In a paper published in The EMBO Journal in 1989, Denis Duboule and Pascal Dollé characterised the structure of the HOX-5 homeogene complex in the house mouse, revealing two new genes. Through in-situ hybridisation experiments on these genes and a member of the HOX-1 complex, as well as comparisons to sequences and domains of expression in Drosophila, Duboule and Dollé found evidence that suggested conservation of the structural and functional organization of the homeobox-containing gene families between insects and vertebrates. Thirty-five years after the paper was published, I spoke with Denis Duboule about the origins of his interest in this topic, his collaboration with Pascal Dollé, his memories of the making of this study, and his reflections, today, on the paper’s findings, interpretations and significance.
Citation: Duboule, D., & Dollé, P. (1989). The structural and functional organization of the murine HOX gene family resembles that of Drosophila homeotic genes. The EMBO Journal, 8(5), 1497-1505.
Interview conducted online on 10 March 2025; Denis Duboule was in Lausanne, Switzerland and Hari Sridhar in Klosterneuburg, Austria.
Hari Sridhar: I’d like to start by asking you to talk a little bit about the motivation for the work presented in this paper, and to relate this to the research that you had done until that point.
Denis Duboule: I’m very happy that you picked this paper because it’s dear to my heart. To understand why we did this work, we need to step back about four years—1984-1985—which is when the homeobox was reported for the first time. That discovery was made in Basel in the Walter Gehring laboratory and also in the States by Matt Scott. Soon after this, Eddy De Robertis, who was actually in Basel at that time, realized that the same sequence was also present in vertebrates. And that was really a shock, because it meant that Drosophila, invertebrates and vertebrates were sharing the same capacity for some proteins to bind DNA. That was the homeobox. Now, as you know from the work of Ed Lewis in 1978, it turned out that these genes in Drosophila, the so called homeotic genes, which are responsible for the organization of the body plan, are in groups; they are in clusters. In the genome of the fly, it’s two little groups with four or five genes each. So, at the time of the report of the homeobox, several laboratories started to say, well, let’s see if they are clustered in invertebrates too, which was sort of a crazy idea. For example, there was the lab of Peter Gruss in Goettingen and Frank Ruddle at Yale. And we also embarked on this. I was just arriving in Strasbourg, and Pierre Chambon, who was the head of the institute, told me: if you want to go for this, I’ll give you a group. I was, at the time, just four months out of my PhD. He also said: I anticipate that these genes will be clustered in vertebrates, so don’t use a lambda library, because fragments are too short. I heard that someone at the EMBL in Germany is just preparing a new type of library which has much larger fragments. That was the cosmid library from Hans Lehrach. So, I went to the EMBL, and Hans, who was a very generous person, and Annemarie Poustka, who had done the work, just gave me the library. No strings attached. That was the old way to do science: you want it, you just get it. So, we screened this particular library instead of lambda libraries. And, in fact, in this way we were able to isolate the first large cluster of Hox genes.
This was a conceptual advance that was 100% dependent on a technical advance, which is a recurrent theme in biology, as you know. At that time, in 1986-1987, I collaborated with a colleague in the UK, Steve Gaunt, who was one of the handful of persons worldwide capable of running in-situ hybridization. It was a new technology with which you could see where the genes were expressed. In your email you asked me: Which is your favourite paper? My favourite paper was this one with Steve Gaunt, published in Development in 1988, where we showed this collinearity organization that Ed Lewis had shown, the fact that the expression of genes depends on their position in the cluster. We were certainly the first to report this in 1988. So, at this point, the conclusion was, well, the mouse have genes that resemble the genes in flies, because they have a homeo domain, and it seems that the functional organization is quite comparable, because they have this collinear system. But, whether the system was actually a copy of that of the flies was unknown. At this time, Pascal Dollé joined the lab as a graduate student. Pascal was an MD who decided to go for a PhD in biology. As an MD, he had a very good knowledge of anatomy, morphology, embryos and so on. And he imported the in-situ hybridization technology to the lab. Pascal was, perhaps, only the third person in Europe who was able to really do fantastic in-situ hybridization. So, we continued to clone these genes. And the reason why the 1989 paper became so popular is because we cloned the last two genes—the first one and the last one. The first one is called labial, which is the gene that makes the labial part of the insect, and the last one is called abdominal B, which makes the genitals at the other end of the body. These two genes were lacking. And without these two genes, you would not be able to say, look, this is a copy of an ancestral situation, one which went into the fly and one which went into a vertebrate. This is what we did in this 1989 paper. We showed that not only is the functional organization maintained, but the genes are the same. In fact, if you read the discussion, there’s a sentence in there about how this would mean that there was an ancestor to both invertebrates and vertebrates, which had one cluster, and the cluster was split in flies and amplified in vertebrates.
This paper actually fixed the next 35 years of research in my laboratory. We did many different things, but I always had in the lab three or four persons who were interested in solving how this functional organization works—what is the mechanism that activates these genes, one after the other, in different parts of the body, and so on. And, last year we got a part of the answer, which we published in Nature Genetics. So, yeah, it was an important paper.
HS: Stepping back a bit, from looking at your publication profile I realized that a lot of your early work was on mouse embryology, and it is after you joined the lab in Strasbourg, when you started becoming interested in molecular work. Could you talk a little bit about that?
DD: Oh my God, you’re touching a sensitive issue here. I did my PhD in Geneva in a very popular lab in mouse embryology. It turned out that the PI of the lab was somewhat very creative with results. Kurt Buerki, a senior scientist in the lab and myself went to him to discuss this and it ended up as a big clash. You can read about it on Wikipedia. It was a big mess. Actually, it was one of the first cases of scientific fraud that went onto the media. Anyway, what this meant was that I did the end of my PhD without a supervisor. There was an international inquiry committee who came to Geneva to investigate this case. And Pierre Chambon, who was an important figure in European science, was a member of this committee. After I got my PhD, I was a little bit depressed, and he told me: Come to recover in Strasbourg. That’s what I did. In Strasbourg, at that time, they were working on transcription and starting to use the mouse system. So, I helped to introduce the mouse system there. But, as I said, after three months, Pierre called me to his office and said: Look, I heard about this homeobox story—that was in 1984—and it seems to me very interesting. If you’d like to go for it, I’ll give you some people, and you just go on your own. I said, Yes, of course. That’s how I got in touch with Pierre Chambon, who actually became a rather close friend, because he trusted me and took me out of this difficult situation.
HS: Can you talk a little bit about the working relationship with Pascal Dollé? Was it like a supervisor-student relationship? What did each of you bring to the study?
DD: Well, look, over the past 40 years, I don’t think I ever had a real supervisor sort of relationship with my colleagues from the lab. It was always quite friendly. But, of course, Pascal was an MD. Initially, he was not trained in genetics, you know, in homeotic genes and so on. He was an extraordinary experimentalist and a brilliant anatomist. This is why he published this series of extraordinary papers, which were at the basis of my future laboratory. I had an excellent relationship with him, and I still do, as is the case with the huge majority of people going through my lab. It’s quite rare that I don’t keep a good relationship. Pascal’s retired now, which tells you something about how old I am!
Pascal was an extraordinary character. He thought that every result was normal. At that time, the frequency of homologous recombination in cells to knock out a gene was around 1%. So, Pascal would get 1% of recombination, and he would just say, Huh, okay. From this 1%, he would produce a mouse, and he would get one chimera out of it. And he’d say, well, that’s okay. And this one chimera would go through a germ line, and you would get one pop with the full genotype. And he’d say: that’s okay. He was a very Alsace character—very quiet, no show off, super fellow.
HS: What other memories do you have of this period?
DD: Oh, that was a fantastic period. Although, when you think about the past, you have a tendency to think it was better, because you eliminate the bad and you keep the good. But it was fun, especially my move in 1988 to the EMBL in Germany. That was a firework! It was all these small groups around, and each group was headed by a star at that time, and it was really a time of discovery for me. This is where I really discovered how science was organized. In Strasbourg, I learnt how science works. Those were real scientists—I mean, Scientists2 —going into molecular details and concentration and so on. Controls, controls, and controls! But at EMBL, I learned how to interact and collaborate, how to hire and these sorts of things, which is essential too.
HS: Was the paper written after you moved to the EMBL?
DD: Yes. The last discovery that led to the publication, which was the isolation of the sequence of the first mouse abdominal B gene, was done in Strasbourg. I remember I was reading a sequencing gel during an internal seminar of Pierre. There were 100 people in the room, and I was in the back. And I was so used to seeing the homeobox on the sequencing gels that I could actually read it without reading the exact bases, you know—I could see the motif. And that was the abdominal B. This I took with me to the EMBL, and a year after that, we wrote the paper. That’s why, in the Acknowledgements, you see that I’m thanking people from the EMBL for the drawings and so on.
HS: This relates to the question I wanted to ask you next. I’m always interested in how figures and images were made at that time, because it so different today. Can you to talk a little bit about that?
DD: Oh yeah; oh my god. There were two kinds of figures. The gels or the experimental figures we would do ourselves. We would take pictures, cut the pictures, stick them, and then we would add the famous Letraset—you know, these stickers with A, B, C? We would buy these sheets of Letraset. And, of course, after three As, there were no As anymore, so we would have to buy a new sheet! Because, of course, we needed to send three or four copies of figures to the editors, and we had to do each of these copies ourselves. The drawings were done by the drawing offices in the institutes. These offices have disappeared now, but at that time, in each institute, there was an office where there was someone making drawings with ink. The drawing in the paper—this sort of the summary figure—is very much influenced by my time in Strasbourg, in the sense that it is not a summary figure. It’s a scientific figure—you see the details; you see the positions of things. And when I look at this, I think I should never have done that. I should have instead made an embryo with colours, with the eyes—you see what I mean—something nice, like what actually went into most textbooks, often even on the cover. That was done, I think, three or four years later, by Eddy De Robertis, in a review paper in the Scientific American. He took this work and reconstituted a very nice embryo with colours and so on. And this went into the textbooks.
No one told me I should do it differently. I didn’t have a PhD supervisor and I didn’t have a postdoc supervisor so I made mistakes. Today, when I look back, I would say it doesn’t matter. But I would do it very differently today.
HS: I’d like to go over the names of the people you acknowledge, just to get a sense of who they were, how you knew them and what they contributed. The first is, of course, Pierre Chambon.
DD: Chambon was for supporting financially. I had a group in Strasbourg, but it was entirely financed by him. We would go to the fridge in his lab, just take the enzymes, and so on. I never needed to write a grant before I went to the EMBL. This is a gentleman type of science that you will not find nowadays—people who just encourage and facilitate young people to become independent and established without expecting anything from them. That was extraordinary.
HS: Then, you thank a number of people for “sharing ideas, unpublished results or technical competences”. The first name there is Robb Krumlauf.
DD: Robb was a competitor on this. He published a similar paper. The story has been quite well reported in the News & Views written by Mike Akam in 1988 on these papers.
HS: The next name is Peter Gruss.
DD: Peter Gruss later became the head of the Max-Planck-Gesellschaft. At that time, he was still interested in developmental biology, and he cloned some of these genes. Some of the positions you see in the big scheme in the paper came from his papers. It’s not something we needed as information to get the paper published, but it’s something that I took, to make a complete organization of the four clusters.
HS: Did you know all these people personally?
DD: Oh yes. They were all people engaged in similar work. That was a time of rush, a rush to get this gene family organized, understanding how they would be expressed, and so on. And, as I said before, all these people were using lambda clones—10 kilobase clones—while we were using 40 kilobase clones.
HS: The next name is Steven Gaunt.
DD: Steve Gaunt and myself initially collaborated with the in-situ hybridization. Before Pascal joined, we couldn’t get it to work in the lab, so we collaborated with Steve who mastered this technology. And, with Steve, we published two important papers: one in 1986 in Nature, which was the first description of the hox expression in the central nervous system, and then in 1988, the description of collinearity. Steve was a great colleague. He’s retired too.
HS: The next name is Walter Gehring.
DD: He cloned the homeobox in Basel. And he’s the person who gave us the little piece of DNA to screen the libraries. That was in 1984, much before we did the work. Walter was a friend too.
HS: And then, Pierre Gerlinger.
DD: Pierre Gerlinger was the colleague who taught me molecular biology in Strasbourg. Nothing to do with this project. When I arrived in Strasbourg, for three months, he was the guy who said: Look, I’m going to show you how you pipette, how you make ligation, how you amplify, how you clone bacteria, and so on. I had no idea of this; I was an embryologist.
HS: Then, Mark Featherstone.
DD: Oh, Mark was, I think, the first postdoc I had in Strasbourg. He’s a Canadian and must have commented on the English writing of the paper, because neither Pascal nor me were English-speaking. You can see that if you read the paper! Mark had a beautiful career, subsequently.
HS: And then, Charles Hart.
DD: Chuck was a postdoc in Chambon’s lab in Strasbourg. I think he too helped us with the writing and with comments and so on.
HS: Then, you thank A. Baron for help in the initial part of this work.
DD: Agnes Baron was my first graduate student in Strasbourg. She is the person who cloned the mouse labial gene. That paper was published in 1986. It’s a very interesting paper, in retrospect.
HS: And then, B.Schuhbaur for technical assistance.
DD: That was a technician who helped Pascal run the in-situ hybridization.
HS: And then, you thank B. Boulay and C. Werle for help in preparing the figures.
DD: They were at the EMBL in Germany when we prepared the figure.
HS: And finally, H. Davies for typing the manuscript.
DD: H. Davies was the secretary of the programme at EMBL. At that time, one had to type the entire manuscript. Oh my God!
HS: This paper was published in The EMBO Journal and communicated by Pierre Chambon. What’s the story here?
DD: The paper was sent to Cell first but it was rejected. For me, at the time, if it was rejected, it was rejected. Today, if I would read those comments and get a rejection, I would certainly write back to the editor, and the paper would go into Cell, eventually. You see what I mean? I had no clue at that time. I was raised in the countryside in Switzerland! I just had no supervision on this; no one ever told me what to do. Normally, you have a supervisor and, as a postdoc, you may not even see the comments. The comments go to the supervisor and they say, look, let’s try to go back with this and this. I thought, if it’s rejected, it’s rejected. To make a long story short, I went to see John Tooze, who was the editor-in-chief of EMBO, and I sent it to EMBO. At that time, you needed a member of EMBO to communicate the paper, so I naturally asked Pierre to be that person. Now, you no longer need to do this. It’s like in PNAS, where you no longer need a member to publish a paper.
In retrospect, I think it’s totally silly. I should have simply called the guy at Cell and said: Look, you don’t realize what this paper is about. It’s going to be heavily cited, I tell you, these sorts of arguments (I must confess that I have never been good at this). But, for me, it was like, when my dad would say no, it was a no!
HS: Do you remember anything about the peer review at EMBO? Was it smooth?
DD: It was quite smooth. It was rapidly accepted.
HS: What do you remember about any responses/reactions you got to the paper? It’s been cited over 1000 times now,
DD: The paper has been heavily cited. I remember getting extraordinary reactions in meetings. I remember a meeting in 1988 or 1989 in Arolla in Switzerland. That was fantastic. But, you know, strangely enough, sometimes you end up with conclusions that are so logical, that are so unexpectedly expected—you see what I mean? —that are completely novel, but, in fact, when you look at it, you say, but of course, it, can’t be otherwise. And that’s the reason why the referees at Cell—and I knew who they were—rejected it. How could we think for a century that an elephant has a trump because it has a gene for a trump? It’s so silly. In fact, many people consider this work as being sort of a natural conclusion out of years of work from many, etc, etc. But it was not. And actually, if people are careful, and if you go back to the literature, if you look at the content of papers on this topic published in 89, 88, 87, 86 etc, you will see that there is not a single sentence, even a speculation or a hypothesis, saying, look, there was an initial cluster early on and it went like this, which means that the ancestral animal already had this genetic system with a head and a tail and so on. I received much more reactions on the paper we published in 1989 in Nature on the limb stuff, which actually led to me eventually meeting Gerd Mueller and Lewis Wolpert and so on, showing that the same genes were actually also used for the limb axis, which was the initial description of molecular pleiotropy in this system. There, I got a lot of reaction because it was in the air. There was this gradient stuff from Lewis Wolpert and the positional information idea and so on. People were very interested. In this case, people switched to the new paradigm very rapidly, and they found it so expected—so normal. I never complain about this. I think it’s fine, it’s great. The value of a paper is not the number of citation but the continuum of citation over years. And so, you know, it’s fine with me. But it wasn’t an explosion of interest, even though the News & Views written by Mike Akam was quite interesting.
HS: In the same journal?
DD: No, it was written in Cell.
HS: You said a little earlier that this paper sort of fixed the research in your lab for the next 35 years. Can you talk a little bit about that? How significant was this for what you did subsequently?
DD: Two things, in fact. The first one is that, at the end of this paper, we ended up with a cluster. We had a full 100 kilobase of DNA in tubes, and therefore we could start looking at the regulation of these genes. And this took us 35 years. Till today, I’m writing things based on this piece of DNA. The second point, which is probably equally important, is that this paper contributed to me being a little known in the field. And, as you know, it brings funding, it brings grants, it sort of gets the ball rolling. If you publish a few papers like this, which are visible, you have an increased chance to get funded. If you get funded, you can hire postdocs. If you hire a good postdoc, etc, etc. So, in this sense, it was very important to me. And it’s partly because of this that I got hired in in Geneva a few years after, and where I got this huge mouse colony for myself. That was fantastic.
HS: And have you stuck with the mouse system since then?
DD: No. But that’s another issue. We run a very intensive mouse genetic program, crossing 1000s of mice and eliminating 1000s of mice. About six to seven years ago, with the emergence of this sort of a pseudo-embryo type of creatures that you can produce with embryonic stem cells, I decided that we should stop using real mice. We no longer work with mouse embryos, for the past seven years.
HS: Why did you decide to stop?
DD: Several reasons. The first one is that I never felt totally easy killing so many animals. I have no regrets, okay. I have no judgment on people doing this. Of course, if you work with the brain, there’s no question about it. But the day I could do otherwise, I decided to do otherwise. Also, the regulations in EU now are so tough—it is so costly to do this—that if you find an alternative system to study your questions, then you should go for it. It doesn’t mean that everyone should stop. Because, of course, there are many questions for which there’s no way you can do it without animals—metabolism, physiology and so on. But that was my choice and I’m happy about it.
HS: I want to read some lines from the abstract and ask you what your thoughts are on them today, 35 years after the paper was published.
DD: Yeah. Oh my god, 35, oh yeah. Thank you. Thanks for reminding me!
HS: You say: “These results clearly support the recently proposed hypothesis that the expression of murine Antp-like homeobox containing genes along the antero-posterior developing body axis follows a positional hierarchy which reflects their respective physical positions within the HOX clusters, similar to that which is found for the Drosophila homeotic genes. Such a structural and functional organization is likely conserved in most vertebrates. Moreover, on the basis of sequence comparisons, we propose that the ordering of homeobox-containing genes within clusters has been conserved between Drosophila and the house mouse. Thus, very different body plans might be achieved, both in insects and vertebrates, by evolutionarily conserved gene networks possibly displaying similar regulatory interactions.”
DD: It’s fine. Actually, I think it is a little bit less pompous than I imagined it to be. Because, you know, as a French speaking person, when I read my old papers, I find them so pompous. I can never reach the end! But this is fine. This is okay. It could essentially be written quite a bit the same today, except the last sentence. Because all these years of work, have shown that the way vertebrates implement this gene network is probably very comparable to a lot of arthropods, like, you know, shrimps and some short germband insects, but certainly not long germband insects like Drosophila. In fact, this system of Hox patterning is associated with animals that have progressively budding segments, like Tribolium—like the grasshopper. Flies don’t do this. They have evolved a completely new system where you produce something, and you cut it, and it’s fixed by the maternal component. So, the last sentence is wrong, to say that there would be the same mechanism in flies—in Drosophila—was a wrong hypothesis. I should have written: “shared mechanism between most arthropods and vertebrates”.
HS: More broadly speaking, what do you see as the place of this paper and its message in the literature on the subject today? What might be a reason for a student or a young researcher should read this paper today?
DD: Well, you know, I don’t think historians should ask scientists about how they see their own papers. That’s the beauty of being a historian. Whenever scientists start to do history, in particular of their own work, it’s often catastrophic. So, historians should do history, and scientists should wait to read that stuff. They may agree or they may disagree. But, you know, if you ask me the question, I would say, it is relevant in terms of evo-devo, but when considering evo-devo from the molecular developmental genetic side. Because, of course, you know—and I’ve had this discussion with Gerd [Mueller] many times—If you go to meetings of evo-devo, you see two crowds. The people coming from the ‘evo’ side, who think that evo-devo is rooted in Ernst Haeckel. And there is the developmental side who think that evo-devo was invented in the mid-80s. And, in fact, the term “evo-devo”, came from Mike Akam. And, it was, initially sort of a negative quote, sort of saying, “this is an evo-devo type of thing”. You see what I mean? Sort of a rapid conclusion taken from developmental biology to apply to evolution—this mechanism means that it evolved like this.. So, I would say, if you look at the molecular development side of evo-devo—I wouldn’t say that in public—it’s an important paper, because it really showed, for the first time, that not only a protein motif was conserved, not only genes were conserved, but the global functional organization of this network was conserved, which, I think, unified all bilateral animals by the existence of a developmental molecular mechanism. So it was, I think, a starting point for this reconstruction of the ancestral organism. Now, whether I would recommend a student to read this, I don’t know, frankly, because I think it is, as I said, a ‘Strasbourg-like’ paper, in the sense that it is detailed and not very readers-friendly. It’s certainly not over-interpreted—I over-interpreted other papers much, much more than this one. But it’s not the flamboyant, you know, there’s no colour scheme, etc. But it’s a good paper. I’m proud of it.
HS: Could you say a little more about how you see your own research in relation to evo-devo? Do you think of yourself as an evo-devo biologist, and has that changed over time?
DD: Oh, yeah. I always did. I remember, in 1993, when we knocked out the first gene in in the limb, we got a phenotype, which, for me, was very much looking like a heterochronic phenotype. So, we published a paper in Cell with the word “heterochronic” in the title. And I remember people in meetings would come to me and say: Denis, why the hell are you using words no one understands? What do you mean by heterochronic? I had to recommend Stephen J Gould’s wonderful Ontogeny and Phylogeny book, which was very important for me. But I never directly worked on this. We worked for something like two decades on the transition from the fin to the limb, but first and foremost at the level of the mechanism. I’ve always been convinced that it’s only the understanding of the mechanism that can make us understand the evolution of the mechanism. If you’re given a Citroën 2CV or a Tesla or a Maserati, you don’t understand the evolution of the engine. You need to understand how the engine works. It’s a classical citation from [Richard] Feynman, which is, “what I can’t create, I don’t understand”. And, as you know—I’ve seen the people you have interviewed, and most of them I know—they have other opinions on this. There are people from the “evo” side who think that the mechanisms of evolution is not the key point to understand. The key point is the classical Darwinian view, which is the selection of the best form and its distribution within a population. And I totally respect this. I think it’s very useful to have these different poles. I was always very interested to attend the European Evo-Devo meeting and was proud to receive the Kovalevsky medal, in this field. I’ve always been interested in evolution and I taught evo-devo as my first course in Collège de France seven years ago, a great exercise for me. It was fun. But, for me, it’s really difficult to think in evolutionary terms without an understanding of the mechanism. This is my bias. I don’t claim it’s a necessity. It’s my own bias.
HS: And has this been your view always?
DD: Yeah, always. I’ve always been reading a lot of stuff on mechanism, but quite little on evolution. I really think you need to understand how things works if you want to understand how they changed, how they evolved.
HS: This is my final question, one which you already answered partially earlier. I wanted to ask you whether you would consider this one of your favourite papers . You earlier said that the favourite is the 1988 paper with Steven Gaunt. Where does this stand?
DD: Oh, look, look, favourite comes with a qualificative: favourite for what? For writing? No. For the number of puns or jokes I put into it? No (I started doing that much later). But it is the 1988 and the 1989 papers that are my favourites. Why? I think mostly because of the point they make about the conservation of the systems. Also, they were not published in big fora. The 1988 paper—because we were in a hurry, we wanted to be first—we published in a supplement to Development. And this one was published in the EMBO Journal And, you know, they’ve been quite well cited. I sometimes tell my colleagues that what you need is a constant production of good work; not one extraordinary thing followed by a piece of nonsense. You need to keep producing good work, over the years, regardless of where it is published. Today, it is more complicated, because if you did that you may not get any grants. Also, this was an important paper for me, mostly because of the talks I gave at meetings, where I presented these results. This is how you make your name, when you go on stage and you talk about something, and people say: of course. I knew this, but where did I know it from? It’s so logical; so obvious. That’s the point. Yeah, I think I was lucky. I was lucky to be involved in this.
0 Comments