In the second week of October, a line from a German university press office went around the world: blood cells from an 80-year-old had been turned into stem cells with a molecular age of under 20.[2] Sixty years, gone. The story was picked up by science outlets through the following fortnight, usually under some variation of scientists reverse 60 years of aging.[3]
That number is real. It is also, oddly, the weakest thing in the paper. The study behind it — published in Aging Cell by Lea Jessica Berg and colleagues at the University Hospital of Bonn, the University of Bonn and RWTH Aachen, in the lab of Oliver Brüstle — contains two findings that are far stranger and far more useful, and neither made the headlines.[1]
Two factors, and no detour through pluripotency
Start with what was actually done, because the method matters more than the number.
The standard way to make a young cell out of an old one is the Yamanaka route: hit a skin or blood cell with four reprogramming factors, push it all the way back to a pluripotent, embryo-like state, then coax it forward into whatever tissue you want. It works, and it resets the aging clock almost completely. It also produces a cell that can form a teratoma — a disorganised tumour — which is one reason the technique has stayed largely in the dish. We covered where that line of work stands in our piece on cellular reprogramming and the hunt for age reversal.
The Bonn team took the other road. They used only two factors, SOX2 and cMYC, overexpressed in erythroid progenitors — the precursor cells that go on to become red blood cells, easily grown from a routine blood draw. Those cells were converted directly into induced neural stem cells (iNSCs), the kind of cell that gives rise to neurons and glia. At no point did they pass through the pluripotent state.[1]
That by itself was known to produce some de-aging. What nobody had done was watch it happen frame by frame, with donors spanning the full human range: from newborn cord blood to a 101-year-old.
The number that beats “80 to under 20”
The paper does not report the result as a subtraction. It reports it as a proportion. Across donors, the converted iNSCs preserved roughly 13% of the original donor age at low passage, and around 5% after extended culture.[1]
That framing carries more information than the press-release version, because a fixed percentage means the reset scales with how old you started. Subtracting sixty years from eighty is a nice headline; retaining five per cent of whatever you brought in is a claim about near-complete erasure. The table below works the published percentages through a few donor ages — the arithmetic is mine, the percentages are the paper's.
| Donor age at blood draw | iNSC, low passage (~13% retained) | iNSC, high passage (~5% retained) |
|---|---|---|
| 101 years | ≈ 13 years | ≈ 5 years |
| 80 years | ≈ 10 years | ≈ 4 years |
| 60 years | ≈ 8 years | ≈ 3 years |
| 30 years | ≈ 4 years | ≈ 1.5 years |
Two things follow. First, the widely quoted “under 20” figure for an 80-year-old donor is, by the paper's own numbers, conservative. Second — and this is the part worth sitting with — the de-aging kept going with time in culture. The cells at high passage were younger than the cells at low passage. Whatever is doing the erasing did not finish when the new cell identity appeared.
Why “slow” is the actual finding
There is a deflationary explanation for epigenetic rejuvenation in a dish that the field has had to rule out repeatedly: dilution. If a cell divides often enough, and the machinery that copies methylation marks onto the new DNA strand is imperfect, the old pattern simply gets watered down. No active rejuvenation required — just arithmetic and a lot of cell divisions.
Berg and colleagues tested this directly. They ran the conversion under proliferation-inhibiting conditions — cells held back from dividing — and the epigenetic de-aging continued anyway.[1] That is the single most important sentence in the paper. It means something is actively remodelling the methylation landscape, not passively losing it.
And it is unhurried. The press release describes molecular age falling gradually across more than 100 days of tracking, with the bulk of the epigenetic reset playing out over roughly the first 50.[2] Compare that with iPSC reprogramming, where the clock crashes to near zero within a few weeks and there is nothing much to observe in between. The authors are explicit about why this matters: the slowness makes iNSC conversion an unusually good model system for dissecting how de-aging works mechanistically. You can take samples along the way.
Age and identity came apart
The second surprise sits in the time-course data. The team tracked DNA methylation and RNA sequencing across the whole conversion, and found that acquiring a genuine neural stem cell signature took considerably longer than the appearance of the first obvious marker — proliferating, PAX6-positive cells that look, by the usual checks, like iNSCs already.[1] Transcriptional differences between young-donor and old-donor cells also faded with extended time in conversion, rather than vanishing at the moment of identity switch.
This speaks to a problem that has shaped the whole partial-reprogramming field. In 2019, Olova and colleagues showed that during partial reprogramming, epigenetic age declines steadily before the cell loses its somatic identity — implying a window in which you could get the rejuvenation without the identity loss.[4] That window is the entire premise of in-vivo reprogramming work, from Ocampo's cyclic expression in mice[5] to the restoration of vision in aged mouse retinas[6] and the multi-omic rejuvenation of human fibroblasts without loss of cell type.[7]
The Bonn data say the two processes are even more loosely coupled than that — they run on different timescales in a system where the cell is explicitly meant to change identity. If age erasure and fate acquisition can be pulled apart in time, they can in principle be targeted separately. That is a far more actionable idea than “we made old cells young.”
What this does not mean
Four caveats, and none of them are small.
Nobody was rejuvenated. The donor's aged blood cells were not restored. They were consumed and replaced by a different cell type that happens to carry a young epigenetic signature. The aged cells in a living body — the ones causing the problems — are untouched by this. The distinction between reversing biological age in an organism and resetting a clock in a dish is the whole ballgame, and it is routinely collapsed in coverage.
The clock is being read outside its training range. Epigenetic clocks were built by fitting methylation patterns to chronological age in human tissue samples.[8] Applying one to a cultured, transcription-factor-converted cell type is an extrapolation. A reading of “under 20” tells you the methylation pattern resembles that of young tissue at the clock's chosen sites; it does not establish that the cell behaves like one from a 20-year-old. If you want the longer version of that caution, see our breakdown of how accurate epigenetic age tests actually are.
cMYC is an oncogene. It is in the toolkit because it is good at forcing cells to change state, and that is precisely why it is a liability in anything heading toward a clinic.
This is a mechanism paper, not a therapy paper. The authors frame it that way themselves. The value is that a slow, observable, non-pluripotent de-aging system now exists to study — which is what the field has been missing. The value is not a treatment.
So where does it leave things
Honestly: further along than it looks, in a less exciting direction than the headline promises.
The realistic near-term application of iNSCs is not rejuvenation at all. It is modelling. Neural stem cells derived from an elderly donor with, say, Alzheimer's disease ought to carry that person's genetic background — but if conversion wipes the epigenetic age, you get a young cell with old genes, which is useful for some questions and actively unhelpful for others. Several labs are trying to make reprogrammed neurons that keep donor age, precisely so that age-related disease can be studied in a dish. This paper tells them how much age survives conversion, and when.[1] Five per cent is a hard number to work with.
The longer-range implication is the decoupling. Every in-vivo rejuvenation strategy depends on erasing epigenetic age without erasing what the cell is. The Bonn time course suggests that gap is wider than assumed — weeks wide, and sustained without cell division. That is a lever, if someone can find the handle. For the rest of us, in the meantime, the uneven pace at which our own tissues age remains a more immediately relevant subject than a flask of very young brain cells in a Bonn incubator.
Common questions
What is epigenetic de-aging?
It is the reversal of the chemical marks — mostly methyl groups attached to DNA — that aging clocks read to estimate how old a cell is. The underlying DNA sequence is untouched; what changes is the pattern governing which genes get read. In this study, cells converted from blood to neural stem cells held onto roughly 13% of their donor's epigenetic age early on, and about 5% after extended culture.[1]
Did scientists really reverse 60 years of aging?
In cultured cells, by the measure they used, yes. Blood cells from an 80-year-old became neural stem cells reading under 20 on an epigenetic clock.[2] But it happened in a plastic flask across roughly 100 days, and it produced a new cell type rather than rejuvenating the donor's existing cells inside the donor. No person got younger.
Could this ever treat brain aging?
Not in this form. Transplanting lab-grown neural stem cells into a human brain is a separate and considerably harder problem, and one of the two factors used here is an oncogene. The plausible route is the opposite direction: use this slow system to work out how the epigenetic reset happens, then try to trigger that mechanism without the full conversion. If you are reading this because you are worried about your own brain, the evidence that currently has outcome data behind it is covered in what actually changes in the brain after 50.