Behavioral operations management (BOM) has gained popularity in the last two decades. The main theme in this new stream of research is to include the human behavior in Operations Management (OM) models to increase the effectiveness of such models. BOM is classified into 4 areas: cognitive psychology, social psychology, group dynamics and system dynamics (Bendoly et al. 2010). This dissertation will focus on the first class, namely cognitive psychology. Cognitive psychology is further classified into heuristics and biases. Tversky and Kahneman (1974) discussed 3 heuristics and 13 cognitive biases that usually face decision makers. This dissertation is going to study 6 cognitive biases under the representativeness heuristic. The model in this dissertation states that cognitive reflection of the individual (Frederick 2005) and training about cognitive biases in the form of warning (Kaufmann and Michel 2009) will help decisions’ makers make less biased decisions. The 6 cognitive biases investigated in this dissertation are insensitivity to prior probability, insensitivity to sample size, misconception of chance, insensitivity to predictability, the illusion of validity and misconception of regression. 6 scenarios in OM contexts have been used in this study. Each scenario corresponds to one cognitive bias. Experimental design has been used as the research tool. To see the impact of training, one group of the participants received the scenarios without training and the other group received them with training. The training consists of a brief description of the cognitive bias as well as an example of the cognitive bias. Cognitive reflection is operationalized using cognitive reflection test (CRT). The survey was distributed to students at University of North Texas (UNT). Logistic regression has been employed to analyze data. The research shows that participants show the cognitive biases proposed by Tversky and Kahneman. Moreover, CRT is significant factor to predict the cognitive bias in two scenarios. Finally, providing training in terms of warning helps participants to make more rational decisions in 4 scenarios. This means that although cognitive biases are inherent in the mind of people, management of corporations has the tool to educate its managers and professionals about such biases which helps companies make more rational decisions.

Pain is a major public health issue throughout the world. Increased understanding of the different forms of pain and identification of susceptibility genes could contribute to improved treatments. The main aims of this thesis were to identify the underlying genetic causes of pain by studying two large families affected with migraine and pain insensitivity, respectively. Migraine is one of the most common neurovascular disorders, affecting over 12% of the western population. The genetic contribution to migraine is about 50% according to family and twin studies. To identify novel susceptibility loci for migraine, we performed a genome-wide screen in a large family with migraine from northern Sweden. Linkage analysis revealed significant evidence of linkage (LOD=5.41) on chromosome 6p12.2-p21.1. A predisposing haplotype spanning 10 Mb was inherited with migraine in all affected members of the pedigree. Further fine-mapping of multiple SNP markers restricted the disease critical region to 8.5 Mb. Nine candidate genes were sequenced, revealing no disease-associated polymorphisms in SLC29A1, CLIC5, PLA2G7, IL17, SLC25A27 and TNFRSF21, but rare novel polymorphisms segregating with the disease haplotype in EFHC1, RHAG and MEP1A. EFHC1 has recently been shown to be involved in epilepsy, which is interesting considering the link between migraine and epilepsy. However, association analysis of EFHC1 revealed no difference between patients and controls, suggesting that this gene is not a risk factor for migraine. The combination of the two polymorphisms in RHAG and MEP1A could, however, not be found in any control individuals, indicating that they might be involved in genetic predisposition to migraine in this family. Disorders with reduced pain sensitivity are very rare, since pain perception is essential for survival. A number of disorders have still been identified with pain insensitivity and peripheral nerve degeneration as major clinical signs, including the hereditary sensory and autonomic neuropathies (HSAN). In order to identify novel susceptibility genes for HSAN V, we performed a genome-wide screen in a large consanguineous pedigree from a small village in northern Sweden. A homozygous region identical-by-descent was identified on chromosome 1p11.2-p13.2 in the three most severely affected patients. Subsequent analysis of candidate genes revealed a missense mutation in a conserved region of the nerve growth factor beta (NGFB) gene, causing a drastic amino acid change (R211W) in the NGF protein. NGF is important for the development and maintenance of the sympathetic and sensory nervous system and is therefore likely to be involved in disease. Functional analysis revealed that mutant NGF failed to induce neurite outgrowth and cell differentiation in PC12 cells. Furthermore, almost no mutant NGF was secreted by COS-7 cells, indicating that the processing and/or secretion of the protein might be disrupted. In conclusion, these findings present a novel migraine locus on chromosome 6 and identification of two rare polymorphisms that might be risk factors for migraine. Furthermore, a mutation in NGFB was found to cause complete loss of deep pain perception, which represents a very interesting model system to study pain mechanisms.

The interactions between Medicago truncatula and Phytophthora medicaginis were examined using skl, a mutant blocked in ethylene perception, and a range of wild accessions of this plant species. P. medicaginis infection of M. truncatula plants resulted in compatible responses, whereas the mutant genotype was found to be hyper-susceptible to the pathogen. Phytophthora reproduction and colonization rates of Medicago tissues supported this conclusion. Infection of skl with different pathogens reinforced this observation. Ethylene production in infected A17 and skl roots showed reduced ethylene evolution in the mutant and suggested that a positive feedback loop, known as autocatalytic ethylene production, amplified the ethylene signal. To complement the study, expression analyses of defense response genes in this interaction were studied by real time RTPCR of Phytophthora-infected and mock-infected roots. The genes analyzed were PAL, CHS, IFR, ACC oxidase, GST, and PR10. The sequences needed for the analysis were found through the scrutiny of the M. truncatula EST database employing phylogenetics and bio-informatics tools. In A17 all the genes studied were up-regulated, although the specific gene expression patterns differed. The comparison of gene expression between A17 and skl genotypes allowed the differentiation between ethylene-dependent and ethylene-independent responses. Discrete results showed that ACC oxidase homologues were downregulated in the ethylene perception mutant, corroborating the ethylene observations. However, the expression of genes involved in the phenylpropanoid metabolism was increased in skl relative to A17, suggestive of an antagonism between the ethylene perception pathway and the regulation of the phenylpropanoid pathway. This result implied that Medicago phytoalexins accumulate in the disease interaction, but raised questions about their role in resistance to Phytophthora infection. This study establishes a link between mechanisms that regulate symbiotic infection and the regulation of disease resistance to Oomycete pathogens, especially P. medicaginis. The results served to identify a series of Phytophthora-induced genes, which remain pathogen-responsive even in the absence of a functional ethylene perception pathway. While it is possible that the products of these genes are involved in resistance to P. medicaginis, the present results demonstrate that ethylene perception is required for resistance.

Un grand nombre d'études abordant le thème de la contrefaçon traite le comportement de consommation de l'article contrefait. Mais peu de recherches s'intéressent au consommateur de la marque de luxe lorsque celle-ci fait l'objet de contrefaçon. Ce travail doctoral vise à contribuer à ce niveau en étudiant les principaux aspects du comportement d'achat de la marque de luxe légitime. Cette recherche passe par l'identification et la mesure de l'insensibilité à la contrefaçon, nouveau concept développé au cours de ce travail, puis par l'étude de l'expérience de la marque et des variables relationnelles à la marque. L'intérêt de cette étude s'inscrit dans la contribution aux recherches sur la contrefaçon et plus précisément dans l'étude du comportement de consommation des articles originaux de la marque de luxe. D'un autre côté, le rapprochement avec l'expérience de marque ou encore avec la relation à la marque n'a pas pu être faite jusqu'à présent dans des recherches traitant de la contrefaçon, d'où l'intérêt de se pencher sur ces deux aspects importants de la marque de luxe. En termes méthodologiques, nous proposons tout d'abord une mesure de la réaction du consommateur de la marque de luxe légitime, à savoir l'échelle de l'insensibilité à la contrefaçon. Ensuite, nous mettons en place une modélisation sur PLS qui englobe les différents concepts étudiés. La capacité du modèle à refléter des relations complexes représente un apport méthodologique important. Au niveau managérial, il s'agit de contrer la contrefaçon en valorisant la marque de luxe. L'apport principal de cette recherche réside donc dans le management de la marque de luxe en elle-même. Nous suggérons aux gestionnaires de la marque de luxe de mieux choisir leurs actions marketing dans un contexte de contrefaçon. Favoriser l'immersion dans l'univers de la marque, impliquer le consommateur et renforcer sa relation avec la marque, ou encore cibler plus particulièrement les prescripteurs et les leaders d'opinion, seraient des moyens efficaces pour renforcer la fidélité du consommateur à sa marque de luxe et ce, malgré la présence de contrefaçon. / A large number of studies on counterfeiting explore consumers behavior in the consumption of counterfeit articles. But few of them consider luxury brand consumers and counterfeiting. Our doctoral study attempts to contribute to this field by studying principal aspects of luxury brand buying behavior. Our research aims to identify and measure insensitivity to counterfeiting which is a new concept developed in this study. Brand experience and relationship brand variables will also be considered. Our study contributes to counterfeiting's researches, and more precisely to researches on consumption of legitimate products. On the other side, the connection with brand experience or the relationships with the luxury brand have not been undertaken in counterfeiting researches which leads us to considerate these two important aspects of the luxury brand. Concerning the methodological facet, we first propose the measure of consumer's reaction toward counterfeiting, namely insensitivity to counterfeiting. Then, we perform a PLS model that includes all concepts studied and specifies complex links between concepts. In a managerial way, we propose to counter counterfeiting by focusing on the luxury brand itself. We suggest for luxury brand managers to choose the appropriate marketing tool. Immersion in brand universe, consumer implication or focusing on opinion leaders, are good ways to enhance consumer's loyalty, despite the presence of counterfeiting.

Made available in DSpace on 2016-08-16T18:18:42Z (GMT). No. of bitstreams: 1 Tese MARCELO SOUZA DE ANDRADE.pdf: 1768433 bytes, checksum: 8051d8f14c6a38ec8166fcb2e817b674 (MD5) Previous issue date: 2012-04-25 / FUNDAÇÃO DE AMPARO À PESQUISA E AO DESENVOLVIMENTO CIENTIFICO E TECNOLÓGICO DO MARANHÃO / Introduction. The androgen insensitivity syndrome (AIS) is a rare disease (1:20,000 to 1:64.000)-linked X chromosome, which generates a disorder of sexual differentiation of the male fetus (XY) with a spectrum of phenotypes ranging from females complete (CAIS) to a male phenotype with discrete signs of androgen insensitivity. An increasing number of mutations have been cataloged and nearly 500 mutations have been related to CAIS and 1000 to the androgen receptor gene. The AR gene is located on Xq11-12, with eight exons, about 919 amino acids. Objective. To characterize the mutations in the AR gene families in the region of the "Bico do Papagaio" in the southwestern state of Maranhao. Methodology. We used molecular biology techniques such as DNA extraction, PCR, electrophoresis, purification of PCR products and sequencing. In addition, we analyzed the clinical and hormonal characteristics of 14 patients and their families. Results. In one family (with two twin affected), we found the mutation R753X and the third molecular diagnosis of CAIS in twins described in the World. In another family, with 12 patients, was identified a new mutation in exon 8, as described P893A, protein AR. Conclusion. This work enabled the application of molecular techniques for the accurate diagnosis of AIS, genetic counseling for relatives of affected patients, and contribute to the formation of more qualified human resources, aiming at the development of biotechnology in the state of Maranhão. / Introdução. A Síndrome de Insensibilidade Androgênica (AIS) é uma doença rara (1:20.000 a 1:64.000), de transmissão ligada ao cromossomo X, que gera um distúrbio da diferenciação sexual do feto masculino (XY) com um espectro de fenótipo que varia desde o feminino completo (CAIS) até um fenótipo masculino com discretos sinais de insensibilidade androgênica. Um número crescente de mutações tem sido catalogadas e quase 500 mutações já foram relacionadas à CAIS e cerca de 1000 ao gene do receptor androgênico. O gene AR localiza-se em Xq11-12, com 8 exons, com cerca de 919 aminoácidos. Objetivo. Caracterizar as mutações no gene AR em famílias da região do Bico do Papagaio , no sudoeste do Estado do Maranhão. Metodologia. Foram utilizadas técnicas de biologia molecular como extração de DNA, PCR, Eletroforese, Purificação de produtos de PCR e Sequenciamento automático. Além disso, foram analisados o quadro clínico e hormonal de 14 pacientes e de seus familiares. Resultados. Em uma das famílias (com duas gêmeas afetadas), foi encontrada a mutação R753X, sendo o terceiro diagnóstico molecular de CAIS em gêmeas descrito no Mundo. Em outra família, com 12 pacientes, foi a identificada uma mutação nova no exon 8, descrita como P893A, na proteína AR. Conclusão. Este trabalho possibilitou a aplicação de técnicas moleculares para o diagnóstico preciso de AIS, aconselhamento genético aos familiares das pacientes afetadas, além de contribuir para a formação de recursos humanos mais qualificados, visando o desenvolvimento da biotecnologia no Estado do Maranhão.

La demande croissante pour les services de cloud computing encourage les opérateurs à optimiser l’utilisation des ressources dans les grappes d’ordinateurs. Cela motive le développement de nouvelles technologies qui rendent plus flexible la gestion des ressources. Cependant, exploiter cette flexibilité pour réduire le nombre d’ordinateurs nécessite aussi des algorithmes de gestion des ressources efficaces et dont la performance est prédictible sous une demande stochastique. Dans cette thèse, nous concevons et analysons de tels algorithmes en utilisant le formalisme de la théorie des files d’attente.Notre abstraction du problème est une file multi-serveur avec plusieurs classes de clients. Les capacités des serveurs sont hétérogènes et les clients de chaque classe entrent dans la file selon un processus de Poisson indépendant. Chaque client peut être traité en parallèle par plusieurs serveurs, selon des contraintes de compatibilité décrites par un graphe biparti entre les classes et les serveurs, et chaque serveur applique la politique premier arrivé, premier servi aux clients qui lui sont affectés. Nous prouvons que, si la demande de service de chaque client suit une loi exponentielle indépendante de moyenne unitaire, alors la performance moyenne sous cette politique simple est la même que sous l’équité équilibrée, une extension de processor-sharing connue pour son insensibilité à la loi de la demande de service. Une forme plus générale de ce résultat, reliant les files order-independent aux réseaux de Whittle, est aussi prouvée. Enfin, nous développons de nouvelles formules pour calculer des métriques de performance.Ces résultats théoriques sont ensuite mis en pratique. Nous commençons par proposer un algorithme d’ordonnancement qui étend le principe de round-robin à une grappe où chaque requête est affectée à un groupe d’ordinateurs par lesquels elle peut ensuite être traitée en parallèle. Notre seconde proposition est un algorithme de répartition de charge à base de jetons pour des grappes où les requêtes ont des contraintes d’affectation. Ces deux algorithmes sont approximativement insensibles à la loi de la taille des requêtes et s’adaptent dynamiquement à la demande. Leur performance peut être prédite en appliquant les formules obtenues pour la file multi-serveur. / The growing demand for cloud-based services encourages operators to maximize resource efficiency within computer clusters. This motivates the development of new technologies that make resource management more flexible. However, exploiting this flexibility to reduce the number of computers also requires efficient resource-management algorithms that have a predictable performance under stochastic demand. In this thesis, we design and analyze such algorithms using the framework of queueing theory.Our abstraction of the problem is a multi-server queue with several customer classes. Servers have heterogeneous capacities and the customers of each class enter the queue according to an independent Poisson process. Each customer can be processed in parallel by several servers, depending on compatibility constraints described by a bipartite graph between classes and servers, and each server applies first-come-first-served policy to its compatible customers. We first prove that, if the service requirements are independent and exponentially distributed with unit mean, this simple policy yields the same average performance as balanced fairness, an extension to processor-sharing known to be insensitive to the distribution of the service requirements. A more general form of this result, relating order-independent queues to Whittle networks, is also proved. Lastly, we derive new formulas to compute performance metrics.These theoretical results are then put into practice. We first propose a scheduling algorithm that extends the principle of round-robin to a cluster where each incoming job is assigned to a pool of computers by which it can subsequently be processed in parallel. Our second proposal is a load-balancing algorithm based on tokens for clusters where jobs have assignment constraints. Both algorithms are approximately insensitive to the job size distribution and adapt dynamically to demand. Their performance can be predicted by applying the formulas derived for the multi-server queue.

La demande croissante pour les services de cloud computing encourage les opérateurs à optimiser l’utilisation des ressources dans les grappes d’ordinateurs. Cela motive le développement de nouvelles technologies qui rendent plus flexible la gestion des ressources. Cependant, exploiter cette flexibilité pour réduire le nombre d’ordinateurs nécessite aussi des algorithmes de gestion des ressources efficaces et dont la performance est prédictible sous une demande stochastique. Dans cette thèse, nous concevons et analysons de tels algorithmes en utilisant le formalisme de la théorie des files d’attente.Notre abstraction du problème est une file multi-serveur avec plusieurs classes de clients. Les capacités des serveurs sont hétérogènes et les clients de chaque classe entrent dans la file selon un processus de Poisson indépendant. Chaque client peut être traité en parallèle par plusieurs serveurs, selon des contraintes de compatibilité décrites par un graphe biparti entre les classes et les serveurs, et chaque serveur applique la politique premier arrivé, premier servi aux clients qui lui sont affectés. Nous prouvons que, si la demande de service de chaque client suit une loi exponentielle indépendante de moyenne unitaire, alors la performance moyenne sous cette politique simple est la même que sous l’équité équilibrée, une extension de processor-sharing connue pour son insensibilité à la loi de la demande de service. Une forme plus générale de ce résultat, reliant les files order-independent aux réseaux de Whittle, est aussi prouvée. Enfin, nous développons de nouvelles formules pour calculer des métriques de performance.Ces résultats théoriques sont ensuite mis en pratique. Nous commençons par proposer un algorithme d’ordonnancement qui étend le principe de round-robin à une grappe où chaque requête est affectée à un groupe d’ordinateurs par lesquels elle peut ensuite être traitée en parallèle. Notre seconde proposition est un algorithme de répartition de charge à base de jetons pour des grappes où les requêtes ont des contraintes d’affectation. Ces deux algorithmes sont approximativement insensibles à la loi de la taille des requêtes et s’adaptent dynamiquement à la demande. Leur performance peut être prédite en appliquant les formules obtenues pour la file multi-serveur. / The growing demand for cloud-based services encourages operators to maximize resource efficiency within computer clusters. This motivates the development of new technologies that make resource management more flexible. However, exploiting this flexibility to reduce the number of computers also requires efficient resource-management algorithms that have a predictable performance under stochastic demand. In this thesis, we design and analyze such algorithms using the framework of queueing theory.Our abstraction of the problem is a multi-server queue with several customer classes. Servers have heterogeneous capacities and the customers of each class enter the queue according to an independent Poisson process. Each customer can be processed in parallel by several servers, depending on compatibility constraints described by a bipartite graph between classes and servers, and each server applies first-come-first-served policy to its compatible customers. We first prove that, if the service requirements are independent and exponentially distributed with unit mean, this simple policy yields the same average performance as balanced fairness, an extension to processor-sharing known to be insensitive to the distribution of the service requirements. A more general form of this result, relating order-independent queues to Whittle networks, is also proved. Lastly, we derive new formulas to compute performance metrics.These theoretical results are then put into practice. We first propose a scheduling algorithm that extends the principle of round-robin to a cluster where each incoming job is assigned to a pool of computers by which it can subsequently be processed in parallel. Our second proposal is a load-balancing algorithm based on tokens for clusters where jobs have assignment constraints. Both algorithms are approximately insensitive to the job size distribution and adapt dynamically to demand. Their performance can be predicted by applying the formulas derived for the multi-server queue.

La demande croissante pour les services de cloud computing encourage les opérateurs à optimiser l’utilisation des ressources dans les grappes d’ordinateurs. Cela motive le développement de nouvelles technologies qui rendent plus flexible la gestion des ressources. Cependant, exploiter cette flexibilité pour réduire le nombre d’ordinateurs nécessite aussi des algorithmes de gestion des ressources efficaces et dont la performance est prédictible sous une demande stochastique. Dans cette thèse, nous concevons et analysons de tels algorithmes en utilisant le formalisme de la théorie des files d’attente.Notre abstraction du problème est une file multi-serveur avec plusieurs classes de clients. Les capacités des serveurs sont hétérogènes et les clients de chaque classe entrent dans la file selon un processus de Poisson indépendant. Chaque client peut être traité en parallèle par plusieurs serveurs, selon des contraintes de compatibilité décrites par un graphe biparti entre les classes et les serveurs, et chaque serveur applique la politique premier arrivé, premier servi aux clients qui lui sont affectés. Nous prouvons que, si la demande de service de chaque client suit une loi exponentielle indépendante de moyenne unitaire, alors la performance moyenne sous cette politique simple est la même que sous l’équité équilibrée, une extension de processor-sharing connue pour son insensibilité à la loi de la demande de service. Une forme plus générale de ce résultat, reliant les files order-independent aux réseaux de Whittle, est aussi prouvée. Enfin, nous développons de nouvelles formules pour calculer des métriques de performance.Ces résultats théoriques sont ensuite mis en pratique. Nous commençons par proposer un algorithme d’ordonnancement qui étend le principe de round-robin à une grappe où chaque requête est affectée à un groupe d’ordinateurs par lesquels elle peut ensuite être traitée en parallèle. Notre seconde proposition est un algorithme de répartition de charge à base de jetons pour des grappes où les requêtes ont des contraintes d’affectation. Ces deux algorithmes sont approximativement insensibles à la loi de la taille des requêtes et s’adaptent dynamiquement à la demande. Leur performance peut être prédite en appliquant les formules obtenues pour la file multi-serveur. / The growing demand for cloud-based services encourages operators to maximize resource efficiency within computer clusters. This motivates the development of new technologies that make resource management more flexible. However, exploiting this flexibility to reduce the number of computers also requires efficient resource-management algorithms that have a predictable performance under stochastic demand. In this thesis, we design and analyze such algorithms using the framework of queueing theory.Our abstraction of the problem is a multi-server queue with several customer classes. Servers have heterogeneous capacities and the customers of each class enter the queue according to an independent Poisson process. Each customer can be processed in parallel by several servers, depending on compatibility constraints described by a bipartite graph between classes and servers, and each server applies first-come-first-served policy to its compatible customers. We first prove that, if the service requirements are independent and exponentially distributed with unit mean, this simple policy yields the same average performance as balanced fairness, an extension to processor-sharing known to be insensitive to the distribution of the service requirements. A more general form of this result, relating order-independent queues to Whittle networks, is also proved. Lastly, we derive new formulas to compute performance metrics.These theoretical results are then put into practice. We first propose a scheduling algorithm that extends the principle of round-robin to a cluster where each incoming job is assigned to a pool of computers by which it can subsequently be processed in parallel. Our second proposal is a load-balancing algorithm based on tokens for clusters where jobs have assignment constraints. Both algorithms are approximately insensitive to the job size distribution and adapt dynamically to demand. Their performance can be predicted by applying the formulas derived for the multi-server queue.

In this paper, we present a new compact hydraulic linear actuator. The concept is developed to change the rolling piston diameter of an active air spring during usage. By doing so, the air spring can actively apply pressure and tension forces. The actuator is designed for small movements at high forces. It is insensitive to side forces, which are introduced by the bellows rolling on the rolling piston of the air spring. A diaphragm sealing is used to minimize friction. Hence a precise adjustment of small displacements at high dynamics is possible and the system is completely leakage-free. We describe the design and development of this actuator and show first measurement results from preliminary tests to show its functionality.

Les cellules de Sertoli (CS) jouent des rôles essentiels pour la régulation de la spermatogenèse, via la signalisation modulée par le récepteur aux androgènes (RA). Les objectifs de ce travail ont été d’identifier les mécanismes moléculaires de la régulation androgénique et des rôles des partenaires moléculaires dans la régulation androgénique des cellules de Sertoli pendant le développement testiculaire, du fœtus à l’âge adulte, pendant différentes stades du développement. Nous avons caractérisé et étudié une nouvelle lignée immortalisée, mature, de CS, ST38c, présentant une expression substantielle de RA endogène, une activation transcriptionnelle du RA induite par les androgènes, ainsi qu’une régulation et une stabilisation de la protéine RA par des mécanismes post traductionnels. Ce modèle a été utilisé afin de tester l’hypothèse que la suppression de l’hormone antimüllérienne (AMH) est modulée directement par les androgènes, via la RA, dans les cellules matures de Sertoli.En parallèle, nous avons testé l’hypothèse que la résistance physiologique aux androgènes du nouveau-né est liée à l’expression différentielle de quelques corégulateurs du RA. Ainsi, nous avons analysé l’expression et la contribution de deux corégulateurs du RA (SRC-2 et HBO1) pendant l’ontogenèse testiculaire humaine et pour des pathologies liées au disfonctionnement de l’action des androgènes ou du RA (syndromes d’insensibilité aux androgènes, hypogonadisme hypogonadotrophique congénital).Nous avons démontré, après des essais de transfection in vitro, que SRC-2 est un coactivateur, alors que HBO1 est in corepresseur du RA dans un modèle de cellules de Sertoli. Nous avons cartographié l’expression testiculaire humaine de SRC-2 et HBO1, pendant différentes stades du développement pré et postnatal et nous avons démontré que SRC-2 présente une expression stable, contrastant avec le profil d’expression différentielle, progressive avec l’âge de RA dans la cellule de Sertoli, suggérant que l’expression du SRC-2 est indépendante de la signalisation androgènique. Par contre, HBO1 et le RA présentent un profil de maturation et d’expression temporelle corrélé, suggérant que l’expression du HBO1 serait liée à une signalisation du RA fonctionnelle dans les cellules de Sertoli. Nous avons aussi démontré que l’expression du HBO1 est induite par les androgènes en présence du RA. Contrairement au SRC-2, HBO1 est non seulement exprimé dans les cellules de Sertoli, mais aussi dans les spermatogonies, pouvant représenter un marqueur potentiellement intéressant des cellules germinales.Enfin, nous avons aussi étudié l’immuno-expression testiculaire de RA et AMH chez des patients post pubères avec un syndrome de déficit en 5α-réductase type 2, et insensibilité minime aux androgènes (MAIS), afin d’étudier la contribution des androgènes (testostérone versus dihydrotéstosterone) pour la spérmatogenèse, la répression de l’AMH ainsi que pour mieux comprendre les perspectives de fertilité chez ces patients. / Sertoli cells (SC) have essential roles in the androgen regulation of spermatogenesis, via the androgen receptor (AR)-mediated signaling. This work aimed at identifying the molecular mechanisms related to the androgenic regulation of the AR and its molecular partners in Sertoli cells during different testicular developmental stages. We first characterized and studied a novel murine mature immortalized Sertoli cell line, called ST38c, which harbors substantial expression of endogenous AR, conserves its androgen-dependent transcriptional activation and exhibits agonist-dependent transcriptional and posttranslational regulation, as well as posttranslational stability.We used this cellular model in order to test the hypothesis that anti Müllerian Hormone (AMH) suppression in mature Sertoli cells would be directly androgen and AR mediated.Next, we hypothesized that the physiological androgen resistance in the neonate would also be related to the differential expression of several AR coregulators. Therefore, we analysed the differential expression and contribution of two AR co-regulators (SRC-2 and HBO1) during human testicular ontogeny, as well as in pathologies associated with androgen action or AR impairment (such as androgen insensitivity syndromes, congenital hypogonadotropic hypogonadism).Using in vitro transfection assays, we showed that SRC-2 is an AR coactivator while HBO1 is an AR corepressor in Sertoli cell models. We provided the cartography of SRC-2 and HBO1 expression during human testicular postnatal different stages and showed that SRC-2 presented a stable expression contrasting with the progressive evolution profile of the AR signaling, suggesting that Sertoli SRC-2 expression was independent of the androgen signaling. Interestingly, HBO1 and AR presented a temporal and positively correlated maturation profile, suggesting that HBO1 expression would be related to a functional AR signaling in the Sertoli cell. Moreover HBO1 expression is induced by androgens in the presence of the AR. Unlike SRC-2, HBO1 is not only expressed in Sertoli cells, but also in spermatogonia, being an interesting germ cell marker.Finally, we also studied AR and AMH immunoexpression in posptubertal cases of 5-α reductase type 2 deficiency and minimal androgen receptor resistance, in respect with the spermatogenesis status of seminiferous tubules, and androgen induced AMH suppression in order to assess the differential contribution of testosterone versus dihydrotestosterone and gather more information about the fertility perspectives in this particular pathologies.