The 'Death Protein' That Forgets to Kill

There is a protein inside your cells whose entire job — as far as anyone knew — was death.

Its name is MLKL, short for Mixed Lineage Kinase domain-Like protein. When a cell receives the signal to die via a form of programmed destruction called necroptosis, MLKL springs into action. It gathers into clusters, migrates to the plasma membrane, punches holes in it, and the cell ruptures. Fast, violent, inflammatory. Necroptosis evolved as an emergency response — a way for cells to die loudly and alert the immune system when quieter forms of death (like apoptosis) are blocked.

That, scientists thought, was the whole story.

It wasn’t.

A study published in April 2026 in Nature Communications by researchers at the University of Tokyo, Tohoku University, and St. Jude Children’s Research Hospital has uncovered a deeply unexpected second life for MLKL — one that doesn’t involve death at all, but instead involves a slow, cumulative damage to mitochondria inside blood stem cells. And this non-lethal mischief, the paper argues, is a major driver of why our blood and immune system declines as we age.

---

Your Blood Is Made Fresh Every Day

Before diving into the discovery, a quick refresher on what’s at stake.

Your body makes roughly 500 billion blood cells every single day. Red blood cells, white blood cells, platelets — all of them trace their lineage to a tiny pool of hematopoietic stem cells (HSCs) tucked deep in the bone marrow. These cells are extraordinary. They can renew themselves indefinitely and produce every type of blood and immune cell the body needs. Without them, blood stops being made.

But HSCs aren’t immortal in any practical sense. As we age, something goes wrong with them. Old HSCs gradually lose two critical abilities: they become less capable of replenishing themselves (reduced regenerative potential), and they start producing an imbalanced mix of cells — specifically, they skew heavily toward myeloid cells (like monocytes and granulocytes) at the expense of lymphoid cells (B and T immune cells). This is a hallmark of aged blood, and it comes with consequences: weaker immune responses, higher risk of blood disorders, and increased susceptibility to myelodysplastic syndromes (MDS) — a category of blood cancers that predominantly strike the elderly.

Researchers have known about this phenomenon for decades. What has remained stubbornly unclear is the molecular mechanism — the precise chain of events that turns a young, capable HSC into an old, dysfunctional one.

---

A Paradox Hidden in Plain Sight

Here is the puzzle that led to this discovery.

Multiple previous studies had flagged the RIPK3–MLKL pathway — the core machinery of necroptosis — as relevant to HSC aging. When MLKL was active in HSCs, those cells performed worse. Conversely, blocking MLKL seemed to protect HSC function. The assumption was that some HSCs were simply being killed by necroptosis during inflammation or other stresses, and those losses accumulated over time.

But there was a problem with this explanation. In aged bone marrow, you actually see more dysfunctional HSCs accumulating, not fewer. If necroptosis was killing them off, you’d expect a depletion. Instead, the bone marrow fills up with old, myeloid-biased stem cells that persist but can’t do their job properly.

The Tokyo team, led by Masayuki Yamashita, decided to look more carefully. What they found upended the standard picture.

Using a sophisticated fluorescent biosensor system in genetically modified mice — called SMART-Tg mice — that reports MLKL activation in real time within individual cells, the researchers confirmed something puzzling: MLKL was being activated in HSCs during inflammation, but those HSCs were not dying.

When they exposed mice to inflammatory signals (polyinosinic-polycytidylic acid, bacterial lipopolysaccharide, or TNF-α), MLKL lit up in HSCs and early progenitor cells. But when they counted HSC numbers and looked for signs of plasma membrane damage — the telltale hallmark of necroptotic death — there was nothing. The cells survived. Their membranes stayed intact. Their numbers didn’t drop.

The death protein was activating, then… not killing.

---

What MLKL Does Instead

If MLKL isn’t blowing up the cells, what is it doing?

The answer turned out to be written on the mitochondria.

When the team examined the mitochondria of aged HSCs using transmission electron microscopy — taking extraordinarily detailed physical snapshots of these organelles — they found something disturbing in old wild-type mice: elongated, swollen mitochondria with disorganized cristae (the internal membrane folds where cellular respiration happens). These are the textbook features of dysfunctional mitochondria, the kind seen in cells that have been through rounds of replication stress.

In aged mice that lacked MLKL entirely (Mlkl–/– mice), this mitochondrial deterioration was dramatically reduced. The mitochondria looked healthier. They maintained better membrane potential — the electrical charge across the inner membrane that drives ATP production. And critically, aged MLKL-deficient HSCs showed improved glycolytic metabolism compared to aged wild-type HSCs.

The team then used immunoelectron microscopy and a technique called proximity ligation assay to ask: where exactly is MLKL going inside the cell? The answer was remarkable. Phosphorylated, activated MLKL accumulated specifically inside the mitochondria of aged HSCs. Not at the plasma membrane — the location that drives cell death — but at the mitochondrial membranes.

To confirm that this was a direct effect of MLKL on mitochondria — not an indirect consequence of inflammation or gene expression changes — they isolated mitochondria from liver cells and added recombinant MLKL protein directly. The result: MLKL’s active N-terminal domain (the part that oligomerizes and forms pores) directly disrupted mitochondrial membrane potential even when no cells were involved. MLKL can physically attack the energy machinery of cells all by itself.

---

Multiple Stresses, One Convergence Point

One of the most elegant aspects of this study is its breadth.

The RIPK3–MLKL axis is activated in HSCs not just by inflammatory signals but also by replication stress (caused here experimentally by the chemotherapy drug 5-fluorouracil, which forces HSCs to proliferate) and by oncogenic stress (a mutant version of RUNX1, the transcription factor whose dysfunction drives myelodysplastic syndrome). In each case:

• MLKL activated in HSCs without killing them
• HSC function declined
• Mice lacking MLKL were significantly protected from these declines

And during natural organismal aging — comparing three-month-old mice to 18-month-old mice — activated MLKL accumulated in HSC mitochondria in a RIPK3-dependent manner. Old MLKL-deficient mice had healthier blood stem cells with better engraftment potential, less myeloid skewing, and fewer of the DNA damage markers (γH2AX foci) that serve as aging signatures.

The picture that emerges is of a mechanism that acts as a common downstream receiver for many different stress signals. Inflammation, replication pressure, oncogenic insult — all roads lead to MLKL activation in HSCs. And activated MLKL, rather than killing those cells outright, silently damages their mitochondria and degrades their function over time.

---

Why Would Biology Do This?

It’s a fair question. If MLKL causes such damage, why does it exist? Why hasn’t evolution eliminated it?

The researchers offer a compelling answer. When the body is under acute stress — fighting an infection, for example — HSCs need to ramp up production of immune cells quickly. The RIPK3–MLKL axis, activated by inflammatory signals, may help force this rapid response. Pro-survival signals (like the NF-κB pathway) can simultaneously prevent the cells from undergoing full necroptotic death, allowing MLKL to activate in a sublethal way. The cells survive and keep producing blood, even at a cost to their long-term function.

In the short term, this trade-off — sustain blood production now, pay the mitochondrial price later — is worth it. Over a lifetime of repeated stress exposures, however, the cumulative damage adds up. Each bout of inflammation, each round of forced proliferation, leaves another deposit of MLKL-mediated mitochondrial damage. The HSCs that survive these episodes are functional enough to persist but progressively less capable of the balanced, robust blood production characteristic of youth.

Interestingly, the naked mole rat — famously long-lived and resistant to age-related tissue dysfunction — appears to have lost functional RIPK3 and MLKL, consistent with the idea that this pathway is part of what drives aging in species that do have it.

---

What This Means for Medicine

The obvious question is whether this knowledge could eventually lead to therapies.

The paper points toward several possibilities. Blocking RIPK3 (which sits upstream of MLKL) already has pharmacological tools in development — a drug called UH15-38, tested in this study, showed a trend toward protecting HSC function during inflammatory stress. MLKL inhibitors are less developed but represent a logical target.

The researchers are careful to note important caveats. This work was done in mice, and human MLKL may behave somewhat differently — the molecular details of how MLKL is regulated and executes necroptosis differ between the two species. Whether the mitochondrial damage mechanism plays out identically in human HSCs remains to be studied.

There is also a complexity worth acknowledging: MLKL appears to have opposing roles depending on context. In aging HSCs, it’s harmful. But other studies suggest MLKL may inhibit leukemia development by promoting the differentiation of leukemia stem cells. Any therapeutic intervention would need to thread a careful needle.

Still, the conceptual contribution here is significant. For decades, research into blood aging has catalogued its features — the myeloid skewing, the mitochondrial dysfunction, the loss of regenerative capacity — without identifying a unifying molecular thread. The RIPK3–MLKL axis now offers a candidate for that thread: a single pathway that receives inputs from many types of cellular stress and converts them into the shared phenotype of an aged blood stem cell.

---

The Protein That Does Too Much

What strikes me most about this discovery is how it reframes what we thought was a simple binary. Kill or don’t kill. Die or survive.

Nature, as usual, is more creative than our categories. MLKL can be activated and not kill its host cell. It can instead redirect its membrane-disrupting machinery inward, to a different cellular membrane, with slower but lasting consequences. It’s a protein doing something it wasn’t “supposed” to do — and doing it reliably, repeatedly, across a lifetime.

This is what makes aging biology so fascinating and so difficult. The mechanisms that decline with age aren’t malfunctions in some obvious sense. They’re often the same mechanisms that work perfectly well in the short term, solving acute problems at the cost of long-term wear. The body isn’t broken. It’s optimized for survival to reproductive age, not for the quiet preservation of blood stem cell function into your eighties.

Understanding that distinction — and learning to intervene at the molecular level — is exactly the kind of knowledge that could, someday, change what growing old feels like.

---

Source: Yamada, Y., Yang, J., Saiki-Tsuchiya, A., et al. “Non-necroptotic MLKL function damages mitochondria and promotes hematopoietic stem cell aging.” Nature Communications 17, Article 71060 (2026). DOI: 10.1038/s41467-026-71060-4