve is a familiar one. Human aging is mostly studied by comparing different people at different ages, not by following the same people over time. According to the authors, it was unclear whether such crosssectional data hold enough information to reveal cause and effect. They report that they do.
Their tool is convergent cross mapping, a method for finding causal links in nonlinear dynamical systems. The team ordered the donors' fibroblast gene data by age. From that, they reconstructed directed gene-to-gene interactions, meaning which gene influences which. The result was a genome-scale causal network. The authors combined its structure with age-related changes in gene activity to predict candidate drivers of aging.
Then they tested the predictions. They boosted single genes in fibroblasts from 55year-old donors and measured the effects with single-cell RNA sequencing. 53.8% of the testable predicted interactions were confirmed. The authors say that rate beat several comparison methods and a random null. It also means that just under half of the testable predictions were not confirmed.
The team then asked whether the network could point to ways of reversing cellular aging. Two independent measures, one based on gene activity and one on highdimensional cell shape, showed that the selected genetic changes moved old fibroblasts toward young-cell states. The two measures were strongly correlated. With the strongest interventions, a majority of the 55-year-old cells came to resemble cells from 25-year-old donors.
There are clear limits. The findings come from one cell type in culture. Resembling young cells on these measures is not the same as working like young cells, and it says nothing about health or lifespan in a person. This work does not bear on any personal health decision today.
Its value, if it holds up, is in method. The authors suggest it could give researchers experimentally testable targets for altering the aging phenotype, and it widens the set of ideas for how cells might be rejuvenated. What to watch next: peer review of the full paper, independent replication, and tests in other cell types and in living organisms.