Toronto researchers have shown that an ancient gene found in all forms of life acts as an executioner, causing cells to die.

The discovery could help researchers develop drug therapies that might limit the damage caused by heart attacks and strokes, and might one day play a role in the battle against cancer, said Dr. Josef Penninger, who oversaw the research.

While Penninger and his team are excited about the findings, published Thursday in the journal Nature, he cautioned that the gene called apoptosis-inducing factor (AIF) might ultimately prove to be too powerful a weapon to unsheathe.

''If we are right with this paper, then this is really the fundamental chisel of life,'' Penninger said in an interview.

''How much we can use it to cure diseases, that's an entirely different question, I think. Because if this is like the chisel of life, that's a very dangerous thing to touch.''

Blocking cell death might extend a person's life - or shorten it dramatically by causing cancer, he said. ''So we have to be very, careful.''

The research team is drawn from Princess Margaret Hospital's Ontario Cancer Institute and the semi-privately funded Amgen Research Institute, which is affiliated with the University of Toronto. The lead author of the article is Nicholas Joza, one of Penninger's students.

Funding for the work was provided by Amgen and Canvac, the national network of vaccine and immunotherapeutic research.

Three years ago, Penninger and Guido Kroemer, a former classmate now living in France, published an article in Nature describing a gene that Kroemer had identified and Penninger had succeeded in cloning.

They theorized that the gene, which they called apoptosis (death) inducing factor, caused programmed cell death.

Scientists had already found another agent for cell death, the protein family caspase, which is the basis of much research to see if it can be used to fight cancer. But from an evolutionary point of view, caspase is believed to be recent - it is not found in plants, for instance.

AIF, on the other hand, is found in all life ''from plants to humans to slime mould,'' Penninger said.

Still, Penninger and his colleagues did not know how important the gene was. So, following common scientific practice, they tried to produce genetically engineered mice lacking AIF to see what would happen.

They couldn't.

When the team sat down to analyse why they couldn't breed the mice, they discovered the mice never grew beyond the embryonic stage. Cells didn't die off to make room for the next stage in development.

All living things, plants, animals and humans, must have cell death to develop beyond an embryo. And fully formed living things must have controlled cell death to stay healthy. For instance, people lose millions of skin cells every day.

All human illnesses boil down to a problem with cell death, Penninger explained. In some diseases, such as cancer, cells don't die when they should and grow out of control. In others - degenerative diseases such as Alzheimer's or Parkinson's - cells die too soon.

If scientists could figure out how to block AIF, they might be able to stop the death of brain cells that accompanies Alzheimer's. If they could find a way to give a targeted dose of AIF, they might be able to kill the rogue cells in a cancerous tumor.

Penninger believes the most promising application would be to try to prevent the cell death that accompanies acute heart incidents such as heart attacks, during which heart muscle cells die, and strokes, which kill brain cells.

''Somebody comes into the clinic with a heart attack and they give them the blocker and you can keep the heart muscle cells alive,'' he suggested.

The team is currently testing that thesis on mouse tissue by removing AIF from heart cells and inducing a heart attack to see what happens.

The prospects for therapies that would require long-term use - cancer or Alzheimer's, for instance - are trickier, he said. AIF is such an effective killing machine that unless you could target its application precisely, the cure might be worse than the initial disease.

''So can you imagine if you give the stuff (in untargeted form) to a person? Within 24 hours, we will just fall apart,'' Penninger said.

''We have to be very careful because it's universal,'' he continued. ''It's in plants, it's in bugs. It might be too universal for life . . . (for science) to touch it at the end of the day.''