Life from the condemned cell: The suicidal tendencies of tadpole tails may signal a new way of treating human diseases, writes Hugh Aldersey-Williams

Where does a tadpole's tail go when the creature grows into a frog? The surprising answer is that the cells in the tadpole's tail commit suicide when given the signal by chemicals released from other cells. The mechanism underlying such suicide commands is not yet fully understood, but the discovery has opened new avenues of research into potential treatments against cancer and Aids.

This form of cell death is different from death caused by external means in which the dying cells swell and dissolve, causing inflammation. Here, the cells shrivel, pulling away from the mesh of surrounding cells in an organism. Micrographs of the process show what look like ball-bearings lost in a honeycomb. The shrunken rounded cells then fragment, leaving the surrounding cell structure intact and without inflammation.

The fact that the process does little to advertise its presence is perhaps one reason why it has been so little studied. Biologists have devoted much time to understanding cell growth. Now it appears that the death-in-life of cells may be equally significant.

Having overcome their surprise at the phenomenon, researchers are turning to ways in which they might exploit cell death to arrest the development of cancer cells and for the treatment of other diseases.

The theory that all cells are pre-programmed to self-destruct was a challenge to conventional biological wisdom. All that keeps the cells from doing so, most scientists now believe, is a flow of molecular signals that constantly convey the message, 'Don't do it'. Similar signals exchanged between cells govern their proliferation as an organism grows. But the idea that cells need such signals to stay alive has only recently gained support.

The theory grew from studies by a group at University College London, led by Professor Martin Raff. The group initially sought to find an explanation for the signals that were known to pass between cells in certain tissue. Then one of Professor Raff's students looked at another tissue, where signals were not thought significant, and found that they were.

The question now became: where do these signals not occur? It was impractical to test all cell types, so the group looked at two types where such signalling was thought least likely. They found signals here, too. Professor Raff now believes there are no cells that can function without these signals. He has even bet pounds 1,000 to anyone who can prove him wrong.

Cell death appears to be the key to a selection procedure in which the best cells - the ones that receive and decode survival signals most efficiently - are chosen from the excess made by an organism. The rest die. 'Cells are cheap,' says Dr Gerard Evan of the Imperial Cancer Research Fund. The growth and survival of multi- cellular organisms from worms to humans seems to rely upon an apparently wasteful process of excessive cell creation followed by tedious weeding out of 'mistakes'.

Better for us that the body throws out many good cells in this way than allow a single cancer cell to prosper. Cell death also provides a fail-safe mechanism whereby a cell prefers to commit suicide for the good of the rest of an organism than to incur risk, for example by accommodating a virus. In the case of strokes and heart attacks, a localised region of the brain or heart dies because it is starved of oxygen. These few cells no longer send out survival signals to their neighbours, promptly inducing suicide in a greater number of cells. If cell death can be headed off, in effect by dosing the body with supplies of the missing signal molecules, then their effects can perhaps be limited as well.

The notion that cells are dependent upon the 'social control' of other cells to stay alive may be easily tested by preparing homogeneous samples of cells, or even by experimenting on just one cell. Results of such experiments suggest that some genetic signal thrusts cells into a transient state where they are only semi-stable. From here, other signal molecules determine which of several things the cell then does. One signal will cause it to multiply; another will put the cell into a limbo where it lives but does not proliferate; another will cause it to die.

The signals normally maintain an equilibrium. Too much cell growth can thus arise either when cells multiply too vigorously or when excess cells are removed insufficiently fast.

Various genes, proteins and simpler chemical compounds have been implicated in cell suicide by inhibiting the passage of the survival signals. If these are absent, the signals get through and the cell lives. What these agents actually do is unclear, however. 'There is still no agreement on a unique effector molecule anywhere in this process,' says Professor Andrew Wylie of Edinburgh University Medical School, an early champion of cell death.

One priority is to discover how the cell dies. Is it the nucleus that is killed? Dr Michael Jacobson, a member of Professor Raff's group, treated cells to remove their nuclei. The residue still experienced cell death. How about the mitochondria - the cell's energy sources? Samples prepared without mitochondrial DNA, and thus unable to 'breathe', also underwent cell death as before. Professor Raff remains confident: 'In the next year or two there is no question we will understand why cells die. It's too early to know how important it will be for diseases, but it will have clinical implications at some point.'

There is already great interest in harnessing cell death, especially for the treatment of cancers. Professor Raff explains: 'No cells have been found that don't survive in this way. I suspect that cancer cells may behave in the same way, but as they become more malignant, they become more autonomous.' In other words, cancer is a disease precisely because tumour cells have lost their social controls. 'This will be a useful way of attacking cancer,' he adds, 'but it won't be a universal way.'

To date, the major cancers remain stubbornly resistant to attack by chemotherapeutic agents, notes Professor Jack Hickman of the University of Manchester. But cell death may be the reason. He tested a range of anti-tumour drugs that do work and found them all to attack the chemical signals that stop the tumour cells committing suicide, not the cells themselves. Other cancers might develop, then, not by excessive proliferation of tumour cells, but because too few of them are killed, due to their survival signals not being stopped.

Naturally created substances may already play a role in survival signalling. Dr Angela Hague and her co-workers from the University of Bristol found that sodium butyrate, a simple chemical formed when dietary fibre is fermented in the intestine, induces death in tumour cells located there. 'This response may, in part, explain the correlation between a high-fibre diet and low incidence of colo-rectal cancer,' they conclude.

(Photograph omitted)