In recent time, it was possible to make to make catalysts from antibodies. You know what antibodies are – antibodies are produced in our immune system to fight invaders... to fight invasion by foreign substances. So they are protein molecules of a different structure.  Much work has been done... actually we know it's about the best organ... we know how it functions... the immune system. And now some years ago Richard Lerner at the Scripps [Research] Institute at La Jolla and Peter Schultz at the University of California at Berkeley have used antibodies to make catalysts out of them. That's a very tricky way, because you have to think what does catalysis mean. Catalysis means you change a molecule. That means usually you have to bring it over some activation barrier. And on the top of the barrier, one calls that the transition state, the molecule has to get through this transition state in order to be... to be transformed, and the catalyst helps it to get over this activation barrier. Now the ingenious idea of Lerner and Schultz was to grow antibodies against transition states. Of course, transition state is not a stable state, so they had to use analogue states which had exactly the same structure as a transition state. And by growing antibodies – which is a procedure of a few weeks, to make antibodies – they were able to make pretty potent and efficient catalysts, by just using this trick, growing them against antibody analogues. And they did the same for making a proteolytic enzyme from antibodies, and here again you find the proton relay. So that is in the protein which evolved in the immune system, that means within a few weeks.

So you see, evolution doesn't need millions or billions of year. Whenever you have a clear problem given to the system, whenever you have optimal conditions, that you need in the immune system... Look, our mutation rate in humans is of the order of 10^-12, so there's almost no change, otherwise we wouldn't be stable, we would simply disappear. Our genome contains three billion nucleotides, so if we copy our genome, that's from generation to generation, you have to be very precise with that and the error rate has been estimated to be as low as 10^-12. But in the immune system you have to produce new substances, you have to optimise them, you have to get the best possible fit in the immune system, there you need high mutation rates and... César Milstein at Cambridge found out that indeed for the somatic mutation in the immune system, the mutation rates are as high as in viruses, so 10^-3 to 10^-4, and there you can within a short time adapt your proteins to a new job.