Bernard Dujon

In Gif-sur-Yvette, Bernard Dujon started to study a strange genetic phenomenon, linked to mitochondrial genetics, whose study was still in infancy [7]. When crossing two haploid yeast species carrying different mitochondrial mutations, conferring the resistance to erythromycin or to chloramphenicol, allele segregation did not follow mendelian rules and recombinants appeared in mysterious proportions. At that time, no one had any idea of the genetic content of mitochondria, except that it contained DNA. Bernard Dujon was studying a particular mitochondrial locus, called omega, that could be present as two different alleles, omega+ or omega-. Genetic crosses between yeast cells carrying different alleles led to highly distorted inheritance in the progeny, since almost all cells ended up carrying the omega+ allele! In 1974, he proposed a model in which gene conversion of the omega- allele to omega+ was achieved by homologous recombination, replacing one allele by the other, copying in the process the flanking erythromycin and chloramphenicol mutations.

At about the same time, recombinant DNA technologies and restriction enzymes were discovered. In 1977, independent researches by Fred Sanger on one side and by Walter Gilbert and Allan Maxam on the other, led to the invention of two different methods to sequence DNA. Later the same year, introns were discovered. Thermal denaturation studies with François Michel, another student of Piotr Slonimski, suggested that omega could be an intron. Bernard Dujon contacted Walter Gilbert at Harvard University about the possibility to come to his lab for a short post-doctoral period, in order to sequence the omega locus. He moved to Harvard the following year and in 1979 achieved the sequence of what would become the first mobile intron to be described [8]. But the most surprising result was not that omega was indeed an intron, but that it contained an open reading frame, putatively encoding a 235 amino acid protein with no homology to any known protein. At that time, no intron was supposed to be coding. Could it be that the encoded protein played a role in the intron mobility between omega- and omega+ strains ? [9]

Back to Gif-sur-Yvette in 1981, Bernard Dujon set up a small team in an old lab space lent by Piotr Slonimski. François Michel rapidly joined him and later on Alain Jacquier, Hugues Blanc, Pierre Dehoux and Laurence Colleaux, as well as sabbatical visitors such as Walt Fangman from the University of Washington. They discovered that the omega intron was present in other yeast species collected in Harvard. Following the sequencing of several other introns, François Michel discovered that these introns could be folded into stem-loops whose structures (if not their sequences) were conserved. This suggested that they could be directly involved in the splicing mechanism by defining exon-intron junctions. In addition, they discovered that two different intronic structures existed, defining what they called group I [10] and group II introns, a nomenclature still in use today. They published their models of intron folding in Biochimie in 1982 and this article quickly became a reference for researchers in the field [11].

But the precise function of the omega-encoded protein was still unknown. Bernard Dujon decided to adapt the mitochondrial gene to the universal genetic code in order to be able to express it in a heterologous system. At that time, it was a real tour de force, since oligonucleotide synthesis and in vitro mutagenesis were uncommon and not available in Gif-sur-Yvette. Fortunately, Bernard Dujon met Francis Galibert, who was working at the Hôpital Saint Louis in Paris and who just came back from Fred Sanger laboratory to set up his own lab. At that time, he was the only one in France to be able to synthesize oligonucleotides. With the help of Francis Galibert's oligonucleotides, Bernard Dujon modified 26 of the 235 codons of the omega reading frame to adapt it to the universal genetic code. Synthesis of the resulting protein in Escherichia coli, in the presence of a plasmid carrying the omega-sequence showed without any ambiguity, in 1985, that the omega protein was a double-strand DNA endonuclease, as was predicted by the model 12 years ago [12]. This nuclease took afterwards the conventional name of I-SceI [13], the first intron homing endonuclease discovered, first of its kind but dozens of others would rapidly follow.

In 1987, a call for a yeast geneticist was published by the Institut Pasteur. Bernard Dujon applied and left Gif-sur-Yvette to move to Paris. In this new scientific environment, he used I-SceI to the purpose of making unique double-strand breaks in complex genomes, such as mouse [14], plants or the human genome, in collaboration with many scientists worldwide. In collaboration with the laboratory of Jean-François Nicolas at the Institut Pasteur, Arnaud Perrin and André Choulika (who would later become two founding members of the biotech company Cellectis) were able to do the first gene replacement in mouse cells using I-SceI [15], at the same time Maria Jasin in United States was doing similar experiments in human cells [16].

In the mid-1980s, advances in cloning and DNA sequencing techniques made it possible to sequence entire genomes of increasing size. Saccharomyces cerevisiae, was the first eukaryotic organism entirely sequenced thanks to an international collaboration . Dujon has been involved in this venture since its inception, coordinating the sequencing and mapping of several chromosomes.[17] The yeast sequence, published in 1996, opened a new area because a very large number of genes whose functions remained totally unknown were discovered in a model organism that had been extensively studied, and the sequences did not reveal any homologues.[18][19] The existence of these "orphans", which are a property of all genomes sequenced to date, combined with other properties of the yeast genome, such as its high redundancy, compactness and homologous recombination efficiency, have made this organism one of the prototypes for "functional genomics" research.

At the end of the 1990s, the acceleration of DNA sequencing methods and advances in silico genome analysis led Dujon, together with other French laboratories, to develop an "evolutionary genomics" of eukaryotic genomes by comparing the genomes of several yeast species in the Hemiascomycetes group.[20] These yeasts reveal at the level of their genomes, large evolutionary distances between them.[21][22] Their comparison makes it possible to develop new theories on the molecular mechanisms of evolution of eukaryotic genomes which, thanks to the power of genetics in S. cerevisiae, can be directly subjected to experimentation. By combining the comparative analysis of genomes in a given evolutionary phylum, that of Hemiasomycetes, with the experimental approaches allowed by these yeasts, Dujon has focused on the mechanisms that govern the evolution of eukaryotic genomes, such as gene formation and loss,[23] horizontal genetic exchanges[24] and the mechanisms of massive alterations in chromosomes resulting from accidents involving DNA replication in cells placed under stress.[25]Dujon is also the author of a non-technical book on genetics.[26]