ACCELERATED EVOLUTION IN CHORDATES AND THE ORIGIN OF LARVACEANS

This TOPPFORSK project aims at studying the concomitant evolutions of tunicate genome and development and ultimately understand how alteration of ancestral chordate genomes have led to a regressive-like change of anatomical complexity. On the genomic side, experimental work is now completed and in addition to three high impact publications already published, two or three manuscripts are programmed. Two will be on the development part of the project and a third one still depends on experiments on splicing mechanisms. (1) The divergence of double strand break (DSB) DNA repair, with Oikopleura and other larvaceans having lost the genes and functions for non-homologous end-joining (NHEJ), as experimentally demonstrated using DNA injections of linear DNA and CRISPR induced mutations. DSBs are repaired using an alternative mechanism exploiting microhomologies on each side of the break (MMEJ-like). Additional and later experiment have tested the effect of knocking down the players of MMEJ mechanism (RNAi and inhibitors), but no conclusion could thus far be reached. Novel projects will be prepared so to identify the players through biochemical identification of molecules associated to the double strand breaks. (2) A correlation of larvacean genome size with the abundance of non-autonomous retrotransposable elements (particulary SINEs) was revealed and published in spring 2019. The ancestral genome was probably small and expanded through invasions by these elements, driven by autonomous counterparts which are not detected in the genome. This is a rather original situation which may reflect specific strategies of transposable elements to operate despite their strict control in genomes refractory to invaders. To pursue on this topic, it has been decided to explore and describe many new related genomes that appeared after the worldwide collection and sequencing of marine organisms (Tara expedition). Already thirty appendicularian genomes have been detected and assembled. (3) The invasion of genes of a family of larvacean species by very large numbers of non-canonical introns. These undoubtedly originate from non-autonomous DNA transposable elements (MITEs), this phenomenon has been observed in a few unicellular organisms but not in animals. Beyond the departure from canonical splice sites, the broad diversity of non-canonical splice sites is striking and raises questions on how they could be spliced out. Various experiments show that this is done by a U2 spliceosome which must have been modified. The exact nature of the modifications is unknown and will be the question for another major grant application. The publication of current results appeared in late 2019. Two strategies have been elaborated, one based on a classical dissection of splicing mechanisms with biochemistry, the other on studying the function of a couple of duplicated splicing factors, that may have evolved to process non-canonical introns. The development side of the project was finally boosted by a good control of the CRISPR technology that we initially established for the above DNA repair project. While it is easy to mutate cells of the embryo, obtaining stable mutant lines for essential development genes is more challenging. This was however possible for four homeobox genes. (4) Mutant lines have been obtained for Pax37B (involved in the development of the house-making epithelium, an innovation of larvaceans) and for Hox1 (the function of Hox genes in tunicates remains essentially unknown). Homozygous mutants show a phenotype, the absence of the largest epithelial cells for Pax37B, those which produce the food concentration filter of the animal house. This seems sufficient for provoking the death of juveniles which would be unable to collect food. More detailed analysis revealed that one of the cells may be present, posing questions on the mechanism of development involved and its genetic variation. (5) As reported earlier, the Hox1 loss of function mutant is also lethal but does not compromise the early development. Mutants show a partial paralysis of larvae and change of the trunk to tail angle. Expression of various markers expressed around the tail trunk junction is examined to understand where precisely the phenotypical change arises from. (6) For both mutants and others in the future, transcriptome studies are engaged in comparison with wild type controls to identify genes that are indirectly or directly regulated by the mutated genes. Such analysis however did not succeed, possibly due to additional expression sites. A solution should be the analysis of transcriptome in single cells and we have begun to establish the method of single cell sequencing for early developmental stages in Oikopleura. This could permit to answer whether other genes of the same ancestral genetic pathways were also co-opted for the epithelium function together with Pax37B, as well for the currently unknown mechanisms to which Hox1 contributes.

The main outcomes are the surprising discoveries of new molecular mechanisms influencing genome evolution. The impact is the necessity for a broad scientific community to reconsider part of the knowledge in textbooks, which basically says that such mechanisms are universally conserved since found in yeast and in human. One example is the loss of an essential DNA repair pathway (NHEJ), the other is the invasion of the genome and process of a multitude of diverse non-canonical introns. The other outcomes are for the establishment of critical technologies in Oikopleura, such as CRISPR, genome sequencing in planktonic animals and currently single cell transcriptome sequencing, but the impact here is restricted to a smaller number of research teams.

The evolution of multicellular organisms is generally seen as a process gradually increasing their anatomical and physiological complexity, through the recurrent addition of lineage-specific innovations. The molecular analysis of phenotypical changes in relatively small taxons (to which belong genetically tractable model systems) suggests that modifications of gene regulation is their main driver, rather than changes in the complement of genes. However, classical zoological studies and/or recent phylogenomics have recognized multiple cases where complexity has been reduced more or less dramatically. In such situations, important changes of the genome from those of more complex ancestors may be observed, including significant gene losses or a relaxation of genome architecture. Despite considerable knowledge of the vertebrates and their historical radiations, the history of the chordate phylum (to which they belong) remains very enigmatic, in part because non-vertebrate chordates have left little and often puzzling fossil records. The present project is on the evolution of larvaceans, a major component found in the plankton of all oceans. Larvaceans belong to the chordate subphylum of tunicates, which we found to be the sister group of vertebrates, despite their strikingly simpler anatomy. Our goal is to reconstruct their history from tunicate and chordate ancestors, during which evolution was strongly accelerated. We propose to examine the correlation between rapid genome changes and the divergence of developmental mechanisms in larvaceans. Our project has three main objectives (i) to better understand how reduction of anatomical complexity and a major lineage-specific innovation could be achieved through changes of developmental mechanisms, (ii) to correlate the changes of developmental mechanisms with accelerated genome modifications, (iii) to facilitate the project by improving methodologies.