Penicillin kills bacteria by binding to some enzymes called, rather curiously, penicillin-binding proteins. When I first heard of them I thought their function was to bind penicillin, it's nothing of the sort. Their function is to make the cell wall, and the penicillin comes in and binds to them and prevents them making the cell wall and the bacterium dies. Now, we find that some of these commensal bacteria are naturally resistant to penicillin. That means to say, if you take an isolate out of the freezer, which was put there before the clinical use of penicillin, back in the 1940s or something, they're already resistant to a drug they've never met, and I think never could have met it. I mean, there's plenty of penicillin being produced in the soil by soil... fungi, and so if I found a soil bacterium which was resistant to penicillin, it would make sense as having happened by natural selection. But it's very hard to explain why something in our throats which could never have met penicillin, I think, should have this resistance. I think it's just a matter of chance as to whether the penicillin could or couldn't bind in an appropriate way to the enzyme. On the other hand, once a piece of that gene was transferred into a... the bacteria in our throats, and penicillin was being used, the selection favouring it would be enormously powerful, of course.

[Q] But was that gene doing anything in the original ones who just were spontaneously resistant?

It was certainly making cell walls, but I think the fact that it was spontaneously resistant was just an accident, that's my guess. It's hard to be sure. And the alternative is that... that it evolved by natural selection in soil bacteria, and then, by some accidental process, it's got transferred into the throat.