TRENDS in Biochemical Sciences Vol.26 No.11 November 2001 http://tibs.trends.com 0968-0004/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0968-0004(01)01979-X 690 Forum Having been trained in physics, my molecular biology was picked up on the streets. The streets, however, were interesting places in the mid-1970s, full of promise and entirely unsolved, even untried, problems. In pursuit of interesting problems, I went to work with Elbert Branscomb, at the Lawrence Livermore laboratory, on error propagation of protein synthesis in Escherichia coli. After showing up in Branscomb’s laboratory and pestering him with endless questions, alternately naïve and unanswerable, about how biology worked, we began working together. After a series of attempts, we observed progressive error effects in protein synthesis in E. coli that, while largely uninterpretable, were provocative and interesting. However, even though we made computer models of the process, protein synthesis feedback remained something of a cipher. Despite this, the work led to two Nature papers [1,2], and I was hooked on acquiring more and more good street smarts in molecular biology. After my initiation with Branscomb I saw great opportunities in biology and I went to several biologists of note to ask advice and wisdom on the future of molecular biology research and my potential role in it. Among these patient scientists were Jim Watson, WallyGilbert, Gene Stanley, Leslie Orgel and Jack Sadler. I am very grateful to all of them for their time and concern, and often, but not always, for what they told me. Watson was particularly accessible and, if I remember well, particularly insightful. On our first meeting one winter day he spent an hour or two with me in his office at the Harvard Biolabs after only an hour notice. He was gracious to see me at all, so it was an added bonus when he was interested, intellectually engaged and engaging. Even with the six-foot tall African fertility god carving in his office, which kept distracting me, I remember certain advice well, for example that I should immediately fly down to Cold Spring Harbor to meet Jeff Miller who was there visiting. That would be the first thread in the story that led to the footprinting method. I could not take Watson’s advice immediately, but caught up with Miller later when I took Cold Spring Harbor summer courses (such as the Yeast Genetics course taught by Fred Sherman, Chris Lawrence and Jerry Fink, in 1975). Miller taught the course in Advanced Bacterial Genetics with Bob Weisberg and Bill Reznikoff in the summer of 1976. Because the lac operon was one of very few systems of gene regulation that was tractable and reasonably well understood at that time, Miller’s masterful genetic analysis of the repressor gene was one of the major intellectual assets at the forefront of research in trying to figure out the mechanisms of genetic regulation. I had all sorts of fantasies about how the repressor might recognize only its own specific operator sequence although, in retrospect, most of these ideas were absurd. The imagined recognition mechanisms were sophisticated in their physics, but the important thing was that this was the real beginning of my interest in understanding the sequence-specific interactions of DNA with proteins. It was not too long after the course that Miller and I agreed that I would go to Switzerland and work with him at the University of Geneva. The first threads from Gilbert’s laboratory Another step towards footprinting came in the spring of 1977 when I spent a couple of months in Wally Gilbert’s Harvard laboratory before going to Geneva, trying to master the Maxam–Gilbert DNA sequencing method. Miller was also in Cambridge at the time, but it was Phil Farabaugh, trying desperately to finish his thesis and head off to Cornell to a postdoc position with Fink, who was saddled with teaching me to sequence. The atmosphere in Gilbert’s laboratory at the time was frenetic, there was so much going on – from Tonogawa and his postdoc, who were sequencing immunoglobin genes, to Greg Sutcliffe sequencing pBR322 [3], and many students and guests working hard, discussing science and technology, and competing for equipment. Frayed nerves often seemed to cause eruptions of emotion in spite of the high sense of intellectual excitement. However, Farabaugh made my time there a pleasure as he had a sense of humor, enthusiasm and intellectual curiosity, as well as great skill with the sequencing method. Farabaugh had, in fact, just finished the sequence of the lacI gene when I arrived in Cambridge. To my great excitement we spent hours, sometimes with Miller, poring over the sequence spread out on the tea room table on bits of paper. It was, after all, the first complete gene ever sequenced. Farabaugh and Miller had found mistakes in the amino acid sequence of the protein by examining the DNA sequence of the gene – now the correct amino acid sequence of the repressor was known for the first time – but there was much more to be done. The long string of letters seemed to need something more than bits of paper for their proper analysis. With a little detective work I discovered a computer in the attic of the Biolabs, seldom used and almost forgotten by most of the molecular biologists. I was able to pull together enough information about the language available on the machine to write a program, on paper punch tape, to look for specific sequence motifs and repeats in the lacI sequence that I had carefully punched in. There was plenty of time in the attic to think about problems of interest – the repeat-finding program, for example, took about eight hours to run through the lacI sequence on this computer – and one always in my mind was how the recognition process worked. A second operator was thought to be located in the early part of the I gene and, indeed, my program found a couple of partially symmetric sequences there that were somewhat similar to the principal operator. Although one of them had been identified by eye already, it intensified my thinking about the problem of how one would determine whether the repressor actually recognized that sequence. Explaining this would open the Forum Historical Perspective The invention of footprinting David J. Galas