Important caveat, stated plainly at the start: this discovery was made in mice. It has not yet been proven that a matching, particularly long-lived stem cell population exists in humans. The researchers say they have initial evidence and that testing this in humans is their next project. Nothing here is medical advice, and no new treatment follows directly from this mouse study yet.
Every day, your body makes billions of new blood cells. Red blood cells carry oxygen. White blood cells fight infection. Platelets help blood to clot. None of these cells lives forever, so the supply must be renewed constantly, from birth until old age.
That renewal comes from blood stem cells in the bone marrow. They can copy themselves, so the pool does not run out, and they can produce every major type of blood cell. That combination makes blood stem cell transplants possible, and it is why these cells matter in diseases such as leukaemia and other severe blood disorders.
But every division carries risk. Copying DNA is never perfect, and the more a cell divides, the more errors can accumulate. If blood stem cells divided all the time, they would wear out early. Nature's answer is dormancy: many stem cells spend most of their lives asleep, waking only when needed. How deep can that sleep go — and is there a reserve that sleeps even longer than the rest?
A reserve hiding in plain sight
A team at the German Cancer Research Center (DKFZ) in Heidelberg and the Heidelberg Institute for Stem Cell Technology and Experimental Medicine (HI-STEM) has now reported such a reserve. Their paper, by Dawn S. Lin and colleagues including Maria C. Puschhof and Andreas Trumpp, appeared in Cell Stem Cell in 2026 (DOI: 10.1016/j.stem.2026.09.016). The DKFZ announced it on 2 November 2026 as “Stem Cells in Deep Sleep as a Secret Reserve.”
Scientists identify blood stem cells partly by proteins on their surface, like name badges. One key badge has been CD150. Cells carrying CD150 were used to pick out long-term stem cells, while cells lacking CD150 were often set aside as ordinary progenitors — cells that make blood for a while, but are not thought to sustain it for life.
The new study shows that rule missed something. Hidden among CD150-negative cells is a population marked by another surface protein, EPCR (endothelial protein C receptor). When isolated, these EPCR-positive, CD150-negative cells did not act like short-lived progenitors. They acted like genuine long-term stem cells. The team named them LT-HSC2, distinct from the conventional EPCR-positive, CD150-positive group, LT-HSC1.
Progress sometimes comes not from a new machine, but from re-examining a group dismissed for failing one familiar test. The cells were there all along; the classification needed updating. A similar rethink appears in our article on the immune protein cGAS found to control cell fate, where one protein turned out to do far more than its first known job.
How deep is “deep sleep”?
Dormancy is measured with a labelling trick. Researchers give cells a marker that fades with each division. Cells that stay asleep keep the label; cells that divide repeatedly lose it. After a long wait, the cells still labelled are the true deep sleepers.
Here the wait was more than 200 days — very long in a mouse's life. Afterwards, only about 2 percent of conventional blood stem cells still carried the quiescence marker. Among the new cells, about 10 to 15 percent were still labelled: roughly five to seven times more likely to have stayed asleep throughout.
That matters because dormancy is protection. A cell that does not divide is not copying DNA, not suffering the wear of constant activity, and not burning its regenerative capacity. The researchers suggest this exceptionally long sleep is what lets these cells keep producing blood for an exceptionally long time. They are a distinct reserve, defined by how stubbornly they stay quiet.
“Dormant” does not mean dead or permanently switched off. Think of a backup generator, maintained and ready but not running. When called, these cells can wake, divide, and rebuild. The skill is balance: sleep long enough to stay fresh, yet remain able to respond.
Cautious under stress, different in speciality
Blood stem cells are often woken by inflammatory signals — alarm calls saying more immune cells are needed. That response is vital in infection, but repeated waking depletes the reserve, like spending savings at every emergency.
In these mice, the new cells responded to inflammatory stimuli much more cautiously than conventional stem cells. They were less easily pushed into dividing. That reluctance may be a feature: a reserve that jumps at every alarm would soon be exhausted, while one that waits keeps its strength for the long term. The paper reports reduced cell-cycle activation under inflammatory stress, and says the cells preserve their stemness for prolonged periods in the living animal.
The groups also differ in tendency once they make blood. CD150-positive cells lean toward red blood cells and platelets. The deep sleepers lean toward immune cells, particularly lymphocytes — the white blood cells including T and B cells. Tendency, however, is not capability. Both groups can make all major blood cell types, as the next test showed.
One cell, a lifetime of blood
The gold-standard test is transplantation: place candidate cells in a recipient mouse whose blood system has been cleared, and ask whether they can rebuild it, and for how long.
Even individual cells from the new population generated all blood cell types and sustained them for life. The ability held after serial transplantation — moving cells from one recipient to another, again and again, each time forcing a full rebuild. Short-lived progenitors fail that test. True long-term stem cells pass it. That is why the team calls this group bona fide, meaning genuine.
As first author Dawn Lin put it, blood stem cells “do not constitute a uniform cell population but rather comprise different subpopulations with distinct properties and functions.” Some respond quickly and favour red cells and platelets. Others sleep deeper, respond cautiously, and favour immune cells. Together they may give blood both speed and endurance.
Rest protecting long-term function is a familiar idea at whole-body level too, as discussed in our article on the science of sleep debt and why your body cannot simply catch up later — though human sleep and stem-cell dormancy are very different processes.
Why it could matter — and what it does not prove
Lead author Andreas Trumpp linked the finding carefully to medicine. Blood stem cells underpin transplants for leukaemias and other severe blood disorders. Understanding the deepest, longest-lasting cells might one day make transplants more durable, and help blood and immunity stay effective as people age.
That is a hope, not a promise. Ageing weakens blood and immunity partly because stem cells lose function. A population resisting that decline in mice is a promising place to look — perhaps for ways to protect, expand, or select such cells in transplants. But none of that has been shown in people yet. Basic maps often precede treatments, as with the UBE2N enzyme that could stop fatty liver disease, which we covered previously: a mechanism understood first, a therapy only afterwards, if at all.
Above all, human proof is missing. Mice and humans share much blood biology, but markers and behaviour can differ, and a mouse lives about two years while a human lives for decades. What counts as “exceptionally long” dormancy in a mouse may not translate directly. Trumpp himself said: “We already have initial evidence of this — that's our next project!” Initial evidence is not proof; the next project is where proof will be sought.
This was also not a clinical trial, drug test, or patient treatment. It was a precise characterisation in mice, using markers, labelling, stress tests, and transplantation. Its strength is mouse biology; its limit is that human biology is still unconfirmed. The paper also refines the early blood-development model: adding EPCR to markers CD150 and CD34 clarifies how stem cells gradually lose dormancy, and how early choices split toward red-cell or immune-cell paths.
What to watch next
Three questions will decide this discovery's importance. Do humans have an equivalent EPCR-positive, CD150-negative population, reliably findable in human bone marrow? Do those cells behave the same way — deep dormancy, cautious stress response, immune leaning, lifelong production? And can that knowledge be turned safely into better transplants or protection against ageing-related decline?
Each answer will take time, and each could change the picture. The human equivalent might behave differently, appear only in youth, or not exist in the same form. Any outcome is progress, because a clearer map replaces guesswork.
For now, the story is about hidden structure: cells overlooked for lacking a famous marker, sleeping far longer than their neighbours, waking cautiously, leaning toward immune cells, and — cell by single cell — rebuilding a lifetime of blood in a mouse. If a matching reserve exists in us, it may help explain how our blood lasts a lifetime. If not, searching for it will still have taught researchers to look more carefully.
Key takeaways
- DKFZ and HI-STEM researchers found a new mouse blood stem cell group, EPCR-positive and CD150-negative, named LT-HSC2.
- After 200+ days, 10–15% still carried a dormancy label, versus about 2% of conventional stem cells.
- They respond cautiously to inflammation, lean toward immune cells, yet can make all blood types.
- Single cells passed serial transplants, sustaining all blood types for life in mice.
- A matching human population is not yet proven; researchers say initial evidence exists and human tests are next.
Source: Lin, D. S. et al., “EPCR expression marks CD150neg deeply dormant long-term hematopoietic stem cells,” Cell Stem Cell (2026), DOI: 10.1016/j.stem.2026.09.016; DKFZ press release, 2 November 2026. General information only, not medical advice.