We published this paper and it seemed to us that we got repeating units and we thought that each unit would bind about five base pairs because there was an interact of 50 pairs, when we published the model. So we showed... so what we proposed was the units folded up, it was two zincs, two cysteines and two histidines, pethidine and you folded up into a compact unit, a module. And so what we had discovered was not just a new fold, because this was a new fold, we hadn't worked out the three dimensional structure, depend upon zinc and the... and there were also... in the sequence, when we noticed in the sequence, we noticed there were three conserved hydrophobic... amino acids, in different positions in the 30 amino acid unit and so McLaughlin's computer analysis showed that there was a link of five amino acids, 25 amino acids and within these 25 amino acids there were in fact seven conserved residues, two cysteines, two histidines, and three conserved hydrophobics, large hydrophobics like phenylalanine, leucine and sometimes methionine, and tyrosine, all large hydrophobic and it was pretty clear that these things must be involved in stabilising this very tight module. You see it's a very... I called it a module, or a structural unit because when you chopped it up with a nuclear... with proteases, this is what was resisted. There were no cuts within the units they were between the units, that was the linker. We were able to define the whole notion of a new principle of DNA recognition, not only a new fold but this was a set of linear domains strung together in a polar fashion, a tandem fashion, which bound a long stretch of DNA. And none of the other transcription factors, people who worked on, worked like this to return to transcription factors, of course a lot of people were working on them, other systems particularly in bacteria and in all cases they work as dimers, homodimers, they bind... they make use of the symmetry of the DNA structure, which is two-fold and moreover the sites which they bind are inverse repeats of sequence, palindromes, which of course have two-fold symmetry, so the helix turned helix, the homo domain, all these have these.

So this was very clearly a new principle and it could bind any sequence of DNA so I actually wrote in the paper at the time... words to the effect, I think I cribbed it from... I didn't want to write a sentence... it would not surprise us, that's a [Francis] Crick fact of sentence, but not surprise that this type of recognition of DNA would be found in many other species other than an lowly frog you see, I didn't write that, I said, 'It would be very likely to occur elsewhere', likely to be used elsewhere. Well, within months papers began to appear. Of course the Brown-Argos paper appeared but they hadn't... they had simply got 12 units and they hadn't really grasped the point about zinc with each unit and so there was a paper on drosophila, a paper from Herbe d'Acula, they wrote and said they found sequences like this, cripple in drosophila Rendite had found it, was it an end rail and things like that, they had all these... they were cloning genes everywhere and within a year the numbers had grown and we now know that 3% of the human genome contained fingers of these TF3A type, that's two cysteine's and two histidine's.

Well, the word 'finger', we used the word finger loosely in the lab, I called it a structural unit, a binding unit, a mini DNA binding domain, but we called it like fingers because we thought of them as fingers gripping and grasping the DNA and each finger, we didn't know how many base pairs we had at the time, that came later from crystal structures, we thought a small number of bases in each finger would bind a small number of bases and because each finger has structurally the same... has structurally the same structure, they were modules in the proper sense of the word that architects use. Unfortunately, the word got debased when you think of the Apollo mission where they had the moon module, moon module is not... totally... it's different part of the thing, but the word module should mean, used to mean the thing is built according to a certain plan, like a building or so on. And still used in that way, after all mathematics is something modular for whatever or multiples of four... so I still call them modules, so the fingers were physical modules, physical units. And I say it now, but unfortunately the term... people began finding zinc-binding sites in other proteins. Now it was known, many of the hundred proteins already known to contain zinc and most of them are enzymes, but it means the alcohol dehydrogenase, which has a zinc site which is bound by four cysteine's, is part of... it's a structural site, zinc is used structurally. Of course, zinc is used structurally and of course zinc is found in quite a number of proteins, enzymes where it's used as a catalytic centre where its four-coordinated but it can adopt a fifth water molecule which is active but this is... but normally zinc is a safe metal.

People used... I started getting invited to meetings by... by organic people and I would always be asked, why zinc? Well, the answer is very simple, zinc has... no other oxidation state. It's very stable so I got to the end of getting... saying it's a dull, grey, boring metal, a safe metal. Now, copper and iron are more glamorous, they're coloured, but iron you see is in involved in ferric and ferrocene and it's involved in, in fact it's used in proteins involved in redox reactions and copper, of course, is found in many proteins which are used in electron transfer reaction. So again, you have more than one oxidation state. And so I did... so this was a, I think, an important discovery. But as I point out, people misunderstand it, we didn't set out to discover the fingers, we didn't know they were there, they simply emerged from exploration which is one of the reasons I find I'm talk... a bit of a digression, there when people say you have to write these new plans, nowadays when you have to, I think it's beginning to be unfashionable... it was in the fashion to write a proposal where you had a hypothesis that you were going to test, you see. I know that the MRC and the EPSRC and people like that began to put in... well, this was... we would never have got a grant... we were trying to explore what these things where you see. So I think it's important to stress that just trying to be curious about something can lead, some phenomenon you have to know... trying to understand it can actually lead unexpectedly to quite important results.