Like so many of the things in my scientific life, it happened pretty much by chance. So we had been studying and still study a particular precursor cell, a cell that when it differentiates becomes a type of supporting cell in the brain called an oligodendrocyte. It makes myelin in the central nervous system. This is a cell that gets killed for reasons that nobody understands in multiple sclerosis so it’s an important cell, but we don’t study it because it’s an important cell, we study it as a model system.

And we had found that this precursor cell divides and then it stops dividing and becomes an oligodendrocyte, it differentiates into an oligodendrocyte that doesn’t divide again. And what we found was that the cell needs signals to stimulate it to divide and if you take the signal away, then the cell immediately stops dividing and differentiates to become an oligodendrocyte. And so we proposed that the differentiation is not a signalled event, it happens automatically when the cell stops dividing and you take away the signal that normally stimulates division. So the proposal was that differentiation happens automatically, it doesn’t require a specific signal to induce it to happen.

Another laboratory in Philadelphia in the States proposed that there was a type of growth factor, a protein signalling molecule that was required for this event, something called IGF1, insulin-like growth factor one. So I asked a graduate student of mine, Ian Hart, at the time to do a simple, simple experiment that should solve this once and for all, whether we’re right or they’re right, and that is to take a single precursor cell, put it alone in a little culture dish, no signalling molecules, it should stop dividing and should become an oligodendrocyte. And if it did that we would know that this is an automatic event, the cell is programmed to do it, you simply have to take away the signal that drives cell division.

So he did the experiment and, lo and behold, what the cell did was die, and I was a little surprised because we give it the right conditions and so on but it died, whereas if you cultured it with its normal neighbours it lived. So when it died, it died with the features that suggested it was going through a form of death that is called apoptosis or programmed cell death, which at the time was thought that you activate an intracellular death or suicide programme and the cell kills itself. So that got me thinking, why should this cell…?          

Anyway, the controversy seemed to me to be solved, that the reason these guys thought you need… oh, what I didn’t tell you is that if you take this cell and put it alone in the well, if you add IGF1, this growth factor, now the cell lives and now it becomes an oligodendrocyte. It stops dividing and becomes an oligodendrocyte, so what it needs the IGF1 for is not to signal it to become an oligodendrocyte, but to keep the damn thing alive so that it can become an oligodendrocyte. So controversy solved, we were both partly right. You do need signals, but you need the signal to keep the cell alive.

So it got me thinking, why should you need a signal to keep the cell alive, that’s a little weird. Now, it was known for some specific cell types, some types of nerve cells, some types of blood cells that they needed signals from other cells to stay alive, but this cell had never been suggested to be one of those special cases. And it got me thinking that maybe all cells need signals to stay alive, which would be pretty weird and pretty interesting, and so I started asking people. Going back to this idea that whenever you get an idea, just discuss it with everybody, and did anyone ever suggest this, that all cells need signals?  And no one ever did as far as I can tell.

And so I started working on other types of cells and asking do they need signals and some other people in the lab, like Ben Barres, joined in and before I knew it more than half the lab was working on cell death and checking out this hypothesis that all cells need signals to avoid killing themselves. I mean, that was the interesting idea and, as far as I can tell, that’s the way it works; every cell in your body is alive only because other cells are constantly signalling them not to kill themselves, which does seem a little bizarre.  And at first… it’s interesting that I’m sort of disappointed because this idea is now just kind of accepted dogma and it became an accepted part of the dogma very, very quickly, long before it was really adequately tested, but now I think enough cells have been looked at that it certainly works that way for most cells. Whether it works for all cells is not clear, but I suspect it’s… it’s right, the idea is right, and so for several years we were into cell death with this… with this idea. So in terms of an idea, this is probably the most important idea I guess I had.