La protéine de liaison à l’ARN double-brin Staufen1 (STAU1) est exprimée dans les cellules de mammifères de manière ubiquitaire. STAU1 est impliqué dans la régulation post-transcriptionnelle de l’expression génique grâce à sa capacité de lier les ARN et moduler leur épissage, leur transport et localisation, leur traduction ainsi que leur dégradation. Des études récentes de notre laboratoire indiquent que l’expression de STAU1 est régulée durant le cycle cellulaire, ayant une abondance maximale au début de la mitose. En prométaphase, STAU1 est lié à des ARNm codant pour des facteurs impliqués dans la régulation de la prolifération, la croissance et la différenciation cellulaires. De plus, des analyses protéomiques menées sur des cellules humaines ont permis d’identifier STAU1 comme un composant de l’appareil mitotique. Cependant, l’importance de cette association n’a pas été investiguée. Par ailleurs, il a été montré qu’une défaillance dans l’expression ou les fonctions de STAU1 pourrait contribuer au développement et l’accélération de plusieurs maladies débilitantes, dont le cancer. Dans cette thèse, nous avons montré la localisation de STAU155 sur le fuseau mitotique dans les cellules de cancer colorectal HCT116 et les cellules non transformées hTERT-RPE1. Nous avons également caractérisé le déterminant moléculaire impliqué dans l’interaction entre STAU155 et les microtubules mitotiques, soit la séquence située dans les 88 premiers acides aminés N-terminaux de RBD2, un domaine qui n’est pas requis pour l’activité de liaison à l’ARN de STAU1. Nous avons ainsi montré que la fraction de STAU1 enrichie sur le fuseau colocalise avec des ribosomes dans des sites actifs de traduction. De plus, notre analyse transcriptomique du fuseau mitotique montre que 1054 transcrits (ARNm, pré-ARNr, lncRNA et snoRNA) sont enrichis sur l’appareil mitotique. De façon intéressante, le knockout de STAU1 entraine la délocalisation des pré-ARNr et de 154 ARNm codants pour des protéines impliquées dans l’organisation du cytosquelette d'actine et la croissance 4 cellulaire. Bien que STAU1 n’est pas essentiel pour la survie et la prolifération des cellules cancéreuses HCT116, nos résultats mettent clairement en évidence l’implication de STAU1 dans la régulation des ARN spécifiques en mitose et suggèrent un nouveau rôle de cette protéine dans la progression mitotique et la cytokinèse par la régulation de la maintenance des pré-ARNr, la ribogenèse et/ou la reconstitution de l’enveloppe nucléaire. / Staufen1 (STAU1) is a double-stranded RNA-binding protein that is ubiquitously expressed in mammals and known for its involvement in the post-transcriptional regulation of gene expression such as splicing, transport and localization, translation, and decay. It has been demonstrated that STAU1 protein expression level is modulated through the cell cycle with peak abundance by the onset of the mitotic phase after which it is degraded. Genome-wide analysis revealed that in prometaphase, STAU1 bound with mRNAs code for factors implicated in cell differentiation, cell growth as well as for cell proliferation. Interestingly, previous large-scale proteomic studies identified STAU1 as a component of the human mitotic spindle apparatus. Altering STAU1 expression patterns or functions may lead to several debilitating human diseases including cancer. In this thesis, we further elucidated the localization of STAU1 at the mitotic spindle of the colorectal cancer HCT116 and the non-transformed hTERT-RPE1 cells. Next, we characterized the molecular determinant required for STAU1/spindle association within the first 88 N-terminal amino acids, a domain that is not required for the RNA binding activity. RNA-Seq analysis of purified mitotic spindles reveals that 1054 mRNAs as well as the precursor ribosomal RNA, lncRNAs and snoRNAs are enriched on spindles compared to cell extracts. Spindle-associated STAU1 partly co-localizes with ribosomes and active sites of translation. Interestingly, the knockout of STAU1 delocalizes pre-rRNA and 154 mRNAs coding for proteins involved in actin cytoskeleton organization and cell growth. Our results highlighting a role for STAU1 in mRNA trafficking to the spindle. These data demonstrate that STAU1 controls the localization of sub-populations of RNA during cell division and suggests a novel role of STAU1 protein in mitotic progression and cytokinesis by regulating pre-rRNA maintenance, ribogenesis and/or nucleoli reassembly.

No description available.

No description available.

Les résultats ont été obtenus avec le logiciel "Insight-2" de Accelris (San Diego, CA) / En 2000, les structures à hautes résolutions des deux sous-unités ribosomiques ont finalement été mises à la disposition du public. L'année suivante, la structure aux rayons X de l'ensemble du ribosome bactérien a été publiée. Ces grandes réalisations ont ouvert une nouvelle ère dans l'étude des mécanismes de la synthèse des protéines. Dès lors, il est devenu possible de relier différents aspects de la fonction du ribosome à des éléments particuliers de sa structure tertiaire. L'établissement de la relation structure-fonction peut toutefois être problématique en raison de l'immense complexité de la structure du ribosome. En d'autres termes, pour que les données cristallographiques sur la structure tertiaire du ribosome soient vraiment utiles à la compréhension du fonctionnement du ribosome, ces données devraient elles-mêmes faire l'objet d'une analyse approfondie. Le travail, présenté ici, peut être vu comme une tentative de ce genre. En appliquant l’analyse systématique des structure cristallographiques du ribosome disponibles, nous avons essayé de résoudre deux problèmes fondamentaux de la biologie ribosomale concernant (1) la nature des réarrangements du ribosome qui ont lieu à différentes étapes de son cycle de fonctionnement et (2) la possibilité de reconstitution de l'évolution du ribosome du monde-à-ARN jusqu’à nos jours. Dans le premier projet, nous avons systématiquement comparé les structures du ribosome disponibles et de sa sous-unité afin d'identifier les domaines rigides, qui ont toujours la même conformation, et les régions flexibles dont la conformation peut varier d'une structure de ribosome à une autre. Il y a deux types de réarrangements structuraux connus dont nous voulions comprendre les mécanismes: le « ratchet-like movement » et la «fermeture de domaines ». Le premier a lieu au cours de la translocation du ribosome et est plus ou moins perçu comme une rotation d'une sous-unité par rapport à l'autre. Le deuxième se produit dans la petite sous-unité et est associé à la reconnaissance codon-anticodon au site A. La comparaison des conformations ribosomales disponibles a révélé les mécanismes spécifiques des deux réarrangements. Bien que la sélection de l'aminoacyl-ARNt appropriée au site A et la translocation du ribosome n'ont jamais été considérés comme ayant quelque chose en commun, nous démontrons ici que les réarrangements de la structure des ribosomes associés au premier processus répète les réarrangements associés au deuxième mais dans l’ordre inverse. En d'autres termes, pendant le cycle d'élongation, la fermeture de domaine et le « ratchet » peuvent ii être considérés comme un mouvement de va-et-vient, qui renvoie finalement le ribosome à sa conformation initiale. Dans le second projet, nous avons fait une tentative de reconstitution de l'évolution de l'ARNr 23S, du monde-à-ARN jusqu`à nos jours. Ici nous nous sommes basés sur la supposition que l'évolution de cette molécule a procédé par des insertions aléatoires des régions relativement courtes dans différentes parties de la chaîne poly-nucléotidique. Pour cela, nous avons élaboré des critères de l'intégrité de la structure ribosomale et présumé que lors de l'évolution, la structure du ribosome s’est toujours adaptée à ces standards. Nous avons examiné l'interaction de type A-mineur, un arrangement fréquent dans la structure de l’ARN ribosomique, constitué d'un empilement d’adénosines non-appariées, attachées à une double hélice. Nous avons supposé que dans toutes les interactions A-mineurs existantes dans le ribosome, la double hélice est apparue avant ou au moins simultanément avec la pile d’adénosines correspondantes. L'application systématique de ce principe à la structure tertiaire de l’ARN 23S a permis d'élucider de manière progressive l'ordre dans lequel les parties différentes de l’ARN 23S ont rejoint la structure. Pris ensemble, les deux projets démontrent l'efficacité de l'analyse systématique in-silico de la structure tertiaire du ribosome et ouvrent la voie à de futures découvertes. / In the year 2000, the first high-resolution structures of the individual ribosomal subunits became available to the public. The following year, the X-ray structure of the complete bacterial ribosome was published. These major achievements opened a new era in studying the mechanisms of protein synthesis. From then on, it became possible to attribute different aspects of the ribosome function to particular elements of its tertiary structure. However, establishing the structure-function relationships is problematic due to the immense complexity of the ribosome structure. In other words, in order to make the crystallographic data on the ribosome tertiary structure really useful for understanding of how the ribosome functions, it must be thoroughly analyzed. Here, based on systematic analysis of the available X-ray conformations of the ribosome we have tried to resolve two fundamental problems of the ribosome biology: concerning (1) the nature of rearrangements in the ribosome that take place at different steps of its functional cycle, and (2) the reconstruction of the ribosome evolution from the RNA world to present time. In the first project, we systematically compared the available structures of the ribosome and its subunits to identify rigid domains, which always have the same conformation, and flexible regions, where the conformation can vary from one ribosome structure to another. There were two known types of structural rearrangements whose mechanisms we wanted to understand: the ratchet-like motion and the so-called domain closure. The ratchet-like motion takes place during the ribosomal translocation and is roughly seen as a rotation of one subunit with respect to the other. The domain closure occurs in the small subunit and is associated with the cognate codon-anticodon recognition in the A-site. Comparison of the available ribosome conformations revealed the detailed mechanisms of both rearrangements. Although the selection of the cognate amino-acyl-tRNA in the A-site and of the ribosomal translocation have never been thought to have anything in common, we demonstrate that the rearrangements in the ribosome structure associated with the first process repeat in reverse order the rearrangements associated with the second process. In other words, during the ribosome elongation cycle, the domain closure and the ratchet-like motion can be seen as a back-and-forth movement, which eventually returns the ribosome to the initial conformation. iv In the second project, we attempted to reconstruct the evolution of the 23S rRNA from the RNA world to present time based on the presumption that the evolutionary expansion of this molecule proceeded though random insertions of relatively short regions into different regions of the polynucleotide chain. We developed criteria for integrity of the ribosome structure and presumed that during the evolutionary expansion, the ribosome structure always matched to these standards. For this, we specifically considered the A-minor interaction, a frequent arrangement in the rRNA structure consisting of a stack of unpaired adenosines tightly attached to a double helix. We presumed that in all A-minor interactions present in the ribosome, the double helix emerged before or at least simultaneously with the corresponding adenosine stack. The systematic application of this principle to the known tertiary structure of the 23S rRNA allowed us to elucidate in a step-vise manner the order in which different part of the modern 23S rRNA joined the structure. Taken together, the two projects demonstrate the effectiveness of the systematic in-silico analysis of the ribosome tertiary structure and pave the way for future discoveries.

Les interactions entre les squelettes sucre-phosphate de nucléotides jouent un rôle important dans la stabilisation des structures tertiaires de larges molécules d’ARN. Elles sont régies par des règles particulières qui gouverne leur formation mais qui jusque là demeure quasiment inconnues. Un élément structural d’ARN pour lequel les interactions sucre-phosphate sont importantes est le motif d’empaquetage de deux doubles hélices d’ARN le long du sillon mineur. Ce motif se trouve à divers endroits dans la structure du ribosome. Il consiste en deux doubles hélices interagissant de manière à ce que le squelette sucre-phosphate de l’une se niche dans le sillon mineur de l’autre et vice versa. La surface de contact entre les deux hélices est majoritairement formée par les riboses et implique au total douze nucléotides. La présente thèse a pour but d’analyser la structure interne de ce motif et sa dépendance de stabilité résultant de l’association optimale ou non des hélices, selon leurs séquences nucléotidiques. Il est démontré dans cette thèse qu’un positionnement approprié des riboses leur permet de former des contacts inter-hélices, par l’entremise d’un choix particulier de l’identité des pairs de bases impliquées. Pour différentes pairs de bases participant à ce contact inter-hélices, l’identité optimale peut être du type Watson-Crick, GC/CG, or certaines pairs de bases non Watson-Crick. Le choix adéquat de paires de bases fournit une interaction inter-hélice stable. Dans quelques cas du motif, l’identité de certaines paires de bases ne correspond pas à la structure la plus stable, ce qui pourrait refléter le fait que ces motifs devraient avoir une liberté de formation et de déformation lors du fonctionnement du ribosome. / Although backbone-backbone interactions play an important role in stabilization of the tertiary structure of large RNA molecules, the particular rules that govern the formation of these interactions remain basically unknown. One RNA structural element for which the backbone-backbone interactions are essential is the along-groove packing motif. This motif is found in numerous locations in the ribosome structure; it consists of two double helices arranged such that the backbone of one helix is packed in the minor groove of the other helix and vice versa. The contact area between the two helices is mostly formed by riboses and totally involves twelve nucleotides. Here we analyze the internal structure of the along-groove packing motif and the dependence of stability of the association of the helices on their nucleotide sequences. We show that the proper positioning of the riboses that allows them to form inter-helix contacts is achieved through the particular choice of the identities of the base pairs involved. For different base pairs participating in the inter-helix contacts the optimal identities can be Watson-Crick, GC/CG, or certain non-Watson-Crick base pairs. The proper choice of the base pairs provides for the stable inter-helix interaction. In some cases of the motif, the identities of certain base pairs do not correspond to the most stable structure, which may reflect the fact that these motifs should break and form during the ribosome function.

Com o advento dos estudos moleculares evolutivos baseados nas sequências dos genes de RNAr 16S em cianobactérias, a taxonomia de Anabaena (Cyanobacteria) tem sido amplamente discutida e uma revisão deste gênero faz-se necessária. Os problemas variam desde o nível genérico, tal como o grupo Anabaena Aphanizomenon, até níveis de diferenciação de linhagens (morfoespécies). Estudos moleculares de linhagens de Anabaena isoladas de ecossistemas brasileiros são inexistentes. A fim de se explorar a diversidade, filogenia e diversificação evolutiva de Anabaena isoladas de ambientes brasileiros, estudos fenotípicos e genotípicos foram realizados no presente estudo. Um total de 43 isolados foram obtidos de corpos d´água do Estado de São Paulo (Reservatórios Billings, Santo Grande, Rio Piracicaba e Lago da ESALQ/USP (Engenharia) e do Estado do Ceará (Lagoa do Povoado Nova Aurora e Rio Camarão). As espécies de Anabaena isoladas foram identificadas como A. aphanizomenoides Forti, A. circinalis Rabenhorst ex Bornet et Flahault, A. crassa (Lemmermann) Komárková-Legnerová et Cronberg, A. cf. fallax Komárek et Komárková-Legnerová e A. planctonica Brunnthaler. Os meios de cultura usados no isolamento das cianobactérias foram AA/4, ASM-1 e BGN, este último também nas variantes com 50% de NaNO3 (BG50) e sem nitrato (BGS), com ou sem adição de vitamina B12. O meio ASM-1 não promoveu o crescimento de Anabaena. As espécies A. circinalis e A. crassa não cresceram no meio de culturaBGN com 17,65 mM de NaNO3, apresentando crescimento no meio BG50. A adição de vitamina B12 favoreceu o crescimento de A. circinalis. Os caracteres morfológicos analisados para 23 isolados foram o comprimento (compr) e diâmetro (diam) da célula vegetativa (V), heterócito (HT), acineto (AC), espira (ES), razões (R) comprimento/diâmetro para V, HT e AC e razão diâmetro/comprimento para ES (RES). O diâmetro da bainha (diamBA) foi também avaliado. As análises de componentes principais (ACP) confirmaram que a espira é uma característica importante para separar as morfoespécies A. circinalis, A.crassa e A. cf. fallax. RV e RHT foram diacríticos para diferenciação de A. cf. fallax. O comprimento do acineto (comprAC) foi importante para diferenciar A. aphanizomenoides de A. planctonica. A bainha diferenciou A. crassa das outras morfoespécies. As análises filogenéticas para o RNAr 16S mostraram os isolados brasileiros de A. circinalis, A. crassa, A. planctonica, A. aphanizomenoides e A. cf. fallax em agrupamentos separados, confirmando os resultados dos caracteres morfológicos. Com exceção da A. planctonica, as sequências de RNAr 16S dos isolados brasileiros não agruparam com linhagens relativas provenientes de outros países, indicando que elas são únicas. As análises filogenéticas dos genes RNAr 16S, rpoC1, rbcL, tufA concatenados corroboraram os resultados de filogenia do gene de RNAr 16S e as identificações morfológicas. A tentativa de obtenção de padrões moleculares para as espécies de Anabaena isoladas do Brasil foi feita utilizando a técnica de PCR-DGGE. Os fragmentos da região 359-781 do RNAr 16S apresentaram banda única para A. circinalis, enquanto que mais de uma banda foi verificado nas outras morfoespécies de Anabaena. / The advent of molecular evolutionary studies based on 16S rRNA genes sequences of cyanobacteria, the taxonomy of Anabaena (Cyanobacteria) has been widely discussed and a revision of the genus is required. The problems range from the generic level, such as Anabaena Aphanizomenon group, to levels of strains differentiation (morphospecies). Molecular studies of Anabaena strains isolated from Brazilian ecosystems are lacking. In order to explore the diversity, phylogeny and evolutionary diversification of Anabaena strains isolated from Brazilian environments, phenotypic and genotypic studies were performed. A total of 43 Anabaena isolates were obtained from water bodies of Sao Paulo State (Billings reservoir, Santo Grande reservoir, Piracicaba river and Engenharia pond Esalq) and Ceara State (Povoado Nova Aurora pond and Camarao river). The isolated Anabaena species were identified as A. aphanizomenoides Forti, A. circinalis Rabenhorst ex Bornet et Flahault, A. crassa (Lemmermann) Komárková-Legnerová et Cronberg, A. cf. fallax Lomárek et Komárková-Legnerová and A. planctonica Brunnthaler. The culture media used for cyanobacterial isolation were AA/4, ASM-1 andBGN, the latter also in the variants with NaNO3 50% (BG50) and without nitrate (BGS), with and without addition of B12 vitamin. The ASM-1 medium did not promote the growth of Anabaena. The A. circinalis and A. crassa species had not grown inBGN culture medium with NaNO3 17.65 mM, showing growth in BG50 medium. The addition of B12 vitamin favored the growth of A. circinalis. The morphological characters analyzed for 23 isolates were the length (L) and diameter (D) of vegetative cell (V), heterocyte (HT), akinete (AK), coil (CO), L/D ratio (R) for V, HT and AK, and D/L ratio for CO (RCO). The sheath diameter (SD) was also evaluated. The principal component analysis (PCA) confirmed that the coil is an important feature to separate the morphospecies A. circinalis, A.crassa and A. cf. fallax. RV and RHT were diacritical for differentiation of A. cf. fallax. The akinete length (LAK) was important to differentiate A. aphanizomenoides from A. planctonica. The sheath differentiated A. crassa from other morphospecies. Phylogenetic analyses of 16S rRNA gene placed A. circinalis and A. crassa, A. planctonica, A. aphanizomenoides and A. cf. fallax Brazilian isolates in separated clades in agreement with morphological characters. With the exception of A. planctonica, the 16S rRNA sequences of the Brazilian isolates did not cluster with relative strains originated from other countries, indicating that they are unique. The phylogenetic analysis of concatenated genes16S rRNA, rpoC1, rbcL, tufA corroborated results of 16S rRNA phylogeny and morphological identifications. The attempt to obtain molecular standards for the Anabaena species isolated from Brazil was made using the PCR-DGGE technique. The fragments of the 359-781 region of 16S rRNA showed only one DNA band for A. circinalis, while more than one band was observed in other Anabaena morphospecies.

Um grande número de proteínas está envolvido no processamento de rRNA, e cada uma delas desempenha uma ação específica, seja como um fator estrutural, regulatório ou catalítico. Apesar da biogênese dos ribossomos ter sido intensamente estudada, ainda não se tem conhecimento claro da função de muitas proteínas envolvidas neste mecanismo. Dentre os snoRNPs, o grupo denominado box C/D é responsável pela metilação e clivagens no pré-rRNA. Através da análise de interação proteína-proteína do sistema do duplo híbrido a proteína aqui denominada Nop17p foi isolada interagindo com a proteína Rrp43p, uma subunidade do exossomo. Estudos de microarray mostraram que o mutante nulo Δnop17 tem o mesmo fenótipo que mutantes de genes envolvidos em tradução. No presente trabalho apresentamos uma análise detalhada da função da Nop17p e a importância da sua interação com a proteína Nop58p, componente do snoRNP de box C/D. Observamos também um defeito na função da Nop58p na ausência da Nop17p e outro dado importante apresentado aqui é que a localização sub-celular de componentes de snoRNPs de box C/D não é correta na ausência de Nop17p. Estes resultados evidenciam um envolvimento direto entre Nop17p e snoRNPs de box C/D, com um papel de regulação da função e/ou na montagem desses complexos. Apresentamos também dados com a proteína homóloga de humanos (hNop17p), que foi expressada na cepa Δnop17 de levedura e que conseguiu suprimir parcialmente o fenótipo termo-sensível dessa cepa, demonstrando uma possível conservação da função de Nop17p ao longo da evolução. / In eukaryotes, pre-rRNA processing depends on cis-acting elements and on a large number of non-ribosomal trans-acting factors, including endonucleases and exonucleases, RNA helicases, rRNA modifying enzymes and components of snoRNPs. The exosome is a conserved eukaryotic protein complex containing multiple 3\'-5\' exonucleases, which has been implicated in pre-rRNA, snoRNA and snRNA processing, as well as in mRNA degradation. In order to identify new proteins involved in rRNA processing, we have screened a yeast two-hybrid cONA library, to isolate proteins interacting with the exosome subunit Rrp43p. In this screen, a novel nucleolar protein, Nop17p, was identified which also interacts with the box C/D snoRNP protein Nop58p. The NOP17 gene is not essential for cell viability but its deletion causes a temperature-sensitive phenotype. Pre-rRNA processing analyses revealed that rRNA formation is affected in the Δnop17 strain subjected to the non-permissive temperature, although it is not blocked completely. In addition, primer extension analyses of RNA isolated from Nop17p-depleted cells subjected to the non-permissive temperature indicates that the pre-rRNA is undergoing different modification or degradation processes in these cells as compared to the parental strain. Nop17p was recently described in the same complex as Nop58p and, interestingly, its depletion leads to mislocalization of Nop1p, Nop56p, Nop58p and Snu13p, which are the core proteins of the box C/D ribonucleoprotein (snoRNP), indicating that Nop17p function is required either for nucleolar retention or for the proper assembly ofthe box C/D snoRNP.

Com o advento dos estudos moleculares evolutivos baseados nas sequências dos genes de RNAr 16S em cianobactérias, a taxonomia de Anabaena (Cyanobacteria) tem sido amplamente discutida e uma revisão deste gênero faz-se necessária. Os problemas variam desde o nível genérico, tal como o grupo Anabaena Aphanizomenon, até níveis de diferenciação de linhagens (morfoespécies). Estudos moleculares de linhagens de Anabaena isoladas de ecossistemas brasileiros são inexistentes. A fim de se explorar a diversidade, filogenia e diversificação evolutiva de Anabaena isoladas de ambientes brasileiros, estudos fenotípicos e genotípicos foram realizados no presente estudo. Um total de 43 isolados foram obtidos de corpos d´água do Estado de São Paulo (Reservatórios Billings, Santo Grande, Rio Piracicaba e Lago da ESALQ/USP (Engenharia) e do Estado do Ceará (Lagoa do Povoado Nova Aurora e Rio Camarão). As espécies de Anabaena isoladas foram identificadas como A. aphanizomenoides Forti, A. circinalis Rabenhorst ex Bornet et Flahault, A. crassa (Lemmermann) Komárková-Legnerová et Cronberg, A. cf. fallax Komárek et Komárková-Legnerová e A. planctonica Brunnthaler. Os meios de cultura usados no isolamento das cianobactérias foram AA/4, ASM-1 e BGN, este último também nas variantes com 50% de NaNO3 (BG50) e sem nitrato (BGS), com ou sem adição de vitamina B12. O meio ASM-1 não promoveu o crescimento de Anabaena. As espécies A. circinalis e A. crassa não cresceram no meio de culturaBGN com 17,65 mM de NaNO3, apresentando crescimento no meio BG50. A adição de vitamina B12 favoreceu o crescimento de A. circinalis. Os caracteres morfológicos analisados para 23 isolados foram o comprimento (compr) e diâmetro (diam) da célula vegetativa (V), heterócito (HT), acineto (AC), espira (ES), razões (R) comprimento/diâmetro para V, HT e AC e razão diâmetro/comprimento para ES (RES). O diâmetro da bainha (diamBA) foi também avaliado. As análises de componentes principais (ACP) confirmaram que a espira é uma característica importante para separar as morfoespécies A. circinalis, A.crassa e A. cf. fallax. RV e RHT foram diacríticos para diferenciação de A. cf. fallax. O comprimento do acineto (comprAC) foi importante para diferenciar A. aphanizomenoides de A. planctonica. A bainha diferenciou A. crassa das outras morfoespécies. As análises filogenéticas para o RNAr 16S mostraram os isolados brasileiros de A. circinalis, A. crassa, A. planctonica, A. aphanizomenoides e A. cf. fallax em agrupamentos separados, confirmando os resultados dos caracteres morfológicos. Com exceção da A. planctonica, as sequências de RNAr 16S dos isolados brasileiros não agruparam com linhagens relativas provenientes de outros países, indicando que elas são únicas. As análises filogenéticas dos genes RNAr 16S, rpoC1, rbcL, tufA concatenados corroboraram os resultados de filogenia do gene de RNAr 16S e as identificações morfológicas. A tentativa de obtenção de padrões moleculares para as espécies de Anabaena isoladas do Brasil foi feita utilizando a técnica de PCR-DGGE. Os fragmentos da região 359-781 do RNAr 16S apresentaram banda única para A. circinalis, enquanto que mais de uma banda foi verificado nas outras morfoespécies de Anabaena. / The advent of molecular evolutionary studies based on 16S rRNA genes sequences of cyanobacteria, the taxonomy of Anabaena (Cyanobacteria) has been widely discussed and a revision of the genus is required. The problems range from the generic level, such as Anabaena Aphanizomenon group, to levels of strains differentiation (morphospecies). Molecular studies of Anabaena strains isolated from Brazilian ecosystems are lacking. In order to explore the diversity, phylogeny and evolutionary diversification of Anabaena strains isolated from Brazilian environments, phenotypic and genotypic studies were performed. A total of 43 Anabaena isolates were obtained from water bodies of Sao Paulo State (Billings reservoir, Santo Grande reservoir, Piracicaba river and Engenharia pond Esalq) and Ceara State (Povoado Nova Aurora pond and Camarao river). The isolated Anabaena species were identified as A. aphanizomenoides Forti, A. circinalis Rabenhorst ex Bornet et Flahault, A. crassa (Lemmermann) Komárková-Legnerová et Cronberg, A. cf. fallax Lomárek et Komárková-Legnerová and A. planctonica Brunnthaler. The culture media used for cyanobacterial isolation were AA/4, ASM-1 andBGN, the latter also in the variants with NaNO3 50% (BG50) and without nitrate (BGS), with and without addition of B12 vitamin. The ASM-1 medium did not promote the growth of Anabaena. The A. circinalis and A. crassa species had not grown inBGN culture medium with NaNO3 17.65 mM, showing growth in BG50 medium. The addition of B12 vitamin favored the growth of A. circinalis. The morphological characters analyzed for 23 isolates were the length (L) and diameter (D) of vegetative cell (V), heterocyte (HT), akinete (AK), coil (CO), L/D ratio (R) for V, HT and AK, and D/L ratio for CO (RCO). The sheath diameter (SD) was also evaluated. The principal component analysis (PCA) confirmed that the coil is an important feature to separate the morphospecies A. circinalis, A.crassa and A. cf. fallax. RV and RHT were diacritical for differentiation of A. cf. fallax. The akinete length (LAK) was important to differentiate A. aphanizomenoides from A. planctonica. The sheath differentiated A. crassa from other morphospecies. Phylogenetic analyses of 16S rRNA gene placed A. circinalis and A. crassa, A. planctonica, A. aphanizomenoides and A. cf. fallax Brazilian isolates in separated clades in agreement with morphological characters. With the exception of A. planctonica, the 16S rRNA sequences of the Brazilian isolates did not cluster with relative strains originated from other countries, indicating that they are unique. The phylogenetic analysis of concatenated genes16S rRNA, rpoC1, rbcL, tufA corroborated results of 16S rRNA phylogeny and morphological identifications. The attempt to obtain molecular standards for the Anabaena species isolated from Brazil was made using the PCR-DGGE technique. The fragments of the 359-781 region of 16S rRNA showed only one DNA band for A. circinalis, while more than one band was observed in other Anabaena morphospecies.

Escherichia coli ribosomal RNA (rRNA) is post-transcriptionally modified by site-specific enzymes. The role of most modifications is not known and little is known about how these enzymes recognize their target substrates. In this thesis, we have structurally and functionally characterized two S-adenosyl-methionine (SAM) dependent 23S rRNA methyltransferases (MTases) that act during the early stages of ribosome assembly in E. coli. RlmM methylates the 2'O-ribose of C2498 in 23S rRNA. We have solved crystal structures of apo RlmM at 1.9Å resolution and of an RlmM-SAM complex at 2.6Å resolution. The RlmM structure revealed an N-terminal THUMP domain and a C-terminal catalytic Rossmann-fold MTase domain. A continuous patch of conserved positive charge on the RlmM surface is likely used for RNA substrate recognition. The SAM-binding site is open and shallow, suggesting that the RNA substrate may be required for tight cofactor binding. Further, we have shown RlmM MTase activity on in vitro transcribed 23S rRNA and its domain V. RlmJ methylates the exocyclic N6 atom of A2030 in 23S rRNA. The 1.85Å crystal structure of RlmJ revealed a Rossmann-fold MTase domain with an inserted small subdomain unique to the RlmJ family. The 1.95Å structure of the RlmJ-SAH-AMP complex revealed that ligand binding induces structural rearrangements in the four loop regions surrounding the active site. The active site of RlmJ is similar to N6-adenine DNA MTases. We have shown RlmJ MTase activity on in vitro transcribed 23S rRNA and a minimal substrate corresponding to helix 72, specific for adenosine. Mutagenesis experiments show that residues Y4, H6, K18 and D164 are critical for catalytic activity. These findings have furthered our understanding of the structure, evolution, substrate recognition and mechanism of rRNA MTases.

En 2000, les structures à hautes résolutions des deux sous-unités ribosomiques ont finalement été mises à la disposition du public. L'année suivante, la structure aux rayons X de l'ensemble du ribosome bactérien a été publiée. Ces grandes réalisations ont ouvert une nouvelle ère dans l'étude des mécanismes de la synthèse des protéines. Dès lors, il est devenu possible de relier différents aspects de la fonction du ribosome à des éléments particuliers de sa structure tertiaire. L'établissement de la relation structure-fonction peut toutefois être problématique en raison de l'immense complexité de la structure du ribosome. En d'autres termes, pour que les données cristallographiques sur la structure tertiaire du ribosome soient vraiment utiles à la compréhension du fonctionnement du ribosome, ces données devraient elles-mêmes faire l'objet d'une analyse approfondie. Le travail, présenté ici, peut être vu comme une tentative de ce genre. En appliquant l’analyse systématique des structure cristallographiques du ribosome disponibles, nous avons essayé de résoudre deux problèmes fondamentaux de la biologie ribosomale concernant (1) la nature des réarrangements du ribosome qui ont lieu à différentes étapes de son cycle de fonctionnement et (2) la possibilité de reconstitution de l'évolution du ribosome du monde-à-ARN jusqu’à nos jours. Dans le premier projet, nous avons systématiquement comparé les structures du ribosome disponibles et de sa sous-unité afin d'identifier les domaines rigides, qui ont toujours la même conformation, et les régions flexibles dont la conformation peut varier d'une structure de ribosome à une autre. Il y a deux types de réarrangements structuraux connus dont nous voulions comprendre les mécanismes: le « ratchet-like movement » et la «fermeture de domaines ». Le premier a lieu au cours de la translocation du ribosome et est plus ou moins perçu comme une rotation d'une sous-unité par rapport à l'autre. Le deuxième se produit dans la petite sous-unité et est associé à la reconnaissance codon-anticodon au site A. La comparaison des conformations ribosomales disponibles a révélé les mécanismes spécifiques des deux réarrangements. Bien que la sélection de l'aminoacyl-ARNt appropriée au site A et la translocation du ribosome n'ont jamais été considérés comme ayant quelque chose en commun, nous démontrons ici que les réarrangements de la structure des ribosomes associés au premier processus répète les réarrangements associés au deuxième mais dans l’ordre inverse. En d'autres termes, pendant le cycle d'élongation, la fermeture de domaine et le « ratchet » peuvent ii être considérés comme un mouvement de va-et-vient, qui renvoie finalement le ribosome à sa conformation initiale. Dans le second projet, nous avons fait une tentative de reconstitution de l'évolution de l'ARNr 23S, du monde-à-ARN jusqu`à nos jours. Ici nous nous sommes basés sur la supposition que l'évolution de cette molécule a procédé par des insertions aléatoires des régions relativement courtes dans différentes parties de la chaîne poly-nucléotidique. Pour cela, nous avons élaboré des critères de l'intégrité de la structure ribosomale et présumé que lors de l'évolution, la structure du ribosome s’est toujours adaptée à ces standards. Nous avons examiné l'interaction de type A-mineur, un arrangement fréquent dans la structure de l’ARN ribosomique, constitué d'un empilement d’adénosines non-appariées, attachées à une double hélice. Nous avons supposé que dans toutes les interactions A-mineurs existantes dans le ribosome, la double hélice est apparue avant ou au moins simultanément avec la pile d’adénosines correspondantes. L'application systématique de ce principe à la structure tertiaire de l’ARN 23S a permis d'élucider de manière progressive l'ordre dans lequel les parties différentes de l’ARN 23S ont rejoint la structure. Pris ensemble, les deux projets démontrent l'efficacité de l'analyse systématique in-silico de la structure tertiaire du ribosome et ouvrent la voie à de futures découvertes. / In the year 2000, the first high-resolution structures of the individual ribosomal subunits became available to the public. The following year, the X-ray structure of the complete bacterial ribosome was published. These major achievements opened a new era in studying the mechanisms of protein synthesis. From then on, it became possible to attribute different aspects of the ribosome function to particular elements of its tertiary structure. However, establishing the structure-function relationships is problematic due to the immense complexity of the ribosome structure. In other words, in order to make the crystallographic data on the ribosome tertiary structure really useful for understanding of how the ribosome functions, it must be thoroughly analyzed. Here, based on systematic analysis of the available X-ray conformations of the ribosome we have tried to resolve two fundamental problems of the ribosome biology: concerning (1) the nature of rearrangements in the ribosome that take place at different steps of its functional cycle, and (2) the reconstruction of the ribosome evolution from the RNA world to present time. In the first project, we systematically compared the available structures of the ribosome and its subunits to identify rigid domains, which always have the same conformation, and flexible regions, where the conformation can vary from one ribosome structure to another. There were two known types of structural rearrangements whose mechanisms we wanted to understand: the ratchet-like motion and the so-called domain closure. The ratchet-like motion takes place during the ribosomal translocation and is roughly seen as a rotation of one subunit with respect to the other. The domain closure occurs in the small subunit and is associated with the cognate codon-anticodon recognition in the A-site. Comparison of the available ribosome conformations revealed the detailed mechanisms of both rearrangements. Although the selection of the cognate amino-acyl-tRNA in the A-site and of the ribosomal translocation have never been thought to have anything in common, we demonstrate that the rearrangements in the ribosome structure associated with the first process repeat in reverse order the rearrangements associated with the second process. In other words, during the ribosome elongation cycle, the domain closure and the ratchet-like motion can be seen as a back-and-forth movement, which eventually returns the ribosome to the initial conformation. iv In the second project, we attempted to reconstruct the evolution of the 23S rRNA from the RNA world to present time based on the presumption that the evolutionary expansion of this molecule proceeded though random insertions of relatively short regions into different regions of the polynucleotide chain. We developed criteria for integrity of the ribosome structure and presumed that during the evolutionary expansion, the ribosome structure always matched to these standards. For this, we specifically considered the A-minor interaction, a frequent arrangement in the rRNA structure consisting of a stack of unpaired adenosines tightly attached to a double helix. We presumed that in all A-minor interactions present in the ribosome, the double helix emerged before or at least simultaneously with the corresponding adenosine stack. The systematic application of this principle to the known tertiary structure of the 23S rRNA allowed us to elucidate in a step-vise manner the order in which different part of the modern 23S rRNA joined the structure. Taken together, the two projects demonstrate the effectiveness of the systematic in-silico analysis of the ribosome tertiary structure and pave the way for future discoveries. / Les résultats ont été obtenus avec le logiciel "Insight-2" de Accelris (San Diego, CA)