L'ARN interférence (ARNi) est un phénomène découvert en 1998 par lequel la présence d'ARN double brin au sein d'une cellule entraîne la dégradation d'ARN de séquence homologue. L'ARNi est effectué par un complexe ribonucléoprotéique contenant des petits ARN double brin et au moins une protéine de la famille Argonaute. Cette thèse a été consacrée à l'étude de l'ARNi chez le protozoaire Trypanosoma brucei. Nous avons d'abord défini les conditions d'utilisation de l'ARNi au niveau de la spécificité et de l'efficacité, paramètres qui ont servi à l'élaboration d'un logiciel permettant la sélection de l'ARN double brin pour les études fonctionnelles. Ensuite, nous avons recherché plusieurs gènes candidats codant pour des protéines participant à l'ARNi. Le meilleur d'entre eux, TbAGO1, appartient à la famille Argonaute et se caractérise par la présence d'un domaine supplémentaire, capable de lier les ARN. Il est essentiel pour l'ARNi chez le trypanosome. Sa délétion produit des défauts significatifs lors de la mitose et nous avons établi que l'ARNi contribue à la formation du fuseau mitotique et à la ségrégation des chromosomes. Un second phénotype observé en l'absence d'ARNi est la surexpression des ARN de deux types de rétroposons (rétrotransposons sans LTR), sans toutefois augmentation de leur activité de rétroposition. Les deux phénotypes sont indépendants l'un de l'autre. Nous avons ensuite démontré que la présence d'ARN double brin entraîne la destruction d'ARN cible de séquence homologue dans le cytoplasme mais peut aussi conduire à une extinction de la transcription du gène correspondant. Ce type de mécanisme pourrait non seulement contrôler l'expression des ARN des rétroposons, mais aussi celle des gènes dans lesquels ils sont insérés.

MicroRNAs are small noncoding RNAs that post-transcriptionally reduce protein output from most human mRNAs by mechanisms that are still obscure. This thesis provides insights into three aspects of microRNA biogenesis and function described below. MicroRNA precursors are excised from primary transcripts by the Microprocessor complex containing Drosha and DGCR8. Although most microRNAs are located in introns of protein-coding and noncoding genes, the mechanisms coordinating microprocessing and splicing are unclear. MiR-211 is a microRNA expressed from intron 6 of melastatin, a suspected melanoma tumor suppressor. We demonstrate that miR-211, and not melastatin, is responsible for the tumor suppressive function of this locus, that Drosha-mediated processing of the miR-211 precursor promotes splicing of melastatin exon 6-exon 7 junctions, and that perturbing 5' splice site recognition by the U1 snRNP reduces Drosha recruitment to intron 6 specifically and intronic microRNA levels globally. Thus we identify a novel physical and functional coupling between microprocessing and splicing. Typically, Agos stabilize mature microRNAs and as a complex stoichiometrically bind to complementary mRNAs. We demonstrate an alternative order of events in which Agos bind and repress pre-formed imperfect microRNA-mRNA duplexes in processing bodies of live cells, and cleave pre-formed perfect microRNA-mRNA duplexes in vitro. Our data support a novel catalytic model whereby Agos first deposit microRNAs onto mRNAs and dissociate, thus priming multiple microRNA-mRNA duplexes for concurrent repression by a single Ago. Despite key roles in development and pathogenesis, effectors and regulators of microRNA-mediated repression are still poorly characterized. An RNAi screen revealed that depletion of ribosomal proteins of either small or large ribosomal subunit dissociates microRNA-containing complexes from mRNAs repressed at translation initiation, increasing their polysome association, translation, and stability relative to untargeted mRNAs. Thus ribosomal proteins globally regulate microRNA function. Another RNAi screen revealed that Akt3 phosphorylates Ago2, which negatively regulates cleavage and positively regulates translational repression of microRNA-targeted mRNAs. Thus Ago2 phosphorylation is a molecular switch between its mRNA cleavage and translational repression activities. The following pages will place these novel insights into biological and disease-relevant context, will describe what was known prior to these studies, and will provide perspectives for future studies.

Genome-wide association studies have implicated autophagy in Crohn’s Disease (CD) pathogenesis. The functional relevance of autophagy in CD remains unknown. I hypothesized that autophagy is involved in microRNA silencing, another process implicated in CD pathogenesis. MicroRNAs are short non-coding RNAs that are loaded onto RNA-induced silencing complex (RISC) and promote degradation and/or repress translation of target mRNAs. RISC formation and turnover occurs on endosomal membranes. Since autophagosomes and endosomes are closely related and RISC components are downstream effectors of microRNA silencing, I hypothesized that autophagy affects RISC, hence modulates microRNA expression. Using immunoblotting and immunofluorescence, I showed that Ago2, a critical component of RISC, is increased in cells with defective autophagy. Using microarray technology, I discovered 5 microRNAs that are differentially expressed in these cells. Taken together, my results propose a compelling mechanism by which autophagy regulates Ago2, thereby affects miRNA expression, which is implicated in the development of CD.

Genome-wide association studies have implicated autophagy in Crohn’s Disease (CD) pathogenesis. The functional relevance of autophagy in CD remains unknown. I hypothesized that autophagy is involved in microRNA silencing, another process implicated in CD pathogenesis. MicroRNAs are short non-coding RNAs that are loaded onto RNA-induced silencing complex (RISC) and promote degradation and/or repress translation of target mRNAs. RISC formation and turnover occurs on endosomal membranes. Since autophagosomes and endosomes are closely related and RISC components are downstream effectors of microRNA silencing, I hypothesized that autophagy affects RISC, hence modulates microRNA expression. Using immunoblotting and immunofluorescence, I showed that Ago2, a critical component of RISC, is increased in cells with defective autophagy. Using microarray technology, I discovered 5 microRNAs that are differentially expressed in these cells. Taken together, my results propose a compelling mechanism by which autophagy regulates Ago2, thereby affects miRNA expression, which is implicated in the development of CD.

Los viroides, los agentes infecciosos más simples de la escala biológica, están constituidos por una molécula circular de RNA monocatenario de aproximadamente 250-400 nucleótios (nt) que no codifica proteína alguna. A pesar de esta simplicidad estructural, los viroides son capaces de replicarse autónomamente, moverse sistémicamente y en muchos casos causar enfermedades en sus plantas huéspedes. Las infecciones producidas por viroides representativos generan la acumulación de pequeños RNAs viroidales (vd-sRNAs) de 21-24 nt con características similares a los pequeños RNA interferentes (siRNAs), la huella dactilar del silenciamiento mediado por RNA. La identificación de los vd-sRNAs implica que los viroides son diana de la primera barrera de silenciamiento mediado por RNA, formada por las RNasas ‘Dicer-like’ (DCLs). Para examinar si los vd-sRNAs se unen a las proteínas AGOs —el componente clave del complejo RISC (‘RNAinduced silencing complex’) que constituye la segunda barrera del silenciamiento mediado por RNA— hojas de Nicotiana benthamiana infectadas por el viroide del tubérculo fusiforme de la patata (PSTVd) se agroinfiltraron con nueve de las diez proteínas AGOs de Arabidopsis thaliana. Inmunoprecipitaciones a partir de los halos agroinfiltrados y análisis ‘Western-’ y ‘Northern-blot’ han mostrado que todas las AGOs se expresaron y, a excepción de AGO6, AGO7 y AGO10, unieron vd-sRNAs: AGO1, AGO2 y AGO3 los de 21 y 22 nt, mientras que AGO4, AGO5 y AGO9 también mostraron afinidad por los de 24 nt. La secuenciación masiva mostró que las AGO1, AGO2, AGO4 y AGO5 agroexpresadas unen los PSTVd-sRNAs en función de su tamaño y nucleótido 5’-terminal, y que los perfiles de los correspondientes vd-sRNAs cargados en las AGOs adoptan una distribución específica a lo largo del genoma viroidal. La agroexpresión de AGO1, AGO2, AGO4 y AGO5 en hojas de N. benthamiana infectadas con PSTVd atenuó la acumulación de los RNAs genómico viroidales, indicando que éstos, o sus precursores, también son diana de RISC. En contraste con los ribovirus, la infección de PSTVd en N. benthamiana no afectó de forma significativa la regulación mediada por miR168 de la AGO1 endógena, que carga vd-sRNAs con especificidad similar a su homóloga de A. thaliana. Mientras se conoce bien la biogénesis de los RNA viroidales, su degradación está restringida a algunos datos que implican al silenciamiento mediado por RNA. En el curso de nuestros estudios sobre el PSTVd, hemos observado consistentemente un patrón de 6-7 RNAs subgénomicos (sgRNAs) de polaridad (+) que aparecen junto con los RNAs monoméricos circulares y lineares en berenjena, un huésped experimental de este viroide. Hibridaciones ‘Northern-blot’ con sondas de tamaño parcial y completo, mostraron que los sgRNAs (+) de PSTVd derivan de diferentes regiones del RNA genómico y que algunos son parcialmente solapantes. Parte de los sgRNAs (+) de PSTVd se observaron también en N. benthamiana y tomate, donde han pasado desapercibidos a causa de su menor acumulación. El análisis por extensión de cebador de sgRNAs (+) de PSTVd representativos excluye que sean productos de terminaciones prematuras de la transcripción, pues carecen del extremo 5’ común que cabría esperar si ésta empezara en una posición específica. Ulteriores análisis mediante 5’- y 3’-RACE indican que los sgRNAs (+) de PSTVd tienen extremos 5’-OH y 3’-P, que probablemente resultan de cortes endonucleolíticos de precursores más largos catalizados por RNasas típicas que generan este tipo de extremos. Análisis de sgRNA (-) de PSTVd, que también se acumulan en berenjena infectada, mostraron que presentan características estructurales muy similares a los sgRNA (+). Nuestros resultados proporcionan una nueva visión de cómo ocurre la degradación in vivo de los RNAs viroidales, posiblemente durante su replicación, y sugieren que síntesis y degradación de las cadenas de PSTVd están conectadas, como se ha observado en los mRNAs. / Minoia, S. (2015). Degradación in vivo de un viroide de replicación nuclear: rutas catalizadas por proteínas Argonauta cargadas con pequeños RNAs viroidales y por otras ribonucleasas que generan RNAs subgenómicos [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/48553 / TESIS

No description available.

No description available.

Les microARN (miARN) sont de petits ARN non-codants qui contrôlent l’expression génique, en s’hybridant, le plus souvent, de manière imparfaite à des séquences spécifiques qui se trouvent généralement dans la région non traduite en 3' (3’UTR) de transcrits cibles. Les miARN guident sur l’ARN messager (ARNm) un complexe protéique appelé RNA-induced Silencing Complex (RISC), composé des protéines Argonaute et TNRC6, qui perturbe l’initiation de la traduction et provoque la déadénylation et la dégradation du transcrit. C'est l’interaction entre le RISC et le complexe de pré-initiation de la traduction 43S (composé de la petite sous-unité ribosomique 40S et des facteurs d’initiation associés) qui entraîne la répression traductionnelle de l’ARNm ciblé. Des résultats récents ont démontré que le RISC perturbe le balayage de la région non traduite en 5’ (5’UTR) par le ribosome, étape qui requiert la présence de 2 facteurs d’initiation qui sont eIF4F qui reconnaît et lie la coiffe ainsi que la protéine PABP, fixée le long de la queue poly(A). Toutefois, les miARN peuvent également induire la stimulation de la traduction des transcrits cibles dans les cellules quiescentes, dans un lysat d’embryons de drosophiles ou encore dans les ovocytes de Xénope. Le mécanisme moléculaire de stimulation de l’expression par les miARN est encore mal connu mais requiert l’absence de queue poly(A) en 3’ des ARN cible et de TNRC6 au sein du complexe RISC. Le Virus de l’Hépatite C (VHC) possède en 5’ de son ARNg un site d’entrée interne du ribosome (IRES) qui recrute la petite sous-unité ribosomique 40S, sans nécessiter la reconnaissance de la coiffe par eIF4F, ni la protéine PABP, ni le balayage de la 5’UTR par le ribosome. Ces caractéristiques singulières nous ont conduits à rechercher l'impact du complexe RISC fixé en 3’ de l’ARNm sur l’initiation de la traduction du VHC. Pour cela, nous avons utilisé des transcrits contenant l'IRES du VHC en 5' et des sites d’hybridation du miARN let-7 en 3’. Ces ARNm ont ensuite été transfectés dans des lignées cellulaires hépatocytaires, ou non. A notre grande surprise, nous avons observé que la fixation du miARN let-7 sur la région 3' du transcrit stimulait fortement l’expression dirigée par l’IRES de VHC. Toutefois, l’augmentation de l’expression n’est pas due à la stabilisation du transcrit mais bien à une hausse significative de la synthèse protéique indépendamment d’un quelconque effet de miR-122. En utilisant d’autres IRES dites 'HCV-like', nous avons pu confirmer ces résultats et démontrer que, l’ajout d’une queue poly(A) en 3’ du transcrit, capable de fixer la PABP, annule cet effet stimulateur suggérant que l’absence de cette protéine est nécessaire pour que le complexe RISC stimule la traduction du VHC. / MicroRNAs (miRNAs) are small non coding RNAs which control gene expression by recognizing and hybridizing to a specific sequence generally located in the 3’UTR of targeted messenger RNA (mRNA). miRNAs serve as a guide for the RNA-Induced Silencing Complex (RISC) that is composed by, at least, the Argonaute proteins and TNRC6. Recent studies have suggested that translation inhibition occurs first and is then followed by deadenylation and degradation of the targeted transcript. The miRNA-induced inhibition of protein synthesis occurs at the level of translation initiation during the ribosomal scanning step and it requires the presence of both the initiation factor eIF4G and the poly(A) Binding Protein (PABP). In this process, the RISC interacts with both PABP and 43S pre-initiation complex (composed by initiation factors and ribosome) and it results in the disruption of linear scanning of the ribosome along the 5’ Untranslated Region (5’UTR). In some specific cases, the binding of miRNAs to their target sequences can upregulate translation initiation. This has notably been demonstrated in G0 quiescent cells, drosophila embryos and Xenopus oocytes. Although the molecular mechanism by which upregulation occurs remains to be precisely determined, it appears that the absence of a poly(A) tail and the lack of availability of the TNRC6 proteins are amongst the major determinants. In the particular case of the Hepatitis C Virus (HCV), the genomic RNA is uncapped and non polyadenylated and harbors an Internal Ribosome Entry Site (IRES) which directly binds to the ribosome with no need for cap-recognition, PABP binding and ribosome scanning. These peculiar features of the HCV IRES prompted us to investigate how viral translation can be regulated by the miRNA machinery. In order to do that, we have used a mRNA that contains the HCV IRES in 5’ and 4 let-7 binding sites in its 3’ extremity. To most of our surprise, we have observed a strong stimulation of the expression of the HCV IRES when the construct is bearing the let-7 sites. This effect is not due to any interference with the miR-122 binding sites although the magnitude of stimulation reached the same level. Our data show that it is the presence of the RISC on the 3' end of the transcript that can stimulate internal ribosome entry at the 5' end. By using other HCV-like IRESes, we could confirm these data and further showed that the absence of a poly(A) tail was an absolute requirement for the stimulation to occur. These effects are not due to an increase of mRNA stability and are rather exerted at the level of translation.

• Argonaute proteins are important effectors in RNA silencing pathways, but they must interact with other machinery to trigger silencing. Ago hooks have emerged as a conserved motif responsible for interaction with Argonaute proteins, but little is know about the sequence surrounding Ago hooks that must restrict or enable interaction with specific Argonautes. • Here we investigated the evolutionary dynamics of an Argonaute-binding platform in NRPE1, the largest subunit of RNA Polymerase V. We compared NRPE1 sequences from more than 50 species, including dense sampling of two plant lineages. • This study demonstrates that the Argonaute-binding platform of NRPE1 retains Ago-hooks, intrinsic disorder, and repetitive character while being highly labile at the sequence level. We reveal that loss of sequence conservation is due to relaxed selection and frequent expansions and contractions of tandem repeat arrays. These factors allow a complete restructuring of the Ago-binding platform over 50-60 million years. This evolutionary pattern is also detected in a second Ago-binding platform, suggesting it is a general mechanism. • The presence of labile repeat arrays in all analyzed NRPE1 Ago-binding platforms indicates that selection maintains repetitive character, potentially to retain the ability to rapidly restructure the Ago-binding platform.

ARGONAUTE proteins (AGOs) are generally accepted as key components of the post transcriptional gene silencing mechanism, also involved in plant antiviral defense. Except for reports on the antiviral roles of AGO1, AGO2 and AGO7 in Arabidopsis, the exact roles played by the individual AGOs in other plant species are largely unknown. This research focused on the identification and characterization of AGOs involved in antiviral RNAi response to various viruses in N. benthamiana. Based on the temporal and spatial distribution of AGO transcripts in 3 and 8-week old plant root, stem and leaf tissues, expressions of NbAGO mRNAs were found to vary with age and tissue specificity. Plant endogenous AGO mRNAs were knocked down through virus induced gene silencing techniques using the Tobacco rattle virus vector system and posteriorly challenged with a GFP-chimeric virus construct deficient of a silencing suppressor. Unlike in control non-silenced plants, the Tomato bushy stunt virus construct deficient of its P19 silencing suppressor was consistently seen to exhibit a strong fluorescence on N. benthamiana plants silenced for NbAGOs 2 and X. Similar results were also obtained upon silencing of NbAGO2 using hairpin vector techniques. Comparable observations were also made when Tobacco mosaic virus GFP constructs were agroinfiltrated on NbAGO2 silenced plants further hinting the antiviral defense roles played by these AGOs. Agroinfiltration of Foxtailmosaic virus, Sunnhemp mosaic virus, and Turnip crinkle virus GFP chimeric constructs on NbAGO2 silenced N. benthamiana plants, however did not result in accumulation of GFP indicating the AGO antiviral defense specificity to TBSV and TMV. The results also hinted at a role for AGO7. Collectively my findings suggest that the expression of AGOs in N. benthamiana is tissue and age dependent, and that unlike in the model plant Arabidopsis where the main antiviral AGO is thought to be AtAGO1; in N. benthamiana, NbAGOs 2 and X seem to be involved in an antiviral defense role against TBSV and TMV with other AGOs perhaps contributing.