The body's method for eliminating damaged cells is called apoptosis. An initiator caspase enzyme activates. Executioner caspases dismantle the cell from inside. The process is irreversible. It has worked this way across species for roughly 600 million years.
A team at the Weizmann Institute of Science found cells that start the sequence and then stop it.
Their study, published in Nature Communications, exposed fruit fly larvae to ionizing radiation and tracked tissue repair at single-cell resolution. Some irradiated cells activated the initiator caspase, beginning the death cascade, and then halted before executioner caspases could finish. Those cells survived. The researchers named them DARE cells.
The Motor
The survival mechanism is molecular. A motor protein called Myo1D physically tethers the initiator caspase Dronc to the cell membrane. Anchored there, Dronc cannot reach the executioner caspases that would complete the cell's destruction. When the team silenced Myo1D, DARE cells died on schedule and tissue regeneration collapsed. When Myo1D was overactivated, cells became substantially harder to kill.
Overactivation of the same motor protein has previously been linked to cancerous tumor growth. The human ortholog is MYO1D, implicated in colorectal carcinogenesis. The human ortholog of the initiator caspase Dronc is caspase-9.
DARE cells did more than survive. Within 48 hours their descendants had repopulated nearly half of the damaged tissue. They recruited a second population, called NARE cells, that had never activated the death cascade at all. The two populations formed a feedback loop: DARE cells secreted growth signals promoting NARE proliferation while NARE cells secreted inhibitory signals back, preventing overgrowth. Tissue repair followed. Then it stopped.
The Second Dose
Then the researchers irradiated the tissue again.
After the second dose, total cell death in the first few hours was half what it had been after the first. Most of the cells that died belonged to the NARE population. DARE cell descendants had become seven times more resistant to the same radiation that their parent cells had barely survived.
"The descendants of DARE cells were found to be exceptionally resistant," said Prof. Eli Arama, who holds the Harry Kay Professorial Chair of Cancer Research at Weizmann and heads the Crown Human Genome Center. "This may help explain why recurrent tumors become more resistant after radiation."
The treatment designed to destroy damaged cells had selected for cells that were harder to destroy.
The Market for Killing Cells
More than half of all cancer patients receive radiation therapy at some point during treatment. The global radiation therapy equipment market is worth approximately $8 billion a year. Siemens Healthineers paid $16.4 billion for Varian Medical Systems in 2021, the largest acquisition in the company's history, because external-beam radiation was supposed to be the growth platform.
The platform has a ceiling. Tumors that return after radiation are often more aggressive and harder to treat. Oncologists have known this for decades but lacked a molecular explanation. The Weizmann study offers one: each course of radiation that does not eliminate every cancer cell selects for descendants carrying Myo1D-mediated resistance. The mechanism is heritable. The advantage compounds.
This is why the money has moved. Bristol-Myers Squibb acquired RayzeBio for $4.1 billion in February 2024 for a radiopharmaceutical platform that delivers radiation directly to cancer cells via targeted molecules. AstraZeneca acquired Fusion Pharmaceuticals for $2 billion the following month. Both bets bypass external-beam radiation entirely. If broad-field radiation creates resistant survivors, the alternative is molecular precision.
Merck's Keytruda, the world's best-selling drug at $31.7 billion in 2025 revenue, represents the other response: combination therapy. On September 18, the same day the Weizmann press release circulated, the European CHMP adopted a positive opinion for Keytruda combined with Padcev in bladder cancer. Adding a checkpoint inhibitor to radiation addresses the resistance problem from the immune side. If radiation alone selects for cells that survive it, you need a second mechanism that kills by a different route.
The global immuno-oncology market was $28 billion in 2022 and is projected to reach $156 billion by 2030. Both are growing faster than the external-beam equipment market they are designed to supplement.
The Constraint
The finding cuts in two directions. Myo1D helps healthy tissue recover from devastating damage. The same protein helps cancer cells survive treatment designed to destroy them. Suppress it and you impair wound healing. Leave it alone and you make tumors harder to kill.
The body uses the same repair machinery regardless of whether the damage came from an accident or a therapy. The cells that survive carry a permanent advantage, and they pass it to every generation that follows.
Radiation therapy will remain a pillar of cancer treatment. The question is whether it can remain the foundation. The Weizmann study explains at a molecular level what the recurrence data has been saying for years: the cells that survive the first course are not the same cells you started with. They are harder, and their children are harder still.