Les Collodaires (Radiolaires) sont des eucaryotes unicellulaires (protistes) marins appartenant au super-groupe des Rhizaria. Tandis que certains sont caractérisés par un mode de vie colonial, d’autres sont observés sous la forme de larges organismes solitaires. Les Collodaires sont des protistes hétérotrophes, prédateurs de plancton, mais également hôtes systématiques de micro-algues photosynthétiques intracellulaires. Les récentes analyses de leur diversité moléculaire dans l’environnement ont démontré leur importante contribution aux communautés planctoniques ainsi que leur distribution globale dans l’océan mondial. Cependant, nos connaissances sur leur diversité, biogéographie et écologie restent paradoxalement parcellaires. La première partie de cette thèse a été dédiée à des études morphologiques détaillées (en microscopie électronique et optique) et combinées à une phylogénie moléculaire élaborée en séquençant les sous-unités 18S et 28S de l’ADN ribosomal pour 75 spécimens, coloniaux ou solitaires. Ce travail a abouti à la réévaluation de la classification des Collodaires et à l’élaboration d’une base de référence morpho-moléculaire robuste. Par la suite, ce cadre de référence morpho-moléculaire a permis d’explorer la biodiversité des Collodaires grâce à une approche de metabarcoding appliquée à une série d’échantillons collectés dans l’océan mondial pendant l’Expédition Tara Océans. La distribution cosmopolite à la surface des océans des différents taxons qui composent les Collodaires, a révélé une diversité plus importante dans les vastes régions océaniques intertropicales et oligotrophiques. Les Collosphaeridae ont été principalement observés en pleine mer alors que les Sphaerozoidae formaient la famille dominante dans les régions côtières, où la biodiversité des Collodaires était plus faible. Les Collophidiidae, formellement décrits au cours de thèse, ont rarement été rencontrés dans les zones photiques, quelque que soit la latitude, suggérant ainsi qu’ils occupent une niche écologique particulière. Enfin, j’ai également employé la technologie d’imagerie in situ Underwater Vision Profiler (UVP5) afin d’explorer de façon quantitative les abondances et biomasses des Collodaires et des Rhizaria, à travers l’océan mondial. Cette approche a révélé que les Rhizaria forment un composant majeur du méso- et macro-plancton, et représentent jusqu’à 4,5% de la biomasse globale des 200 premiers mètres de l’océan mondial. Plus particulièrement dans les 100 premiers mètres, les Collodaires constituent le groupe le plus important des Rhizaria et leur distribution suggère que la photosymbiose pourrait influencer leur succès dans les régions oligotrophiques où ils sont particulièrement abondants. Au-delà d’améliorer notre compréhension de la diversité, la biogéographie et l’écologie des Collodaires dans l’océan mondial, ce travail de thèse souligne la pertinence de combiner et d’utiliser des approches alternatives d’échantillonnage et d’analyses tel que le séquençage haut-débit et l’imagerie in situ dans l’étude des protistes marins dans leur environnement. / Collodaria (Radiolaria) are unicellular marine eukaryotes (protists) belonging to the super-group Rhizaria. Collodarian species contribute to planktonic communities as large solitary cells or can form large gelatinous colonies. They are heterotrophic organisms feeding on other plankton, which also systematically harbour intracellular symbiotic microalgae. Recent environmental molecular diversity surveys demonstrated their important contribution to planktonic communities and their worldwide occurrence in the global ocean. However, knowledge on their diversity, biogeography and ecology is paradoxically very poor. In the first part of this thesis I performed detailed morphological analyses (electron and optical microscopy) combined with a molecular phylogeny based on the 18S and 28S rRNA genes, sequencing for a total of 75 distinct colonial and solitary specimens. Ultimately, this work led to the revision of the Collodaria classification and to the construction of a robust morpho-molecular reference database. Then, this morpho-molecular framework allowed the exploration of Collodaria biodiversity through a metabarcoding approach across samples collected in the global ocean during the Tara Ocean expedition. The cosmopolitan distribution of the different collodarian taxa in the surface oceans revealed a higher biodiversity in the vast oligotrophic inter-tropical open oceans. Collosphaeridae were predominantly found in the open oceans while the Sphaerozoidae were the dominant family in the less diverse coastal regions. The newly defined Collophidiidae were rarely encountered in the photic zones at all latitudes, suggesting that they inhabit a different ecological niche. Finally, I also used the in situ imaging system Underwater Vision Profiler (UVP5) to quantitatively explore the abundances and biomasses of collodarian and rhizarian in the global ocean. This approach revealed that the Rhizaria were a major component of the meso- and macro-plankton, constituting up to 4.5% of the global carbon standing stock in the upper 200 m of the world oceans. More specifically, Collodaria were the most important rhizarian groups in the first 100 m of the oceans, and their distribution suggested that photosymbiosis might be an important factor explaining their success in oligotrophic regions where they are particularly abundant. Besides the improvement of our knowledge on the diversity, biogeography and ecology of Collodaria in the global ocean, this thesis highlights the relevance to combine and/or use alternative sampling and analytical procedures such as high-throughput sequencing and in situ imaging technologies to study marine protists in their environment.

In the late 1990s, a series of landmark publications described RNA interference (RNAi) and related RNA silencing phenomena in nematodes, plants and fungi. By manipulating RNA silencing, biologists have been able to create tools for specifically inactivating genes. In organisms from trypanosomes to insects, RNA silencing is now indispensible for studying gene function. RNA silencing has been used in a project aimed at systematically knocking out all genes in the model plant Arabidopsis thaliana. RNA silencing has a natural role in defending eukaryotic cells against virus replication. By assembling virus DNA sequences in a form that triggers RNA silencing, biologists have created plants resistant to specific viruses. In this study, we set out to test if a similar approach would protect plants against infection by the agriculturally important Brassica pathogen, Plasmodiophora brassicae. P. brassicae is an obligate intracellular biotroph, from the little studied eukaryotic supergroup, the Rhizaria. To identify the gene sequences that would be starting material for P. brassicae RNA silencing, new P. brassicae genes were gathered by cDNA cloning or genomic PCR-walking. Using suppression subtractive hybridisation (SSH) and oligo-capping cloning of full-length cDNAs, 76 new gene sequences were identified. A large proportion of the cDNAs were predicted to contain signal peptides for ER translocation. In addition to the new cDNA identified here, partial sequences for the P. brassicae actin and TPS genes were published by other researchers close to the beginning of this study. Using PCR-walking, full-length genomic DNA sequences from both genes were obtained. Later, genomic DNA sequences spanning or flanking a total of 24 P. brassicae genes were obtained. The P. brassicae genes were rich in typical eukaryotic spliceosomal introns. Transcription of P. brassicae genes also appears likely to begin from initiator elements rather than TATA-box-containing promoters. A segment of the P. brassicae actin gene was assembled in hairpin format and transformed into Arabidopsis thaliana. Observation of simultaneous knockdown of the GUS marker gene as well as detection of siRNAs indicated that the hpRNA sequences induced RNA silencing. However, inoculation of these plants with P. brassicae resulted in heavy club root infection. We were unable to detect decreases in actin gene expression in the infecting P. brassicae, at either early or late stages of infection. We conclude that, within the limits of the techniques used here, there is no evidence for induction of RNA silencing in P. brassicae by in planta produced siRNAs.

In the late 1990s, a series of landmark publications described RNA interference (RNAi) and related RNA silencing phenomena in nematodes, plants and fungi. By manipulating RNA silencing, biologists have been able to create tools for specifically inactivating genes. In organisms from trypanosomes to insects, RNA silencing is now indispensible for studying gene function. RNA silencing has been used in a project aimed at systematically knocking out all genes in the model plant Arabidopsis thaliana. RNA silencing has a natural role in defending eukaryotic cells against virus replication. By assembling virus DNA sequences in a form that triggers RNA silencing, biologists have created plants resistant to specific viruses. In this study, we set out to test if a similar approach would protect plants against infection by the agriculturally important Brassica pathogen, Plasmodiophora brassicae. P. brassicae is an obligate intracellular biotroph, from the little studied eukaryotic supergroup, the Rhizaria. To identify the gene sequences that would be starting material for P. brassicae RNA silencing, new P. brassicae genes were gathered by cDNA cloning or genomic PCR-walking. Using suppression subtractive hybridisation (SSH) and oligo-capping cloning of full-length cDNAs, 76 new gene sequences were identified. A large proportion of the cDNAs were predicted to contain signal peptides for ER translocation. In addition to the new cDNA identified here, partial sequences for the P. brassicae actin and TPS genes were published by other researchers close to the beginning of this study. Using PCR-walking, full-length genomic DNA sequences from both genes were obtained. Later, genomic DNA sequences spanning or flanking a total of 24 P. brassicae genes were obtained. The P. brassicae genes were rich in typical eukaryotic spliceosomal introns. Transcription of P. brassicae genes also appears likely to begin from initiator elements rather than TATA-box-containing promoters. A segment of the P. brassicae actin gene was assembled in hairpin format and transformed into Arabidopsis thaliana. Observation of simultaneous knockdown of the GUS marker gene as well as detection of siRNAs indicated that the hpRNA sequences induced RNA silencing. However, inoculation of these plants with P. brassicae resulted in heavy club root infection. We were unable to detect decreases in actin gene expression in the infecting P. brassicae, at either early or late stages of infection. We conclude that, within the limits of the techniques used here, there is no evidence for induction of RNA silencing in P. brassicae by in planta produced siRNAs.

In the late 1990s, a series of landmark publications described RNA interference (RNAi) and related RNA silencing phenomena in nematodes, plants and fungi. By manipulating RNA silencing, biologists have been able to create tools for specifically inactivating genes. In organisms from trypanosomes to insects, RNA silencing is now indispensible for studying gene function. RNA silencing has been used in a project aimed at systematically knocking out all genes in the model plant Arabidopsis thaliana. RNA silencing has a natural role in defending eukaryotic cells against virus replication. By assembling virus DNA sequences in a form that triggers RNA silencing, biologists have created plants resistant to specific viruses. In this study, we set out to test if a similar approach would protect plants against infection by the agriculturally important Brassica pathogen, Plasmodiophora brassicae. P. brassicae is an obligate intracellular biotroph, from the little studied eukaryotic supergroup, the Rhizaria. To identify the gene sequences that would be starting material for P. brassicae RNA silencing, new P. brassicae genes were gathered by cDNA cloning or genomic PCR-walking. Using suppression subtractive hybridisation (SSH) and oligo-capping cloning of full-length cDNAs, 76 new gene sequences were identified. A large proportion of the cDNAs were predicted to contain signal peptides for ER translocation. In addition to the new cDNA identified here, partial sequences for the P. brassicae actin and TPS genes were published by other researchers close to the beginning of this study. Using PCR-walking, full-length genomic DNA sequences from both genes were obtained. Later, genomic DNA sequences spanning or flanking a total of 24 P. brassicae genes were obtained. The P. brassicae genes were rich in typical eukaryotic spliceosomal introns. Transcription of P. brassicae genes also appears likely to begin from initiator elements rather than TATA-box-containing promoters. A segment of the P. brassicae actin gene was assembled in hairpin format and transformed into Arabidopsis thaliana. Observation of simultaneous knockdown of the GUS marker gene as well as detection of siRNAs indicated that the hpRNA sequences induced RNA silencing. However, inoculation of these plants with P. brassicae resulted in heavy club root infection. We were unable to detect decreases in actin gene expression in the infecting P. brassicae, at either early or late stages of infection. We conclude that, within the limits of the techniques used here, there is no evidence for induction of RNA silencing in P. brassicae by in planta produced siRNAs.