Activités cachées et mobiles du génome / The dynamic genome: hidden and mobile activities
Séminaire organisé par Bernard de Massy (IGH, Montpellier, France) et Déborah Bourc’his (Institut Curie, Paris, France) du 2 au 7 mai 2017
Peter Refsing Andersen, Mireille Betermier, Déborah Bourc’his, Julius Brennecke, Vincent Colot, Jérôme Déjardin, Laurent Duret, Anas Fadloun, Ian Henderson, Todd Macfarlan, Harmit Malik, Robert Martienssen, Bernard de Massy, Didier Mazel, Maria-Elena Torres Padilla, Didier Trono, Jonathan Weitzman
Les activités du génome dépassent largement l’information codée par les gènes. L’importance des régions non codantes et de séquences répétées et parfois mobiles est de plus en plus largement reconnue mais ces activités sont encore mal comprises. Elles ont la particularité d’être soumises à des contraintes évolutives différentes de celles qui s’exercent directement sur des régions codantes. L’objet de cette conférence a été de rassembler des experts qui étudient différents aspects de ces propriétés du génome mais sous des angles différents. Les présentations ont révélé les multiples facettes qui illustrent à quel point notre génome (ou celui d’autres espèces) est en interaction avec d’autres éléments génétiques et avec son environnement. Ces interactions qui peuvent être conflictuelles, peuvent être source de diversité également. Elles sont par essence dynamiques et comprendre leur fonctionnement implique d’analyser leur évolution. Les participants ont discuté comment les coûts dus à l’invasion d’un génome par des éléments égoïstes peuvent être transformés pour finalement ouvrir de nouvelles possibilités d’évolution. Les différentes études présentées ont permis de confronter les différentes stratégies moléculaires impliquées, allant de modifications de l’ADN et de la chromatine à l’organisation des chromosomes. L’évolution moléculaire et comment la sélection agit sur ces activités ont été au centre de ces débats, confrontant des expertises de biologie moléculaire et de génétique des populations. Il est clair que les nombreux aspects non codants ou répétés dans le génome et qui peuvent en composer la majorité sont à la base de nombreuses activités de la cellule et de son interaction avec l’environnent. Les recherches dans ce domaine ont des perspectives particulièrement intéressantes.
Compte-rendu (en anglais)
Repressing selfish elements
One of the challenges faced by many genomes is to control the activity of genetic elements that have a selfish behavior, such as transposons. Several strategies have been developed to repress the activities of transposable elements (TEs) but at the same time, some of those elements or part of them can be high jacked by the host for performing other functions. In this case, these elements become a source of genetic diversity and innovation.
Déborah Bourc’his has presented her discovery of a relatively recent DNA methyltransferase, DNMT3C, which selectively methylates the most active transposons, specifically their promoter regions and only in the context of fetal spermatogenesis. This activity is absolutely essential for male fertility in the mouse, yet the Dnmt3C gene is not conserved in all mammals, but emerged by tandem duplication of the Dnmt3B gene specifically in Muroidea rodents some 46 MYr ago. D. Bourc’his discussed the principles by which DNMT3C may sense active transposons, how it may have evolved divergent function relative to its paralog DNMT3B and how non-Muroidea mammals- men included- may utilize DNMT3B to produce a DNMT3C-like protein.
Another major pathway for repressing transposable elements involves small RNAs. Rob Martienssen and Julius Brennecke, – by studying plants, mammalian cells and drosophila – have revealed exciting properties of these pathways. In mouse embryonic stem cells, R. Martienssen has discovered that epigenetically induced states of transposon reactivation lead to the production of active small RNA species, namely 18 nucleotide-long tRNA fragments (tRFF). Although their biogenesis remains unknown, they specifically target ETn transposons, and inhibit their ability to mobilize by interfering with their reverse transcriptase activity. J. Brennecke reported novel insights into the piRNA pathway in the follicle cells of Drosophila and has addressed the important question of how this pathway achieves specificity in the selection of piRNA precursors. According to his exciting findings, J. Brennecke proposes that basically all accessible single stranded transcripts are scanned by a machinery, which has the capacity to feed a single stranded RNA to the nuclease Zucchini for processing it into piRNAs. Processes like active translation or other features of endogenous mRNAs would inhibit this scanning process, hence the system would process preferentially non-mRNA transcripts, for example those emerging from transposon enriched piRNA clusters. Another essential component of this pathway is the production of piRNAs in the germ line, for which Peter Andersen finally resolved this long-sought conundrum: He demonstrated that a platform of non-canonical transcription proteins is assembled at the chromatin of piRNA clusters, including paralogs of the transcription initiation factors TFIIA and TBP. Multiple and concerted events of gene duplication and functional reconnection have therefore been instrumental during evolution to exploit the chromatin identity of piRNA-producing loci in order to achieve productive transcription from these loci, a prerequisite for efficient piRNA biogenesis.
One amazing strategy to deal with multiple invading elements in the genome has been developed in ciliates. Mireille Bétermier who is studying Paramecia has shown recent advances in the remarkable molecular program that allows the removal of these otherwise deleterious elements from the genome to allow a proper gene expression program. M. Bétermier has recently gained interesting insight into the principles of PiggyMac, the domesticated transposase that catalyzes DNA elimination: its activity is dependent upon nine different PiggyMac-like cofactors that have apparently lost catalytic activity but promote PiggyMac stabilization and/or translocation into the nucleus.
Maintaining the repression of transposable elements
A major question in repressing such activities is whether this repression is stable, trans generational, or needs to be reestablished at each generation. Vincent Colot uses the plant Arabidopsis thaliana as a model to study the transgenerational inheritance of transposon-mediated phenotypes. His experimental model relies on the genetic induction of DNA methylation loss and subsequent segregation of induced differentially methylated regions (DMRs) over generations (Epigenetic Recombinant Inbred Lines, EpiRILs). He has demonstrated that indeed DMRs can be stable over generations. Remarkably, this experimental system perfectly recapitulates the spectrum of DNA methylation variants, transposon transcripts and spontaneous insertions that exist among wild Arabidopsis accessions.
Allowing a window of expression, the road to domestication
Epigenetic regulation is by nature reprogrammable, and this is specifically crucial in multicellular organisms for allowing cell fate diversification. One question faced in this field is how to erase epigenetic marks while maintaining repeat elements repressed. The very early mouse embryo shows dynamic regulation of transposon expression, with a massive upregulation at the time of zygotic activation and then silencing towards the formation of the pluripotent blastocyst. Maria Elena Torres Padilla suggests that these transposon regulatory changes may provide important functions. M.-E. Torres Padilla presented support for the hypothesis that these may impact on chromatin activity and accessibility of the early embryonic genome, another strategy to take advantage of this hidden fraction of our genome. A pathological context where some transposons are reactivated is cancer cells. Anas Fadloun proposed to take advantage of this characteristics to kill cancer cells, by driving the expression of a drug-induced suicide gene that would only be expressed upon transposition of an Alu element. This system is being optimized and may prove to be useful for targeting cancer cells or for new screening strategy for controllers of transposition.
The arm race between transposons and repressors
Mammals rely on an impressive molecular machinery for targeting transposons, the family of KRAB Zinc Finger Proteins (KZFPs). These proteins perform their function by recruiting heterochromatic complexes to their targets. As a consequence of their involvement in an evolutionary arm race against transposons, many of these proteins are rapidly evolving, using the high mutability of their zinc finger domains. This can lead to the emergence of new sequence targets and their co-option as epigenetic regulators for other genomic sequences than transposons.
The groups of Didier Trono and Todd MacFarlan have illustrated the evolution and diversity of functions of ZFPs. D. Trono has performed a comprehensive analysis of the phylogeny, genomic and molecular properties of a very large number of KZFP proteins. This study highlights a remarkable domestication of TEs by KZFPs, which have contributed to the diversification of regulatory networks with remarkable specificity in different lineages in vertebrates. T. MacFarlan is taking a systematic genetic approach of deleting individual or cluster of ZFP genes to understand the spectrum of their functions. One striking example of functional rewiring from transposon to cellular gene control is provided by the ZFP568, which has evolved specifically in placental mammals to control an alternative placenta-specific promoter of the growth-promoting Igf2 gene in the mouse.
An unlimited plasticity
Didier Mazel – who studies antibiotic resistance in various gram negative bacteria- has unraveled several key aspects of the process that has led to multidrug resistance in bacteria in response to the rise in antibiotic treatments. An outstanding molecular system has evolved and generated integrons, a structure that is made of mobile cassettes that can be acquired by various mechanisms – including conjugative plasmids-, thus allowing the expression of various antibiotic resistance genes. Didier Mazel presented the sophisticated site-specific recombination mechanism that allows the cassettes to be inserted, mobilized and expressed in the host genome (Vibrio cholera). The outstanding plasticity of this system is yet another example of the dynamic genome and interactions between a host and its environment.
Genetic conflicts due to selfish elements
Among the many ways selfish elements can promote their propagation, one is meiotic drive, where a selfish element favors its transmission upon meiosis. Harmit Malik and Sarah Zanders have unraveled a fascinating example of such meiotic drive in Schizosaccharomyces pombe. A specific locus and allele on S. kambucha genome produces a poison that can diffuse among spores and kill all spores after meiosis, unless they carry an antidote that is expressed from the same locus in a spore-autonomous manner. These two opposing genetic information – poison and antidote- emanate from two different transcripts of the same locus. This locus named wtf4 is thus a selfish element that distorts segregation in its favor and generates genetic incompatibilities. This highlights how selfish, parasitic elements can propagate, even at the expense of fertility of the species.
The PRDM9 protein is known to determine genome-wide the localization of meiotic recombination events in humans and mice. However, the behavior of this protein is puzzling as its binding sites are eroded during evolution and this process is linked to genetic incompatibility between some mouse strains. In fact, Prdm9 has been proposed to a speciation gene. Bernard de Massy has presented new features of PRDM9 activity, which involves its lysine methyltransferase activity. In addition, based on protein interaction analysis, B. de Massy has proposed a model that could explain the role of H3K4me3 and be important to understand how heterozygosity could lead to hybrid sterility.
Controlling genetic interactions by genetic linkage
The coexistence along chromosomes of selfish elements, regulatory regions and genes, raises the question as to how they became genetically linked and how this could impact the efficiency of selection. Ian Henderson presented the highest resolution recombination map of A. thaliana using Spo11-oligo profiling and showed an enrichment for DNA breaks at the 5’ ends and surprisingly also at the 3’ ends of genes. Moreover, DSB were found to be suppressed in centromeric regions but not in met1 mutants (deficient for DNA methyltransferase), revealing a DNA methylation-mediated regulation of DSB repair in these chromosomal regions. Laurent Duret has analyzed an important feature of genetic diversity: the effect of GC-biased gene conversion (GCBGC), its quantitative impact on genome evolution and its occurrence among eukaryotes. The strength of the bias measured in several species is remarkably strong, and its correlation with recombination further demonstrated. However, the GC bias itself appears to be modulated by population size and/or recombination rate, which eventually limits the increase in GC content resulting from this bias. The recent analysis by L. Duret on a wide variety of species suggests that this phenomenon is very largely spread among eukaryotes.
Organization of non-genic genomic regions
Understanding the chromatin and protein-based factors acting on repeated elements is a major challenge that Jérôme Déjardin is trying to tackle using a technique he developed, called PICh. He first presented an improved PICh-derived strategy, which allows a significant gain in yield and purification efficiency (by 10 to 100-fold compared to regular PICh), and therefore the identification of most factors bound to repetitive genomic regions from a low amount of starting material. In the second part of his talk, he described how heterochromatin forms at the extremities of chromosomes (telomeres), and how this heterochromatin- which harbors unusual features- is used to control telomere transcription and maintenance.
Taking advantage of neighbors
Jonathan Weitzman studies Theileria annulata, which exemplifies that genetic innovation can be gained in the absence of transposons. This parasite has emerged from at least two endosymbiotic events and contains a nuclear, a mitochondrial and an apicoplast (algae-derived) genomes. Among the different apicomplexan species, Theileria is remarkable by its ability to transform the host (ruminant blood cells) towards a cancer-like phenotype. Jonathan showed that this characteristics is linked to the acquisition of specialized machineries: 1) signal peptides, such the one added onto the Pin1 propyl isomerase that inhibits host ubiquitin-based degradation and promotes the upregulation of the c-jun transforming factor and 2) epigenetic components, such as SET-endowed methyltransferases, which were acquired by horizontal gene transfer and may confer a specific chromatin state to the Theileria genome. “Stealing, hacking and hijacking”: this may be the key for evolving and specializing compact, transposon-less genomes.
The meeting allowed building very constructive and interesting interactions between researchers who do not often meet in regular international conferences. The format was fantastic for discussions and each presentation stimulated very lively discussions between participants, several of which also extended during the allocated social times. Participants were researchers (15) and postdocs (2).