In a lab at the University of Zurich’s Institute for Regenerative Medicine, the experiment that mattered took one hour. A batch of human blood stem cells, the kind pulled from a donor and destined for a transplant, sat in a solution of a protein called activated protein C. Then they went into mice. The cells that had taken the one-hour bath held onto the one quality that makes a blood stem cell worth transplanting: the capacity to renew itself and rebuild a whole blood system for the long haul. The untreated cells did not hold it as well. Treated cells divided less, matured less readily, and stayed in a resting state, and when they were pushed through a second round of transplantation, the brutal standard test of whether a cell is truly a self-renewing stem cell and not a short-lived imitation, the treated ones were still producing human blood. That is the finding Simon Pöllmann, Ute Modlich and colleagues report in EMBO Molecular Medicine, with collaborators at the University of California, Santa Cruz.
The protein in that solution is not a newcomer, and that is where the story gets interesting. Activated protein C, or aPC, already had a career in medicine, and it ended badly. In recombinant form it was sold as Xigris, a treatment for severe sepsis, authorized in Europe in 2002 and marketed for the better part of a decade on the theory that its anti-inflammatory, anticoagulant properties would save the sickest patients. Then the confirmatory trial came in. In PROWESS-SHOCK, 28-day mortality was 26.4 percent with the drug versus 24.2 percent with placebo across 1,680 patients, a difference that pointed the wrong way and was not statistically significant. Eli Lilly withdrew the product worldwide on October 25, 2011. The molecule that could not prove it helped dying adults is now the promising agent in a dish of stem cells.
| Group | Value (percent) |
|---|---|
| Xigris | 26.4 |
| Placebo | 24.2 |
That is not the contradiction it looks like, and the reason is worth spelling out. The hard problem in modern blood stem cell transplants and gene therapies is not getting cells into a patient. It is keeping them from spoiling before they get there. The cells are harvested, cultured, sometimes genetically edited over days in a dish, and the dish is exactly where they betray you. Outside the bone marrow’s protective niche, they do what cells do when prodded: they divide, they differentiate, they burn through their potential and turn into ordinary blood. By the time they are infused, the rarest and most useful ones, the self-renewing cells that rebuild an entire blood system, have thinned out. A one-hour treatment that tells those cells to sit still and stay primed is, if it holds up, a useful trick.
The mechanism rewards a close look. Activated protein C works through a receptor called PAR1. So does thrombin, the clotting protein. Same receptor, opposite instructions: thrombin pushes the cells to divide and differentiate, and aPC does the reverse, a split the signaling literature calls biased agonism, where the same doorbell rings two different alarms depending on who leans on the button. The Zurich group did not stumble into this blind. Back in 2019, Modlich’s team, then at the Paul-Ehrlich-Institut, identified the receptor EPCR as a marker of the blood stem cells that actually engraft, work published in Blood. EPCR is a co-receptor for the PAR1 pathway and aPC is its natural ligand, so the new paper is the payoff on that lead: having found the lock, they went and tried the key.
Now the discipline, because the way this result is being summarized, “enhances blood stem cell transplantation outcomes,” is a phrase doing more work than the data can bear. No patient was transplanted. The improved engraftment, the better output of human blood, the survival through serial transplantation, all of it happened in immunodeficient mice carrying human cells. That is a legitimate and demanding model, the one the field uses precisely because it is hard to fool. What it cannot tell you is whether the trick survives the thing that would actually use it: a clinical manufacturing run, where a patient’s graft is cultured and edited under GMP conditions on its way to a real bloodstream, and where the test is not how well a mouse refills its marrow but how well a leukemia patient rebuilds a blood system that stays rebuilt. That experiment has not been run, and the distance to it is measured in years and in trials.
What would change, if the trick survives that trip, is narrow and worth being precise about. It is a brief, pre-transplant conditioning step for donor grafts and for the cultured, edited cells used in gene therapy, the setting where keeping the cells from spoiling in the dish is the whole game. No new infusion into patients, no systemic exposure, just a treated graft that arrives with more of its best cells intact. That is a far smaller claim than “better transplants,” and it is the one the evidence points at. It is also why the Xigris history is reassurance rather than warning: aPC does many things in a living bloodstream, and the failure in sepsis says nothing about what an hour in a dish does to a cell before it is washed and infused.
Modlich has spent the better part of a decade following EPCR from a marker on a mouse stem cell to a druggable pathway on a human one, and she has now shown that leaning on that pathway the right way keeps the cells that matter from throwing themselves away in culture. In mice. Whether the cells in a Zurich dish behave like the cells in a patient’s graft is the one experiment that will settle it, and it is the one still waiting to be done.
Sources
- EMBO Molecular Medicine, Pöllmann et al., “Activated protein C promotes hematopoietic stem and progenitor cell quiescence and engraftment” (2026)
- News-Medical / University of Zurich IREM, “New signaling pathway enhances blood stem cell transplantation outcomes” (Oct 9, 2026)
- Paul-Ehrlich-Institut (via idw), Kohlscheen, Modlich et al., EPCR supports stem-cell engraftment, Blood (2019)
- European Medicines Agency, Xigris (drotrecogin alfa activated) withdrawal notice (2011)