I had a new research student coming, a chap called Jonathan Miller who had a degree from Yale and also a higher diploma in biophysics from them, he turned out to be a very good mathematician, he landed up... it's an interesting story himself, he used to walk around with books on abstract mathematics, I don't know if you remember him, John? He could never decide what he wanted to do and just as an aside I discovered later on he'd been enrolled as a PhD student here in Cambridge and I was his supervisor, but in his third year he was also enrolled in Princeton physics without telling me... because he still then couldn't decide, I think he's rather sorry now because he could get onto a good thing if he had realised where this was going to lead. So he had had practical experience although he was good at mathematics, both his parents were doctors and they wanted him to work in medicine or biomedicine so he'd spent every summer in the last four years working in labs so he knew how to do bio chemical operations but he was very terribly messy, very messy, one of the messiest people I've ever known... never cleared up, which figures in the story later. So I said, well, the thing to do, you can follow the recipe and there should be... and so we began to make TF3A, there were recipes for this and it was interesting, none of the people later on, Louise Fairall and Daniella joined... and they couldn't bear themselves to kill a frog so I had to pith the frogs and kill them because I used to do that when I was a medical student and extract the oocytes. So they followed the recipe and they got very low yields, looking at gels where you could run the protein on, it was all described by Reuter and Brown and the binding sites. We got very low yields so there was clearly something wrong with it, so I thought obviously the extraction procedure they describe are very inefficient, but you see those two were molecular biologists and they work with gels and they work in microgram quantities, we wanted to work in milligram quantities, which we knew to be there. Because Hugh Pelham had shown me gels which show that there must be able to be milligrams... you should be able to get dozens of milligrams from a set of ovaries which were ovulated. So now we were going to change the buffer conditions and now I looked at their buffers and what they had done was to use EDTA which is a chelating agent which takes up metals, which was very sensible to do because if you have any metals around, you're working with RNA or DNA they tend to hydrolyse the nucleic acid. So they put in EDTA.

Now the sequence of the protein wasn't known, the sequence of the gene wasn't known, TF3A, but amino acid analysis was known, was known to contain quite a lot of cysteines and histidines. Now cysteines, there were a lot of... there were... about 20 cycsteines in it so they assumed there must be SS bridges in it, so they put in reducing agents, dithiothreitol to protect it and so the... now both of these things were removing metals, because dithiothreitol because of the sulphur groups in it will also, and in fact has a binding constant for metals, for zinc which turned out to be the metal, was ten to the ten, ten to the eleven for other metals as well. So what they were doing, and we showed that what was happening was the... we were trying to make a complex of the TF3A and the DNA target, we knew the sequence of the DNA target, everything was known, it was well set up problem, the things associated, so we gradually changed the buffer conditions and I came to the conclusion that there couldn't be any SS bridges because it didn't make any difference. And indeed, sorry I won't describe everything, let me go back a bit. I thought there'd be a test if there's SS bridges, so I knew enough chemistry to know that sodium borohydride was a very strong reducing agent so I got Jonathan to add sodium borohydride to the TF3A, it had no effect at all and this meant that there couldn't be any SS bridges in it. So the only thing it could be would be a metal. And so what we did was to gradually omit the dithiothreitol, start omitting EDTA, we began to get higher yields. That was just... but by this time I deduced there must be a metal there. So then to find out what the metal was we did send the material to a lab in medical school who were involved in, had absorption spectral photometry to test metals in different specimens, but the sub wasn't pure enough, we had to purify it still further... so what we did eventually to... it's quite a long story, was to start adding EDTA specifically, titrating it and using other collating agents for example phenthramine which takes out copper and other things which take out iron to see if they acted. But the EDTA and these things, we found that... and then added metals back to the preparation that Jonathan was doing, titrate metals back and we found that zinc and only zinc... when added back was able to bind back to... in other words, it restored binding to DNA which was the target and also to RNA.