Serous ovarian cancer is the 5th leading cause of cancer-related deaths in women, with a 30% survival rate when spread into the highly hypoxic and visceral peritoneal cavity. Despite efforts to treat this highly metastatic disease, traditional chemotherapeutic and cytoreductive therapies are unable to diminish or induce cell death of circulating metastases from colonizing secondary sites due to their genetic and histologic heterogeneity and development of drug resistance. The dissemination route for primary metastasis, however, is most often conserved to the peritoneal cavity, which is low in nutrients and hypoxic (1-2% O2). Cells exfoliated from the primary tumor will aggregate during migration, which elicits a survival signal to maintain viability in this environment. The underlying cellular and molecular changes involved with aggregation have yet to be determined. We have previously found that aggregation of murine ovarian surface epithelial (MOSE) cells present a more suppressed metabolic phenotype upon aggregation. My research sought to identify how the mitochondria were internally regulated to support malignant transformation, migration, and invasion through modulation of quality control, mitochondrial dynamics, mitophagy, and mitobiogenesis. We have shown that aggregation of cancer cells supports increased mitochondrial fragmentation localized to the hypoxic core of our spheroid models. Further, aggregation supports enhanced viability through an upregulation of cancer genetic pathways associated with cell death, proliferation, stemness, and epithelial mesenchymal transition (EMT). Nutrient deprivation during migration further enhanced mitochondrial fragmentation and induction of mitophagy to prevent activation of apoptosis. Additionally, we have identified a phenotypic switch from enhanced mitophagy during peritoneal dissemination that supports survival of ovarian cancer cell aggregates to mitochondrial biogenesis during secondary tissue colonization that enables proliferation upon invasion. We have associated these changes with an increased bioenergetic proliferative niche through inhibition of proliferation, migration, and mitochondrial translation. This research has contributed to the understanding for the role of mitophagy as a survival rather than apoptotic signal in cancer cells as adaptation to nutrient-deprived environments, while also identifying how these processes can be reversed upon adhesion to support invasion and metastatic capacity during secondary colonization. This research is significant because it will identify molecular adaptations associated with the viability of disseminating cancer metastases as well as promote novel preventative therapeutics that can be used to limit the mortality of highly aggressive ovarian cancer in women. / Doctor of Philosophy / Ovarian cancer continues to be one of the highest contributors of gynecologic cancer-related deaths in women. This is due to limited symptomology, biomarker availability, and screenings for patients. Women are mostly diagnosed when the disease has already spread throughout the abdominal cavity which makes treatment much more difficult and, accordingly, the survival rate is much lower. Ovarian metastases mostly spread throughout the peritoneal cavity. Interestingly, this cavity has been identified to being limited in nutrients and oxygen that are essential for survival thus suggesting that these cancer cells must adapt to these harsh conditions to remain viable. We have previously observed that the cancer cells are able to clump together, and form 3D structures known as spheroids which have drastically reduced their proliferation and appear highly resistant tor treatment than single cells. In this project, we wanted to determine how the mitochondria (primary energy producers) were structurally changing in response to the formation of these spheroids and in nutrient- and oxygen-starved conditions. We have found that these organelles become much smaller and circular in low-oxygen conditions, especially in the center of the spheroids. Further, we found changes in cancer- and mitochondrial-related pathways during spheroid formation which could further support survival. Finally, we found that key functions related to the mitochondrial quality control and enhanced mitochondrial content and activity are switched when changing nutrient availability from low oxygen and nutrient conditions to oxygenated and nutrient-rich conditions and generate conditions that allow the spheroids to attach to abdominal organs and form secondary tumors. This research is important because it suggests new possible markers that can be used as therapeutic targets to prevent these aggressive functions associated with more terminally staged disease.

In this thesis, three parts, each with several chapters, are respectively devoted to hydrostatic, viscous and inertial fluids theories and applications. In the hydrostatics part, the classical Maclaurin spheroids theory is generalized, for the first time, to a more realistic multi-layer model, which enables the studies of some gravity problems and direct numerical simulations of flows in fast rotating spheroidal cavities. As an application of the figure theory, the zonal flow in the deep atmosphere of Jupiter is investigated for a better understanding of the Jovian gravity field. High viscosity flows, for example Stokes flows, occur in a lot of processes involving low-speed motions in fluids. Microorganism swimming is such typical a case. A fully three dimensional analytic solution of incompressible Stokes equation is derived in the exterior domain of an arbitrarily translating and rotating prolate spheroid, which models a large family of microorganisms such as cocci bacteria. The solution is then applied to the magnetotactic bacteria swimming problem and good consistency has been found between theoretical predictions and laboratory observations of the moving patterns of such bacteria under magnetic fields. In the analysis of dynamics of planetary fluid systems, which are featured by fast rotation and very small viscosity effects, three dimensional fully nonlinear numerical simulations of Navier-Stokes equations play important roles. A precession driven flow in a rotating channel is studied by the combination of asymptotic analyses and fully numerical simulations. Various results of laminar and turbulent flows are thereby presented. Computational fluid dynamics requires massive computing capability. To make full use of the power of modern high performance computing facilities, a C++ finite-element analysis code is under development based on PETSc platform. The code and data structures will be elaborated, along with the presentations of some preliminary results.

Le mélanome métastatique est le cancer de la peau le plus agressif. Bien que son taux d’incidence soit inférieur à 1%, plus de 75% des décès associés à un cancer de la peau lui sont attribués. Au cours des dernières années, de nouvelles stratégies thérapeutiques ont permis d’améliorer la survie des patients. Cependant, des mécanismes de résistance à ces traitements se développent dans la majorité des cas, conduisant à une phase de rechute, et une survie à 5 ans inférieure à 20%. Des modèles d’étude expérimentaux sont nécessaires afin de comprendre les mécanismes impliqués dans l’apparition de ces résistances et développer de nouvelles stratégies thérapeutiques. Différents modèles in vitro sont actuellement utilisés pour le développement de drogues anti-tumorales, tels que celui du sphéroïde. Bien qu’il permette de reproduire l’organisation tridimensionnelle d’une tumeur, l’absence de microenvironnement tumoral empêche l’étude des interactions entre les cellules tumorales et celui-ci alors que ces facteurs jouent un rôle primordial dans la croissance tumorale et le développement de métastases. Dans ce contexte, mes travaux ont porté sur le développement et la caractérisation d’un modèle ex vivo de mélanome humain complet permettant l’étude de l’évolution d’une tumeur dans le tissu sain et l’évaluation de composés pharmacologiques. Les travaux réalisés ont tout d’abord conduit au développement d’un modèle de cancer cutané basé sur la combinaison d’un modèle de sphéroïde de lignée cellulaire de mélanome humain et du modèle de peau humaine ex vivo NativeSkin®, développé par la société Genoskin. Une procédure a été développée et validée pour permettre l’implantation reproductible d’un sphéroïde dans le derme des explants de peau. Parallèlement, j’ai développé une approche d’imagerie in situ par microscopie à feuille de lumière après transparisation des modèles. J’ai également développé une stratégie d’analyse d’images permettant la caractérisation quantitative de l'évolution du sphéroïde implanté en 3 dimensions et de suivre la dispersion des cellules du tumorales au sein de l’explant de peau. La caractérisation histologique du modèle implanté a révélé de façon très inattendue une perte progressive de l’intégrité du sphéroïde après implantation, associée à une diminution rapide de la prolifération des cellules le constituant et l’apoptose massive des cellules situées à sa périphérie. Ce phénomène a été observé de façon similaire lors de l’implantation de sphéroïdes produits à partir de différents types cellulaires. Afin de comprendre ces résultats, j’ai étudié l’implication potentielle de différents paramètres dans l’induction de la mortalité cellulaire observée tels que les conditions d’implantation, les facteurs synthétisés par le modèle et la contrainte mécanique exercée par le derme. Les résultats obtenus suggèrent que les facteurs sécrétés par les modèles après implantation du sphéroïde ont un effet antiprolifératif sur les sphéroïdes de mélanome et qu’ils induisent la mortalité des cellules situées à sa périphérie. Par ailleurs, l’application d’une contrainte mécanique extérieure sur les sphéroïdes de mélanome entraîne la perte de la cohésion de leur structure. Enfin, l’implantation de sphéroïdes dans le derme de biopsies de peau préalablement desséchées, induisant une perte de la viabilité cellulaire, a conduit à des résultats opposés à ceux observés avec de la peau normale : la structure des sphéroïdes reste cohésive et la prolifération des cellules est maintenue en périphérie du sphéroïde sans qu’aucune apoptose massive ne soit observée. L'ensemble de ces travaux semble suggérer que la mortalité du sphéroïde pourrait être, en partie, la conséquence d’une contrainte mécanique exercée par la peau sur le sphéroïde et/ou de facteurs produits par la peau durant sa culture. Ces données ouvrent des perspectives intéressantes dans le domaine de l’ingénierie tissulaire pour l’évaluation pharmacologique de composés thérapeutiques. / Malignant melanoma is the most aggressive form of skin cancer. Although it only occurs in less than 1%, it is responsible for more than 75% of skin cancer-related deaths. Furthermore, melanoma incidence has constantly increased during the last decades. New therapies such as targeted therapy and immunotherapy have emerged over the past years, significantly improving the overall survival rates of patients with advanced melanoma stages. However, resistance to those treatments develops in most cases, leading to relapse with a 5-years survival of those patients under 20%. Experimental models are needed in order to better understand the molecular events underlying these resistance mechanisms, and to develop new therapeutic strategies. MultiCellular Tumor spheroid is an increasingly recognized 3D in vitro model for pharmacological evaluation. Although this model accurately reproduces the 3D architecture, cell-cell interaction and cell heterogeneity found in microtumor in vivo, spheroids lack tumor-microenvironment interactions, which play a key role in tumor growth and metastasis development. In this context, the aim of my project was to develop and characterize a fully ex vivo human melanoma model for the study of tumor growth within the skin and the evaluation of antitumor drugs. Our approach relies on the combination of human melanoma cell lines grown in Multicellular Tumor Spheroids and the NativeSkin® model, an ex vivo human skin model produced by the biotechnology company Genoskin. Hence, I developed and validated a method to reproducibly implant one spheroid into the dermal compartment of skin explants cultured ex vivo. In parallel I have developed in situ imaging strategies based on light-sheet microscopy (SPIM, “Selective Plane Illumination Microscopy”) after optical clearing of the implanted skin biopsies. I also developed analytic methods to allow for the quantitative characterization of the spheroids evolution in 3 dimensions as well as tumor cells dispersal within the dermis of skin explants. Histological characterization of the implanted models over time revealed a progressive loss of the spheroids integrity after implantation associated with a rapid decrease in cell proliferation and massive apoptosis of the cells located in the peripheral layers. These results were shared by implanted spheroids made from different cell types. Further experiments were conducted in order to better understand these results and evaluate the impact of different parameters on the implanted microtumors viability such as the implantation procedure conditions, factors synthesized by the model after spheroid implantation and external mechanical stress. Results suggest that factors produced by the implanted models have an antiproliferative effect on melanoma spheroids and induce mortality in the peripheral layers of the spheroids. Moreover, results show that mechanical stress applied on melanoma spheroids induces loss of their cohesion. Finally, implantation of spheroids within the dermis of previously dessicated biopsies for 7 days, causing loss of skin cells viability, led to opposite results than in normal skin: spheroids maintain both a cohesive structure and proliferation in the peripheral cells without any massive apoptosis. Overall, this work led to the validation of a methodology to reproducibly implant spheroids into an ex vivo skin explant and the setup of an optical clearing technique necessary for in situ imaging of the implanted spheroid. Histological characterization unexpectedly revealed spheroids cells death following their implantation. Results suggest that this mortality could be partly related to mechanical stress exerted on the spheroids by the skin and/or by factors produced by the skin during culture. These data open new perspectives in the research field of tissue engineering for antitumoral pharmacology.

In this Thesis, we have developed a microtechnology-based method for culturing and visualizing high numbers of individual cells and cell-cell interactions over extended periods of time. The foundation of the device is a silicon-glass multiwell microplate (also referred as microchip) directly compatible with fluorescence microscopy. The initial microchip design involved thousands of square wells of sizes up to 80 µm, for screening large numbers of cell-cell interactions at the single cell level. Biocompatibility and confinement tests proved the feasibility of the idea, and further investigation showed the conservation of immune cellular processes within the wells. Although the system is very reliable for screening, limitations related to synchronization of the interaction events, and the inability to maintain conjugations for long time periods, led to the development of a novel ultrasonic manipulation multiwell microdevice. The main components of the ultrasonic device is a 100-well silicon-glass microchip and an ultrasonic transducer. The transducer is used for ultrasonic actuation on the chip with a frequency causing half-wave resonances in each of the wells (2.0-2.5 MHz for wells with sizes 300-350 µm). Therefore, cells in suspension are directed by acoustic radiation forces towards a pressure node formed in the center of each well. This method allows simultaneous aggregation of cells in all wells and sustains cells confined within a small area for long time periods (even up to several days). The biological target of investigation in this Thesis is the natural killer (NK) cells and their functional properties. NK cells belong to the lymphatic group and they are important factors for host defense and immune regulation. They are characterized by the ability to interact with virus infected cells and cancer cells upon contact, and under suitable conditions they can induce target cell death. We have utilized the ultrasonic microdevice to induce NK-target cell interactions at the single cell level. Our results confirm a heterogeneity within IL-2 activated NK cell populations, with some cells being inactive, while others are capable to kill quickly and in a consecutive manner. Furthermore, we have integrated the ultrasonic microdevice in a temperature regulation system that allows to actuate with high-voltage ultrasound, but still sustain the cell physiological temperature. Using this system we have been able to induce formation of up to 100 solid tumors (HepG2 cells) in parallel without using surface modification or hydrogels. Finally, we used the tumors as targets for investigating NK cells ability to infiltrate and kill solid tumors.  To summarize, a method is presented for investigating individual NK cell behavior against target cells and solid tumors. Although we have utilized our technique to investigate NK cells, there is no limitation of the target of investigation. In the future, the device could be used for any type of cells where interactions at the single cell level can reveal critical information, but also to form solid tumors of primary cancer cells for toxicology studies. / <p>QC 20150113</p>

京都大学 / 新制・課程博士 / 博士(医学) / 甲第23062号 / 医博第4689号 / 新制||医||1048(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 武藤 学, 教授 妹尾 浩, 教授 小川 誠司 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Le glioblastome est un cancer du cerveau très agressif dont les thérapies actuelles n’augmentent que très peu la durée de vie. Dans cette thèse, nous étudions un nouveau traitement par effet magnéto-mécanique de particules (TEMMP). Un champ magnétique rotatif à faible fréquence (20 Hz) est appliqué pour faire vibrer des particules magnétiques en contact avec les cellules cancéreuses. Les particules développées sont produites par une approche top-down en salle blanche. Les disques de permalloy utilisés présentent une configuration en vortex avec une faible rémanence et une bonne dispersion en suspension. Des particules multicouches de Co/Pt avec une anisotropie perpendiculaire et des vortex de permalloy en forme d’ellipses sont aussi étudiés. L’efficacité du TEMMP est évaluée in-vitro sur des cellules de glioblastome et les différents paramètres sont optimisés. Une forte diminution du nombre de cellules après traitement est alors observée et le comportement des cellules restantes est affecté. Le TEMMP est ensuite adapté pour une étude in-vivo dans un modèle orthotopique de glioblastome chez la souris nude. L’injection des particules en intra-tumoral est mise au point. Les tissus sont peu affectés par le TEMMP comparé à une injection de particules, et une faible augmentation de la survie est observée. Pour mimer les propriétés mécaniques du cerveau de manière plus pertinente, un modèle in-vitro 3D est alors développé et validé. Conçu avec des sphéroïdes de cellules pris dans un gel d’agarose, ce modèle apporte des pistes d’optimisation. / Glioblastoma is a brain cancer with a very poor prognosis. Existing therapies improve only slightly the median survival. In this work, we study a new treatment by magneto-mechanical actuation of particles (TMMAP). A low frequency (20 Hz) rotating magnetic field is applied to stimulate magnetic particles localized near cancer cells. Magnetic particles are produced by a top-down approach in clean room. Permalloy disks with a vortex configuration have a low remanence and are well dispersed in suspension. Multilayers of Co/Pt with a perpendicular anisotropy and permalloy vortex particles with an ellipse shape are also studied. TMMAP efficiency is tested in-vitro on glioblastoma cell line and the parameters are optimized. A huge diminution of living cells and an affected behavior of the remaining cells are observed after treatment. TMMAP is then adapted to an in-vivo study on glioblastoma orthotopic model on nude mice and the intratumoral injection of the particles is developed. Few differences are observed between tissues submitted to TMMAP or injected with particles, and survival is slightly increased. To mimic mechanical properties of the brain in a more relevant model, an in-vitro 3D model is proposed and validated. Based on cells grown as a spheroid and encapsulated in an agarose gel, this model brings optimization tracks.

No description available.

No description available.

Immunohistological (IHC) stained images occupy a fundamental role in the pathologist's diagnosis and monitoring of cancer development. The manual process of monitoring such images is a subjective, time consuming process that typically relies on the visual ability and experience level of the pathologist. A novel and comprehensive system for the automated quantification of antibody inside stained cell nuclei in immunohistochemistry images is proposed and demonstrated in this research. The system is based on a cellular level approach, where each nucleus is individually analyzed to observe the effects of protein antibodies inside the nuclei. The system provides three main quantitative descriptions of stained nuclei. The first quantitative measurement automatically generates the total number of cell nuclei in an image. The second measure classifies the positive and negative stained nuclei based on the nuclei colour, morphological and textural features. Such features are extracted directly from each nucleus to provide discriminative characteristics of different stained nuclei. The output generated from the first and second quantitative measures are used collectively to calculate the percentage of positive nuclei (PS). The third measure proposes a novel automated method for determining the staining intensity level of positive nuclei or what is known as the intensity score (IS). The minor intensity features are observed and used to classify low, intermediate and high stained positive nuclei. Statistical methods were applied throughout the research to validate the system results against the ground truth pathology data. Experimental results demonstrate the effectiveness of the proposed approach and provide high accuracy when compared to the ground truth pathology data.

Le cancer épithélial de l’ovaire (EOC) est le plus mortel des cancers gynécologiques. Cette maladie complexe progresse rapidement de façon difficilement décelable aux stades précoces. De plus, malgré une chirurgie cytoréductive et des traitements de chimiothérapie le taux de survie des patientes diagnostiquées aux stades avancées demeurt faible. Dans le but d’étudier l’EOC dans un contexte ex vivo, l’utilisation de modèles cellulaires est indispensable. Les lignées cellulaires d’EOC sont un outil pratique pour la recherche cependant, la façon dont l'expression des gènes est affectée en culture par comparaison à la tumeur d'origine n'est pas encore bien élucidée. Notre objectif était donc de développer et de caractériser de nouveaux modèles de culture in vitro qui réflèteront plus fidèlement la maladie in vivo. Nous avons tout d’abord utiliser des lignées cellulaires disponibles au laboratoire afin de mettre au point un modèle 3D de culture in vitro d’EOC. Des sphéroïdes ont été générés à l’aide de la méthode des gouttelettes inversées, une méthode pionnière pour la culture des cellules tumorales. Nous avons ensuite procédé à une analyse des profils d’expression afin de comparer le modèle sphéroïde au modèle de culture en monocouche et le modèle xénogreffe in vivo. Ainsi, nous avons identifié des gènes stratifiant les modèles tridimensionnels, tant in vivo qu’in vitro, du modèle 2D monocouche. Parmi les meilleurs candidats, nous avons sélectionné S100A6 pour une caractérisation ultérieure. L’expression de ce gène fût modulée afin d’étudier l’impact de son inhibition sur les paramètres de croissance des sphéroïdes. L’inhibition de ce gène a comme effet de réduire la motilité cellulaire mais seulement au niveau du modèle sphéroïde. Finalement, toujours dans l’optique de développer des modèles d’EOC les plus représentatifs de la maladie in vivo, nous avons réussi à développer des lignées cellulaires uniques dérivées de patientes atteintes d’EOC du type séreux, soit le plus commun des EOC. Jusque là, très peu de lignées cellulaires provenant de ce type de cancer et de patientes n’ayant pas reçu de chimiothérapie ont été produites. De plus, nous avons pour la première fois caractérise des lignées d’EOC de type séreux provenant à la fois de l’ascite et de la tumeur solide de la même patiente. / The epithelial ovarian cancer (EOC) is the most lethal of gynecological cancers. This complexe and heterogenous disease progresses rapidly and is almost asymptomatic in early stages. The survival rate of patients with late stage diagnosis remains low albeit cytoreductive surgery and chemotherapy. In order to study the EOC disease in an ex vivo context, the use of different cellular models is necessary. EOC cell lines derived from long-term passages of malignant ovarian cancers are useful tools for molecular and cellular research but it is not clear how culture conditions affect overall gene expression and oncogenic potential as compared to the original tumor. The main goal of this research was to develo and characterize new in vitro model systems that will recapitulate more closely some of the growth conditions encountered by tumor cells in vivo. In order to develop an in vitro tridimensional EOC spheroid model, we have used cell lines previously established in our laboratory. Spheroids were generated using the hanging droplet method, which was innovative for the culture of cancer cells. Comparative gene expression profile analysis of monolayer cultures, 3D spheroids and in vivo xenografts were performed and we have shown that the spheroid transcriptome more closely reflects expression patterns of the in vivo model compared to that of monolayer cultures. Among the best candidates, S100A6 gene over-expressed in the 3D models versus monolayer cultures was chosen for further analysis. To begin to address how S100A6 might affect EOC growth parameters, we have inhibited its expression in our in vitro models. The loss of S100A6 in the spheroid model results in an reduction of cellular migration, which seems to be in line with previous in vivo results published by other researchers. Always with the objective of developing the most relevant to the in vivo disease model systems, we have also succeeded in developing a unique EOC cell lines derived from patients with the most frequently diagnosed serous type of cancer. Very few cell lines derived from this type of cancers and from chemotherapy naïve patients are available. Moreover, we characterize for the first time EOC serous type cell lines derived from the ascites and the solid tumor of the same patient.