On the 30th Anniversary of DNA Sequencing in Population Genetics

30 years ago today, the “struggle to measure genetic variation” in natural populations was finally won. In a paper entitled “Nucleotide polymorphism at the alcohol dehydrogenase locus of Drosophila melanogaster” (published on 4 Aug 1983), Martin Kreitman reported the first effort to use DNA sequencing to study genetic variation at the ultimate level of resolution possible. Kreitman (1983) was instantly recognized as a major advance and became a textbook example in population genetics by the end of the 1980s. John Gillespie refers to this paper as “a milestone in evolutionary genetics“. Jeff Powell in his brief history of molecular population genetics goes so far as to say “It would be difficult to overestimate the importance of this paper”.

Arguably, the importance of Kreitman (1983) is greater now than ever, in that it provides both the technical and conceptual foundations for the modern gold rush in population genomics, including important global initiatives such as the 1000 Genomes Project. However, I suspect this paper is less well know to the increasing number of researchers who have come to studying molecular variation from routes other than through a training in population genetics. For those not familiar with this landmark paper, it is worth taking the time to read it or Nathan Pearson‘s excellent summary over on Genomena.

As with other landmark scientific efforts, I am intrigued by how such projects and papers come together. Powell’s “brief history” describes how Kreitman arrived at using DNA to study variation in Adh, including some direct quotes from Kreitman (p. 145). However, this account leaves out an interesting story about the publication of this paper that I had heard bits and pieces of over time. Hard as it may be to imagine in today’s post-genomic sequence-everything world, using DNA sequencing to study genetic variation in natural populations was not immediately recognized as being of fundamental importance, at least by the editors of Nature where it was ultimately published.

To better understand the events of the publication of this work, I recently asked Richard Lewontin, Kreitman’s PhD supervisor, to provide his recollections on this project and the paper. Here is what he had to say by email (12 July 2013):

Dear Casey Bergman

I am delighted that you are commemorating Marty’s 1983 paper that changed the whole face of experimental population genetics. The story of the paper is as follows.

It was always the policy in our lab group that graduate students invented their own theses. My view was (and still is) that someone who cannot come up with an idea for a research program and a plan for carrying it out should not be a graduate student. Marty is a wonderful example of what a graduate student can do without being told what to do by his or her professor. Marty came to us from a zoology background and one day not very long after he became a member of the group he came to me and asked how I would feel about his investigating the genetic variation in Drosophila populations by looking at DNA sequence variation rather than the usual molecular method of looking at proteins which then occupied our lab. My sole contribution to Marty’s proposal was to say “It sounds like a great idea.”  I had never thought of the idea before but it became immediately obvious to me that it was a marvelous idea.  So Marty went over on his own initiative, to Wally Gilbert’s lab and learned all the methodology from George Church who was then in the Gilbert lab.

After Marty’s work was finished and he was to get his degree, he wrote a paper based on his thesis and, with my encouragement, sent the paper to Nature. He offered to make me a co-author, but I refused on long-standing principle. Since the idea and the work were entirely his, he was the sole author, a policy that was general in our group. I had no doubt that it was the most important work done in experimental population genetics  in many years and Nature was an obvious choice for this pathbreaking work.

The paper was soon returned by the Editor saying that they were not interested because they already had so many papers that gave the DNA sequence of various genes that they really did not want yet another one! Obviously they missed the point. My immediate reaction was to have Marty send the paper to a leading influential British Drosophila geneticist who would obviously understand its importance, asking him to retransmit the paper to Nature with his recommendation. He did so and the Editor of Nature then accepted it for publication. The rest is history.

Our own lab very quickly converted from protein electrophoresis to DNA sequencing, and I spent a lot of time using and updating the computer interface with the gel reading process, starting from  Marty’s original programs for reading gels and outputting sequences. We never went back to protein electrophoresis. While protein gel electrophoresis certainly revolutionized population genetics, Marty’s introduction of DNA sequencing as the method for evolutionary genetic investigation of population genetic issues was a much more powerful one and made possible the testing of a  variety of evolutionary questions for which protein gel electrophoresis was inadequate. Marty deserves to be considered as one of the major developers of evolutionary and population genetics studies.

Yours ,

Dick Lewontin

Some may argue that Kreitman (1983) did not reveal all forms of genetic variation at the molecular level (e.g. large-scale structural variants) and therefore does not truly represent the “end” of the struggle to measure variation. What is clear, however, is that Kreitman (1983) does indeed represent the beginning of the “struggle to interpret genetic variation” at the fundamental genetic level, a struggle that may ultimately take longer then measuring variation itself. According to Maynard Olson interpreting (human) genomic variation will be a multi-generational effort “like building the European cathedrals“. 30 years in, Olson’s assessment is proving to be remarkably accurate. Here’s to Kreitman (1983) for laying the first stone!

Related Posts:

8 thoughts on “On the 30th Anniversary of DNA Sequencing in Population Genetics

  1. I just gave a brief introduction to Kreitman 1983 and to Lewontin & Hubby 1966 (“We then have a dilemma.” p. 607) just this past Wednesday in my undergrad evol. gen. course, but I didn’t notice the 30-year anniversary. I coulda brought a cake! I forwarded your blog to the class, a tad belatedly.

    I also enjoyed Lewontin’s comments on grad student research, authorship and the review process. Also worth sharing with students.

  2. Dick Lewontin’s memory is perfect. So, what else is new? But he doesn’t give himself proper credit. I was on sabbatical leave as a guest in his lab in the fall of 1979 when Marty Kreitman began this painstaking work, so I had a ringside seat to history. What was Lewontin’s role? He was the draw, the intellectual force that attracted so many extraordinarily talented people to work in his lab. They didn’t need Dick to tell them what to do. But they stimulated each other as they interacted and argued passionately about science every single day, and night. If Dick hadn’t been there, neither would they have been, including Marty.

  3. I am, of course, deeply flattered, but couldn’t resist adding to Dick Lewontin’s recollection about how I ended up deciding to study DNA sequence polymorphism. First, Bruce Grant is entirely correct in describing the freewheeling intellectual atmosphere Dick set for his lab. We all encouraged one another not to retreat from thinking about and ultimately working on ambitious hard problems, and no evolutionary biologist has ever been better than Dick in enunciating them.

    I went to Wally Gilbert’s lab seeking expertise in purifying the E. coli lac repressor protein. The idea was to characterize the electrophoretic properties of various available protein variants. At the time I didn’t know how to do molecular biology at all. I kindly received help from a couple of senior postdocs, but ended up giving up on the project because the native-form repressor was just too difficult to isolate on a gel (and even in over-expressing strains, was not very abundant). One idea I had (i.e., someone suggested to me) was to use labeled DNA to the repressor to stain a gel image of partially purified repressor, transferred to membrane. I almost succeeded with this approach long before Western blotting was invented (I remember how excited some people in the Gilbert lab were for reasons I didn’t understand), but decided to give up simply because the intellectual value of the research didn’t merit the difficulty of the experimental work that lay ahead.

    This was a couple of years after Maxam and Gilbert invented DNA sequencing, and the lab was just completing the sequence of the plasmid pBR322, at 4361pb then the longest assembled sequence. Several in the lab knew I was interested in polymorphism and introduced me to the idea of doing a DNA sequence study to measure it. At the time, Dick Lewontin was starting a sabbatical and taking his place was none other than Allan Wilson. Allan was supportive and I remember spending some part of the summer in the library pouring through molecular biology journal articles reading methods sections to try to figure out how molecular biology was done. The “Maniatis” manual was, of course, not yet available. At some point, Allan had returned from a conference and he excitedly told me that the Adh locus had been cloned, and that I should immediately commence my study on it. I had already published a paper on Adh (my first one as a graduate student with Dick, under the supervision of Jerry Coyne). As good fortune would have it, Tom Maniatis, who had left Harvard when a ban on recombinant DNA was instituted by Cambridge, was returning because the ban had been lifted. And, David Goldberg, his graduate student, was one of the two groups that cloned the gene. David was very accommodating and gave me material to get started. But it was a graduate student in Wally’s lab, Winship Herr, who took me under wing and made it possible for me to make initial progress. Subsequently, many people in Wally’s lab, notably George Church, chipped in and contributed to my success with the project.

    Lewontin, for his part, was from the get-go enthusiastic about the project, and quickly became extraordinarily so. Wally Gilbert also took some interest in the work, having established his place as a molecular evolution, and he paid for the study in its entirety. Neither Wally nor Dick suggested their names be on my paper, and Dick in fact removed his from a draft. He was determined not to take any credit for the work, whereas nothing could be further from the truth. For my part, the grandson of a carpenter, I knew how to work with my hands and in this case put them to good use.

  4. This landmark paper opened up an entirely new area of research, and the talk and paper were utterly memorable to me for that reason. I also agree that Dick Lewontin, whom I’ve recently got to know personally, was highly influential. As his grant funded Marty Kreitman’s work, he would have been a coauthor under today’s customs, and as Bruce Grant argues, participated extensively in lab discussions about the project.

    However, let us also consider how far we’ve come since then.

    The paper showed for the first time, I think, that in a gene (Adh) non-synonymous coding polymorphisms were rare, indicating strong constraints on coding variation. The only coding variant found was the fast-slow amino acid polymorphism, but there were many synonymous changes, suggesting less selective constraint on coding variants. The paper also showed that the 3′ non-coding region was deficient in polymorphisms, suggesting for the first time that non-protein coding regions could also be under selective constraint.

    However, the big Kimura/Lewontin/Hubby question was: is most evolutionary change selected or neutral? It seems to me that the 1983 paper did not answer that, although my memory of the Lewontin conclusion to the joint SSE talk was that that selection had been proved for the fast-slow polymorphism.

    By showing that most variation was synonymous or non-coding, the work did show that natural selection against coding variation was important.

    But it did not prove whether fast-slow was under selection. It was still possible to argue that the fast-slow polymorphism was a rare amino-acid change which was neutral.

    Although synonymous sites were inferred to be under weaker selection, the results did not show whether the synonymous sites were neutral. Studies of codon bias since then suggest that even synonymous sites can be under conservative selection.

    Finally, the Kreitman paper did not (and of course could not, with a single gene) test the conflicting ideas that the majority of genetic variation was selected or neutral.

    We now have better data from whole genome analyses. Selection seems clearly more important in the genome than believed under the Kimura neutral hypothesis. The Kreitman and Lewontin intuition was correct.

    Kreitman’s 1983 paper was a landmark. But it did not quite answer the major questions it was think intended to answer. Instead, it raised and answered questions that had not been thought of and opened up entirely new areas of enquiry.

  5. Eleven DNA sequences and an initially underapreciated scientific revolution | simonymous

  6. Evolutionary Classics Reading Group 2014 | gcbias

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s