Submitted by Eduarda Figueiredo (eduarda.ffigueiredo@ufpe.br) on 2015-03-11T15:05:41Z No. of bitstreams: 2 TESE DOUTORADO YANA_2012.pdf: 5122730 bytes, checksum: a798a9fb98cb3241c68deaf639f4bc36 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) / Made available in DSpace on 2015-03-11T15:05:41Z (GMT). No. of bitstreams: 2 TESE DOUTORADO YANA_2012.pdf: 5122730 bytes, checksum: a798a9fb98cb3241c68deaf639f4bc36 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Previous issue date: 2012-05-04 / Existe uma gama de processos que podem ser empregados no tratamento de compostos orgânicos tóxicos como o fenol presente em efluentes líquidos. Todavia, restrições como alto custo de instalação e tempo de contato relativamente longo devem ser levadas em consideração para a escolha do processo a ser utilizado. Uma nova técnica de tratamento de efluentes contaminados por compostos fenólicos, denominada de tratamento térmico por contato direto (Direct Contact Thermal Treatment - DiCTT), vem atraindo o interesse de vários grupos de pesquisa. O Departamento de Engenharia Química da UFPE possui um protótipo de planta experimental deste processo. O método DiCTT tem como atrativo principal a utilização do gás natural como fonte energética, a capacidade demonstrada de oxidar compostos fenólicos a baixas temperaturas e pressão atmosférica, e sua natureza em empregar uma configuração de reator muito compacta. Nesta pesquisa, foram avaliados os efeitos das variáveis operacionais: vazão de alimentação do efluente líquido, concentração inicial de fenol, a taxa de reciclagem dos gases de combustão, razão molar estequiométrica fenol/peróxido de hidrogênio, vazão do gás natural e excesso do ar. Após estudos preliminares, envolvendo duas etapas (Etapa 1 e Etapa 2) e dois modos operatórios do processo (MO1 e MO2), foi efetuado um planejamento experimental do tipo Delineamento Composto Central Rotacional com fatorial (k) de (23) considerando a taxa de evaporação da fase líquida (inferior à 12%) e a temperatura do efluente líquido (75-790C) (Etapa 3). Com os resultados obtidos, foi realizado um planejamento fatorial (k) completo de (32) para identificação das condições operacionais ótimas do processo, nas faixas de variação estudadas (Etapa 4). As análises envolveram a oxidação termoquímica do fenol e dos seus respectivos intermediários de degradação até a mineralização do composto orgânico e formação de ácidos, com monitoramento das concentrações de fenol e seus intermediários por Cromatografia em fase Líquida de Alta Eficiência, medições do teor de Carbono Orgânico Total com um analisador COT e do potencial hidrogeniônico, com um medidor de pH, respectivamente. Aplicando a superfície de respostas e curvas de contorno com uso do software Statistica versão 8.0, foram identificadas as condições ótimas, nos intervalos de variação estudados, para a degradação do fenol, até 100%, e a conversão de COT, até 40%: vazão de gás natural de 4 m3/h, excesso de ar de 10%, razão molar estequiométrica fenol/peróxido de hidrogênio de 75% e taxa de reciclagem dos gases de combustão de 100%, valores satisfatórios para a degradação do fenol, porém, necessita-se ainda de melhorar as taxas de mineralização. Na etapa final desta pesquisa, foi usada a modelagem via Redes Neurais Artificiais (RNAs), com o software Statistica 8.0 e o módulo “Neural Networks” para predizer os resultados experimentais da degradação do fenol, conversão do COT e da velocidade de degradação do fenol, em função do tempo. Com os resultados obtidos, possibilitou-se concluir que o modelo mais representativo é o Modelo de Regressão e a forma da rede é a Multi Layer Perceptron. Verifica-se ainda que as correlações observadas variaram entre 95,33% até 99,58%, indicando um modelo satisfatório na predição dos resultados experimentais.

Os ácidos ribonucleicos (RNAs) podem ser classificados em duas classes principais: codificante e não codificante de proteína. Os codificantes, representados pelos RNAs mensageiros (mRNAs), possuem a informação necessária à síntese proteica. Já os RNAs não codificantes (ncRNAs) não são traduzidos em proteínas, mas estão envolvidos em várias atividades celulares distintas e associados a várias doenças tais como cardiopatias, câncer e desordens psiquiátricas. A descoberta de novos ncRNAs e seus papéis moleculares favorece avanços no conhecimento da biologia molecular e pode também impulsionar o desenvolvimento de novas terapias contra doenças. A identificação de ncRNAs é uma ativa área de pesquisa e um dos correntes métodos é a classificação de sequências transcritas utilizando sistemas de reconhecimento de padrões baseados em suas características. Muitos classificadores têm sido desenvolvidos com este propósito, especialmente nos últimos três anos. Um exemplo é o Coding Potential Calculator (CPC), baseado em Máquinas de Vetores de Suporte (SVM). No entanto, outros algoritmos robustos são também reconhecidos pelo seu potencial em tarefas de classificação, como por exemplo Random Forest (RF). O método mais utilizado para avaliação destas ferramentas tem sido a validação cruzada k-fold. Uma questão não considerada nessa forma de validação é a suposição de que as distribuições de frequências dentro do banco de dados, em termos das classes das sequências e outras variáveis, não se alteram ao longo do tempo. Caso essa premissa não seja verdadeira, métodos tradicionais como a validação cruzada e o hold-out podem subestimar os erros de classificação. Constata-se, portanto, a necessidade de um método de validação que leve em consideração a constante evolução dos bancos de dados ao longo do tempo, para proporcionar uma análise de desempenho mais realista destes classificadores. Neste trabalho comparamos dois métodos de avaliação de classificadores: hold-out temporal e hold-out tradicional (atemporal). Além disso, testamos novos modelos de classificação a partir da combinação de diferentes algoritmos de indução com características de classificadores do estado da arte e um novo conjunto de características. A partir dos testes das hipóteses, observamos que tanto a validação hold-out tradicional quanto a validação hold-out temporal tendem a subestimar os erros de classificação, que a avaliação por validação temporal é mais fidedigna, que classificadores treinados a partir de parâmetros calibrados por validação temporal não melhoram a classificação e que nosso modelo de classificação baseado em Random Forest e treinado com características de classificadores do estado da arte e mais um novo conjunto de características proporcionou uma melhora significativa na discriminação dos RNAs codificantes e não codificantes. Por fim, destacamos o potencial do algoritmo Random Forest e das características utilizadas, diante deste problema de classificação, e sugerimos o uso do método de validação hold-out temporal para a obtenção de estimativas de desempenho mais fidedignas para os classificadores de RNAs codificantes e não codificantes de proteína. / Ribonucleic acids (RNAs) can be classified into two main classes: coding and non-coding of protein. The coding, represented by messenger RNAs (mRNAs), has the necessary information for protein synthesis. Non-coding RNAs (ncRNAs) are not translated into proteins but are involved in several distinct cellular activities associated with various diseases such as heart disease, cancer and psychiatric disorders. The discovery of new ncRNAs and their molecular roles favors advances in the knowledge of molecular biology and may also boost the development of new therapies against diseases. The identification of ncRNAs is an active area of research and one of the current methods is the classification of transcribed sequences using pattern recognition systems based on their characteristics. Many classifiers have been developed for this purpose, especially in the last three years. An example is the Coding Potential Calculator (CPC), based on Supporting Vector Machines (SVM). However, other robust algorithms are also recognized for their potential in classification tasks, such as Random Forest (RF). The most commonly used method for evaluating these tools has been cross-validation k-fold. An issue not considered in this form of validation is the assumption that frequency distributions within the database, in terms of sequence classes and other variables, do not change over time. If this assumption is not true, traditional methods such as cross-validation and hold-out may underestimate classification errors. The need for a validation method that takes into account the constant evolution of databases over time is therefore needed to provide a more realistic performance analysis of these classifiers. In this work we compare two methods of evaluation of classifiers: time hold-out and traditional hold-out (without considering the time). In addition, we tested new classification models from the combination of different induction algorithms with state-ofthe-art classifier characteristics and a new set of characteristics. From the hypothesis tests, we observe that both the traditional hold-out validation and the time hold-out validation tend to underestimate the classification errors, that the time validation evaluation is more reliable, than classifiers trained from parameters calibrated by time validation did not improve classification and that our Random Forest-based classification model trained with state-of-the-art classifier characteristics and a new set of characteristics provided a significant improvement in the discrimination of the coding and non-coding RNAs. Finally, we highlight the potential of the Random Forest algorithm and the characteristics used, in view of this classification problem, and we suggest the use of the time hold-out validation method to obtain more reliable estimates of the protein coding and non-coding RNA classifiers.

The C4 cycle represents a series of biochemical and anatomical modifications that are targeted to overcome the effects of photorespiration caused by the oxygenase capability of Ribulose Bisphosphate Carboxylase/Oxygenase (RuBisCO). The cycle has evolved independently in over 60 lineages, which suggests that recruitment of genes into the C4 cycle is a relatively easy process. However, the mechanisms by which the anatomy and cell-specificity of the components of the C4 cycle is achieved is poorly understood. Preliminary work in maize indicated several components of the C4 cycle may be targeted by microRNAs (miRNAs). To explore this, a library of sRNA sequences from mature leaf tissue of the model C4 species Cleome gynandra L. was generated and then searched against a list of expressed sequence tag sequences for candidate genes of the C4 cycle. To complement this, transgenic C. gynandra containing the viral p19 protein, which is capable of suppressing miRNA activity, were produced. A limited subset of the candidate C4 genes showed a high level of sRNA read alignment. In C. gynandra plants expressing p19 photosynthesis was compromised and transcripts of several genes (most notably RbcS and RCA) were upregulated. These data were complemented by examining the effect of illumination on developing C. gynandra cotyledons, and attempts to generate a hybrid between C. gynandra and the C3 C. hassleriana Chodat. RbcS also showed elevated abundance in etiolated cotyledons, although this rapidly declined after illumination. The remainder of the C4 genes profiled accumulated in etiolated tissue, but were upregulated within 6 hours of illumination. Therefore, this study has illustrated that miRNA activity may play a role in maintaining the C4 photosynthetic cycle at optimum efficiency, although it has not been possible to identify at which point(s) this regulation is applied. Secondly, RbcS appears to be subject to multiple regulatory mechanisms in C. gynandra, and it is possible that miRNAs have a role in negatively regulating expression of RbcS.

L’anévrisme de l’aorte abdominale (AAA) représente un problème majeur de santé publique et est associé à des taux extrêmement élevés de mortalité en cas de rupture aortique. Il est classiquement associé à l’athérosclérose et aux facteurs de risque cardiovasculaires, à l’exception du diabète qui jouerait un rôle protecteur dans la maladie. Bien que certaines études aient montré une infiltration de macrophages et une modification de l’expression des micro-ARNs dans la paroi anévrismale, leurs rôles respectifs dans le développement de l’AAA ne sont pas encore totalement élucidés. Même si les modèles expérimentaux classiques sont utiles pour adresser cette question, ils ne miment pas parfaitement la physiopathologie humaine. Nous avons développé un nouveau modèle d’AAA associant application topique d’élastase et neutralisation systémique du TGFβ permettant de reproduire les principales caractéristiques de l’AAA humain et induisant une rupture aortique. Les objectifs de ce travail étaient de: -1/ Caractériser le phénotype des monocytes/ macrophages dans ce modèle murin d’AAA. -2/ D’étudier l’expression des micro-ARNs dans ce modèle. -3 / Les mécanismes impliqués dans la relation négative entre diabète et AAA étant à ce jour peu connus, le 3ème objectif était de mettre en place une étude clinique afin d’étudier l’expression des micro-ARNs chez des patients diabétiques ayant un AAA.L’application d’élastase associée à la neutralisation du TGFβ chez des souris C57/Bl6j entrainait une augmentation significative de la densité de macrophages infiltrés dans le tissu aortique anévrismal. L’analyse des tissus aortiques a montré des modifications significatives des marqueurs des macrophages pro-inflammatoires dits « M1 » et des marqueurs des macrophages impliqués dans la réparation tissulaire, dits « M2 ». Afin de mieux comprendre le rôle des macrophages dans ce modèle murin, des injections de clodronate ont été réalisées pour dépléter ces cellules. L’injection de clodronate diminuait de façon significative la dilatation anévrismale et prévenait la rupture. Ceci était associé à une préservation de la matrice extracellulaire et une modification de l’expression de certains marqueurs des macrophages, dont l’arginase-1 (ARG1), molécule impliquée dans la réparation tissulaire. La proportion de macrophages exprimant l’ARG1 augmentait en fonction de la sévérité de l’AAA. Enfin, la neutralisation du TGFβ induisait une diminution significative d’un type particulier de monocytes dits « Sat-Mono», impliqués dans la fibrose. Cette étude a ainsi montré le rôle des macrophages dans le développement et la rupture anévrismale avec une modification de leur phénotype. 752 micro-ARNs ont ensuite été analysés dans le tissu aortique, ce qui a permis d’identifier les micro-ARNs dont l’expression était modifiée dans ce modèle par rapport au groupe contrôle. Enfin, l’expression des micro-ARNs a été recherché chez l’homme en mettant en place une étude clinique. Bien que l’AAA chez l’homme soit classiquement associé à l’athérosclérose et aux facteurs de risque cardiovasculaire, il est paradoxalement négativement associé au diabète. Les mécanismes en cause sont encore peu connus. Nous avons comparé l’expression des micro-ARNs entre des patients diabétiques et non-diabétiques présentant un AAA. Cette étude pilote a permis d’identifier des cibles potentielles impliquées dans l’association négative entre diabète et AAA. / Abdominal aortic aneurysm (AAA) is a major public health concern and is associated with extremely high rates of mortality in case of aortic rupture. AAA is most often associated with atherosclerosis and cardiovascular risk factors, except diabetes that may rather play a protective role in the disease. Even though several studies have highlighted an infiltration of macrophages and changes of the expression of micro-RNAs in the aneurysmal wall, their role in AAA is still not fully understood. While experimental animal models are very useful to address this question, none of them perfectly mimics human pathophysiology. We recently created a new murine model of AAA based on topic application of elastase on the aorta associated with systemic TGFβ neutralization which reproduces the main human features of AAA and leads to fatal aortic rupture. The aims of this study were: -1/ Characterize the phenotype of monocytes/ macrophages in this murine model of AAA. -2/ Study the expression of micro-RNAs in this model. -3/ As the mechanisms involved in the negative association between diabetes and AAA are still poorly known, the third goal was to mount a clinical study to compare the expression of micro-RNAs between diabetic and non-diabetic patients with AAA. Topic application of elastase associated with systemic TGFβ neutralization in C57/Bl6j male mice led to a significant increase of macrophage infiltration in the aneurysmal tissue. This was associated with changes of the gene expression of the markers of the pro-inflammatory macrophages, called “M1” and of the macrophages involved in wound healing, called “M2”. To investigate the role of macrophages in this model, we used liposomes containing clodronate injections to deplete these cells. This led to significant decrease of the aortic dilatation and prevented rupture. This was associated with a better preservation of the extracellular matrix and significant changes in the gene expression of the markers of macrophages including arginase-1 (ARG1), a molecule involved in would healing. The proportion of macrophages expressing ARG1 increased with the severity of the AAA. At last, TGFβ neutralization led to a significant decrease of a population of macrophages involved in fibrosis, called “Sat-Mono”. This study highlighted the role and the phenotypic changes of macrophages during AAA development. We then analyzed the expression of 752 micro-RNAs in the aneurysmal aortic tissue which allowed identifying the micro-RNAs whose expression varied in the murine model. At last, the expression of micro-RNAs was investigated in patients with AAA. We compared the expression of micro-RNAs between diabetic and non-diabetic patients with AAA. This pilot study led to the identification of micro-RNAs that could potentially represent new targets involved in the negative association between diabetes and AAA.

[EN] Transcriptomics is one of the most important and relevant areas of bioinformatics. It allows detecting the genes that are expressed at a particular moment in time to explore the relation between genotype and phenotype. Transcriptomic analysis has been historically performed using microarrays until 2008 when high-throughput RNA sequencing (RNA-Seq) was launched on the market, replacing the old technique. However, despite the clear advantages over microarrays, it was necessary to understand factors such as the quality of the data, reproducibility and replicability of the analyses and potential biases. The first section of the thesis covers these studies. First, an R package called NOISeq was developed and published in the public repository "Bioconductor", which includes a set of tools to better understand the quality of RNA-Seq data, minimise the impact of noise in any posterior analyses and implements two new methodologies (NOISeq and NOISeqBio) to overcome the difficulties of comparing two different groups of samples (differential expression). Second, I show our contribution to the Sequencing Quality Control (SEQC) project, a continuation of the Microarray Quality Control (MAQC) project led by the US Food and Drug Administration (FDA, United States) that aims to assess the reproducibility and replicability of any RNA-Seq analysis. One of the most effective approaches to understand the different factors that influence the regulation of gene expression, such as the synergic effect of transcription factors, methylation events and chromatin accessibility, is the integration of transcriptomic with other omics data. To this aim, a file that contains the chromosomal position where the events take place is required. For this reason, in the second chapter, we present a new and easy to customise tool (RGmatch) to associate chromosomal positions to the exons, transcripts or genes that could regulate the events. Another aspect of great interest is the study of non-coding genes, especially long non-coding RNAs (lncRNAs). Not long ago, these regions were thought not to play a relevant role and were only considered as transcriptional noise. However, they represent a high percentage of the human genes and it was recently shown that they actually play an important role in gene regulation. Due to these motivations, in the last chapter we focus, first, in trying to find a methodology to find out the generic functions of every lncRNA using publicly available data and, second, we develop a new tool (spongeScan) to predict the lncRNAs that could be involved in the sequestration of micro-RNAs (miRNAs) and therefore altering their regulation task. / [ES] La transcriptómica es una de las áreas más importantes y destacadas en bioinformática, ya que permite ver qué genes están expresados en un momento dado para poder explorar la relación existente entre genotipo y fenotipo. El análisis transcriptómico se ha realizado históricamente mediante el uso de microarrays hasta que, en el año 2008, la secuenciación masiva de ARN (RNA-Seq) fue lanzada al mercado y comenzó a desplazar poco a poco su uso. Sin embargo, a pesar de las ventajas evidentes frente a los microarrays, resultaba necesario entender factores como la calidad de los datos, reproducibilidad y replicabilidad de los análisis así como los potenciales sesgos. La primera parte de la tesis aborda precisamente estos estudios. En primer lugar, se desarrolla un paquete de R llamado NOISeq, publicado en el repositorio público "Bioconductor", el cual incluye un conjunto de herramientas para entender la calidad de datos de RNA-Seq, herramientas de procesado para minimizar el impacto del ruido en posteriores análisis y dos nuevas metodologías (NOISeq y NOISeqBio) para abordar la problemática de la comparación entre dos grupos (expresión diferencial). Por otro lado, presento nuestra contribución al proyecto Sequencing Quality Control (SEQC), una continuación del proyecto Microarray Quality Control (MAQC) liderado por la US Food and Drug Administration (FDA) que pretende evaluar precisamente la reproducibilidad y replicabilidad de los análisis realizados sobre datos de RNA-Seq. Una de las estrategias más efectivas para entender los diferentes factores que influyen en la regulación de la expresión génica, como puede ser el efecto sinérgico de los factores de transcripción, eventos de metilación y accesibilidad de la cromatina, es la integración de la transcriptómica con otros datos ómicos. Para ello se necesita generar un fichero que indique las posiciones cromosómicas donde se producen estos eventos. Por este motivo, en el segundo capítulo de la tesis presentamos una nueva herramienta (RGmatch) altamente customizable que permite asociar estas posiciones cromosómicas a los posibles genes, transcritos o exones a los que podría estar regulando cada uno de estos eventos. Otro de los aspectos de gran interés en este campo es el estudio de los genes no codificantes, especialmente los ARN largos no codificantes (lncRNAs). Hasta no hace mucho, se pensaba que estos genes no jugaban ningún papel fundamental y se consideraban como simple ruido transcripcional. Sin embargo, suponen un alto porcentaje de los genes del ser humano y se ha demostrado que juegan un papel crucial en la regulación de otros genes. Por este motivo, en el último capítulo nos centramos, en un primer lugar, en intentar obtener una metodología que permita averiguar las funciones generales de cada lncRNA haciendo uso de datos ya publicados y, en segundo lugar, generamos una nueva herramienta (spongeScan) que permite predecir qué lncRNAs podrían estar secuestrando determinados micro-RNAs (miRNAs), alterando así la regulación llevada a cabo por estos últimos. / [CA] La transcriptòmica és una de les àrees més importants i destacades en bioinformàtica, ja que permet veure quins gens s'expressen en un moment donat per a poder explorar la relació existent entre genotip i fenotip. L'anàlisi transcriptòmic s'ha fet històricament per mitjà de l'ús de microarrays fins l'any 2008 quan la tècnica de seqüenciació massiva d'ARN (RNA-Seq) es va fer pública i va començar a desplaçar a poc a poc el seu ús. No obstant això, a pesar dels avantatges evidents enfront dels microarrays, resultava necessari entendre factors com la qualitat de les dades, reproducibilitat i replicabilitat dels anàlisis, així com els possibles caires introduïts. La primera part de la tesi aborda precisament estos estudis. En primer lloc, es va programar un paquet de R anomenat NOISeq publicat al repositori públic "Bioconductor", el qual inclou un conjunt d'eines per a entendre la qualitat de les dades de RNA-Seq, eines de processat per a minimitzar l'impact del soroll en anàlisis posteriors i dos noves metodologies (NOISeq i NOISeqBio) per a abordar la problemàtica de la comparació entre dos grups (expressió diferencial). D'altra banda, presente la nostra contribució al projecte Sequencing Quality Control (SEQC), una continuació del projecte Microarray Quality Control (MAQC) liderat per la US Food and Drug Administration (FDA) que pretén avaluar precisament la reproducibilitat i replicabilitat dels anàlisis realitzats sobre dades de RNA-Seq. Una de les estratègies més efectives per a entendre els diferents factors que influïxen a la regulació de l'expressió gènica, com pot ser l'efecte sinèrgic dels factors de transcripció, esdeveniments de metilació i accessibilitat de la cromatina, és la integració de la transcriptómica amb altres dades ómiques. Per això es necessita generar un fitxer que indique les posicions cromosòmiques on es produïxen aquests esdeveniments. Per aquest motiu, en el segon capítol de la tesi presentem una nova eina (RGmatch) altament customizable que permet associar aquestes posicions cromosòmiques als possibles gens, transcrits o exons als que podria estar regulant cada un d'aquests esdeveniments regulatoris. Altre dels aspectes de gran interés en aquest camp és l'estudi dels genes no codificants, especialment dels ARN llargs no codificants (lncRNAs). Fins no fa molt, encara es pensava que aquests gens no jugaven cap paper fonamental i es consideraven com a simple soroll transcripcional. No obstant això, suposen un alt percentatge dels gens de l'ésser humà i s'ha demostrat que juguen un paper crucial en la regulació d'altres gens. Per aquest motiu, en l'últim capítol ens centrem, en un primer lloc, en intentar obtenir una metodologia que permeta esbrinar les funcions generals de cada lncRNA fent ús de dades ja publicades i, en segon lloc, presentem una nova eina (spongeScan) que permet predeir quins lncRNAs podríen estar segrestant determinats micro-RNAs (miRNAs), alterant així la regulació duta a terme per aquests últims. / Furió Tarí, P. (2020). Development of bioinformatic tools for massive sequencing analysis [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/152485 / TESIS

Beyond the haploid genome, mammalian sperm carry a payload of epigenetic information with the potential to modulate offspring phenotype. Morphologically mature sperm exit the testes, but cannot swim or interact with the oocyte without extensive remodeling during epididymal transit; this includes modifications to the lipid composition of the sperm membrane, gain of necessary proteins, and a dramatic shift in sperm RNA content. Epididymal maturation has also been linked to changes in the sperm methylome suggesting that the epididymis might play a broader role in shaping the sperm epigenome. First, we characterized the genome-wide methylation landscape in seven germ cell populations from throughout the male reproductive tract. Our data emphasize the stability of cytosine methylation in mammalian sperm, and identify a surprising, albeit transient, period during which sperm are associated with extracellular DNA. Second, given our interest in the small RNA repertoire of sperm we set out to address known bias in sequencing protocols by comparing several small RNA cloning protocols. We found a protocol recently developed by Kathleen Collins’ lab (OTTR) to be superior to commercially available kits in providing an accurate representation of tRNA fragment levels as compared to Northern blotting. These results not only provide a more accurate representation of tRNA fragments, but also more complexity than previously seen allowing us to reassess the true sperm small RNA content. Taken together, these results provide significant insight into the mechanisms and factors modulating sperm epigenomics during post-testicular sperm maturation.

When the human genome project was completed, it revealed a surprising result. 98% of the genome did not code for protein of which more than 50% are repeats— later known as ”Junk DNA”. However, comparative genomics unveiled that many noncoding elements are evolutionarily constrained; thus luckily to have a role in genome stability and regulation. Though, their exact functions remained largely unknown. Several large international consortia such as the Functional Annotation of Mammalian Genomes (FANTOM) and the Encyclopedia of DNA Elements (ENCODE) were set to understand the structure and the regulation of the genome. Specifically, these endeavors aim to measure and reveal the transcribed components and functional elements of the genome. One of the most the striking findings of these efforts is that most of the genome is transcribed, including non-conserved noncoding elements and repeat elements. Specifically, we investigated the evolution and epigenetic properties of noncoding elements. 1. We compared genomes of evolutionarily distant species and showed the ubiquity of constrained noncoding elements in metazoa. 2. By integrating multi-omic data (such as transcriptome, nucleosome profiling, histone modifications), I conducted a comprehensive analysis of epigenetic properties (chromatin states) of conserved noncoding elements in insects. We showed that those elements have distinct and protective sequence features, undergo dynamic epigenetic regulation, and appear to be associated with the structural components of the chromatin, replication origins, and nuclear matrix. 3. I focused on the relationship between enhancers and repetitive elements. Using Cap Analysis of Gene Expression (CAGE) and RNASeq, I compiled a full catalog of active enhancers (a class of noncoding elements) during myogenesis of human primary cells of healthy donors and donors affected by Duchenne muscular dystrophy (DMD). Comparing the two time-courses, a significant change in the epigenetic landscape in DMD was observed that lead to global dysregulation of enhancers and associated repetitive elements.

MicroRNAs are small RNA molecules that could possibly play a major role in fertility. In the experiment, spermatozoa were extracted from bovine followed by an extraction of total RNA. Bovine spermatozoa were extracted from two bulls of different fertility, high and low fertility. An expression array was done to compare the expression levels of the microRNAs. It was shown that thousands of microRNAs are present in bovine spermatozoa but only a small amount was significantly expressed. The microRNAs from low fertility bulls were more highly expressed than those in high fertility bulls. A Bioanalyzer gel was used to confirm the results of the microarray data. The microRNAs were present in the bull’s spermatozoa at 25 nucleotides. The functions of the significantly expressed microRNAs are not known but there is a great possibility that their functions affect fertility.

Os genes MGC16121 e CR596471 localizam-se no cromossomo X (Xq26) entre os loci HPRT1 e PLAC1, uma região rica em genes associados com a reprodução humana. A importância de tais genes reside na possibilidade de estarem envolvidos no desenvolvimento placentário e fetal e de serem expressos em poucos tecidos normais. Camundongos portadores de deleções próximas do gene ortólogo HPRT1 de humanos apresentam cerca de um terço do tamanho dos camundongos selvagens ou em alguns casos são natimortos. No entanto, este fenótipo não é observado quando o gene está mutado. Assim, pode-se supor que o fenótipo anormal das cobaias não é resultado da deficiência do HPRT1, mas sim de genes e/ou microRNAs (miRNAs) próximos a ele. Estes resultados abrem perspectivas em relação ao estudo dos genes MGC16121, CR596471 e miRNAs das vizinhanças. O objetivo deste trabalho foi caracterizar a estrutura, a expressão e o mecanismo de regulação por metilação dos genes MGC16121 e CR596471. Adicionalmente foram analisados quanto ao perfil de expressão e regulação por metilação os miRNAs das vizinhanças (miR-424, 503, 450a, 450b-5p e 542-3p). O gene MGC16121 mostrou-se específico de placenta e também expresso em 50% das 18 linhagens tumorais analisadas. Já CR596471 e os miRNAs das vizinhanças foram mais expressos em placenta do que qualquer outro tecido normal analisado, sendo o primeiro expresso também em 100% das linhagens tumorais avaliadas. Houve correlação positiva e significativa entre todos os genes e miRNAs em relação à expressão em tecidos normais, porém o mesmo não foi observado para linhagens tumorais. A respeito da regulação, os genes CR596471 e MGC16121 e os miRNAs miR-424, 503 e 450a foram regulados negativamente por metilação do DNA em pelo menos uma das três linhagens tratadas com o agente demetilante 5-aza-2-deoxicitidina. Apoiando este fato, os dinucleotídeos CpG das ilhas CpGs situadas próximas às regiões 5 dos genes CR596471 e MGC16121 foram pelo menos em parte desmetilados após o mesmo tratamento.Os dados relativos à estrutura primária dos genes indicam que os transcritos, apesar de serem lncRNAs apresentaram características de mRNAs. Para MGC16121 foi determinado um transcrito composto de 3 éxons e, para CR596471, um transcrito composto de 3 éxons e outro composto de 2 éxons. Os transcritos aqui determinados são relativamente conservados quando comparados a sequências de RNA encontradas em outros mamíferos, principalmente em primatas. Adicionalmente, o transcrito de MGC16121 possui subestruturas secundárias visivelmente semelhantes com aquelas dos transcritos homólogos encontrados em alguns primatas. De acordo com os resultados, o gene MGC16121 pode ser considerado um possível bom marcador para diagnóstico, prognóstico e talvez para terapias contra cânceres. Todavia, mais experimentos devem ser realizados para verificar a função dos genes MGC16212 e CR5976471, além de avaliar mais robustamente a capacidade do gene MGC16121 ser utilizado como ferramenta na medicina contra o câncer. / CR596471 and MGC16121 genes lie on chromosome X (Xq26) between the HPRT1 and PLAC1 loci, a region rich in genes associated with human reproduction. The importance of such genes is the possibility that they might be involved in placental and fetal development, aware that they are expressed in few normal tissues. Deletions in mice around the orthologous gene of human HPRT1 affect their development or lead to stillbirth. However, this phenotype is not observed when this gene is mutated. So we can assume that the abnormal phenotype of mice cannot be due to HPRT1 deficiency, but to genes and/or microRNAs (miRNAs) nearby. These results support the idea of investigating the mechanisms involved in the regulation of the MGC16121 and CR596471 genes, and their neighbor miRNAs. This study aimed to characterize the structure, expression and regulation mechanism by methylation of genes MGC16121 and CR596471. In addition, the expression profile and methylation regulation of the neighbor miRNAs (miR-424, 503, 450a, 450b-5p and 542-3p) were analyzed. MGC16121 was demonstrated to be placenta specific and expressed in 50% of 18 tumor cell lines analyzed. CR596471 and the neighbor miRNAs were more expressed in placenta than in any other normal tissue analyzed. The former was also expressed in all tumor cell lines evaluated. There was significant and positive correlation between all genes and miRNAs regarding normal tissue expression. However, the same was not observed for the tumor cell lines. With respect to regulation, the genes CR596471 and MGC16121, and miRNAs miR-424, 503 and 450a were negatively regulated by DNA methylation at least in one of the three cell lines treated with the demethylating agent 5- aza-2-deoxycytidine. Supporting these results, the CpG dinucleotides from CpG islands located near the CR596471 and MGC16121 5 regions were at least partially demethylated after the same treatment. The data concerning to genes primary structures indicate that the transcripts, despite of being considered lncRNAs, presented mRNAs characteristics. It was determined one transcript for MGC16121 gene which consisted of three exons, and for CR596471 gene, two transcripts were found, one with three exons and other composed of two exons. The transcripts herein determined are relatively conserved when compared to RNAs sequences found in other mammals, mostly in primates. Besides, the MGC16121 transcript presents similar secondary substructures to those found in homologous transcripts from other primate species. According to the results, MGC16121 gene could be considered a possible good biomarker to diagnosis, prognosis and perhaps to therapies against cancers. Nevertheless, more experiments must be accomplished in order to verify the functions of MGC16121 and CR596471 genes, in addition to evaluate more robustly the competence of MGC16121 gene to be used as a tool in medicine against cancer.

Tnt1 is a transcriptionally active LTR-retrotransposon, present in over 600 copies in the Nicotiana tabacum genome. Under normal growth conditions, Tnt1 expression is limited to basal levels, but its expression is further induced under biotic and abiotic stresses. Transgenic tobacco plants (HP plants) expressing a Tnt1 reverse transcriptase hairpin were generated. These showed pleiotropic phenotypes such as cell death spots on the leaves and callose deposition and other severe abnormal development in aerial and underground portions. RNA sequencing of leaves with cell death spots revealed a rewiring of transcriptional regulatory networks related to stress responses exclusive to HPs. Among the positively modulated genes were ethylene synthesis and response cascade genes. The objective of the present work was to unravel the relation observed between Tnt1 and ethylene, generating a model. The results obtained suggest that HP seedlings and plants have increased ethylene synthesis when compared to the wildtype. Folding prediction of Tnt1 messenger RNA allowed the identification of ethylene-responsive sequences in putative stem loop locations. Thus it is possible that Tnt1 expression can produce small RNAs targeted to sequences present in the Tnt1 retrotransposon itself as well as at the promoter region of other ethylene responsive genes. Quantification of the expression of Tnt1 and ethylene related genes revealed \"phase opposition\" expression kinetics in the HPs, which led us to hypothesize that there might be an antagonistic relationship between the expression of Tnt1 and the expression of ethylene responsive genes involved in plant defense responses. Our findings suggest that Tnt1 could generate sRNAs that exerts transcriptional control over itself as well as other genes. Our model establishes a completely new biological role for a retrotransposon: Tnt1 would provide feedback control to ethylene-mediated gene regulation in tobacco defense responses, bringing the system back to a homeostatic condition and turning the defense responses down. / Tnt1 é um retrotransposon com LTR transcricionalmente ativo, e está presente em mais de 600 cópias no genoma de Nicotiana tabacum. Em condições normais de crescimento Tnt1 é expresso em níveis basais. No entanto, sua expressão é induzida pelo estímulo de estresses bióticos e abióticos. Plantas de tabaco transgênicas (chamadas de HP) expressando um grampo da transcriptase reversa de Tnt1 foram geradas. Estas apresentaram fenótipos como: pontos de morte celular e deposição de calose nas folhas e severas anomalias de desenvolvimento severas nas porções aérea e radicular das plantas. Sequenciamento de RNA de folhas com os pontos de morte celular revelou uma reorganização de redes de regulação transcricional relacionadas a resposta a estresses. Essas novas redes surgiram exclusivamente nas plantas HP. Entre os genes modulados positivamente estavam genes de síntese e de resposta ao etileno. O presente trabalho teve como objetivo elucidar a relação observada entre Tnt1 e o fitormônio etileno gerando um modelo de atuação. Os resultados obtidos permitiram demonstrar que plântulas e plantas HP adultas tem um aumento na síntese de etileno quando comparadas à selvagem. A predição do dobramento do RNA mensageiro de Tnt1 permitiu a identificação de sequências responsivas ao etileno localizadas em posição potencial para formar grampos. Desta forma, é possível que a expressão de Tnt1 leve à produção de pequenos RNAs que tem como alvo sequências responsivas a etileno presentes tanto no próprio elemento quanto em regiões promotoras de outros genes. A quantificação da expressão de Tnt1 versus genes relacionados ao etileno revelou um padrão em \"oposição de fase\" nas HPs, o que nos levou a hipotetizar que talvez ocorra uma relação antagonista entre a expressão de Tnt1 e a expressão de genes responsivos ao etileno envolvidos em respostas de defesa vegetais. Nossos resultados sugerem que Tnt1 pode gerar pequenos RNAs que exercem controle transcricional sobre Tnt1 e outros genes endógenos. Nosso modelo estabelece um novo papel biológico para um retrotransposon: Tnt1 agiria como um modulador da indução de genes mediada por etileno nas respostas de defesa de tabaco, trazendo o sistema de volta à condição homeostática e encerrando as respostas de defesa.