O comportamento materno é essencial para a sobrevivência e desenvolvimento do filhote mamífero. Durante a prenhez, as fêmeas recebem estímulos sensoriais e hormonais capazes de modificar e preparar o cérebro da mãe para o início dos padrões de comportamento materno (por exemplo, aumentando o número neurônios produtores de oxitocina no hipotálamo). Estudos têm identificado o hipotálamo como o principal responsável por estas mudanças, porém outras áreas do cérebro também estão envolvidas no processo do comportamento materno. Peg3, um gene marcado paternalmente expresso, é conhecido por controlar o comportamento materno em camundongos. Fêmeas nocautes para o gene Peg3 falham em aumentar a ingestão de alimentos, na ejeção de leite e em algumas atividades maternais, como placentofagia e construção do ninho. Este estudo teve como objetivo determinar os padrões de metilação da região diferencialmente metilada de Peg3 (Peg3DMR) de animais da raça Nelore de bovinos em diversas áreas do cérebro. Amostras foram coletadas das seguintes áreas: córtex frontal, occipital, temporal e parietal, hipocampo e hipotálamo, num total de 8 animais (4 machos e 4 fêmeas). O padrão de metilação destas amostras foi analisado pelo protocolo COBRA (do inglês, Combined Bisulfite-Restriction Analysis), que combina a modificação do DNA por bissulfito de sódio, amplificação por PCR e digestão por enzima de restrição. Foram encontrados diferentes padrões de metilação entre as amostras, ocorrendo uma predominância de hipometilação entre as amostras do sexo masculino, e padrões mais variados nas amostras do sexo feminino. As variações nos padrões de metilação ocorreram de maneira mais marcante entre as amostras de uma mesma região cerebral de diferentes animais, do que entre as amostras de várias regiões de um mesmo animal. Os resultados indicam que pode haver uma variação no status de imprinting em nível populacional, porém estudos com um número maior de amostras são necessários para a verificação da significância estatística destas variações. / The maternal behavior is essential to survival and development of mammalian offspring. Throughout pregnancy, females receive sensory and hormonal stimuli which promote modifications and prepare the mothers brain to the onset of maternal behavior patterns (for example, by increasing numbers of neurons producing oxytocin in the hypothalamus). Studies have identified the hypothalamus as the main responsible for these changes, but other areas of the brain are also involved in the maternal behavior process. Peg3, an imprinted paternally expressed gene, is known to control maternal behavior in mice. Peg3 knockout females failed in increasing food intake, milk ejection and some maternal activities as placentofagia and nest building. This study aimed to determine the methylation patterns of the differently methylated region of Peg3 (DMR-Peg3) of animals from Nellore cattle breed in several areas of the brain. Samples were collected from the following areas of cattle brain: the frontal, occipital, temporal and parietal cortices, hippocampus and hypothalamus, in a total of 8 animals (4 males and 4 females). The methylation pattern of these samples was analyzed by the protocol COBRA (Combined Bisulfite-Restriction Analysis), which combines DNA modification by sodium bisulfite, PCR amplification and digestion by restriction enzymes. It was found different methylation patterns among the samples. There was a predominance of hypomethylation among male samples, while different patterns were found among the female samples. Variation in the methylation patterns was more markedly observed among samples of the same cerebral region among different animals, then among samples of several regions within an animal. The results suggest that there may be a variation in the imprinting status at a population level, but further assays, with an increased number of samples are needed to verify the statistical significance of this variation.

Abstract The Prader-Willi (PWS) and Angelman (AS) syndromes are clinically distinct developmental disorders which are caused by genetic defects in the imprinted domain at chromosome 15q11-q13, resulting in the loss of paternal (PWS) or maternal (AS) gene function. In this study, the genetic changes of 15q11-q13 and their possible inheritance in Finnish PWS (n=76) and AS (n=47) patients are described. The diagnosis could be confirmed by laboratory methods in all PWS and in 43 (91%) AS patients. A deletion of 15q11-q13 accounted for 76% of the PWS and 67% of the AS patients in whom a specific genetic defect had been established. The origin of deletion was always paternal in PWS and maternal in AS. In PWS, deletions of four different sizes were detected, while in AS only type I or II deletions were found. The smallest overlap of deletions/critical region detected was from locus D15S13 to locus D15S10 in PWS and from locus D15S128 to locus D15S12 in AS. A rare recurrence of del(15)(q11q13) due to maternal germ line mosaicism is described. Maternal uniparental disomy of chromosome 15 accounted for 21% of PWS patients and paternal UPD for 2% of AS patients. In PWS, most UPD cases were due to errors in maternal meiosis (87%), most commonly leading to maternal heterodisomy (MI error). In AS, a rare error in the second segregation of paternal meiosis was found. UPD was associated with advanced maternal age, the mean being 34.6 years. Imprinting defects were found in 3% of PWS (two non-IC-deletions) and 11% of AS (IC deletion in one sib pair and three non-IC-deletions) patients. In the case with IC deletion, the mutation was seen in several generations. The non-deletion cases were thought to be due to a de novo prezygotic or postzygotic error. In the non-deletion PWS cases, the maternally imprinted paternal chromosome region was shown to have been inherited from the paternal grandmother, while in AS the paternally imprinted maternal chromosome region had been inherited from either the maternal grandfather or the maternal grandmother. The region of IC involved in AS was defined to be 1.15 kb. Five (11%) AS patients with normal DNA methylation test results had a UBE3A mutation. One of the two novel missense mutations (902A→C) had been inherited from the mosaic mother, while the mutation 975T→C was a new one. De novo deletions 1930delAG and 3093delAAGA have also been described previously, suggesting that these sites may be mutation hotspots in UBE3A. Identification of different genetic aetiologies with different recurrence risks is essential for genetic counselling, and close co-operation between clinicians and the laboratory is required both for diagnosis and for the detection of possible inheritance.

京都大学 / 0048 / 新制・課程博士 / 博士(医科学) / 甲第21023号 / 医科博第84号 / 新制||医科||6(附属図書館) / 京都大学大学院医学研究科医科学専攻 / (主査)教授 斎藤 通紀, 教授 萩原 正敏, 教授 小川 誠司 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

DNA methylation refers to an addition of a methyl group to the 5 position of the cytosine pyrimidine ring. As the best characterized epigenetic mark, DNA methylation plays an important role in a plethora of biological functions, including gene repression, genomic imprinting, silencing of retro-transposons and X chromosome inactivation. Genomic imprinting refers to the mono-allelic expression of certain genes according to their parent-of-origin. In mammals, the expression of imprinted genes is controlled by the cis-acting regulatory elements, termed imprinted control regions (ICRs). ICRs are marked by parent-of-origin-specific DNA methylation and loss of DNA methylation at ICRs also causes aberrant expression of imprinted genes. Therefore it is believed that the genomic imprinting is a DNA methylation-associated epigenetic phenomenon. As accurate expression of imprinted genes is essential for normal embryonic growth, energy homeostasis, development of the brain and behaviour and abnormal expression of imprinted genes leads to numerous clinical phenotype and human disorders, it is important to investigate how the imprinted DNA methylation is stably maintained in mammals. DNA methyltransferases (DNMTs) are the main enzymes that play a in the establishment and maintenance of imprinted DNA methylation. In primordial germ cells (PGCs), DNMT3A and DNMT3L are involved in the establishment of imprinted DNA methylation. Whereas once established, the imprinted DNA methylation is maintained by DNMT1, DNMT3A and DNMT3B, but mainly by DNMT1. In addition, some other enzymes and DNA binding proteins also play a role in this process. One of the best examples is ZFP57, which forms a complex with KAP1 and SETDB1. ZFP57 maintains imprinted DNA methylation by recognizing a methylated hexa-nucleotide and recruits DNMTs to the ICRs in mammalian embryonic stem (ES) cells. Interestingly, DNA methylation analysis combined with promoter microarrays carried out in our lab suggested that imprinted DNA methylation is absent from some of the maternal ICRs in ES cells genetically null for G9a, a histone H3 lysine 9 methylase. This indicates that G9a might also play a role in the maintenance of imprinted DNA methylation. In my work, I found that the repressive H3K9me2 and imprinted DNA methylation are absent from several analysed ICRs in embryonic stem (ES) cells genetically null for either G9a or its partner histone methyltransferase GLP. A knockdown of G9a in ES cells reproduced these observations suggesting that G9a/GLP complex is required for the maintenance of imprinted DNA methylation. I also found that neither wild type nor catalytically inactive G9a can restore the loss of imprinted DNA methylation in G9a-/- ES cells. Chromatin immunoprecipitation (ChIP) combined with bisulfite DNA sequencing showed that imprinted DNA methylation was present on the H3K9me2-marked allele indicating a direct role for G9a in maintenance of genomic imprinting. Using a pharmacological inhibitor of G9a and mutagenesis analyses, I found that G9a maintains the imprinted DNA methylation independently of its catalytic activity and recruits DNMTs to the ICRs via its ankyrin repeat domain. Dimerization of G9a with GLP is also essential for the maintenance of genomic imprinting in ES cells. In summary, in addition to establish H3K9me2, histone methyltransferases G9a and GLP also play an essential role in the maintenance of genomic methylation imprints in ES cells.

L'empreinte génomique est un mécanisme épigénétique qui conduit à l'expression d'un seul des deux allèles parentaux pour une centaine de gènes autosomaux chez les mammifères. La majorité des gènes soumis à l'empreinte est regroupée en clusters et tous ces gènes sont sous le contrôle de séquences discrètes appelées ICR (Imprinting Control Region). Les ICRs sont marquées épigénétiquement par une méthylation d'ADN et des modifications des histones alléliques. La méthylation d'ADN au niveau de ces ICRs est un facteur clé de l'empreinte et va être établie dans les lignées germinales suivant le sexe de l'embryon. Après fécondation, le nouvel embryon portera les empreintes paternelles et maternelles, ces empreintes devront alors être maintenues pendant tout le développement et interprétés dans le but de conduire à l'expression allélique des gènes soumis à l'empreinte. Cependant, la méthylation d'ADN ne peut expliquer à elle seule tous les aspects de l'empreinte génomique. Ainsi, d'autres marques épigénétiques doivent agir dans le maintien et la lecture de ces empreintes. Nous avons mis en évidence dans un premier temps que le contrôle de l'expression allélique dans le cerveau de Grb10 repose sur la résolution d'un domaine bivalent allélique spécifiquement dans le cerveau. Ces résultats mettent en avant pour la première fois un domaine bivalent dans le contrôle de l'expression des gènes soumis à l'empreinte et propose un nouveau mécanisme dans l'expression tissu spécifique de ces gènes. D'autre part, bien que des études en cellules ES aient démontré un rôle de G9a dans le maintien des empreintes au cours du développement embryonnaire, nos données suggèrent que G9a ne serait pas essentielle a ce maintien dans un contexte in vivo. / Genomic imprinting is a developmental mechanism which leads to parent-of-origin-specific expression for about one hundred genes in mammals. Most of imprinted genes are clustered and all are under control of sequence of few kilobases called Imprinting Control Region or ICR. ICRs are epigenetically marked by allelic DNA methylation and histone modifications. DNA methylation on ICRs is a key factor which is established in germ cells according to the sex of the embryo. After fecundation, the new embryo will harbored both paternal and maternal imprints which have to be maintained during the development and read to lead to allelic expression of imprinted genes. However, allelic DNA methylation alone cannot explain every aspect of genomic imprinting. Thus, there should be other epigenetic marks which act in the maintaining and reading of the imprints.Our data first indicate that bivalent chromatin, in combination with neuronal factors, controls the paternal expression of Grb10 in brain, the bivalent domain being resolved upon neural commitment, during the developmental window in which paternal expression is activated. This finding highlights a novel mechanism to control tissue-specific imprinting. On an other hand, although previous studies in ES cells show a role for G9a in the maintaining of imprints during embryonic development, our data suggest that G9a would not be essential in an in vivo model.

O imprinting genômico, processo regulado epigeneticamente segundo o qual os genes se expressam de acordo com sua origem parental, está envolvido no crescimento e desenvolvimento placentário. Na região 11p15.5 encontram-se vários genes regulados por duas regiões controladoras de imprinting (ICR1 e ICR2), onde se encontram as regiões diferencialmente metiladas H19DMR e KvDMR1. Acredita-se que o padrão de imprinting seja dinamicamente regulado durante o desenvolvimento da placenta. Em humanos, há poucas informações sobre imprinting genômico e desenvolvimento placentário, principalmente para estágios precoces do desenvolvimento devido às dificuldades técnicas de obtenção dessas placentas. A descrição de mosaicismo do padrão de metilação restrito a placenta ou entre a placenta e o feto evidencia um perfil epigenético único deste órgão. A 5-hidroximetilação, a qual não tem um papel de silenciamento gênico, pode ser confundida com a metilação do DNA nas análises moleculares. O objetivo principal do presente estudo foi o de verificar a influência da idade gestacional (IG) no perfil de metilação do DNA das ICRs 1 e 2 em vilosidade coriônica, bem como a existência de mosaicismo do perfil de metilação intra-placentário. Neste trabalho também foi investigada a presença de hidroximetilação na KvDMR1. Foram coletadas amostras de tecido placentário, sendo 25 de vilosidades coriônicas (VC) (15 de 3° trimestre gestacional e 10 do 1° trimestre) e nove de cordão umbilical (UC) de 1° trimestre (pareadas com a VC). Quatro placentas de 3° trimestre foram analisadas em separado para o estudo de mosaicismo. O perfil de metilação do DNA das regiões foi verificado por PCR Específica para a Metilação (MS-PCR), Análise Combinada de Bissulfito e Restrição Enzimática (COBRA) e Método de Digestão Enzimática Sensível à Metilação Associada à PCR em Tempo Real (DESM-RT), além do ensaio para hidroximetilação na KvDMR1. Com os ensaios qualitativos (MS-PCR e COBRA) foi observado um perfil de metilação monoalélico, sendo que na H19DMR foi identificada a presença de CpGs diferentemente metilados. Para a H19DMR foram observadas médias de 0,43 de metilação em VC e 0,31 em UC de 1° trimestre, e de 0,41 em VC de 3° trimestre. Para a KvDMR1, foram encontradas médias de 0,47 em VC e 0,57 em UC de 1° trimestre, e de 0,41 em VC de 3° trimestre. A presença de hidroximetilação na KvDMR1 foi excluída. Não foram observadas diferenças significativas entre as médias das diferentes IGs ou entre tecidos pelos testes t e F para ambas as regiões. Não foi observada correlação positiva no perfil de metilação para H19DMR e KvDMR1 entre os tecidos. Em relação ao mosaicismo, não houve diferenças significativas no perfil de metilação entre os diferentes cotilédones amostrados numa mesma placenta. Os resultados demonstram uma discordância entre tecido embrionário (UC) e extraembrionário (VC). Apesar de não serem observadas alterações significantes nos perfis de metilação da H19DMR e KvDMR1 em diferentes IGs, as informações apresentadas são importantes para as pesquisas sobre a dinâmica do fenômeno de imprinting genômico ao longo da gestação, para os estudos de mosaicismo intraplacentário bem como o perfil epigenético da placenta em relação a outros tecidos. / Genomic imprinting, an epigenetically regulated process by which genes are expressed accordingly to their parental origin, is involved in placental growth and development. In 11p15.5 region, there are many genes regulated by two Imprinting Control Regions (ICR1 and ICR2), in which are found two Differentially Methylated Regions, H19DMR and KvDMR1, respectively. Imprinting patterns seem to be adjusted during placenta development. In humans, there is little information on genomic imprinting and placental development, especially for early stages of development due to technical difficulties in obtaining these placentas. The description of mosaicism in methylation pattern restricted to placenta or between placenta and fetus shows a unique epigenetic profile of this organ. The 5-hidroxymethylation, which has no role in gene silencing, can be confused with DNA methylation in molecular analysis. The main aim of our study was to verify the influence of gestational age (GA) in DNA methylation profile of ICRs 1 and 2 in chorionic villi, as well as the existence of intra-placental methylation profile mosaicism. The presence of hydroximethylation in the KvDMR1 was also investigated. Samples were collected from placentas, 25 from chorionic villi (CV) (15 of the 3rd gestational trimester and 10 of the 1st trimester) and nine from umbilical cord (UC) in 1st trimester (paired with the CV samples). Four 3rd trimester placentas were separately analyzed for mosaicism. DNA methylation profile was verified by Methylation Specific PCR (MS-PCR), and Combined Bisulfite Restriction Analysis (COBRA) and Methylation-Sensitive Enzyme Digestion Method associated with Real-Time PCR (DESM-RT), in addition to hydroximethylation test in the KvDMR1 region. With qualitative assays (MS-PCR and COBRA), it was observed a monoallelic methylation pattern, and, only for the H19DMR, differently methylated CpGs were observed. For the H19DMR, we observed methylation means of 0.43 in CV and 0.31 in UC of 1st trimester, and 0.41 in CV of 3rd trimester. For KvDMR1, we observed means of 0.47 in CV and 0.57 in UC of 1st trimester, and 0.41 in CV of 3rd trimester. No hydroximethylation in the KvDMR1 was observed. There were no significant differences between the means of different GAs or between tissues by F and t tests for both regions. No positive correlation was found on methylation profile for H19DMR and KvDMR1 between tissues. In relation to mosaicism, there were no significant differences in methylation profile between different cotyledons sampled in the same placenta. The results showed a discrepancy between embryonic (UC) and extra-embryonic (CV) tissues. Although it was not observed significant changes in methylation profiles of H19DMR and KvDMR1 in different GAs, the presented results are important to research on dynamics of genomic imprinting phenomenon during pregnancy, studies of intra-placental mosaicism and placenta epigenetic profile in relation to other tissues.

A síndrome de Silver-Russell (SRS) é caracterizada principalmente por grave retardo de crescimento intrauterino e pós-natal e face típica, pequena e triangular, entre outras características variáveis. A SRS é geneticamente heterogênea, ocorrendo em geral de forma esporádica. Mutações genéticas e epigenéticas em regiões sujeitas a imprinting genômico nos cromossomos 7 e 11 são detectadas em cerca de 50% dos pacientes. Mais frequentemente, a SRS é causada pela alteração da expressão gênica na região 11p15 devido à hipometilação do centro de imprinting telomérico (ICR1) que ocorre em pelo menos 40% dos afetados. Duplicações cromossômicas de origem materna incluindo o centro de imprinting centromérico (ICR2) estão presentes em 1-2% dos casos. A dissomia uniparental materna do cromossomo 7 (matUPD7) é responsável por 5-10% dos casos. Mais recentemente microdeleções e microduplicações cromossômicas foram detectadas em um grupo pequeno de pacientes, algumas delas se mostrando com possível efeito patogênico. Com a identificação da hipometilação de ICR1 em 11p15, matUPD(7) e desequilíbrios (sub)microscópicos, a confirmação molecular para o diagnóstico clínico da SRS tornou-se possível em ~50% dos pacientes, o que deixa metade dos casos sem causa genética determinada. A amostra foi constituída por 64 pacientes brasileiros não aparentados, com suspeita clínica da síndrome de Silver-Russell. O número de cópias de DNA e o padrão de metilação do cromossomo 11p15 foram investigados em 49 pacientes utilizando MS-MLPA, e 21 (43%) deles apresentaram hipometilação de ICR1. Em um desses pacientes (2%), ambos os centros, ICR1 e ICR2, estavam hipometilados, alteração complexa que já foi relatada em ~4% dos pacientes com SRS que apresentavam hipometilação de ICR1. Em outro paciente (2%), foi detectada uma microduplicação de origem materna que incluía o domínio ICR2, mas não ICR1. Essa microduplicação segrega em três gerações de uma família e a manifestação da síndrome depende da transmissão via materna: houve quatro casos de transmissões paternas da microduplicação de um único homem uniformemente resultando em prole normal, e cinco transmissões maternas, de duas irmãs clinicamente normais, com todas as crianças apresentando SRS. Outra microduplicação de origem materna restrita ao domínio ICR2 e associada com SRS em um menino foi descrita anteriormente. Entre os genes duplicados nos dois casos, CDKN1C aparece como candidato para o fenótipo da SRS, uma vez que codifica para um inibidor de quinase dependente de ciclina que regula negativamente o crescimento celular e tem papel crucial no desenvolvimento fetal humano. Esse novo caso familial vem confirmar que a duplicação restrita ao domínio ICR2, de herança materna, está causalmente associada com a SRS; mostra também que nenhuma alteração fenotípica aparente está presente, quando a duplicação é herdada via paterna. Entre os 64 pacientes da amostra, três (4,7%) foram identificados apresentando matUPD(7), pela genotipagem de microssatélites do cromossomo 7. As frequências de hipometilação de ICR1 (43%) e matUPD(7) (4,7%) entre os nossos pacientes, concordantes com o de outros estudos semelhantes, apontam para a seleção adequada dos pacientes com SRS, do ponto de vista clínico. A investigação de microrrearranjos cromossômicos por a-CGH foi realizada em 19 pacientes, que previamente tiveram afastadas alterações (epi)genéticas em 11p15 e a matUPD(7). A maioria dos pacientes não apresentou alterações (n=7) ou possuía apenas CNV frequentes em indivíduos normais da população e consideradas polimorfismos (n=8). Quatro microdeleções potencialmente patogênicas foram detectadas, em 2p23.3 (~320 Kb), 13q24 (~94,3 Kb), 15q11.2 (~320 Kb) e 16p13.11 (~95,8 Kb). Em nenhum dos casos foi possível estabelecer relação direta com o fenótipo da SRS, porque não foi possível investigar ambos os genitores ou a alteração estava presente em um genitor clinicamente normal ou já tinha sido relatada em indivíduo normal da população, não havendo, entretanto, indicação de ser polimórfica. A penetrância incompleta ou a manifestação de alelo recessivo patogênico no cromossomo homólogo são duas possíveis explicações para o efeito patogênico das microdeleções herdadas de genitor clinicamente normal. Três estudos recentes que utilizaram microarrays na busca de genes ou regiões cromossômicas associadas com a SRS, em que a causa genética era desconhecida, detectaram microduplicações e microdeleções, algumas potencialmente patogênicas: uma microdeleção em 15q26.3, incluindo o gene IGF1R, foi identificada em dois pacientes; outras microdeleções incluíam os genes IGF2BP3 em 7p15, GPC5 em 13q31.3, o MAPK1 em 22q11.2 e o HMGA2 em 12q14, considerados candidatos, possivelmente influenciando o crescimento. Esse conjunto de resultados indica que a investigação de microrrearranjos deve estender-se a um número maior de pacientes com SRS, na busca regiões cromossômicas e genes que possam estar causalmente associados com a síndrome. Em 30 pacientes com SRS, buscamos mutações no gene CDKAL1, por sequenciamento direto das regiões codificadoras. Esse gene foi considerado candidato para a síndrome, após ter sido interrompido em um dos nossos pacientes com SRS, portador de uma translocação t(5;6). Nenhuma alteração patogênica foi detectada, indicando que mutações de ponto na região codificadora do gene CDKAL1 não é causa comum da SRS. Em 18 dos 30 pacientes, investigamos a presença de microdeleções e microduplicações por a-CGH e não encontramos alteração que incluísse esse gene. Considerando o pequeno tamanho amostral, não podemos excluir definitivamente a possibilidade de que alterações no gene CDKAL1 possam contribuir para a etiologia da SRS. / Silver Russell syndrome (SRS) is characterized by severe intrauterine and postnatal growth retardation in association with a typical small triangular face and other variable features. Most cases are sporadic. Genetic and epigenetic disturbances on imprinted regions at chromosomes 7 and 11 are detected in about 50% of the patients. Most frequently, SRS is caused by altered gene expression on chromosome 11p15 due to hypomethylation of the telomeric imprinting center (ICR1) that is present in at least 40% of the patients. Maternally inherited duplications encompassing the centromic imprinting center (ICR2) domains at 11p15 are present in about 1-2% of cases. Maternal uniparental disomy of chromosome 7 (mUPD7) is identified in 5-10% of patients. More recently, chromosomal microdeletions and microduplications were detected in a small group of SRS patients, some of them with possible pathogenic effect. This leaves approximately half of the SRS cases without a genetic cause determined. Our cohort consisted of 64 unrelated Brazilian patients with clinical diagnosis of SRS. DNA copy number changes and the methylation pattern on chromosome 11p15 were investigated in 49 patients by MS-MLPA, and 21 (43%) presented with hypomethylation of ICR1. In one patient (2%), both centers (ICR1 and ICR2) were hypomethylated, a complex alteration that has been reported in ~4% of SRS patients that shows hypomethylation of ICR1. In a further patient (2%), we detected a ~1.6 Mb microduplication encompassing the whole ICR2 domain, but not the ICR1. This microduplication was shown to segregate in a three-generation family, and was associated with SRS whenever maternally transmitted: there were four instances of paternal transmissions of the microduplication from a single male uniformly resulting in normal offspring, and five maternal transmissions, via two clinically normal sisters, with all the children exhibiting SRS. A maternally inherited microduplication also restricted to the ICR2 domain and associated with SRS in a boy was described previously. Among the duplicated genes in both cases, CDKN1C is a likely candidate for the SRS phenotype, because it encodes a cyclin-dependent kinase inhibitor that negatively regulates cell proliferation and growth, and plays a crucial role in human fetal development. This new case brings confirmatory evidence that microduplications restricted to the ICR2 domain result in SRS when maternally transmitted. It also shows that no apparent phenotypic change is present when ICR2 duplication is paternally inherited. By genotyping chromosome 7 microsatellites, we identified three patients (4.7%) with mUPD(7), in the cohort of 64 patients. The frequencies of hypomethylation of ICR1 (43%) and mUPD(7) (4.7%) among our patients are in accordance with the literature, and point to a proper selection of patients with SRS, from the clinical point of view. The investigation of submicroscopic chromosomal imbalances by a-CGH was performed in 19 patients in whom (epi)genetic mutations at 11p15 and mUPD(7) had been excluded. Most patients showed no changes (n = 7) or had only CNV considered to be polymorphic (n = 8). Four potentially pathogenic microdeletions were detected, on chomosomes 2p23.3 (~320 Kb), 13q24 (~94.3 Kb), 15q11.2 (~320 Kb) and 16p13.11 (~95.8 Kb). In neither case we could establish a direct relationship between the imbalance and the phenotype, because it was not possible to investigate both parents or the change was present in a clinically normal parent or it had been reported in normal individuals, without, however, indication of being polymorphic. Incomplete penetrance or unmasking of a pathogenic recessive allele on the homologous chromosome are two possible explanations to the pathogenic effect of a microdeletion inherited from a clinically normal parent. Three recent studies that used microarrays to identify genes or chromosomal regions associated with SRS, wherein the genetic cause was unknown, detected microdeletions and microduplications, some of them potentially pathogenic: a microdeletion at 15q26.3, including the IGF1R gene, was identified in two patients; other microdeletions included the IGF2BP3 gene at 7p15, GPC5 gene at 13q31.3, MAPK1 gene at 22q11.2 and HMGA2 gene at 12q14, which were considered candidates, possibly influencing growth. This set of results, including ours, indicates that the investigation of submicroscopic chromosomal imbalances should be extended to a larger cohort of SRS patients, in the search for chromosomal regions and genes that may be causally associated with the syndrome. In 30 SRS patients, we searched for point mutations in the CDKAL1 gene by direct sequencing of coding regions. This gene was considered a candidate for SRS, after being disrupted in one of our SRS patients with a t(5;6). No pathogenic mutation was detected and, therefore, point mutations in the coding region of CDKAL1 do not appear to be a common cause of SRS. In 18 of the 30 patients, we investigated the presence of microdeletions and microduplications by a-CGH and found no changes encompassing CDKAL1 gene. Considering the small cohort size, we cannot definitely exclude the possibility that changes in CDKAL1 gene may contribute to the etiology of SRS.

Le gène H19, soumis à l’empreinte parentale, est fortement exprimé durant le développement embryonnaire, cependant, son expression est réprimée après la naissance dans l’ensemble des tissus à l’exception du muscle squelettique, et plus particulièrement des cellules souches musculaires : les cellules satellites. L’objectif de ma thèse a été de déterminer le rôle du gène H19 dans la mise en place et dans la fonction de ces cellules souches durant la myogénèse adulte. En utilisant un modèle murin présentant une délétion du gène H19, les souris H19∆3, notre laboratoire avait montré que le gène H19 est capable de moduler, dans le muscle embryonnaire, l’expression de neuf gènes appartenant à un réseau de gènes soumis à l’empreinte parentale (IGN) impliqué dans la croissance. Au cours de ma thèse, j’ai étudié le phénotype des muscles de ces souris mutantes qui présentent une hyperplasie et une hypertrophie des fibres musculaires. Ce phénotype est accompagné d’une diminution du nombre de cellules satellites qui apparait lors de l’entrée en quiescence de ces cellules. De façon étonnante, nous avons observé une meilleure capacité de régénération, malgré le nombre réduit de cellules satellites, dans les muscles H19∆3 comparée à celle des muscles wt. Cela indique que la capacité d’auto-renouvellement des cellules satellites n’est pas influencée par l’absence du gène H19. De même, nous avons observé une surexpression de plusieurs gènes appartenant à l’IGN lors de la régénération musculaire des muscles mutants comparés aux muscles wt. Ces résultats indiquent que le gène H19 module l’expression des gènes de l’IGN durant l’embryogénèse et par la suite, durant les étapes de régénération de la myogénèse adulte. / The imprinted H19 gene is highly expressed during embryonic development. H19 is fully repressed after birth in all tissues, with the exception of skeletal muscle, and especially of the muscle stem cells: the satellite cells. The aim of my thesis was to define the function of the H19 gene in the satellite cells establishment and function during adult myogenesis. Using loss-of-function H19∆3 mice, the laboratory had shown that the H19 gene was able to modulate the expression of several genes belonging to an imprinted gene network (IGN) in the embryonic muscle. During my thesis, I studied the muscle phenotype of these adult mice, which present both fiber hyperplasia and hypertrophy. This phenotype is accompanied by an important reduction of the satellite cell number, probably due to a delay in their entry into quiescence. Unexpectedly, despite the reduction in the number of satellite cells in mutant mice, the self-renewal capacity of the satellite cells is fully retained. In addition, we observe a better regeneration potential of the mutant muscles compared with wt muscles. This is accompanied by the enhanced expression of several genes from the IGN. These results indicate that H19 gene can modulate IGN gene expression both during embryogenesis and after birth, in adult myogenesis.

Epigenética é o ramo da biologia que estuda as características herdáveis não associadas a alterações na seqüência de nucleotídeos do DNA. Um dos principais processos epigenético estudados é a metilação do DNA, a qual está associada a diversos mecanismos de regulação gênica, entre eles o imprinting (marcação) genômico. Esse tipo de regulação caracteriza-se pela expressão parental específica dos loci associados e à metilação diferencial em regiões regulatórias conhecidas como centros de imprinting (ICs). Alterações desse mecanismo estão relacionadas com síndromes de hipo e hipercrescimento em humanos e animais domésticos, desenvolvimento de tumores, doenças associadas com alterações de comportamento e já foram detectadas em indivíduos concebidos por técnicas de reprodução assistida e em células-tronco embrionárias derivadas de diferentes espécies. Essas duas últimas evidenciam que genes marcados são particularmente lábeis ao estresse induzido por manipulação celular in vitro. As possíveis causas dessas epimutações não estão completamente esclarecidas. Os bovinos parecem ser um melhor modelo comparativo no estudo dessas alterações, evitando a utilização de embriões humanos. No entanto, existem poucas seqüências descritas de genes marcados nessa espécie. No presente estudo, duas regiões diferencialmente metiladas (H19DMR e KvDMR1) foram caracterizadas em bovinos em termos de elementos conservados (EC), enriquecimento de elementos repetitivos (ERs) e padrões de metilação. A análise de ECs e ERs foi realizada utilizandose os programas VISTA e RepeatMasker, respectivamente. Os padrões de metilação para ambas as DMRs foram analisados utilizando-se o ensaio de COBRA (do inglês COmbined Bisulfite Restriction Analysis) em DNA de sangue periférico e espermatozóides em amostras de animais concebidos in vivo. Também foi pesquisada a possível ocorrência de perda de imprinting em uma amostra de quatro animais clonados. A análise dos resultados indicou que os padrões de imprinting observados nas DMRs bovinas estudadas são semelhantes aos descritos para regiões homólogas em outras espécies de mamíferos. As características genômicas mostraram uma maior similaridade nas regiões analisadas entre bovinos e humanos do que entre humanos e camundongos. Não foram encontradas diferenças entre o padrão de imprinting de animais gerados naturalmente ou por transferência nuclear. Os resultados desse trabalho poderão auxiliar em futuras pesquisas de genes marcados em bovinos, além de contribuir para o melhoramento na utilização dessa espécie como modelo de comparação para desenvolvimento humano. / Epigenetics is the branch of biology which studies heritable changes in genome function that occur without a change in nucleotide sequence within the DNA. One of the most studied epigenetic process is the DNA methylation, which is associated with several gene regulation mechanisms such as genomic imprinting. This type of regulation is characterized by parental specific gene expression and differential methylation of the associated loci in regulatory sequences named imprinting centers (ICs). Alterations of this mechanism has been related to hypo and hypergrowth syndromes in humans and domestic animals, tumor development, behavior disorders, and it has also been associated with epimutations in individuals conceived by assisted reproduction (AR) techniques and stem cells derived from different species. These last two evidences are indicatives of the imprinted genes lability to in vitro cell manipulation. The possible causes of these epimutations are not completely clear. Cattle seem to be a better comparative model in the study of this epigenetic alterations, and it can avoid the use of human embryos. However, there is few description of imprinted gene sequences this species. In the present work, two differently methylated regions (H19DMR and KvDMR1) were characterized in terms of conserved elements (CEs), enrichment of repetitive elements (RE) and methylation patterns. The CEs and REs analysis was carried out using the VISTA and RepeatMasker softwares, respectively. The methylation patterns for both DMRs were analyzed by COBRA (COmbined Bisulfite Restriction Analysis) assay in DNA from peripherical blood and sperm samples of in vivo conceived animals. It also was investigated the loss of imprinting in samples of four cloned animals. The results indicated that the imprinting patterns of the studied bovine DMRs are similar to the other homologue regions in mammals. The genomic features demonstrated a bigger similarity of the analyzed regions between cattle and humans than between humans and mice. Differences between the imprinting patterns of in vivo conceived versus cloned animals were not found. The results of this work can help future studies of imprinted genes in cattle, and, in addition, can contribute for the improvements of this animal model as a comparative to the human development.

Parcela considerável de pacientes com distúrbios de crescimento não têm a causa de seus quadros clínicos estabelecida, incluindo aproximadamente 50% dos pacientes com diagnóstico clínico de síndrome de Silver−Russell (SRS) e 10-20% dos pacientes com síndrome de Beckwith-Wiedemann (BWS). O objetivo deste estudo foi investigar as causas genéticas e epigenéticas de distúrbios de crescimento, de etiologia desconhecida, numa contribuição para o entendimento de mecanismos que regulam o crescimento. O estudo compreendeu: (1) a investigação de microdesequilíbrios cromossômicos, por aCGH; (2) a análise do perfil de expressão alelo-específica de genes sujeitos a imprinting (IG), por pirossequenciamento (PSQ) ou sequenciamento de Sanger; (3) a investigação do padrão de metilação global em pacientes com restrição de crescimento, utilizando microarray de metilação. A casuística constituiu-se de 41 pacientes não aparentados, com distúrbios de crescimento, de etiologia desconhecida: (1) 25, com hipótese diagnóstica de SRS; (2) seis, com restrição de crescimento intrauterino e peso ao nascimento abaixo do 10º percentil, associados a outros sinais clínicos; (3) sete, com hipótese diagnóstica de BWS; e (4) três, com macrossomia pré-natal ou pós-natal, associada a outros sinais. A investigação de microdesequilíbrios cromossômicos foi realizada em 40 pacientes. Foram detectadas 58 variantes raras em 30/40 pacientes (75%): 40 foram consideradas provavelmente benignas (18 pacientes, 45%), 12, com efeito patogênico desconhecido (11 pacientes, 27,5%), duas, provavelmente patogênicas (um paciente, 2,5%) e quatro, patogênicas (três pacientes, 7,5%). Essas frequências são comparáveis àquelas descritas em estudos que investigaram CNV em grupos de pacientes com distúrbios de crescimento e outras alterações congênitas, incluindo SRS, e mostram a importância da investigação de microdesequilíbrios cromossômicos nesses pacientes. A diversidade dos microdesequilíbrios cromossômicos identificados é reflexo da heterogeneidade clínica das casuísticas. Neste estudo, muitos dos pacientes com hipótese diagnóstica de SRS e BWS apresentavam sinais clínicos atípicos, explicando a ausência neles das alterações (epi)genéticas que causam essas síndromes. A identificação de CNV características de outras síndromes reflete a sobreposição de sinais clínicos com BWS e SRS. A análise do perfil de expressão alelo-específica de IG foi realizada em um subgrupo de 18 pacientes com restrição de crescimento. Trinta IG com função em proliferação celular, crescimento fetal ou neurodesenvolvimento foram inicialmente selecionados. Após seleção de SNP transcritos com alta frequência na população, genotipagem de pacientes, genitores e indivíduos controle, determinação da expressão dos IG em sangue periférico e seu padrão de expressão (mono ou bialélico), 13 IG, expressos no sangue, tiveram a expressão alelo-específica avaliada, sete deles por PSQ e seis por sequenciamento de Sanger. Alterações no perfil de expressão de dois genes, de expressão normalmente paterna, foram detectadas em 4/18 pacientes (22%). Este estudo é o primeiro a utilizar pirossequenciamento e sequenciamento de Sanger na avaliação do perfil de expressão alelo-específica de IG, em pacientes com restrição de crescimento. Apesar de terem limitações, ambas as técnicas mostraram-se robustas e revelaram alterações de expressão alélica interessantes; entretanto, a relação dessas alterações com o quadro clínico dos pacientes permanece por esclarecer. A investigação da metilação global do DNA foi realizada em subgrupo de 21 pacientes com restrição de crescimento e em 24 indivíduos controle. Dois tipos de análise foram realizados: (1) análise diferencial de grupo e (2) análise diferencial individual. Na primeira análise, em que foi comparado o padrão de metilação do grupo de pacientes com quadro clínico sugestivo de SRS (n=16) com o do grupo controle (n=24), não houve indicação de hipo ou hipermetilação global no grupo SRS. Na segunda análise, foi comparado o padrão de metilação de cada um dos 21 pacientes com restrição de crescimento e dos 24 indivíduos controle, com o padrão de metilação do grupo controle. O número médio de CpG hipermetilados e de segmentos diferencialmente metilados (SDM) foi significativamente maior nos pacientes. Foram identificados 82 SDM hipermetilados, estando 57 associados a gene(s) (69,5%), em 16 pacientes, e 51 SDM hipometilados, 41 deles associados a gene(s) (80,4%), em 10 pacientes. A análise de ontologia genética dos 61 genes associados aos SDM hipo ou hipermetilados nos pacientes destacou genes que atuam no desenvolvimento e na morfogênese do sistema esquelético e de órgãos fetais, e na regulação da transcrição gênica e de processos metabólicos. Alterações de metilação em genes que atuam em processos de proliferação e diferenciação celulares e crescimento foram identificadas em 9/20 dos pacientes (45%), sugerindo implicação clínica. Não foi detectada alteração epigenética comum aos pacientes com diagnóstico clínico de SRS, explicável provavelmente pela heterogeneidade clínica. A investigação de metilação global, utilizando microarray, produziu novos dados que podem contribuir para a compreensão de mecanismos moleculares que influenciam o crescimento pré- e pós-natal. Na translocação aparentemente equilibrada - t(5;6)(q35.2;p22.3)dn, detectada em paciente com suspeita clínica de SRS, a interrupção de um gene, pela quebra no cromossomo 6, pode ser a causa do quadro clínico; alternativamente, a translocação pode ter impactado a regulação de genes de desenvolvimento localizados próximos aos pontos de quebra. A análise de expressão em sangue periférico mostrou que os níveis de cDNA do gene, interrompido pelo ponto de quebra da translocação, estavam reduzidos à metade. Além de sinais típicos da SRS, a paciente apresentava algumas características clínicas sugestivas de displasia cleidocraniana. Assim, a translocação t(5;6) pode ter alterado a interação de genes de desenvolvimento e seus elementos reguladores, levando à desregulação de sua expressão espaço-temporal, e resultando num fenótipo atípico, com características sobrepostas de mais de uma síndrome genética / A large number of patients with growth disorders do not have the cause of their clinical phenotype established, including about 50% of patients with Silver-Russell syndrome (SRS), and 10-20% of patients with Beckwith-Wiedemann syndrome (BWS). The aim of this study was to investigate the (epi)genetic causes of growth disorders of unknown etiology, in a contribution to the understanding of growth regulation. The study included: (1) the investigation of submicroscopic chromosomal imbalances, by aCGH, (2) the analysis of the allele-specific expression profile of imprinted genes (IG), by pyrosequencing (PSQ) or Sanger sequencing, in patients with growth restriction; (3) the investigation of global methylation pattern in patients with growth restriction, using methylation microarray. The cohort consisted of 41 unrelated patients with growth disorders: (1) 25, with the diagnostic hypothesis of SRS; (2) six, with intrauterine growth restriction and birth weight below the 10th centile, associated with other clinical signs; (3) seven, with the diagnostic hypothesis of BWS; and (4) three, with prenatal or postnatal macrosomia, associated with other clinical signs. Chromosomal microdeletions and microduplications were investigated in 40 patients. Fifty-eight rare variants were detected in 30/40 patients (75%): 40 were considered likely benign (18 patients, 45%), 12, of unknown pathogenic significance (11 patients, 27.5%), two, likely pathogenic (one patient, 2.5%), and four, pathogenic (three patients, 7.5%). These frequencies are similar to those described in studies investigating CNVs in groups of patients with growth disorders and other congenital abnormalities, including SRS, and show the importance of investigating chromosomal microimbalances in these patients. The diversity of CNVs identified can be attributed to the clinical heterogeneity of these cohorts. In this study, many of the patients, with the diagnostic hypothesis of SRS or BWS, had atypical clinical signs, thus explaining the absence of specific SRS/BWS (epi)genetic mutations. The identification of CNVs, known to be causative of other syndromes, reflected the overlapping of some of their clinical features with those of SRS and BWS. The analysis of IG allele-specific expression profile was performed in a subgroup of 18 patients with growth restriction. Thirty IGs were initially selected, based on their association with cell proliferation, fetal growth or neurodevelopment. Transcribed SNPs with high frequency in the general population were selected for the genotyping of patients, parents and control subjects, determination of IG expression in peripheral blood, and of the monoallelic or biallelic expression pattern. The allele-specific expression of 13 IGs expressed in blood was then investigated in patients (seven of them by PSQ and six by Sanger sequencing). Expression alterations of two normally paternally expressed genes were detected in 4/18 patients (22%). This study is the first to use pyrosequencing and Sanger sequencing in the evaluation of IG allele-specific expression profile, in patients with growth restriction. Despite the limitations, both techniques have proved to be robust, and revealed interesting alterations in allelic expression; however, the causal relationship of these alterations with the clinical phenotypes remained unclear. The investigation of the global DNA methylation was performed in a subgroup of 21 patients with growth restriction, and in 24 control subjects. Two types of analysis were performed: (1) group differential analysis, and (2) individual differential analysis. In the first analysis, the methylation pattern obtained for the group of patients with the diagnostic hypothesis of SRS (n=16) was compared to that of the control group (n=24); no bias towards DNA hypo or hypermethylation was detected in the SRS group. In the second analysis, the methylation patterns of each of the 21 patients with growth restriction, and each of the 24 control subjects were compared to the methylation pattern of the control group. The average numbers of hypermethylated CpGs and of differentially methylated segments (DMSs) were significantly higher in the patients. In total, 82 hypermethylated DMSs - 57 associated with gene(s) (69.5%), in 16 patients, and 51 hypomethylated DMSs - 41 associated with gene(s) (80.4%), in 10 patients, were identified. Gene ontology analysis of the 61 DMS-associated genes highlighted genes involved in development and morphogenesis of the skeletal system and fetal organs, and also in the regulation of gene transcription and metabolic processes. Methylation changes in genes involved with cellular proliferation and differentiation, and growth were identified in 9/20 patients (45%), suggesting clinical implications; an epigenetic mutation common to SRS patients was not detected, likely due to the clinical heterogeneity of the cohort. The data generated by this global methylation analysis, using microarray, might contribute to the understanding of molecular mechanisms in growth restriction. In an apparently balanced translocation -t(5;6)(q35.2;p22.3)dn, detected in a patient with the diagnostic hypothesis of SRS, a gene found to be disrupted by the chromosome 6 breakpoint might explain the phenotype; alternatively, the translocation might have impacted the regulation of developmental genes in the vicinity of breakpoints. Expression analysis showed a significant decrease in the disrupted gene cDNA levels in the patient\'s blood cells, as expected. In addition to the SRS typical signs, the patient presented clinical features suggestive of cleidocranial dysplasia. Thus, the translocation t(5;6) might have altered the interaction of developmental genes and regulatory elements, leading to misregulation of spatiotemporal gene expression, thus resulting in an atypical phenotype, with overlapping features of more than one genetic syndrome