Allopolyploids are formed by the combination of evolutionarily-diverged genomes, the union of which leads to dynamic changes in gene expression and genome organization. Expression patterns of orthologous genes are rapidly and stochastically established in newly created allotetraploids, where gene silencing is maintained by microRNAs, DNA methylation, and other chromatin modifications. Among them, DNA methylation has been known as an important mechanism of epigenetic regulation of gene expression and chromatin structure. However, it is unclear how DNA methylation affects genome-wide expression of homoeologous genes in the natural polyploid Arabidopsis suecica that contains genome of both A. thaliana and A. arenosa. To understand the role that DNA methylation plays in the polyploidization process, a comparative analysis was performed comparing up- or down-regulated genes in met1-RNAi A. suecica lines with the non-additively expressed genes in the synthetic allotetraploids, i.e., different from the mid-parent value. The previous studies indicated that decreased DNA methylation in A. suecica induces A. arenosa-specific demethylation in centromere regions and differentially alters expression of >200 genes encoding many transposons, unknown proteins and some other functional proteins that are located along chromosomes, whereas >1,300 non-additively expressed genes in the synthetic allotetraploids are distributed randomly along the chromosomes and encode various proteins in metabolism, energy, cellular biogenesis, cell defense and aging, and hormonal regulation. The origins of the progenitors of the genes whose expressions are altered in both met1-RNAi A. suecica and resynthesized allotetraploid were analyzed with single strand conformation polymorphism (SSCP) analysis. Reactivated genes in met1-RNAi A. suecica lines were predominately derived from the A. thaliana genome in euchromatic regions, whereas the suppressed genes were mainly derived from the A. Arenosa genome, indicating that changes in DNA methylation are genome-sensitive. The data suggest that allotetraploids incidentally display chromosome-specific changes and genomedependent regulation of homoeologous genes in response to DNA methylation perturbations.

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第20956号 / 理博第4408号 / 新制||理||1633(附属図書館) / 京都大学大学院理学研究科生物科学専攻 / (主査)准教授 渡辺 勝敏, 教授 曽田 貞滋, 教授 中川 尚史 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

Allopolyploids form through the hybridization of two or more diploid genomes. A challenge to reproduction in allopolyploids is that pairing can occur between homologous chromosomes or homeologous chromosomes (i.e.different subgenomes.). Crossover between homeologous chromosomes can result in chromosome rearrangements that lower fertility and overall fitness. Rearrangements can alter the dosage of either entire chromosomes or just parts of chromosomes. Understanding the frequency and extent of rearrangements will help to explain the evolution and genome stabilization of agriculturally important allopolyploid species. Pyrosequencing is a useful tool in the study dosage changes in allopolyploids because it allows quantification of the relative contribution from each progenitor species at any given locus. Here we use pyrosequencing to analyze resynthesized Brassica napus allopolyploids and their progeny. Targets for pyrosequencing were identified using a bioinformatic approach taking advantage of recently-released Brassica genome sequence. SNPs identified through bioinformatics were confirmed through molecular biology. Markers along the A3/C3 homeolog pair were used to identify the occurrence of novel homeologous exchanges during meiosis in the parent plant, and segregation patterns arising from dosage changes in the parent. We identify a higher frequency of homeologous rearrangements at the distal end of the chromosomes. We also observe that the presence of a dosage change in a parent increases the likelihood that the chromosome bearing the dosage change will undergo subsequent rearrangements in neighboring loci.

<p>Endopolyploid cells (hereafter - polyploid cells), which contain whole genome duplications in an otherwise diploid organism, play vital roles in development and physiology of diverse organs such as our heart and liver. Polyploidy is also observed with high frequency in many tumors, and division of such cells frequently creates aneuploidy (chromosomal imbalances), a hallmark of cancer. Despite its frequent occurrence and association with aneuploidy, little is known about the specific role that polyploidy plays in diverse contexts. Using a new model tissue, the Drosophila rectal papilla, we sought to uncover connections between polyploidy and aneuploidy during organ development. Our lab previously discovered that the papillar cells of the Drosophila hindgut undergo developmentally programmed polyploid cell divisions, and that these polyploid cell divisions are highly error-prone. Time-lapse studies of polyploid mitosis revealed that the papillar cells undergo a high percentage of tripolar anaphase, which causes extreme aneuploidy. Despite this massive chromosome imbalance, we found the tripolar daughter cells are viable and support normal organ development and function, suggesting acquiring extra genome sets enables a cell to tolerate the genomic alterations incurred by aneuploidy. We further extended these findings by seeking mechanisms by which the papillar cells tolerated this resultant aneuploidy.</p> / Dissertation

A cana-de-açúcar (Saccharum ssp.) é uma cultura valiosa na produção de alimento, fibra e energia para o Brasil e, especialmente, para o estado de São Paulo. Com o advento da biotecnologia, alternativas de melhoramento genético têm despertado a atenção da comunidade científica, sendo etapas cruciais para tais avanços o sequenciamento e a caracterização do genoma das espécies cultivadas. Dada sua natureza poliploide, com frequente aneuploidia, a cana-de-açúcar apresenta dificuldades às práticas convencionais em genômica, de maneira que é vantajoso fazer uso de recursos de sequenciamento de nova geração e de espécies próximas para elucidar de forma mais efetiva o genoma da gramínea. Uma contribuição interessante, nesse sentido, é a caracterização funcional de polimorfismos genéticos existentes entre genótipos do gênero Saccharum, auxiliando investigações relacionadas à genômica de poliploides complexos, desenvolvendo um recurso a ser utilizado futuramente por melhoristas. Esse trabalho realizou a caracterização da variabilidade genômica a partir de dados genotípicos de indivíduos do Painel Brasileiro de Genótipos de Cana-de-Açúcar, obtidos via genotipagem-por-sequenciamento, utilizando como referência o genoma já sequenciado do sorgo. Os sítios variantes (sobretudo polimorfismos de nucleotídeo único) foram detectados com o software FreeBayes e suas possíveis funções e posições foram anotadas com o programa SnpEff. Utilizaram-se estatísticas de genética de populações, como a frequência alélica para várias classes de polimorfimo, o Teste de McDonald & Kreitman (busca de evidêcias de evolução adaptativa) e a heterozigosidade combinada (busca de regiões genômicas com assinatura de seleção), de modo a identificar regiões genômicas potencialmente envolvidas em eventos evolutivos. Os resultados demonstraram a perda de variabilidade entre os genótipos melhorados em relação aos ancestrais, com evidências de assinaturas de seleção, envolvendo questões sensíveis ao funcionamento da maquinaria celular (como respiração e fotossíntese) e a características valoradas para a cultura (destacando-se a resistência a patógenos e a biossíntese da sacarose). Tais indícios fornecem subsídios à compreensão do genoma e ao melhoramento genético desse poliploide. / Sugarcane (Saccharum ssp.) is a valuable crop for food, fiber and energy production in Brazil, especially to the São Paulo State. With the advent of biotechnology, alternatives to breeding have enticed attention of the scientific community, with genome sequencing and characterization being crucial steps to these advances. Because sugarcane is polyploid, with frequent aneuploidy, it presents difficulties to the application of standard practices in genomics, such that it is advantageous to make use of next generation sequencing alternatives and resources from related species to more effectively elucidate the genome of this grass. Thus, an interesting contribution is the functional characterization of genetic polymorphisms from the Saccharum genus, aiding investigations related to genomics of complex polyploids, developing a resource to be used in the future by breeders. Our goal was to perform this characterization with genotypic data from individuals of the Brazilian Panel of Sugarcane Genotypes, obtained by genotyping-by-sequencing (GBS), using as reference the previously sequenced sorghum genome. We called the variants (mainly single nucleotide polymorphisms) with FreeBayes and annotated their functions and positions with SnpEff. We used population genetics statistics, such as the allele frequency, the McDonald & Kreitman Test and the pooled heterozygosity, to identify genomic regions potentially involved in evolutionary events. The results showed a loss of variability between bred genotypes in relation to the ancestors, with evidences of selective sweeps, involving regions related to the cellular machinery (such as respiration and photosynthesis) and specific crop traits (especially disease resistance and sucrose biosynthesis). These results support understanding of the genome and breeding efforts in this polyploid grass.

Crops, such as sugarcane, potato, and several forages and berries, are autopolyploids. In plant breeding studies, a key subject is the pairwise relatedness between cultivars. This information can be incorporated in GWAS (genome-wide association) and GS (genomic selection) studies allowing powerful genetic analysis. Despite the autopolyploid importance in agriculture, they lack genetic studies considering ploidy level, and, therefore, polyssomic inheritance. Our objective in this work is: i) to investigate, through simulations, different molecular pairwise relatedness estimators; ii) to recommend which one is the best to be used in different scenarios regarding characteristic as number of loci, number of alleles, ploidy level, double-reduction, and inbreeding level. Our recommendations may guide autopolyploid breeding programs will be able to choose the best strategies and estimators based their reality. / Culturas agrícolas como cana-de-açúcar, batata, batata doce, diferentes forrageiras e frutas são autopoliploides. Em melhoramento de plantas, o conhecimento do parentesco entre cultivares é crítico. Essa informação pode ser incorporada em estudos de GWAS (genome-wide association studies) e GS (genomic selection) permitindo uma análise genética e predição de QTLs e valores genéticos de maior poder estatístico. No entanto, faltam estudos genéticos que considerem parentesco em um cenário autopoliploide. Nossos objetivos neste trabalho são: i) investigar, através de simulações, diferentes estimadores de parentesco entre indivíduos; ii) recomendar o melhor estimador basedo em cenários específicos considerando diferentes número de loci, número de alelos, ploidia, nível de dupla-redução e endogamia. Nossas recomendações poderão orientar estratégias de programas de melhoramento genético de espécies autopoliploides para escolha de estimadores baseados em seu cenário específico.

Although polyploidy is believed to be a major source of evolutionary novelty, few studies have examined the genetic basis of phenotypic variation in wild polyploids. In this thesis I have studied the genetic basis of flowering time variation in the wild tetraploid crucifer Capsella bursa-pastoris, as well as the evolutionary history of this species. First, phylogenetic methods were employed to test hypotheses on the origin of C. bursa-pastoris. Based on DNA sequences from two chloroplast DNA loci and three independent nuclear genes, we found no support for the notion of C. bursa-pastoris as an autopolyploid of the diploid C. grandiflora, or an allopolyploid of C. grandiflora and C. rubella, even though some C. bursa-pastoris accessions shared alleles with C. rubella at nuclear loci. Using divergence population genetic methods, a larger sample of accessions and data for six duplicated nuclear genes, we found that allele sharing in sympatry was better explained by introgressive hybridization than by multiple origins of the tetraploid. The genetic basis of flowering time variation was examined using three approaches. A gene expression microarray study revealed that early- and late-flowering accessions differ in circadian rhythm, as well as in the gibberellin pathway affecting flowering time. Second, two QTL (Quantitative Trait Loci) for flowering time map to duplicated linkage groups. Third, polymorphisms at the candidate genes CRYPTOCHROME1 (CRY1), in one of the QTL regions, and FLOWERING LOCUS C (FLC) are associated with natural flowering time variation. Different FLC splice site polymorphisms are associated with flowering time in samples from Western Eurasia and China. The CRY1 association is only found in Europe, where alleles introgressed from C. rubella have an effect on flowering time. In conclusion, duplicated genes, introgressive hybridization and splicing variation may all have played a role in the evolution of flowering time variation in C. bursa-pastoris.

Hybridization is an important mechanism in diversification. It often makes taxonomy difficult. Lack of strong supported intrageneric classification in genus Pteris (Pteridaceae) could be caused by natural hybridization. Most hybridization documnted in Pteris was based on limited evidence. This study focuses on Pteris cadieri complex, the taxon with putative hybridization. The species complex displayed significant morphological variation and was associated with hybrid origin. Reproductive biology revealed variation in spore number per sporangium, spore size, spore shape and apogamous reproduction, which imply its hybrid origin. Cytology analysis using chromosome counting and flow cytometry identified diploids, triploids, and tetraploids. CpDNA and nuclear DNA supported that Pteris cadieri complex is hybrid origin: paternal and maternal lineages were inferred and 11 taxa were identified. Furthermore, comparing materials form Hainan and Taiwan, , it is clear that the species complex is composed by taxa arisen from multiple hybridization. Systematic inconsistency existed between chloroplast and nuclear phylogenies in Pteris impled that other taxa might have involved in hybridization events, in addition to the Pteris cadieri complex. Hybridization may be very common in Pteris. To infer intrageneric taxonomy of Pteris, effect of reticulate evolution should never be neglected. Finally, based on morphological and evolutionary traits, the taxonomy of Pteris cadieri complex is revised. There are Pteris cadieri Christ, Pteris dimorpha Copel. var. dimorpha, Pteris dimorpha var. plumbea (Christ) Y.-S. Chao, H.-Y. Liu & W.-L. Chiou, Pteris grevilleana Wall. ex Agardh var. grevilleanan, Pteris grevilleana Wall. ex Agardh var. ornata Alderw., and Pteris hainanensis Ching.

Ferns are one of the three main clades of vascular plants. They have few easily studied morphological characters, reflected in a historically unstable classification. The fern genus Woodsia is known to have a complex evolutionary history including numerous polyploid taxa and hybrids. It is a cosmopolitan group of small rock loving ferns mainly found in montane areas. This thesis aims at analyzing the patterns of diploid and polyploid evolution in Woodsia and to resolve and classify the relationships of Woodsiaceae and the other families in the large fern clade Eupolypods II. The Eupolypods II family relationships were inferred with DNA sequences from 81 specimens representing all major lineages. This resulted in the first well supported phylogeny of this clade and revealed Woodsiaceae to be non-monophyletic. The genera previously placed in this family were reclassified into five new or resurrected families. Swedish fern genera that have changed family classification are Woodsia (hällebräknar), now in the monogeneric family Woodsiaceae, Athyrium (majbräknar), now  in Athyriaceeae and Cystopteris (stenbräknar) and Gymnocarpium (ekbräknar) now in Cystopteridaceae. To analyze the evolution of Woodsia, phylogenies were produced from five plastid and two nuclear regions sequenced from 188 specimens. The results show that most taxa in Woodsia are polyploid. Polyploidization is the most common mode of speciation in the genus with an estimated polyploid speciation rate of 54%. The polyploids are mostly young and many of the polyploid taxa seem to have formed multiple times. The results also address several taxonomic and biogeographic questions. In the process of the work we made methodological advancements and developed 20 new low copy nuclear marker regions as well as a software pipeline for finding primers in transcriptome datasets. The alignment editor software AliView was developed for handling the increasing size datasets in a user friendly way. In conclusion this thesis provides new insights into the complexities of the evolution of a fern genus in which much of the diversity is accommodated in young species formed through polyploidization. It provides a framework of phylogenetic relationships at different levels that both answers long standing questions and generates new ones.

Submitted by Cláudia Bueno (claudiamoura18@gmail.com) on 2015-10-20T16:00:07Z No. of bitstreams: 2 Dissertação - Daniel Garcia Silva - 2012.pdf: 2191471 bytes, checksum: 68bc7d63efc7307ce6b62a63489d372d (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2015-10-21T10:02:07Z (GMT) No. of bitstreams: 2 Dissertação - Daniel Garcia Silva - 2012.pdf: 2191471 bytes, checksum: 68bc7d63efc7307ce6b62a63489d372d (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2015-10-21T10:02:07Z (GMT). No. of bitstreams: 2 Dissertação - Daniel Garcia Silva - 2012.pdf: 2191471 bytes, checksum: 68bc7d63efc7307ce6b62a63489d372d (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2012-06-28 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Sugarcane is an important crop, cultivated in more than 90 countries, occupying an area of approximately 20 million of hectares. Modern varieties (Saccharum spp.) are highly heterozygous interspecific hybrids, polyploids and often aneuploids, with chromosome numbers between 100 and 130. Such characteristics explain the common opinion that the genome of sugarcane is the most complex among cultivated species, posing a challenge to breeding programs. As a contribution to the understanding of this complex genomic architecture, this study aimed to build the first linkage maps using exclusively DArT markers in sugarcane. The maps were built using a progeny derived from the cross between varieties largely used in the Brazilian breeding program of RIDESA (RB97327 x RB72454). The initial mapping population comprised 186 individuals. Total genomic DNA was extracted from axial buds, following the protocol of Al-Janabi et al. (1999). Using the DArT P/L core facility to generate DArT data, a total of 7680 markers were analyzed, of which 850 were polymorphic. The analysis of segregation patterns in the progeny revealed that 47% of the individuals in the progeny were in fact derived from selfing of the female parent RB97327. These individuals were analyzed as a distinct generation. Linkage analyses were then performed on two populations (from selfing and crossing) separately. The software OneMap was used to construct the maps. The established linkage criteria for linkage analysis were LOD-score ≥ 3.5 and recombination fraction ≤ 0.4. In the first map, built using data from individuals originated from selfing, from 850 polymorphic markers, 392 markers (segregating in a 3:1 manner) were used to create 80 linkage groups related to the variety RB97327. For the population derived from the biparental crossing, four linkage maps were built: an integrated map composed of 98 linkage groups including 632 markers (1:1 and 3:1); an integrated framework map, using a more conservative ordering criteria for the linkage groups, which was composed of 94 linkage groups; and two other linkage maps, one for each parent (RB97327 and RB72454), built to estimate the genome size of the varieties involved in this study. The total length of the linkage map built using data from individuals derived from selfing of the variety RB97327 was 828 cM. The total length of the integrated linkage map was 2848 cM. The lengths of the maps built for each parent, using data from individuals derived from crossing, were 1465 cM (RB97327) and 1976 cM (RB72454). Using the methodology of Hulbert et al. (1988), the estimated genome sizes for these varieties were 2811 cM e 3471 cM, respectively. The maps obtained in these cases covered a low percentage of the estimated genome sizes (52% and 57%). In spite of the low polymorphism, DArT markers showed to be an efficient technique to perform genotyping of sugarcane. Hundreds of polymorphic markers were generated in only one assay, using two methods of genome complexity reduction. These markers represent a new tool for genetic studies in sugarcane, especially if the low cost (USD/marker) involved in data production is considered. / A cana-de-açúcar é uma importante cultura, cultivada em mais de 90 países, ocupando uma área total de aproximadamente 20 milhões de hectares. As variedades modernas (Saccharum spp.) são híbridos interespecíficos altamente heterozigóticos, poliploides e frequentemente aneuploides, com número cromossômico variando de 100 a 130. Tais características proporcionaram ao genoma da cana-de-açúcar o título de mais complexo entre as espécies cultivadas, o que representa um desafio para os programas de melhoramento genético da cultura. No intuito de contribuir com dados que auxiliem na compreensão dessa complexa arquitetura genômica, o presente estudo objetivou a construção dos primeiros mapas de ligação para cana-de-açúcar utilizando exclusivamente marcadores DArT, avaliados na progênie derivada do cruzamento de variedades amplamente utilizadas nos programas de melhoramento da RIDESA (RB97327 x RB72454). A população inicial de mapeamento foi composta por 186 indivíduos. O DNA genômico foi extraído de gemas axiais, seguindo o protocolo proposto por Al-Janabi et al. (1999). Após a extração, quantificação e homogeneização da concentração de DNA das amostras, o material foi enviado para a empresa DArT P/L para a geração dos marcadores DArT. Um total de 7680 locos foi analisado, dos quais 850 se apresentaram polimórficos. A análise dos padrões de segregação obtidos na progênie revelou que 47% dos indivíduos da progênie avaliada foram provenientes de autofecundação do genitor feminino RB97327. Os indivíduos identificados como provenientes de autofecundação foram analisados como uma geração distinta. As análises de ligação foram realizadas nas duas populações separadamente. O software OneMap foi utilizado para a construção dos mapas. Os critérios estabelecidos para proceder com as análises de ligação foram LOD-score ≥ 3,5 e fração de recombinação ≤ 0,4. No primeiro mapa, originário da população de autofecundação, dos 850 marcadores polimórficos, 392 marcadores com segregação 3:1 foram utilizados para originar 80 grupos de ligação referentes à variedade RB97327. Para a população derivada do cruzamento biparental foram construídos quatro mapas de ligação: um mapa integrado composto por 98 grupos de ligação a partir da análise de 632 marcadores (com segregações 1:1 e 3:1); um mapa framework integrado, construído a partir de uma ordenação mais refinada dos marcadores dentro de cada um dos grupos de ligação, o qual foi composto por 94 grupos de ligação; e, com o objetivo de se estimar o tamanho do genoma das variedades envolvidas neste estudo, dois mapas de ligação, um para cada genitor (RB97327 e RB72454). O comprimento total do primeiro mapa, referente à variedade RB97327, foi de 828cM. O comprimento total do mapa integrado foi de 2848 cM. Os comprimentos totais dos mapas obtidos para cada um dos genitores, gerados a partir de dados da população de cruzamento biparental, foram de 1465Cm (RB97327) e de 1976 cM (RB72454). Utilizando a metodologia de Hulbert et al. (1988), os tamanhos estimados dos genomas das variedades RB97327 e RB72454 foram 2811 cM e 3471 cM, respectivamente. Assim, pode-se afirmar que os mapas obtidos neste caso apresentaram baixa cobertura (52% e 57%), perante o tamanho estimado dos genomas. Apesar do baixo polimorfismo, os marcadores DArT se mostraram eficientes na genotipagem de progênies de cana-de-açúcar, pois, centenas de marcas polimórficas foram geradas em apenas um ensaio, com dois métodos de redução de complexidade. Estes marcadores representam uma nova ferramenta para o desenvolvimento de estudos genéticos em cana-de-açúcar, principalmente se considerado o baixo custo (R$/marcador) envolvido na obtenção dos genótipos.