Este estudo avaliou os efeitos de extratos de culturas líquidas de Leifsonia xyli subsp. xyli (Lxx) cultivada sob condições de estresse osmótico e na presença de inibidores da rota não-mevalonato (MEP/DOXP) de síntese de pigmentos isoprenóides, na germinação de sementes de alface. Também avaliou a relação entre o efeito tóxico dos extratos e a produção de pigmentos pela bactéria. Os extratos foram obtidos do sobrenadante de culturas de Lxx usando acetato de etila como solvente. Ensaios com sementes de alface embebidas nos extratos indicaram uma ação inibitória na germinação das sementes e no alongamento das radículas quando a bactéria foi cultivada na presença de 7% de PEG 6000 ou 100 mM de NaCl. O extrato foi submetido a vários tratamentos térmicos (temperatura ambiente; 30, 60 e 90oC em banho-maria e 121oC por autoclavagem) e sua ação inibitória em sementes mostrou-se termoestável. A presença de pigmentos isoprenóides de células cultivadas com PEG ou NaCl e de células incubadas sem estes agentes estressantes foi avaliada por espectrofotometria após extração dos pigmentos com metanol. Em células cultivadas na ausência de estresse, notou-se um pico de absorbância bem definido no comprimento de onda de 400 nm, ao passo que este pico foi sensivelmente reduzido em células cultivadas sob estresse. Esse efeito foi mais pronunciado na presença de PEG. Para confirmar a natureza isoprenóide do composto, Lxx foi cultivada na presença de fosmidomicina ou difenilamina, que são inibidores da rota metabólica não-mevalonato destes compostos em Microbacteriaceae. Ambas as substâncias reduziram o pico de absorbância no espectro em relação a células cultivadas sem adição destes antibióticos, confirmando a natureza do composto e sua via de síntese. No entanto, a redução foi mais acentuada no caso da fosmidomicina. Por fim, a fosmidomicina foi utilizada com a finalidade de verificar seu efeito na toxidez dos extratos. Foi observado que o efeito tóxico do extrato foi reduzido quando Lxx foi cultivada na presença do inibidor. Ao se avaliar o conteúdo das células, notou-se também redução expressiva no conteúdo relativo de compostos isoprenóides, como esperado. Os resultados indicaram que a bactéria secreta um produto cuja ação é similar à do ABA e que há uma correlação entre a produção do mesmo e a síntese de pigmentos isoprenóides. Em conjunto, o estudo dá suporte à hipótese levantada anteriormente por Monteiro- Vitorello et al. (2004) com base em análise in silico do genoma de Lxx, segundo a qual a bactéria estaria apta a secretar um composto análogo ao ácido abscísico (ABA). / This study evaluated the effects of extracts of liquid cultures of Leifsonia xyli subsp. xyli (Lxx) grown under osmotic stress and in the presence of inhibitors of the non-mevalonate pathway (MEP/DOXP) of synthesis of isoprenoid pigments on the germination of lettuce seeds. The relationship between the extracts toxic effect and the production of pigments was evaluated as well. The extracts were obtained from the supernatant of Lxx cultures using ethyl acetate as a solvent. Essays with lettuce seeds soaked in the extracts indicated an inhibitory action on germination and on the elongation of the radicles when the bacterium was cultivated in the presence of 7% PEG 6000 or 100 mM NaCl. The extract was submitted to different thermal treatments (room temperature; 30, 60 and 90oC water bath and 121oC autoclaving) and its inhibitory action was shown to be thermostable. The presence of isoprenoid pigments in cells cultivated with PEG, NaCl or in the absence of these stressing agents was evaluated by spectrophotometry after extracting the pigments with methanol. In cells incubated in the absence of stress, a well-defined absorbance peak at 400 nm wavelength was noted, whereas this peak was sensibly reduced in cells cultivated under stress. This effect was more pronounced in the presence of PEG. To confirm the isoprenoid nature of the compound, Lxx was incubated in the presence of fosmidomycin and diphenylamine, which are inhibitors of the nonmevalonate metabolic pathway of these compounds in Microbacteriaceae. Both substances reduced the absorbance peak in the spectrum compared to cells cultivated in the absence of these antibiotics, confirming the nature of the compound and its synthesis pathway. Moreover, the peak reduction was more accentuated in the presence of fosmidomycin. Lastly, fosmidomycin was used to evaluate its effect on the toxicity of the extracts. It was observed that the toxic effect of the extract was reduced when Lxx was cultivated in the presence of this inhibitor. When the cell content was evaluated, an expressive reduction on the relative content of carotenoid compounds has also been noticed, as expected. The results indicated that the bacterium secretes a compound whose action is similar to ABA and that there is a correlation between the production of this compound and the synthesis of isoprenoid pigments. The results support the hypothesis proposed earlier by Monteiro-Vitorello et al. (2004) based upon an in silico analysis of the genome of Lxx, according to which the bacterium would be able to secrete a compound analogous to abscisic acid (ABA)

Orientador: Juliana Regina Peiró / Coorientador:Sérgio Moraes Aoki / Banca:José Paes de Oliveira Filho / Banca:Flávia de Almeida Lucas / Resumo: Endotoxemia é distúrbio sistêmico que se origina da resposta do hospedeiro a um componente da membrana celular das bactérias Gram- negativas. Esta resposta se dá através da exposição dos receptores celulares TLR-4 e TLR-2 ao LPS. Os objetivos deste estudo foram investigar as alterações na expressão gênica da exposição ao LPS em leucócitos de equinos utilizando a técnica de microarray, avaliar a eficiência do modelo ex vivo para os estudos envolvendo a endotoxemia pela mesma técnica, avaliar a expressão global de genes em vias envolvidas, identificar componentes da cascata metabólica com potenciais para novas terapias, e fornecer subsídio para futuros estudos. Amostras de sangue total (15mL) de cavalos saudáveis (n=6) foram incubadas durante 4 horas a 37°C com 5% de CO2 na presença (1 ou 10 ng/mL) ou ausência de LPS (0 ng/mL). Alíquotas de 500µL de sangue foram coletadas nos momentos 0, 2 e 4 horas após o estímulo do LPS. O RNA, extraído das mostras, foi utilizado para a transcrição do cDNA. A hibridização do cDNA marcado com Cy-3 foi realizada em lâminas 4x44K v2 de humanos contendo sequências homólogas com a espécie equina para os genes de interesse. Após a leitura das lâminas, com filtro para genes 30x mais ou menos expressos, verificou-se um aumento da expressão do TLR-2 na concentração de 10 ng LPS/mL no momento 4 em relação ao momento 2. Este resultado sugere que, embora o receptor TLR-4 seja o principal receptor no reconhecimento do LPS, o receptor TLR-2 também tem um papel no reconhecimento destas moléculas / Abstract: Endotoxemia is systemic disturbance that origins from the host response to a component of the cellular membrane of Gram-negative bacterias. This response occurs through exposure of cellular receptors TLR-4 and TLR-2 to LPS. The objectives of this study were to investigate changes in gene expression of exposure to LPS in horses leukocytes using the microarray technique, to evaluate the efficiency of the ex vivo model for studies of endotoxemia by the same technique, to evaluate the global expression of genes in the metabolic pathways involved, identify components of the metabolic cascade with potential for new therapies, and provide allowance for future studies. Whole blood samples (15mL) of healthy horses (n=6) were incubated for 4 hours at 37°C with 5% of CO2 in the presence (1 or 10 ng/ml) or absence of LPS (0 ng/ml). Aliquots of 500μL of blood were collected at 0,2 and 4 hours after LPS stimulation. The RNAs, extracted from the samples, were used for the transcription of the cDNAs. Hybridization of labeled cDNAs with Cy-3 were performed in 4x44K v2 slides containing human sequences homologous to the equine species for the genes of interest. After the reading of the slides with filter for genes 30x up regulation or down regulation expression, it was observed an increased expression of the TLR-2 concentration of 10 ng LPS/mL at time 4 compared to time 2. This result suggests that, although the TLR-4 receptor is the main LPS recognition receptor, the receptor TLR-2 also plays a role in recognition of these molecules / Mestre

Submitted by Cleide Dantas (cleidedantas@ufpa.br) on 2014-02-10T11:41:12Z No. of bitstreams: 2 license_rdf: 23898 bytes, checksum: e363e809996cf46ada20da1accfcd9c7 (MD5) Dissertacao_BasesMolecularesResistencia.pdf: 941177 bytes, checksum: ee3acd908afddff5f4d0f8ea3e686640 (MD5) / Approved for entry into archive by Ana Rosa Silva (arosa@ufpa.br) on 2014-03-31T16:49:02Z (GMT) No. of bitstreams: 2 Dissertacao_BasesMolecularesResistencia.pdf: 941177 bytes, checksum: ee3acd908afddff5f4d0f8ea3e686640 (MD5) license_rdf: 23898 bytes, checksum: e363e809996cf46ada20da1accfcd9c7 (MD5) / Made available in DSpace on 2014-03-31T16:49:02Z (GMT). No. of bitstreams: 2 Dissertacao_BasesMolecularesResistencia.pdf: 941177 bytes, checksum: ee3acd908afddff5f4d0f8ea3e686640 (MD5) license_rdf: 23898 bytes, checksum: e363e809996cf46ada20da1accfcd9c7 (MD5) Previous issue date: 2004 / Mercúrio é um dos elementos mais tóxicos tanto para seres humanos como os demais animais. Em ambientes naturais seus compostos podem ter origem em fontes naturais ou em decorrência da ação do homem. A atividade microbiana tem papel crítico na biorremediação. A resistência ao mercúrio nas bactérias está associada a um conjunto de genes organizados no operon mer. Considerando este cenário realizamos um estudo com bactérias Gram-negativas isoladas de sedimento de rios de duas áreas com distintos graus de atividade antropogênica, Barreiras e Caxiuanã. A resistência ao mercúrio foi avaliada em ensaios de crescimento in vitro da bactéria em meio contendo Hg. A presença do operon mer foi determinada por PCR para os genes RTPCABD, componentes do operon. O ensaio in vitro, determinou que apenas 2 isolados dos 107 avaliados, Acinetobacter baumannii de Caxiuanã e Pseudomonas stutzeri de Barreiras, apresentavam resistência ao Hg. Um operon mer foi identificado e caracterizado apenas do isolado Acinetobacter baumannii. Este apresenta os genes RTPCAD e sua organização e seqüência nucleotídica possui identidade total com o operon mer de um isolado de Acinetobacter calcoaceticus da Rússia. / Mercury is one of the most toxic elements, for both human and animals. Its compounds in natural environments can arise from natural and anthropogenic sources. Microbial activities play a critical rule in bioremediation. Mercury resistance in bacteria is associated with a gene cluster present on mer operon. Considering that picture we carried on a study with Gram-negative bacteria isolates from freshwater sediment from two areas with distinct anthropogenic impact, Barreiras and Caxiuanã. Resistance to Hg was evaluated by in vitro growing in medium containing Hg. The presence of mer operon was evaluated by PCR targeting RTPCABD mer gene sequences. The in vitro assay revealed that only two strains, from 107 isolates, were resistant to Hg the bacteria Acinetobacter baumannii from Caxiuanã and a Pseudomonas stutzeri from Barreiras. A mer operon was identified in the Acinetobacter baumannii strain and its sequencing revealed a organization as merRTPCAD and a complete nucleotide identity with a mer operon identified in a Acinetobacter calcoaceticus from Russia.

Submitted by Jorge Silva (jorgelmsilva@ufrrj.br) on 2017-05-04T18:50:24Z No. of bitstreams: 1 2016 - Fernando Lima Aires Gon?alves.pdf: 4917479 bytes, checksum: d7c7e6d879e9495bcc0db5d59f03c813 (MD5) / Made available in DSpace on 2017-05-04T18:50:24Z (GMT). No. of bitstreams: 1 2016 - Fernando Lima Aires Gon?alves.pdf: 4917479 bytes, checksum: d7c7e6d879e9495bcc0db5d59f03c813 (MD5) Previous issue date: 2016-02-26 / Three studies were conducted to evaluate the efficiency of seedling crowning using cardboard in reforestation with arboreal species of the Atlantic. The first study aimed to test the effect of the crowning with cardboard on the growth of Atlantic Forest species planted in a pasture area with dominance of Andropogon bicornis. The experimental design was randomized blocks with two treatments and 12 repetitions. Treatments consisted of crowning with cardboard or crowning with hoe applied to 11 tree species. We evaluated the growth of plants by measuring the height (H) and diameter of the collar height (DAC) at 6, 12 and 18 months after planting beyond survival rate. It was also carried out an assessment of the costs of each treatment. The results showed that the board did not interfere significantly in any forest species assessed. However, the survival rate was higher in the treatment crown of cardboard (80.7%) compared to the capping treatment with hoe (73.1%). In the economic aspect, the crowning with cardboard obtained material cost and labor up to 50% lower than the crowning with hoe. In the second study we evaluated the crowning efficiency with cardboard in the suppression of four species of grasses commonly found in forest areas in the Atlantic Forest. Four experiments were set up in places with dominant species A. bicornis, Urochloa decumbens, Urochloa umidicola and Panicum maximum. The experimental design was completely randomized with five replications. The treatments were the culmination of cardboard and without crowning. The cardboard efficiency was monitored by sampling the total biomass and the level of senescence of grasses in the area crowned. The cardboard was efficient in suppressing the four species of grasses evaluated, because after the evaluation period (100- 170 days) had hardly live plant material in the field of crowns. In the third study was rated as the crowning with cardboard influences the temperature and moisture in the surface layers in the soil. The experimental design was a randomized complete block design with four treatments and eight repetitions. The treatments were 1 - crown with cardboard (cardboard disposed directly on the braqui?ria); 2 - no crowning; 3 - with cardboard crown after crown with hoe (willing cardboard on the exposed soil); 4 - crowning with hoe. The moisture in the layers from 0 to 5 cm and 5 to 10 cm of soil, and the temperature in the layers 0 and 2,5 cm, 2.5 to 5 cm and 5 to 10 cm of the soil were monitored during periods 19 and subsequent 7 days, respectively, at a rain simulation 130 mm for 24 hours and natural rainfall of 133 mm for 13 days. The results indicated that treatment 1 (crowning with cardboard) decreased soil drying speed while maintaining higher water content in the crown area compared to the other treatments. Both capping treatments showed cardboard soil temperature in the region of the crown to 8.5?C lower than the capping treatment with cardboard. The set of results of the three studies demonstrate that the use of the capping technique with cardboard in reforestation deployment can significantly reduce operating costs after planting, either by reducing hand labor used in capping operations, as the lower need for seedlings for replanting. / Foram conduzidos tr?s estudos objetivando avaliar a efici?ncia do coroamento de mudas com papel?o em reflorestamentos com esp?cies arb?reas da Mata Atl?ntica. O primeiro estudo teve como objetivo testar o efeito do coroamento com papel?o sobre o crescimento de esp?cies da Mata Atl?ntica plantadas em uma ?rea de pastagem com domin?ncia de Andropogon bicornis. O delineamento experimental foi em blocos casualizados com dois tratamentos e 12 repeti??es. Os tratamentos consistiram de coroamento com papel?o ou coroamento com enxada aplicados a 11 esp?cies arb?reas. Avaliou-se o crescimento das plantas atrav?s da mensura??o da altura (H) e di?metro a altura do coleto (DAC) aos 6, 12 e 18 meses ap?s o plantio al?m da taxa de sobreviv?ncia. Tamb?m foi realizada uma avalia??o dos custos de cada tratamento. Os resultados encontrados mostraram que o papel?o n?o interferiu de maneira significativa em nenhuma esp?cie florestal avaliada. Entretanto, a taxa de sobreviv?ncia foi maior no tratamento de coroamento com papel?o (80,7%) comparado ao tratamento de coroamento com enxada (73,1%). No aspecto econ?mico, o coroamento com papel?o obteve custo de material e m?o de obra at? 50% inferior ao do coroamento com enxada. No segundo estudo avaliou-se a efici?ncia do coroamento com papel?o na supress?o de quatro esp?cies de gram?neas forrageiras comumente encontradas em ?reas de reflorestamento no bioma Mata Atl?ntica. Foram montados quatro experimentos em s?tios com domin?ncia das esp?cies A. bicornis, Urochloa decumbens, Urochloa umidicola e Panicum maximum. O delineamento experimental foi o inteiramente casualizado com cinco repeti??es. Os tratamentos foram o coroamento com papel?o e sem coroamento. A efici?ncia do papel?o foi monitorada atrav?s de coletas da biomassa total e do n?vel de senesc?ncia das gram?neas na ?rea coroada. O papel?o se mostrou eficiente na supress?o das quatro esp?cies de gram?neas avaliadas, pois ap?s o per?odo de avalia??o (100 a 170 dias) praticamente n?o havia material vegetal vivo na ?rea das coroas. No terceiro estudo foi avaliado como o coroamento com papel?o influencia na temperatura e umidade nas camadas superficiais no solo. O delineamento experimental foi o de blocos casualizados com quatro tratamentos e oito repeti??es. Os tratamentos foram 1 ? coroamento com papel?o (papel?o disposto diretamente sobre a braqui?ria); 2 ? sem coroamento; 3 ? coroamento com papel?o ap?s coroamento com enxada (papel?o disposto sobre o solo exposto); 4 ? coroamento com enxada. A umidade, nas camadas de 0 a 5 cm e 5 a 10 cm do solo, e a temperatura, nas camadas de 0 a 2,5 cm, 2,5 a 5 cm e 5 a 10 cm do solo, foram monitorados durante per?odos de 19 e 7 dias subsequentes, respectivamente, a uma simula??o de chuva de 130 mm durante 24 hs e a chuvas naturais de 133 mm durante 13 dias. Os resultados indicaram que o tratamento 1 (coroamento com papel?o) diminuiu a velocidade de ressecamento do solo, mantendo maior teor de ?gua na ?rea da coroa em compara??o aos demais tratamentos. Ambos os tratamentos de coroamento com papel?o apresentaram temperatura do solo na regi?o da coroa at? 8,5?C menor do que no tratamento de coroamento com papel?o. O conjunto de resultados dos tr?s estudos realizados demonstram que o uso da t?cnica de coroamento com papel?o na implanta??o de reflorestamentos pode reduzir significativamente os custos operacionais ap?s o plantio, tanto pela redu??o de m?o-de-obra empregada em opera??es de coroamento, quanto pela menor necessidade de replantio de mudas.

Le point de départ de cette thèse est un problème posé par le groupe électromagnétisme de EADS-CCR : comment résoudre plusieurs systèmes linéaires avec la même matrice mais différents seconds membres ? Pour l'application voulue, les matrices sont complexes, denses et de grande taille. Un problème standard comporte environ quelques millions d'inconnues. Comme de telles matrices ne peuvent être ni calculées, ni stockées dans un processus industriel, l'utilisation d'un produit matrice-vecteur approché est la seule alternative. En l'occurrence, le produit matrice-vecteur est effectué en utilisant la méthode multipôle rapide. Dans ce contexte, le but de cette thèse est d'adapter les méthodes itératives de type Krylov de telle sorte qu'elles traitent efficacement les nombreux seconds membres. Des travaux préliminaires avec un seul second membre ont montré que la méthode GMRES est particulièrement efficace et robuste pour cette application. En conséquence dans cette thèse nous abordons uniquement les variantes de GMRES. Les schémas d'orthogonalisation que nous avons implantés dans GMRES sont des variantes de l'algorithme de Gram-Schmidt. <br /><br />Dans une première partie, nous nous intéressons à l'influence des erreurs d'arrondi dans les algorithmes de Gram-Schmidt. Nos résultats répondent à des questions vieilles de vingt-cinq ans. Nous donnons l'explication théorique de ce qui était communément observé et accepté : <br /><br /> - l'algorithme de Gram-Schmidt modifié génère un ensemble de vecteurs bien conditionné ;<br /> - l'algorithme de Gram-Schmidt itéré deux fois fabrique un ensemble de vecteurs orthonormé.<br /><br />Ces deux propositions reposent sur l'hypothèse que la matrice de départ est "numériquement non singulière" en un sens qui est clairement défini. D'autre part, quand l'algorithme de Gram-Schmidt est itéré avec un critère de réorthogonalisation, nous proposons un nouveau critère. Nous montrons que l'algorithme obtenu est robuste alors que le critère communément utilisé est mis en défaut dans certains cas. Finalement, nous généralisons des résultats standards sur les normes en terme de valeurs singulières pour l'algorithme de Gram-Schmidt modifié. Ceci nous permet de dériver un schéma de réorthogonalisation a posteriori utilisant une matrice de rang faible. Ces résultats ont plusieurs applications directes. Nous en donnons des exemples avec les méthodes de Krylov pour résoudre des problèmes linéaires avec plusieurs seconds membres.<br /><br />Dans la deuxième partie, nous avons implémenté des variantes de la méthode GMRES pour les arithmétiques réelle et complexe, simple et double précisions. Cette implémentation convient pour des ordinateurs classiques, à mémoire partagée ou distribuée. Le code en résultant satisfait aux critères de qualité des librairies standards et son implémentation est largement détaillée. Pour des besoins de simplicité, flexibilité et efficacité, les solveurs utilisent un mécanisme de reverse communication pour les produits matrice-vecteur, les étapes de préconditionnement et les produits scalaires. Différents schémas d'orthogonalisation sont implémentés pour réduire le coût de calcul des produits scalaires, un point particulièrement important pour l'efficacité des méthodes de Krylov dans un environnement parallèle distribué. Le critère d'arrêt implémenté est basé sur l'erreur inverse normalisée. Les variantes disponibles sont GMRES-DR, seed-GMRES et block-GMRES. Ces codes s'ajoutent aux variantes déjà existantes (GMRES, flexible GMRES et SQMR). Un produit matrice-vecteur avec une décomposition LU est utilisé dans GMRES-DR de telle sorte que le stockage des approximations des vecteurs propres se fasse sur les premiers vecteurs de l'espace de Krylov. Un restart implicite et une étape de préconditionnement implicite ont été implémentés dans seed-GMRES. Nous supprimons ainsi un produit matrice-vecteur et une étape de préconditionnement par second membre et par cycle de GMRES. La version de block-GMRES permet à l'utilisateur de sélectionner différents modes de déflation. Pour terminer, des résultats reliant la norme du résidu de GMRES à la plus petite valeur singulière de l'espace construit par la méthode de Krylov ont été généralisés à la méthode block-GMRES.<br /><br />La troisième partie est consacrée à l'amélioration des techniques standards pour la résolution des systèmes linéaires dans le cadre des problèmes électromagnétiques. Après une présentation approfondie du code, nous étudions l'influence de la non-symétrie sur la convergence de l'algorithme SQMR. Nous étudions aussi le comportement de GMRES-DR sur nos problèmes. Ceci correspond à deux méthodes avec un seul second membre, le reste de cette partie concerne les cas comportant plusieurs seconds membres. Tout d'abord, nous examinons en détail les techniques qui permettent d'adapter les méthodes utilisées pour un second membre unique aux cas comportant plusieurs seconds membres. Par exemple, on peut améliorer la qualité du préconditionneur, avoir une stratégie de solution initiale, grouper les opérations de plusieurs résolutions ou encore paralléliser plusieurs résolutions. Dans le contexte du calcul de surface équivalente radar monostatique, nous avons montré que l'espace des seconds membres du problème continu était de dimension finie. La dimension donnée par notre théorie est proche de celle que nous observons en pratique. Cette propriété nous permet de réduire considérablement le nombre de systèmes linéaires à résoudre. Dans ce contexte, une version de la méthode block-GMRES est donnée. Ensuite, nous abordons certains problèmes spécifiques des méthodes seed-GMRES et block-GMRES pour lesquels nous proposons des solutions. Pour finir, des résultats plus prospectifs sont donnés. Plusieurs stratégies pour extraire et ajouter de l'information spectrale d'un cycle de GMRES à l'autre sont proposées et comparées. Puis nous utilisons le fait que la méthode multipôle rapide est un produit matrice-vecteur inexact dont la précision est réglable. Moins précis est le produit matrice-vecteur, plus rapide il est. Nous montrons comment tirer partie de cette propriété en utilisant un schéma relâché (méthode de Krylov inexacte) ou des itérations emboîtées (flexible GMRES). Enfin, le critère d'arrêt basé sur l'erreur inverse normalisée dans le cadre du calcul d'une surface équivalente radar est remis en question.

Les principaux objectifs de ce travail de recherche étaient (i) l'étude de la composition lipidique de quatre bactéries Gram-négatives hydrocarbonoclastes Marinobacter hydrocarbonoclasticus, Marinobacter aquaeolei, Acinetobacter calcoaceticus et Pseudomonas oleovorans en fonction de la source de carbone, (ii) la détermination des différents voies métaboliques de dégradation des hydrocarbures comportant une chaîne alkyle et (iii) mieux connaître la composition en Β-hydroxy acides lipopolysaccharidiques de leurs membranes extérieures. La première partie de ce travail présente une description qualitative et quantitative des lipides "non liés", labiles en milieu basique et labiles en milieu acide de M. hydrocarbonoclasticus en fonction de l'hydrocarbure (alcanes normaux et ramifiés, phényl et cyclohexyl-alcanes, et 1-alcène) utilisé comme source unique de carbone et d'énergie. Ces résultats ont montré que la structure chimique des lipides identifiés suit celle de l'hydrocarbure fourni: apparition de proportions significatives et dans certains cas très importantes d'acides gras issus de la dégradation de ces composés. Les Β-hydroxy acides des LPS sont moins sensibles au changement de la structure de l'hydrocarbure. Le n-Β-12:0 est toujours dominant alors que les Β-hydroxy acides provenant de la dégradation des hydrocarbures peuvent être incorporés aux LPS mais ne constituent que des composés mineurs. La présence d'intermédiaires métaboliques de chaque hydrocarbure dans les lipides "non liés" a permis la détermination des différents voies métaboliques de dégradation. La deuxième partie de ce travail porte sur la composition lipidique des bactéries M. aqueolei, A. calcoaceticus et P. oleovorans en fonction de la source de carbone (acétate ou alcane). Cette partie a montré le degrés de variabilité des acides gras chez les bactéries Gram-négatives M. aquaeolei et A. calcoaceticus: substitution des acides gras normaux pairs par des acides gras de même parité et même squelette que l'alcane utilisé. Chez P. oleovorans l'utilisation d'alcanes courts se manifeste aussi bien sur la distribution des acides gras membranaire que sur la composition en Β-hydroxy acides des LPS. Enfin, les Β-hydroxy acides issus des quatre bactéries étudiées sont essentiellement normaux pairs de C10 à C14 dont le composé prédominant est dépendant de l'espèce bactérienne.

Natural Language Processing (NLP) develops computational approaches to processing language data. Supervised machine learning has become the dominant methodology of modern NLP. The performance of a supervised NLP system crucially depends on the amount of data available for training. In the standard supervised framework, if a sequence of words was not encountered in the training set, the system can only guess at its label at test time. The cost of producing labeled training examples is a bottleneck for current NLP technology. On the other hand, a vast quantity of unlabeled data is freely available. This dissertation proposes effective, efficient, versatile methodologies for 1) extracting useful information from very large (potentially web-scale) volumes of unlabeled data and 2) combining such information with standard supervised machine learning for NLP. We demonstrate novel ways to exploit unlabeled data, we scale these approaches to make use of all the text on the web, and we show improvements on a variety of challenging NLP tasks. This combination of learning from both labeled and unlabeled data is often referred to as semi-supervised learning. Although lacking manually-provided labels, the statistics of unlabeled patterns can often distinguish the correct label for an ambiguous test instance. In the first part of this dissertation, we propose to use the counts of unlabeled patterns as features in supervised classifiers, with these classifiers trained on varying amounts of labeled data. We propose a general approach for integrating information from multiple, overlapping sequences of context for lexical disambiguation problems. We also show how standard machine learning algorithms can be modified to incorporate a particular kind of prior knowledge: knowledge of effective weightings for count-based features. We also evaluate performance within and across domains for two generation and two analysis tasks, assessing the impact of combining web-scale counts with conventional features. In the second part of this dissertation, rather than using the aggregate statistics as features, we propose to use them to generate labeled training examples. By automatically labeling a large number of examples, we can train powerful discriminative models, leveraging fine-grained features of input words.

Wir versuchen herauszufinden, ob das subjektive Maß der Cloze-Vorhersagbarkeit mit der Kombination objektiver Maße (semantische und n-gram-Maße) geschätzt werden kann, die auf den statistischen Eigenschaften von Textkorpora beruhen. Die semantischen Maße werden entweder durch Abfragen von Internet-Suchmaschinen oder durch die Anwendung der Latent Semantic Analysis gebildet, während die n-gram-Wortmaße allein auf den Ergebnissen von Internet-Suchmaschinen basieren. Weiterhin untersuchen wir die Rolle der Cloze-Vorhersagbarkeit in SWIFT, einem Modell der Blickkontrolle, und wägen ab, ob andere Parameter den der Vorhersagbarkeit ersetzen können. Unsere Ergebnisse legen nahe, dass ein computationales Modell, welches Vorhersagbarkeitswerte berechnet, nicht nur Maße beachten muss, die die Relatiertheit eines Wortes zum Kontext darstellen; das Vorhandensein eines Maßes bezüglich der Nicht-Relatiertheit ist von ebenso großer Bedeutung. Obwohl hier jedoch nur Relatiertheits-Maße zur Verfügung stehen, sollte SWIFT ebensogute Ergebnisse liefern, wenn wir Cloze-Vorhersagbarkeit mit unseren Maßen ersetzen. / We try to determine whether it is possible to approximate the subjective Cloze predictability measure with two types of objective measures, semantic and word n-gram measures, based on the statistical properties of text corpora. The semantic measures are constructed either by querying Internet search engines or by applying Latent Semantic Analysis, while the word n-gram measures solely depend on the results of Internet search engines. We also analyse the role of Cloze predictability in the SWIFT eye movement model, and evaluate whether other parameters might be able to take the place of predictability. Our results suggest that a computational model that generates predictability values not only needs to use measures that can determine the relatedness of a word to its context; the presence of measures that assert unrelatedness is just as important. In spite of the fact, however, that we only have similarity measures, we predict that SWIFT should perform just as well when we replace Cloze predictability with our measures.

The engineering of bacterial surfaces has in recent yearsattracted a lot of attention with applications in manydifferent areas of bioscience. Here we describe the use of twodifferent surface display systems for the gram-positivebacteria Staphylococcus carnosus and Staphylococcus xylosus invarious biotechnological applications. Environmental microbiology currently attracts a lot ofattention since genetically engineered plants and bacteriamight be used as bioadsorbents for sequestration of toxicmetals. Bacterial surface display of metal-binding peptidesmight enable recycling of the biomass by desorption ofaccumulated heavymetals. In an attempt to recruitstaphylococcal display systems for bioremediation purposes,polyhistidyl peptides were successfullly displayed on thesurface of recombinant S. carnosus and S. xylosus cells.Whole-cell Ni2+-binding assays demonstrated that therecombinant cells had gained metal-binding capacity compared towild-type cells. Tailor-made, metal-binding staphylococci was created using apreviously constructed phage-display combinatorial proteinlibrary based on a fungal cellulose-binding domain (CBD)derived from the cellobiohydrolase Cel7A of Trichoderma reseii.Novel metal-binding CBDs were generated through a phagemediated selection procedure. Selected CBD variants, now devoidof cellulose binding, were randomly selected and sequenceanalysis of selected variants revealed a marked preference forhistidine residues at the randomized positions. Surface displayof these novel CBD variants resulted in recombinantstaphylococci with increased metal-binding capacity compared tocontrol strains, indicating that this could become a generalstrategy to engineer bacteria for improved binding to specificmetal ions. Directed immobilization of cells with surface displayedheterologous proteins have widespread use in modernbiotechnology. Among other things they could provide aconvenient way of generating biofilters, biocatalysts orwhole-cell diagnostic devices. It was therefore investigatedwhether directed immobilization of recombinant staphylococci oncotton fibers could be achieved by functional display of afungal cellulose-binding domain (CBD). Recombinant S. carnosuscells with surface anchored CBDs from Trichoderma reseii Cel6Awere found to efficiently bind to cotton fibers creating almosta monolayer on the fibrous support. The co-expression of thisCBD together with previously described metal-binding proteinson the surface of our staphylococci would create means fordeveloping effective bioadsorbents for remediationpurposes. The original plasmid vector, designed for heterologoussurface display on recombinant S. carnosus cells has exhibitedproblems related to structural instability, possibly due to thepresence of a phage f1 origin of replication in the vectorsequence. This would be a problem if using the vector systemfor library display applications. Therefore, novel surfacedisplay vectors, lacking the phage ori were constructed andevaluated by enzymatic and flow cytometric whole-cell assays.One such novel vector, pSCXm, exhibited dramatically increasedplasmid stability with the retained high surface density ofexpressed heterologous proteins characteristic for the originalS. carnosus display vector, thus making it potentially moresuitable for library display applications. The successful engineering of our staphylococcal displaysystem encouraged us to further evaluate the potential to usethe staphylococcal system for display of combinatorial proteinlibraries and subsequent affinity based selections using flowcytometric cell sorting. A model system of recombinant S.carnosus cells with surface displayed engineered protein Adomains was constructed. It was demonstrated that target cellscould be sorted essentially quantitatively from a moderateexcess of background cells in a single sorting-step.Furthermore, the possibility of using staphylococcal surfacedisplay and flow cytometric cell sorting also for specificenrichment of very rare target cells by multiple rounds ofcell-sorting and in between amplification was demonstrated. <b>Key words:</b>affibody, albumin binding protein, bacterialsurface display, cell immobilization, bioremediation,combinatorial protein engineering, flow cytometry,Gram-positive, metal binding, staphylococcal protein A,Staphylococcus carnosus, Staphylococcus xylosus, whole-celldevices

Speech recognition in Turkish Language is a challenging problem in several perspectives. Most of the challenges are related to the morphological structure of the language. Since Turkish is an agglutinative language, it is possible to generate many words from a single stem by using suffixes. This characteristic of the language increases the out-of-vocabulary (OOV) words, which degrade the performance of a speech recognizer dramatically. Also, Turkish language allows words to be ordered in a free manner, which makes it difficult to generate robust language models. In this thesis, the existing models and approaches which address the problem of Turkish LVCSR (Large Vocabulary Continuous Speech Recognition) are explored. Different recognition units (words, morphs, stem and endings) are used in generating the n-gram language models. 3-gram and 4-gram language models are generated with respect to the recognition unit. Since the solution domain of speech recognition is involved with machine learning, the performance of the recognizer depends on the sufficiency of the audio data used in acoustic model training. However, it is difficult to obtain rich audio corpora for the Turkish language. In this thesis, existing approaches are used to solve the problem of Turkish LVCSR by using a limited audio corpus. We also proposed several data selection approaches in order to improve the robustness of the acoustic model.