Prosodic boundaries (cued by pitch variations, final lengthening, pause) have been consistently demonstrated to have an immediate influence on parsing in a variety of syntactic structures cross-linguistically. For example, in sentences with temporary ambiguities such as Early and Late closure (EC/LC), which contain two potential boundary positions – the first (#1) compatible with EC and the second (#2) compatible with LC (e.g., Whenever the bear was approaching #1 the people #2 (EC): …would run away; (LC): …the dogs would run away), without the benefit of prosodic information, the preferred (or default) interpretation is LC, which consequently leads to processing difficulties (garden-path effects) in EC structures. The majority of studies on spoken sentence processing has focused on the impact of a single boundary on the closure or attachment preference of a specific phrase or clause. However, more recently, several influential hypotheses have emerged that aim to account for the interplay between two boundaries in a sentence, specifically in terms of size and location; the most influential of these argue that processing is either (i) local, with large boundaries independently integrated, which serve as strategic cues to syntactic closure (Watson & Gibson, 2005), or (ii) global, with the relative difference between the magnitude of boundaries across an utterance modulating interpretation (Clifton, Carlson, & Frazier, 2002). Although differing in details, these hypotheses suggest that listeners process boundary information at the sentence level in a categorical manner. In contrast, there is some data to suggest that boundaries can differ in a gradient quantitative manner, and that listeners are sensitive to this range of boundary sizes. The aims of the current dissertation were therefore to use behavioral and event-related potential (ERP) measures: (i) to contrast the predictions of the opposing theoretical accounts using temporary syntactic ambiguities, and (ii) to test whether gradient differences in boundary size impact listeners' parsing decisions in a gradient or categorical manner.Using an innovative paradigm, I conducted two behavioral experiments (Study 1), and one ERP experiment (Study 2), where listeners were presented with highly controlled digitally-manipulated EC/LC sentences, each containing two prosodic boundaries (as in the example above), which differed only in terms of their relative sizes. The outcomes of the three experiments reveal an initial, profound bias of boundaries on syntactic preference, which was nearly impossible to override. In addition, subtle differences between prosodic boundaries are detected by the brain and affect the degree of processing difficulty. Finally, the effect of boundaries on parsing is far more intricate than previously assumed. These outcomes cannot be accommodated by a purely categorical account, and cast serious doubts on most current models of prosodic online processing. We present the extended Boundary Deletion Hypothesis (eBDH), an alternative account for prosodic phrasing, based on the results of all three experiments. / L'influence des frontières prosodiques (caractérisées par des variations d'intensité, d'allongement final et des pauses) sur l'analyse de structures syntaxiques a été démontrée de façon systématique dans plusieurs langues. Un exemple se retrouve dans les phrases à ambiguïtés temporaires, lesquelles contiennent deux positions potentielles : la première (#1) étant compatible avec une clôture précoce (CP), la deuxième (#2) avec une clôture tardive (par exemple, Aussitôt que l'ours s'approcha #1 les gens #2 (CP) :… se sauvèrent ; (CT) : … les chiens se sauvèrent). En l'absence d'information prosodique, l'interprétation privilégiée (ou par défaut) correspond à une structure en CT, ce qui a pour effet de provoquer des difficultés de traitement (effets « cul de sac ») des structures CP. Une majorité d'études de phrases parlées se sont concentrées sur l'impact d'une seule frontière sur les préférences de clôture ou de jonction d'un syntagme ou d'une proposition spécifique. Toutefois, plusieurs théories importantes ont récemment vu le jour et tentent d'expliquer l'interaction s'établissant entre deux frontières dans une même phrase, prenant leur taille et leur position comme principaux facteurs. Selon les approches les plus importantes sur la question, le traitement est soit (i) local, les frontières plus importantes étant intégrées indépendamment et servant d'indices stratégiques à la clôture syntaxique (Watson & Gibson, 2005), soit (ii) global, la différence relative entre la magnitude des frontières à travers la phrase modulant l'interprétation (Clifton, Carlson & Frazier, 2002). Quoique différentes dans le détail, ces théories suggèrent que les personnes traitent l'information liée à la frontière au niveau de la phrase de manière catégorielle. Cependant, plusieurs données suggèrent que les frontières peuvent différer de manière graduelle et quantitative, et que les personnes sont sensibles à la taille des frontières. Les buts de la présente thèse étaient dès lors d'utiliser des mesures comportementales et électrophysiologiques (Potentiels évoqués ou PÉs) afin de (i) contraster les prédictions propres à chacune de ces théories au moyens d'ambiguïtés syntaxiques temporaires, et (ii) de tester si les différences graduelles de tailles des frontières ont un impact graduel ou catégoriel sur les décisions d'analyse des participants. Nous avons conduit deux expériences comportementales (Étude 1) et une expérience PÉs (Étude 2) au moyen d'un paradigme novateur, au cours duquel des phrases CP/CT manipulées numériquement et rigoureusement contrôlées étaient présentées aux participants. Ces phrases contenaient deux frontières prosodiques (comme dans l'exemple donné plus haut) différant exclusivement au niveau de leurs tailles relatives. Les trois expériences révèlent que les frontières induisent un biais initial, profond et impossible à surpasser sur les préférences syntaxiques. En outre, des différences subtiles entre les frontières prosodiques sont détectées par le cerveau et affecte le degré de difficulté du traitement. Enfin, l'effet des frontières prosodiques sur l'analyse syntaxique est bien plus sophistiqué qu'assumé précédemment. Ces résultats échappent à une hypothèse strictement catégorielle et mettent sérieusement en doute la plupart des modèles actuels de traitement prosodique en temps réel. Sur la base des présentes données, nous présentons l'Hypothèse étendue de Délétion de Frontières (HeDF), offrant une explication alternative du phrasé prosodique.

Many children misarticulate more sounds than expected for their age and present with a Developmental Phonological Disorder (DPD). Children with DPD are at risk for later academic and socio-emotional difficulties. Successful intervention for these children is dependent on a good understanding of the underlying nature of their disorder (Stackhouse & Wells, 1997). Whereas speech-language pathologists have been assessing the surface manifestations of communication disorders for many decades, knowledge of the underlying causes of speech and language impairments is more limited but growing. In addition, while there is a large body of research on the surface manifestations of DPD in English-speaking children, there is currently a paucity of information regarding the surface manifestation of the disorder in French-speaking children. The purpose of the research described in this dissertation was to investigate the underlying psycholinguistic profiles and the surface manifestations of DPD in a large group of French-speaking preschoolers. More precisely, Study one aimed to determine whether the psycholinguistic profiles of French-speaking children with DPD are similar to those of English-speaking children with DPD as reported in the literature. Seventy-two French-speaking children with DPD, aged4 to 6 years, were assessed on measures of articulation accuracy, receptive vocabulary, nonverbal intelligence, phonological processing and structure and function of the oral-speech mechanism. Ten typically developing French-speaking children completed the articulation accuracy, receptive vocabulary, phonological awareness, and syllable repetition tasks. Results indicated that the vast majority of French-speaking children with DPD presented with phonological processing difficulties; furthermore, measures of phonological processing explained significant variance in speech production accuracy after controlling for individual differences in receptive vocabulary skills and maternal education. Study two aimed to determine whether surface speech errors aremanifested differently in French-speaking children with DPD in comparison to a very similar group of English-speaking children with DPD. Twenty-four French-speaking children with DPD were matched on percentage of consonants correct in conversation, age, and receptive vocabulary to English-speaking children with DPD. By comparing these children's productions of consonants on a single-word test of articulation, we found that the surface manifestations of DPD are differentin these two languages. The French-speaking children obtained low match ratios for the major sound class features [+consonantal] and [+sonorant], reflecting a high frequency of omission errors. In contrast, English-speaking children obtained high match ratios for the major sound class features [+consonantal] and [+sonorant] and produced more substitution errors.The results of Study one add to the growing body of literature supporting a core deficit in phonological processing in children with DPD. In English-speaking children, this underlying deficit in phonological processing remains present regardless of environmental changes and/or the child's maturation, whereas the surface speech errors produced by young children with DPD change as they get older and in response to environmental changes. In the studies described here, asimilar core deficit in phonological processing was found in children speaking French, whereas the surface speech errors were different than those of English-speaking children, due to the differences in the phonological systems of these two languages. These findings point to the importance of assessing the phonological processing skills of children with a current or past history of DPD, as well as to the need to use test instruments with French-speaking children that reflect thephonological characteristics of the language at multiple levels of the phonological hierarchy. / Beaucoup d'enfants ont plus de difficultés à prononcer clairement les sons que les autres enfants de leur âge, et présentent un trouble phonologique (TP). Les enfants qui ont un TP sont à risque de présenter des difficultés académiques et socio-émotives. Une bonne compréhension de la nature sous-jacente de leur TP est nécessaire afin de fournir une intervention efficace (Stackhouse & Wells, 1997). Tandis que les orthophonistes évaluent les manifestations de surface des troubles de la communication depuis plusieurs décennies, la connaissance des causes sous jacentesdes TP et des troubles de langage est plus limitée, mais grandissante. De plus, alors qu'il y a un large corpus de recherche au niveau des manifestations de surface des TP chez les enfants anglophones, il y a actuellement très peu de données au niveau des manifestations de surface de ce trouble chez les enfants francophones.Le but des études décrites dans cette thèse était d'examiner les profils psycholinguistiques sous-jacents et les manifestations de surface du TP chez un grand nombre d'enfants francophones d'âge préscolaire. Plus précisément, l'Étude un avait pour but de déterminer si les profils psycholinguistiques d'enfants francophones avec un TP sont semblables à ceux des enfants anglophones avec un TP, tel que rapporté dans la littérature. Soixante-douze enfants avec un TP, âgés de 4 à 6 ans, ont été évalués à l'aide de mesures de précision de l'articulation, du vocabulaire réceptif, de l'intelligence non-verbale, du traitement phonologique et de la structure et de la fonction du mécanisme oral-périphérique. Dix enfants avec un développement typique ont également complété les mesures de précision de l'articulation, du vocabulaire réceptif, de la conscience phonologique, et de répétition de syllabes. Les résultats ont indiqué que la vaste majorité des enfants francophones avec un TP présentaient des difficultés au niveau du traitement phonologique; en outre, les mesures de traitement phonologique ont expliqué une portion significative de la variance au niveau de la précision de l'articulation, même en tenant compte de l'effet de la variation du vocabulaire réceptif et du niveau d'éducation maternelle. L'Étude deux avait pour but de déterminer si les erreurs de production des sons de la parole se manifestent différemment chez les enfants francophones avec un TP, en comparaison à un groupe très semblable d'enfants anglophones. Vingt-quatre enfants francophones avec un TP ont été jumelés au niveau du pourcentage de consonnes correctes en conversation, de l'âge, et du vocabulaire réceptif, à des enfants anglophones avec un TP. En comparant la production de consonnes de ces enfants, obtenues à l'aide d'un test d'articulation de mots simples, nous avonstrouvé que les manifestations de surface du TP sont différentes dans ces deux langues. Les enfants francophones ont obtenu de bas ratios de jumelage des traits pour les classes majeures de traits [+consonantique] et [+sonant], reflétant une haute fréquence d'omission de consonnes. Au contraire, les enfants anglophones ont obtenu des ratios élevés de jumelage des traits phonologiques pour les classes [+consonantique] et [+sonant], et ont produit plus d'erreurs de substitutions de consonnes. Les résultats de l'Étude un viennent s'ajouter à un nombre croissant d'études qui appuient un déficit principal au niveau du traitement phonologique chez les enfants avec un TP. Dans les études décrites dans cette thèse, nous avons trouvé un déficit principal au niveau du traitement phonologique chez des enfants francophones avec un TP, alors que les erreurs de prononciation des sons qu'ils ont produites sont différentes des erreurs produites par des enfants anglophonesavec un TP puisque les systèmes phonologiques de ces deux langues diffèrent l'une de l'autre. Ces conclusions mettent en évidence l'importance d'évaluer les habiletés de traitement phonologique des enfants avec des antécédents courants ou antérieurs de TP.

Vitamin D is very well known for its classical role in the maintenance of calcium and phosphorus homeostasis as well as in the prevention of rickets. More recent findings of its ability to inhibit cell proliferation, induce apoptosis, induce differentiation, inhibit angiogenesis, and modulate the immune system have made it a current topic of intense research, particularly in the field of cancer research. We used a murine model of breast cancer metastasis to bone to investigate the effect of vitamin D deficiency on the growth of breast cancer tumour cells within bone. We also established that these breast cancer tumour cells express the enzyme CYP27B1 (1α-hydroxylase) which is able to convert the inactive vitamin D precursor 25-hydroxyvitamin D (25(OH)D) to the active metabolite 1,25-dihydroxyvitamin D (1,25(OH)2D). We next examined the effect of the local activation of vitamin D by tumoral CYP27B1 on the growth of these tumour cells within bone. Although we did not see a significant difference in the growth of breast cancer tumour cells in the bones of vitamin D deficient mice as compared to vitamin D sufficient mice, we have demonstrated that breast cancer tumour cells that do not express CYP27B1 grow much more aggressively within bone than breast cancer tumour cells which express CYP27B1. This suggests a very important role for the local activation of vitamin D by extra-renal CYP27B1 on the growth of breast cancer tumour cells within the bone microenvironment. These findings suggest a potential use for 25(OH)D as a treatment for breast cancer metastasis to bone either alone or in combination. / La vitamine D est bien connue pour son rôle dans le maintien des concentrations de calcium et du phosphore dans la circulation ainsi que dans la prévention du rachitisme. La découverte plus récente de sa capacité d'inhiber la prolifération cellulaire, induire leur différentiation ainsi que l'apoptose cellulaire, inhiber l'angiogenèse, et moduler le système immunitaire rend son étude un sujet de recherche très intéressant surtout dans le domaine de la recherche sur le cancer. Nous avons étudié l'effet de la carence en vitamine D sur la croissance tumorale dans un modèle murin de métastases osseuses du cancer du sein. Nous avons aussi établi que ces cellules expriment l'enzyme CYP27B1 (1α-hydroxylase) et sont donc capables d'activer la vitamine D en son métabolite actif la 1,25-dihydroxyvitamine D (1,25(OH)2D) à partir du métabolite inactif, la 25-hydroxyvitamine D (25(OH)D). Nous avons ensuite examiné l'effet de l'activation locale de la vitamine D par les cellules tumorales dérivées du sein sur la croissance de ces cellules dans le microenvironnement osseux. Nous n'avons constaté aucune différence significative entre la croissance des cellules tumorales du cancer du sein dans l'os chez les souris carencées en vitamine D en comparaison aux souris non carencées en vitamine D. Cependant, nous avons démontré que les cellules tumorales du cancer du sein qui expriment le CYP27B1 croissent beaucoup moins vite dans l'os que les cellules tumorales qui n'expriment pas le CYP27B1. Ces résultats suggèrent un rôle très important de l'activation extra-rénale de la vitamine D par les cellules tumorales du cancer du sein pour inhiber la croissance de ces cellules dans l'os. En conclusion, ces travaux indiquent que le précurseur inactif 25(OH)D pourrait être utilisé seul ou en combinaison pour le traitement des métastases osseuses du cancer du sein.

Uveal Melanoma (UM) is the most common intraocular tumor in adults. Liver metastasis is the leading cause of death in patients affected by this disease and in approximately 40% of the cases, metastasis occurs 10 years after initial diagnosis. Tumor dormancy has been considered as a leading theory for the delay of the manifestation of metastatic disease and it has been the subject of numerous studies, which includes investigating metastasis suppressor genes (MSG). Studies have shown that the MSG KISS1 plays a role in various human malignancies, including melanoma, and it seems to be involved in the dormancy phase of the metastatic cascade. Previous studies from our laboratory reported that loss of KISS1 expression is related to worse prognosis in UM. In this light, mechanisms that involve increase of KISS1 expression are of interest for UM researchers. Interestingly, pathways involved in UM progression include upregulation of c-KIT, a cell surface molecule normally found in melanoma cells. This process seems to occur at the same time that KISS1 is downregulated and overt metastasis become clinically detectable. Therefore, we verified if inhibition of c-KIT could be related to KISS1 expression in metastatic UM. In addition, considering that a newly identified class of small non coding RNAs (miRNAs) are master regulators of gene expression, we sought to investigate if miRNAs were involved in metastasis of UM. Therefore, the objective of this thesis was to identify mechanisms that increase the expression of KISS1 and to better understand UM metastasis. To address these questions, we used a c-KIT inhibitor, imatinib mesylate (IM), and miRNAs in this study. Human UM cell lines with different metastatic potential showed increased levels of KISS1, by real-time reverse transcriptase polymerase chain reaction (RT-PCR), after treatment with IM. Notably, an increase in KISS1 was observed at the protein level in a dose response manner when the most aggressive UM cell line (92.1) was treated with different concentrations of IM. Increased levels of KISS1 expression were confirmed in vivo using an experimental animal model (albino rabbits). Using a miRNA array, we were able to identify miRNAs with prometastatic and antimetastatic effects in different UM cell lines and under diverse conditions. The miRNAs 10a, 10b, 21, and let-7 were upregulated in a liver metastatic cell line compared to the primary UM cell line from the same patient. Additionally, the UM cell line with aggressive potential transfected with KISS1 showed decreased expression of the miR-221 and increased expression of miR-146 compared to the non-transfected control. Similar results were obtained when the same cell line was treated with IM. In order to translate these results to a clinical setting, in situ hybridization of miR-221 was performed in 15 human UM FFPE tissues; increased expression of miR-221 was positively correlated with metastasis in UM. In conclusion, KISS1 was upregulated following treatment with IM. In addition, down regulation of miR-221 was found in all miRNA arrays with increases in KISS1 and treatment with IM. To the best of our knowledge, this is the first study to show that treatment with a c-KIT inhibitor causes an upregulation of KISS1, and that miR-221 may promote metastasis in UM. Consequently, induction of KISS1 expression downregulates miR-221 and should be considered as potential targets for adjuvant therapy in UM metastasis. / Parmi les tumeurs intraoculaires malignes, le plus fréquent chez l'adulte est le mélanome de l'uvée (MU). Les métastases hépatiques représentent la principale cause de décès chez les patients affectés par cette maladie et dans environ 40 % des cas, les métastases apparaissent plus de 10 ans après le diagnostic initial. La latence de la tumeur est considérée comme étant la théorie principale qui expliquerait le retard de l'apparition des métastases. Cela fut le sujet de nombreuses études dont un certain nombre intéressaient par l'analyse des Gènes Suppresseurs de Métastases (GSM)Il a été démontré que le GSM KISS1 joue un rôle dans différentes maladies malignes, dont le mélanome. Ce gène parait être impliqué dans la phase de latence de la cascade métastatique. Des études antérieures conduites dans notre laboratoire ont conclu que la perte d'expression du gène KISS1 est associée à un mauvais pronostic chez les patients atteints de MU. Ainsi, les mécanismes qui impliquent l'augmentation d'expression de KISS1 présentent un intérêt pour les chercheurs travaillant sur MU.Curieusement, les voies de signalisations impliquées dans la progression de MU incluent la régulation positive de c-KIT une molécule de surface cellulaire. Ce phénomène semble coïncider avec la régulation négative de KISS1 et le moment où les métastases déclarées deviennent cliniquement détectables. Nous avons donc vérifié si l'inhibition de c-KIT pourrait-être liée à l'expression de KISS1. De plus, considérant qu'une nouvelle classe de petits ARN non codés (miARN), qui sont les principaux régulateurs de l'expression génique, vient d'être identifiée, nous avons cherché à examiner si les miARNs étaient impliqués dans les métastases de MU.Par conséquent, l'objectif de cette thèse était d'identifier les mécanismes qui augmentent l'expression de KISS1 afin de mieux comprendre l'évolution des métastases de MU. Pour aborder ces questions, nous avons utilisé un inhibiteur de c-KIT, l'Imatinib mesylate (IM), et des miARNs.Les lignées cellulaires de MU aux différents potentiels métastatiques présentent un niveau élevé de KISS1 après un traitement à l'IM. Les niveaux accrus de KISS1 ont été confirmés in vivo en utilisant le model expérimental animal (lapins albinos). En utilisant des puces à miARN, nous avons réussis à identifier les miARNs aux effets pro-métastatiques et anti-métastatiques dans différentes lignées cellulaires de MU et sous diverses conditions. Une régulation positive des miARNs 10a, 10b, 21 et let-7 fut identifiée dans une lignée cellulaire de métastase hépatique par rapport à la lignée cellulaire de la tumeur primaire du même patient. De plus, dans une lignée cellulaire de MU où une transfection de KISS1 fut réalisée, on a pu observer une diminution de l'expression de miR-221 et une augmentation de l'expression de miR-146, par rapport au control non-transfecté. Des résultats semblables ont été obtenus, quand la même lignée cellulaire a été traitée avec IM. Afin d'utiliser ces résultats dans un environnement clinique, nous avons exécuté une hybridation in situ de miR-221 sur 15 tissus humains (fixés dans le formol puis inclus dans la paraffine) de MU dans lesquels une expression accrue de miR-221 est corrélé positivement avec des métastases dans le MU.En conclusion, KISS1 est régulé positivement sous l'action d'IM. De plus, une régulation négative de miR-221 fut constatée dans toutes les puces de miARN qui présentaient une augmentation de KISS1 ainsi que pendant le traitement avec IM. À notre connaissance, ceci est la première étude à démontrer que le traitement avec un inhibiteur de c-KIT cause une régulation positive de KISS1 et que le miR-221 peut avoir un effet métastatique sur le MU. Par conséquent, l'augmentation de la expression de KISS ainsi que la suppression de miR-221 peuvent potentiellement être considérées comme des cibles potentielles dans la thérapie adjuvante de MU métastatique.

Oil and gas exploration in Northern Canada uses explosive-based seismic techniques to locate hydrocarbon reserves beneath waterbodies not frozen to the bottom. The use of explosives in, or near, waterbodies has the potential to harm fishes, primarily through instantaneous pressure changes (IPCs) generated from the detonations. These IPCs can damage soft tissues through the rapid compression and expansion of the swimbladder as the pressure wave passes. In Canada, a document entitled Guidelines for the Use of Explosives In or Near Canadian Fisheries Waters recommends that peak pressures not exceed 100 kPa for the protection of fish, however damage has been reported below this level. To simulate seismic exploration and examine potential pathological changes surrounding the current Guideline, fish across different developmental stages and with varying degrees of swimbladder presence were exposed to a variety of explosive based IPC levels in field experiments. Early life stages of rainbow trout (Oncorhynchus mykiss) including eyed eggs, sac fry, and juveniles were caged and exposed to discrete detonations from 0 to 280 kPa in the Mackenzie Delta, NWT. These fish were subsequently examined for both gross pathological and histological changes to cranial structures as well as swimbladder, kidney, liver and gill tissue. Results showed changes in both the area and circumference of the cranial region of eyed eggs, as well as swimbladder, ocular and kidney damage in juveniles. Additionally, caged adult swimbladder bearing lake trout (Salvelinus namaycush) and non-swimbladder bearing slimy sculpin (Cottus cognatus) were exposed to explosive based IPCs ranging in peak pressure from 0 to 127 kPa at the Experimental Lakes Area, Ontario. Fish were later examined grossly and blood, liver, kidney, intestine, and spleen were examined to determine the presence of any traumatic based pathological changes. Results indicated the occurrence of swimbladder hemorrhage in lake trout exposed to IPCs near the current Guideline level. Finally, a risk assessment for lake trout of the Mackenzie Delta exposed to IPCs was undertaken; to examine the potential for adverse risk to individuals and populations, and the likelihood of populations being unable to recover. Based on the findings of the aforementioned studies, the recommended Guideline level is not protective of early life stages of rainbow trout and furthermore represents the threshold at which damage to the swimbladder in adult lake trout does not occur, as such, a re-examination of the recommended Guideline level is warranted.

Septoria tritici blotch (Stb) is a major foliar disease of wheat worldwide caused by the fungal pathogen Mycosphaerella graminicola. This project mapped the chromosomal locations of Stb resistance, which will be useful for wheat cultivar enhancement. In the first study of this project, Stb resistance and grain yield quantitative trait loci (QTL) were identified in a winter wheat doubled-haploid (DH) population produced from the cultivars Pastiche (resistant) and Torfrida (susceptible). A genetic map of the population was constructed using 104 microsatellite and 202 DArT markers. Separate disease and yield field trials were conducted in north-western Europe to measure natural Stb infection and grain yield, respectively. A resistance QTL from Pastiche was identified (QStb.jic-2A) that reduced flag leaf pycnidial density by 31.2% compared to the population mean. A second QTL from Pastiche (QYld.jic-2A) co-localized with QStb.jic-2A, and conferred an increase in grain yield of approximately 2.0% (0.19 t ha-1). The second study determined the map location of the third multiple-isolate resistance gene in the hexaploid landrace Salamouni. A population was developed using the resistant DH line 98S05B*13 (Salamouni/Katepwa) crossed to the susceptible cultivar Katepwa. The parent 98S05B*13 was found to contain the resistance gene Stb14 on chromosome 3BS, therefore, to minimize the effect of this gene, individual plants of the F2 population were selected for self-pollination based on presence or absence of Stb14 using flanking microsatellite markers. The Stb14(+) F2:3 families were screened with M. graminicola isolate MG96-36. Bulked segregant analysis identified a possible linkage to the gene of interest on chromosome 3AS. Polymorphic microsatellite markers on 3AS were used to construct a linkage group. The markers barc321 and barc12 were found to flank the resistance gene at genetic distances of 1.9 and 2.5 cM, respectively. This position on 3AS has not been previously linked to Stb resistance and this gene will be designated StbSm3. The third population was developed to fine map the isolate-specific resistance gene Stb14 located on chromosome 3B of Salamouni. A large population of 84 fixed recombinant F4 families was developed using the co-dominant microsatellite markers flanking Stb14. Attempts to identify clearly polymorphic molecular markers within the interval containing Stb14 were unsuccessful.

Gene therapy has a great potential for the treatment of a wide range of diseases. However, the development of a safe and efficient delivery vector is the major obstacle for gene therapy. Recently synthesized 2 - (dimethylamino) ethyl methacrylate 2-(methacryloxloxyethyl phosphorylcholine) (DMA-MPC) diblock copolymer was investigated in this work as a novel non-viral vector for gene delivery. It has been previously demonstrated that the cationic DMA block can condense DNA efficiently. The zwitterionic PC head groups are found naturally in the outer leaflet of biomembranes and are extremely biocompatible. It is thus proposed here that the MPC can act as a new steric stabilizer to the system. Different compositions of DMA-MPC diblock copolymers were evaluated. The MPC block with minimum length 30 monomeric units can successfully provide steric stabilization to the system, and reduce nonspecific cellular interaction by providing a steric barrier to the DNA complexes. However, long MPC chain can hinder the interaction between cationic DMA and DNA, leading to the formation of loosely condensed complexes which were more susceptible to enzymatic degradation. Therefore the composition of the copolymer must be carefully adjusted so that the DNA condensing and steric stabilization effect are well balanced. In order to investigate the cellular uptake mechanism DMA homopolymerDNA complexes, the effect of different endocytosis inhibitors was examined. Microtubules and actin filaments were involved in the uptake of DNA complexes, suggesting that the complexes were internalised by endocytosis. Both the clathrin- and caveolae- mediated pathway were responsible for the uptake of DNA complexes, and the former appeared to be the main route of entry. Finally, folic acid ligand was incorporated into the DMA-MPC copolymer in order to improve the specific targeting. Initial data showed that there was selective uptake of the folate conjugated system in folate receptor expressing cells possibly via receptor mediated endocytosis. However, parameters such as the optimum length of MPC component, number of ligands per DNA complex and the composition of the system need to be further investigated in order to maximize the specificity and transfection efficiency.

This thesis describes the search for a practical skin dosimetry method for cardiac catheterization procedures, and the application of an optimisation strategy in barium enema imaging. Kodak EDR2 film was characterised across the range of exposure conditions used in the cardiac catheterization laboratory. Its dose-response curve was modelled using a novel equation, and overall uncertainty in film response was estimated. The film saturated at 1 Gy, limiting its usefulness for skin dosimetry. Its performance was found to be strongly dependent on beam filtration, an aspect that had not previously been studied. The film was then used to measure skin doses to patients undergoing coronary angiograms and angioplasties. For angiograms, all skin doses were well below 1 Gy. For angioplasties, 23% of films showed localised saturation, indicating peak skin doses of at least 1 Gy. Dose-area-product was shown to be a poor predictor of high peak skin dose. A mathematical model was developed and software written, to calculate patient skin dose maps from exposure and projection data stored in the image files. This offered a practical method for assessing the magnitude and approximate location of the peak skin dose. Accuracy was limited by a lack of information regarding fluoroscopic exposures, couch position and beam limitation. After including an estimated contribution from fluoroscopy, the model successfully identified those patients whose skin doses exceeded 1 Gy. Following a baseline survey of local barium enema practice, several dose reduction methods were considered. It was decided to introduce copper filtration. 0.1 mm copper reduced mean patient DAP by 37%, without any measurable difference in contrast detail detectability. A detailed phantom study determined the optimal copper thickness as 0.3 mm. This reduced mean patient DAP by 55%, relative to the baseline survey. A visual grading analysis study showed no significant difference in clinical image quality.

CD55 is one of the complement regulatory/inhibitory proteins and is over-expressed on a wide range of solid tumours. CD55 is also known to be deposited within tumour stroma and is secreted in an active soluble form, mediated by matrix metalloproteinase-7. The complement cascade forms part of the innate immune system and culminates in cell lysis of targeted cells. As a complement regulatory protein, the primary function of CD55 is to accelerate the decay of complement components preventing formation of the membrane attack complex. CD55 is also known to be a ligand for the T cell early activation antigen CD97, and their interaction has been shown to inhibit the proliferation of activated T cells. This project aimed to develop anti-tumour immunotherapeutics aimed at exploiting CD55 as a tumour associated antigen. Initial strategies were to develop monoclonal antibodies, specific to identified epitopes from within the CD55 protein sequence, capable of binding, and neutralising CD55s decay accelerating activity. Developed antibodies would also have the potential to induce antibody dependent cell cytotoxicity, thus blocking CD55 protection of tumours and mediating an active anti-tumour response. Antibodies were raised specific to CD55 derived linear peptides, which have been used for the assessment of CD55 expression in breast tumour sections. Monoclonal antibodies failed to recognise natively expressed protein on viable tumour cells and alternate strategies were developed. An effective immunotherapy for the treatment of cancer would engage both cellular and humoral mediated responses for effective clearance of target cells. In order to achieve this, a DNA vaccine incorporating a human IgG Fc tail was developed expressing the active sites of CD55, containing HLA-A*201 restricted heteroclitic epitopes. The vaccines were used to immunise HLA-A*201 HHDII transgenic mice and CD55 specific responses were assessed. One of the vaccines analysed, elicited CD55 specific antibodies capable of recognising tumour cells in vitro and also generated epitope specific CD8+ T cell mediated lysis of epitope bearing cells. The frequency of CD55 specific T cells was obtained via antigen specific IFN gamma release ELISPOT assays and the cytokine profile of responses generated was assessed via luminex analysis. In conclusion, CD55 remains a viable target for immunotherapies aimed at CD55 bearing tumours. DNA vaccines encoding modified epitopes are capable or raising cellular and humoral responses to this antigen and further studies should be completed in order to determine anti-cancer effects in tumour bearing models.

No description available.