Includes bibliographic references. / 1 v. (various foliations) : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Details the results of in vivo and in vitro studies into the effects of zinc deficiency, hypervitaminosis A, alcohol, nicotine and salicyclic acid, at the morphological and ultrastructural levels in the developing embryo. / Thesis (Ph.D.)--University of Adelaide, Dept. of Anatomy and Histology, 1992

Malgré l’amélioration des prises en charge diabétologiques et obstétricales, la grossesse chez la patiente ayant un diabète pré-gestationnel ou gestationnel reste à ce jour à haut risque pour la mère et pour l’enfant. Chez l’enfant, les anomalies de croissance, macrosomie, mais parfois Retard de Croissance Intra-Utérin (RCIU) restent à ce jour très fréquentes avec des conséquences à court et à long terme. La croissance fœtale est un processus complexe mettant en jeu la susceptibilité génétique fœtale mais surtout le milieu intra-utérin à savoir l’environnement métabolique maternel et placentaire. Les mécanismes physiopathologiques en lien avec ces anomalies de croissance dans ce contexte de diabète restent encore incompris et mal expliqués par l’hyperglycémie maternelle seule. A l’interface entre la mère et le fœtus, le placenta exerce plusieurs fonctions influençant le métabolisme maternel et fœto-placentaire donc le développement de l’unité fœto-placentaire. Le placenta, acteur crucial de la programmation fœtale, va s’adapter à son environnement afin de permettre la survie fœtale.L’objectif de ce travail de thèse était d’étudier le compartiment placentaire en analysant l’expression des gènes impliqués dans la croissance fœto-placentaire afin de déterminer des facteurs prédictifs des anomalies de croissance au cours du diabète maternel. Pour répondre à cet objectif, nous avons d'abord utilisé un modèle de rate gestante rendue diabétique par la streptozotocine seule ou associée avec la nicotinamide et validé certains de nos résultats dans des placentas issus de patientes diabétiques de type 1. L’analyse du transcriptome placentaire a mis en évidence l’implication prépondérante de certains gènes appartenant à la famille prolactine (PRL), au système rénine-angiotensine et aux métalloprotéases. La caractéristique phénotypique de ces ratons était de présenter un RCIU à la naissance avec sur le plan histologique une hypovascularisation placentaire associée.Nous nous sommes surtout intéressés aux gènes placentaires appartenant à la famille PRL, non décrits auparavant dans la littérature dans le diabète, comme prl8a2, connu aussi sous le nom de Dprp (Decidual Prolactin Related-Protein). La PRL dans sa forme native de 23-kDa a des propriétés pro-angiogéniques alors que clivée en vasoinhibines par la Bone morphogenetic protein1 (BMP1), la cathepsine D, a des propriétés anti-angiogéniques. Chez nos 2 modèles de rates, nous confirmons une surexpression par qPCR de Dprp, et de Bmp1 et une augmentation du rapport du clivage de la PRL et donc des vasoinhibines par rapport aux contrôles.Nous avons pu valider ces résultats dans des placentas de patientes diabétiques de type 1 dont la caractéristique chez les nouveaux nés était un petit poids de naissance. Enfin, nous nous sommes intéressés à la cinétique de ces anomalies concernant la famille PRL dans nos modèles animaux. Nous avons pu montrer chez la rate gestante diabétique que le RCIU était présent dès le 14ème jour de gestation et que la quantité en vasoinhibines et l’expression des gènes Bmp1 et Dprp n'étaient modifiées qu'à partir du 17ème jour de gestation.Ces travaux sont en faveur d’une implication de la PRL placentaire et de ses vasoinhibines dans le diabète maternel laissant leur supposer un rôle dans l’hypovascularisation placentaire, mise en évidence à la fois chez l'homme et l'animal. En perspective, nous envisageons de poursuivre ces travaux avec une approche plus fonctionnelle. Il convient de préciser l’implication de la BMP1 en confirmant sa responsabilité dans le clivage de la PRL, en analysant plus finement la relation entre vasoinhibines et hyperglycémie en tenant compte du degré et de la durée d’exposition de l'hyperglycémie. Enfin, il serait intéressant de regarder l’implication de la PRL placentaire non plus au cours du RCIU mais plutôt au cours de la macrosomie fœtale, qui reste l’anomalie de croissance la plus fréquente au cours du diabète maternel. / Despite the improvement of obstetrical and diabetological care, the pregnancy of the patient presenting a gestational or pregestational diabetes remains ourdays at a high risk for the mother and for its child. For the child, fetal growth disorders such as macrosomia but also intra-uterine growth restriction (IUGR) are still very frequent with short and long-term consequences. Fetal growth is a complex process involving the fetal genetic susceptibility but also the intra-uterine environment especially in its maternal and placental metabolic aspects. The link between the physiopathological mechanisms of these disorders and fetal growth in this context of maternal diabetes remains unclear and partially explained by maternal hyperglycemia only. At an interface between the mother and the fetus, the placenta employes multiples functions that influence maternal, fetal and placental metabolisms and consequently the fetoplacental unit development. The placenta, as crucial actor of fetal programming, must adapt to its environnment for the survival of the fetus.The objectives of this thesis were to study the placental compartment with an analysis of expression of genes involved in feto-placental growth to determine the predictive factors of these growth disorders during maternal diabetes. To bring a response to these objectives, we used initially a model of gestant rat diabetes induced by streptozotocin alone or in combination with nicotinamide and we validated some of our results in the placenta from type 1 diabetic mothers.The placental transcriptomic analysis pointed out the involvment of some genes of the prolactin (PRL) family, of the renine-angiotensin-aldosterone system and of metalloproteinase family. The principal phenotypical characteristic of the pups at birth was an IUGR with an histological aspect of a placental hypovascularization associated.We focused especially to the placental genes of the PRL familly, non described before in the litterature in diabetes, such as prl8a2 also known as Dprp (decidual prolactin related-protein). PRL in its native form of 23 kDa is proangiogenic but when processed by Bone morphogenetic protein 1 (BMP-1) or cathepsin D (CTSD) to vasoinhibins has antiangiogenic properties. In our 2 rat models, we demonstrated by qPCR an upregulation of Bmp-1 and Dprp with an increase amount of vasoinhibins when compared to controls.We could validate some of our results in the placenta from diabetic type 1 women with a characteristic of small birth weight of the newborns.Finally, we interested in the course of these disorders concerning PRL family in our animal models during their pregnancy. We could demonstrate that IUGR was present by 14th day of gestation. Bmp-1 or Dprp gene expression and the vasoinhibin amount were not different between groups at the 14th day of gestation but modified by 17th day of gestation.These studies highlighted a placental involvment of PRL and its vasoinhibins during maternal diabetes suggesting a role in placental hypovascularisation in animal and women.The perspectives will be in continuing these studies with a more functional approach. We have to bring more details about the involvment of BMP-1 in this PRL process with an in-depth analysis of the link between hyperglycemia and vasoinhibins among the degree and the time of exposition to hyperglycemia. Finally, it would be interesting to study the involvment of placental PRL not only in the cases of IUGR but also in that of macrosomia, that remains the most frequent fetal growth disorder during maternal diabetes.

MSCAGR (Animal Science) / Department of Animal Science / The main purpose of this study was to assess the efficiency of ultrasonography in monitoring reproductive organs, pregnancy diagnosis, and foetal gender identification and to verify its reliability by laparoscopy and laparotomy, where applicable. Reproductive organs, pregnancy diagnosis and gender of the foetus were examined by A-mode ultrasound using 3.0 - 8.0 MHz trans-rectal transducer. A Sony Olympus Model laparoscope with a camera transducer was used to monitor the reproductive organs and pregnancy diagnosis. In monitoring the follicular dynamics, daily ultrasonography (ULTS) scanning was done for 17 days in sheep and for 21 days in both goats and cattle. Follicles of diameter ≥ 3 mm were selected for analysis of growth, ovulation and regression. For determining the efficiency of the techniques, laparoscopy (LAPSC) and laparotomy (LAPT) were used on days 3 and 10 of the goats and sheep oestrous cycle. The follicles were grouped into three categories according to their diameter as 3 - 4.9 mm, 5 - 7.9 mm and ≥ 8 mm, whereas the follicles of cattle were grouped as 3 - 4.9 mm, 5 - 9.9 mm and ≥ 10 mm. Early pregnancy diagnosis examinations were carried out from day 18 post insemination until pregnancy was confirmed. Foetal gender examinations were conducted from day 40 of pregnancy until the day the gender of the foetus was confirmed. Follicular development was accompanied by the occurrence of waves of follicular growth at different period of the oestrous cycle. The first follicular wave emerged on day 1.0 ± 0.4 in goats, 1.2 ± 0.4 in sheep and 2.2 ± 0.4 in cattle. The maximum diameter of the dominant follicles of observed follicular waves in goats was 7.3 ± 0.4 mm, 6.6 ± 0.2 mm, 7.3 ± 0.2 mm; in sheep was 6.4 ± 0.4 mm, 6.6 ± 0.4 mm and 6.7 ± 0.7 mm and in cattle was 13.1 ± 0.8 mm, 14.2 ± 0.6 mm and 15.7 ± 0.6 mm in wave 1, 2 and 3, respectively. However, the maximum size of the dominant follicle of the ovulatory wave in cattle was larger than the dominant follicles of both first and second waves, but in goats and sheep the dominant follicles were of similar size throughout the waves. In cattle, the ovulatory wave was shorter (p ˂ 0.05) than the duration of the first and second waves, while in sheep and goats were similar throughout the waves. In goats the total number of follicles counted in right and left ovaries under category 3 - 4.9 mm was lower with ULTS and LAPSC than with LAPT method (p ˂ 0.05). In sheep the mean number of follicles between 3 - 4.9 mm category in both right and left ovaries were different (p ˂ 0.05) between ULTS and LAPT. However, for categories 5 - 7.9 mm and ≥ 8 mm in both goats and sheep the mean numbers of follicles observed by all techniques were similar (p ˃ 0.05). In goats, pregnancy diagnosis accuracy improved from zero percent on day 18 to 100% on day 26 - 28, in sheep pregnancy diagnosis was 40% on day 18 and improved to 100% on day 20 - 22 vi of gestation. In cattle accuracy of pregnancy diagnosis was not possible at day 18 and gradually increased to 100% on day 30 - 32 of gestation. Out of 5 (100%) goat’s foetuses whose gender was determined, the diagnosis was correct in 100% (3/3) of the male foetuses and 100% (2/2) of the female foetuses. In sheep two foetuses were sexed as males while the other three were sexed as females and were both 100%. Out of 60% (3/5) of foetuses examined in cattle, 1 (100%) was identified as male and the remaining 2 (100%) were identified as females. The results obtained confirmed that the accuracy for foetal gender by ultrasonography was 100% in all foetuses observed. The current study demonstrated that trans-rectal ultrasonography examination is an efficient method for monitoring follicular dynamics, diagnosing pregnancy and foetal gender identification and that it is as reliable as laparoscopy and laparotomy where they were applied together. / NRF