Individual differences in gene expression : insights from transcriptional control of the CSL gene and from human population studies

CSL is a key transcription factor, mostly acting as a repressor, which has been shown to have a highly context-dependent function. While known as the main effector of Notch signaling, it can also exert Notch-independent functions. The downstream effects of the Notch/CSL signaling pathway and its involvement in several biological processes have been intensively studied. We recently showed that CSL is important to maintain skin homeostasis, as its specific deletion in mouse dermal fibroblasts -or downmodulation in human stromal fibroblasts- creates an inducing environment for squamous cell carcinoma (SCC) development, possibly due to the conversion of stromal fibroblasts into cancer associated fibroblasts (CAFs). Despite the wide interest in CSL as a transcriptional regulator, the mechanism of its own regulation has so far been neglected. We show here that CSL expression levels differ between individuals, and correlate among others with genes involved in DNA damage response. Starting from this finding we show that in dermal fibroblasts CSL is under transcriptional control of stress inducers such as UVA irradiation and Reactive Oxygen Species (ROS) induction, and that a main player in CSL transcriptional regulation is the transcription factor p53. In a separate line of work, we focused on individual variability, studying the differences in gene expression between human populations in various cancer types, particularly focusing on the Caucasian and African populations. It is indeed widely known that these populations have different incidences and mortalities for various cancers, and response to cancer treatment may also vary between them. We show here several genes that are differentially expressed and could be of interest in the study of population differences in cancer. -- CSL est un facteur de transcription agissant essentiellement comme répresseur, et qui a une fonction hautement dépendant du contexte. C'est l'effecteur principal de la voie de signalisation de Notch, mais il peut également exercer ses fonctions dans une façon Notch- indépendante. Nous avons récemment montré que CSL est important pour maintenir l'homéostasie de la peau. Sa suppression spécifique dans les fibroblastes dermiques de la souris ou dans les fibroblastes stromales humaines crée un environnement favorable pour le développement du carcinome épidermoïde (SCC), probablement en raison de la conversion des fibroblastes en fibroblastes associé au cancer (CAF). Malgré le grand intérêt de CSL comme régulateur transcriptionnel, le mécanisme de sa propre régulation a été jusqu'ici négligée. Nous montrons ici que dans les fibroblastes dermiques CSL est sous le contrôle transcriptionnel de facteurs de stress tels que l'irradiation UVA et l'induction des ROS dont p53 est l'acteur principal de cette régulation. Nous montrons aussi que les niveaux d'expression de CSL varient selon les individus, en corrélation avec d'autres gènes impliqués dans la réponse aux dommages de l'ADN. Dans une autre axe de recherche, concernant la variabilité individuelle, nous avons étudié les différences dans l'expression des gènes dans différents types de cancer entre les populations humaines, en se concentrant particulièrement sur les populations africaines et caucasiennes. Il est en effet bien connu que ces populations montrent des variations dans l'incidence des cancers, la mortalité, ainsi que pour les réponses au traitement. Nous montrons ici plusieurs gènes qui sont exprimés différemment et pourraient être digne d'intérêt dans l'étude du cancer au sein de différentes populations.

Molecular Physics of Elementary Processes relevant to Hypersonics: atom-molecule, molecule-molecule and atom-surface processes.

In the present chapter some prototype gas and gas-surface processes occurring within the hypersonic flow layer surrounding spacecrafts at planetary entry are discussed. The discussion is based on microscopic dynamical calculations of the detailed cross sections and rate coefficients performed using classical mechanics treatments for atoms, molecules and surfaces. Such treatment allows the evaluation of the efficiency of thermal processes (both at equilibrium and nonequilibrium distributions) based on state-to-state and state specific calculations properly averaged over the population of the initial states. The dependence of the efficiency of the considered processes on the initial partitioning of energy among the various degrees of freedom is discussed.

Mechanistic population Models in Biology: Model Derivation and Application

In general, models of ecological systems can be broadly categorized as ’top-down’ or ’bottom-up’ models, based on the hierarchical level that the model processes are formulated on. The structure of a top-down, also known as phenomenological, population model can be interpreted in terms of population characteristics, but it typically lacks an interpretation on a more basic level. In contrast, bottom-up, also known as mechanistic, population models are derived from assumptions and processes on a more basic level, which allows interpretation of the model parameters in terms of individual behavior. Both approaches, phenomenological and mechanistic modelling, can have their advantages and disadvantages in different situations. However, mechanistically derived models might be better at capturing the properties of the system at hand, and thus give more accurate predictions. In particular, when models are used for evolutionary studies, mechanistic models are more appropriate, since natural selection takes place on the individual level, and in mechanistic models the direct connection between model parameters and individual properties has already been established. The purpose of this thesis is twofold. Firstly, a systematical way to derive mechanistic discrete-time population models is presented. The derivation is based on combining explicitly modelled, continuous processes on the individual level within a reproductive period with a discrete-time maturation process between reproductive periods. Secondly, as an example of how evolutionary studies can be carried out in mechanistic models, the evolution of the timing of reproduction is investigated. Thus, these two lines of research, derivation of mechanistic population models and evolutionary studies, are complementary to each other.

Population structure in a highly pelagic seabird, the Cory's shearwater (Calonectris diomedea): an examination of genetics, morphology and ecology

Increasing evidence suggests oceanic traits may play a key role in the genetic structuring of marine organisms. Whereas genetic breaks in the open ocean are well known in fishes and marine invertebrates, the importance of marine habitat characteristics in seabirds remains less certain. We investigated the role of oceanic transitions versus population genetic processes in driving population differentiation in a highly vagile seabird, the Cory"s shearwater, combining molecular, morphological and ecological data from 27 breeding colonies distributed across the Mediterranean (Calonectris diomedea diomedea) and the Atlantic (C. d. borealis). Genetic and biometric analyses showed a clear differentiation between Atlantic and Mediterranean Cory"s shearwaters. Ringing-recovery data indicated high site fidelity of the species, but we found some cases of dispersal among neighbouring breeding sites (<300 km) and a few long distance movements (>1000 km) within and between each basin. In agreement with this, comparison of phenotypic and genetic data revealed both current and historical dispersal events. Within each region, we did not detect any genetic substructure among archipelagos in the Atlantic, but we found a slight genetic differentiation between western and eastern breeding colonies in the Mediterranean. Accordingly, gene flow estimates suggested substantial dispersal among colonies within basins. Overall, genetic structure of the Cory"s shearwater matches main oceanographic breaks (Almería-Oran Oceanic Front and Siculo-Tunisian Strait), but spatial analyses suggest that patterns of genetic differentiation are better explained by geographic rather than oceanographic distances. In line with previous studies, genetic, phenotypic and ecological evidence supported the separation of Atlantic and Mediterranean forms, suggesting the 2 taxa should be regarded as different species.

Demographic Modelling of Human Population Growth

This work presents models and methods that have been used in producing forecasts of population growth. The work is intended to emphasize the reliability bounds of the model forecasts. Leslie model and various versions of logistic population models are presented. References to literature and several studies are given. A lot of relevant methodology has been developed in biological sciences. The Leslie modelling approach involves the use of current trends in mortality,fertility, migration and emigration. The model treats population divided in age groups and the model is given as a recursive system. Other group of models is based on straightforward extrapolation of census data. Trajectories of simple exponential growth function and logistic models are used to produce the forecast. The work presents the basics of Leslie type modelling and the logistic models, including multi- parameter logistic functions. The latter model is also analysed from model reliability point of view. Bayesian approach and MCMC method are used to create error bounds of the model predictions.

Adaptive Markov chain Monte Carlo and Bayesian filtering for state space models

This thesis is concerned with the state and parameter estimation in state space models. The estimation of states and parameters is an important task when mathematical modeling is applied to many different application areas such as the global positioning systems, target tracking, navigation, brain imaging, spread of infectious diseases, biological processes, telecommunications, audio signal processing, stochastic optimal control, machine learning, and physical systems. In Bayesian settings, the estimation of states or parameters amounts to computation of the posterior probability density function. Except for a very restricted number of models, it is impossible to compute this density function in a closed form. Hence, we need approximation methods. A state estimation problem involves estimating the states (latent variables) that are not directly observed in the output of the system. In this thesis, we use the Kalman filter, extended Kalman filter, Gauss–Hermite filters, and particle filters to estimate the states based on available measurements. Among these filters, particle filters are numerical methods for approximating the filtering distributions of non-linear non-Gaussian state space models via Monte Carlo. The performance of a particle filter heavily depends on the chosen importance distribution. For instance, inappropriate choice of the importance distribution can lead to the failure of convergence of the particle filter algorithm. In this thesis, we analyze the theoretical Lᵖ particle filter convergence with general importance distributions, where p ≥2 is an integer. A parameter estimation problem is considered with inferring the model parameters from measurements. For high-dimensional complex models, estimation of parameters can be done by Markov chain Monte Carlo (MCMC) methods. In its operation, the MCMC method requires the unnormalized posterior distribution of the parameters and a proposal distribution. In this thesis, we show how the posterior density function of the parameters of a state space model can be computed by filtering based methods, where the states are integrated out. This type of computation is then applied to estimate parameters of stochastic differential equations. Furthermore, we compute the partial derivatives of the log-posterior density function and use the hybrid Monte Carlo and scaled conjugate gradient methods to infer the parameters of stochastic differential equations. The computational efficiency of MCMC methods is highly depend on the chosen proposal distribution. A commonly used proposal distribution is Gaussian. In this kind of proposal, the covariance matrix must be well tuned. To tune it, adaptive MCMC methods can be used. In this thesis, we propose a new way of updating the covariance matrix using the variational Bayesian adaptive Kalman filter algorithm.

A new model for the population pharmacokinetics of didanosine in healthy subjects

Didanosine (ddI) is a component of highly active antiretroviral therapy drug combinations, used especially in resource-limited settings and in zidovudine-resistant patients. The population pharmacokinetics of ddI was evaluated in 48 healthy volunteers enrolled in two bioequivalence studies. These data, along with a set of co-variates, were the subject of a nonlinear mixed-effect modeling analysis using the NONMEM program. A two-compartment model with first order absorption (ADVAN3 TRANS3) was fitted to the serum ddI concentration data. Final pharmacokinetic parameters, expressed as functions of the co-variates gender and creatinine clearance (CL CR), were: oral clearance (CL = 55.1 + 240 x CL CR + 16.6 L/h for males and CL = 55.1 + 240 x CL CR for females), central volume (V2 = 9.8 L), intercompartmental clearance (Q = 40.9 L/h), peripheral volume (V3 = 62.7 + 22.9 L for males and V3 = 62.7 L for females), absorption rate constant (Ka = 1.51/h), and dissolution time of the tablet (D = 0.43 h). The intraindividual (residual) variability expressed as coefficient of variation was 13.0%, whereas the interindividual variability of CL, Q, V3, Ka, and D was 20.1, 75.8, 20.6, 18.9, and 38.2%, respectively. The relatively high (>30%) interindividual variability for some of these parameters, observed under the controlled experimental settings of bioequivalence trials in healthy volunteers, may result from genetic variability of the processes involved in ddI absorption and disposition.

TP53 codon 72 polymorphism as a risk factor for cardiovascular disease in a Brazilian population

TP53, a tumor suppressor gene, has a critical role in cell cycle, apoptosis and cell senescence and participates in many crucial physiological and pathological processes. Identification of TP53 polymorphism in older people and age-related diseases may provide an understanding of its physiology and pathophysiological role as well as risk factors for complex diseases. TP53 codon 72 (TP53:72) polymorphism was investigated in 383 individuals aged 66 to 97 years in a cohort from a Brazilian Elderly Longitudinal Study. We investigated allele frequency, genotype distribution and allele association with morbidities such as cardiovascular disease, type II diabetes, obesity, neoplasia, low cognitive level (dementia), and depression. We also determined the association of this polymorphism with serum lipid fractions and urea, creatinine, albumin, fasting glucose, and glycated hemoglobin levels. DNA was isolated from blood cells, amplified by PCR using sense 5'-TTGCCGTCCCAAGCAATGGATGA-3' and antisense 5'-TCTGGGAAGGGACAGAAGATGAC-3' primers and digested with the BstUI enzyme. This polymorphism is within exon 4 at nucleotide residue 347. Descriptive statistics, logistic regression analysis and Student t-test using the multiple comparison test were used. Allele frequencies, R (Arg) = 0.69 and P (Pro) = 0.31, were similar to other populations. Genotype distributions were within Hardy-Weinberg equilibrium. This polymorphism did not show significant association with any age-related disease or serum variables. However, R allele carriers showed lower HDL levels and a higher frequency of cardiovascular disease than P allele subjects. These findings may help to elucidate the physiopathological role of TP53:72 polymorphism in Brazilian elderly people.

A morphometric study of the lumbar spinous process in the Chinese population

Our goal was to analyze the anatomical parameters of the lumbar spine spinous process for an interspinous stabilization device designed for the Chinese population and to offer an anatomical basis for its clinical application. The posterior lumbar spines (T12-S1) of 52 adult cadavers were used for measuring the following: distance between two adjacent spinous processes (DB), distance across two adjacent spinous processes (DA), thickness of the central spinous processes (TC), thickness of the superior margin of the spinous processes (TS), thickness of the inferior margin of the spinous processes (TI), and height of the spinous processes (H). Variance and correlation analyses were conducted for these data, and the data met the normal distribution and homogeneity of variance. DB decreased gradually from L1-2 to L5-S1. DA increased from T12-L1 to L2-3 and then decreased from L2-3 to L4-5. The largest H in males was noted at L3 (25.45±5.96 mm), whereas for females the largest H was noted at L4 (18.71±4.50 mm). Usually, TS of the adjacent spinous process was lower than TI. Based on the anatomical parameters of the lumbar spinous processes obtained in this study, an “H”-shaped coronal plane (posterior view) was proposed as an interspinous stabilization device for the Chinese population. This study reports morphometric data of the lumbar spinous processes in the Chinese population, which provides an anatomical basis for future clinical applications.

Incidence of resistance to benzimidazole fungicides in a field population of Venturia inaequalis /

The allele-specific polymerase chain reaction (PCR) was used to screen for the presence of benomyl resistance, and to characterize their levels and frequencies in field populations of Venturia inaequalis during two seasons. Three hundred isolates of V. inaequalis were collected each season from infected leaves of MalusX domestica. Borkh c.v. Mcintosh. The trees used were sprayed in the year prior to collection with five applications of benomyl, its homologue Azindoyle, or water. Monoconidial isolates of V. inaequalis were grown on 2% potato dextrose agar (PDA) for four weeks. Each isolate was taken from a single lesion from a single leaf. Total genomic DNA was extracted from the four week old colonies of V. inaequalis, prepared and used as a template in PCR reactions. PCR reactions were achieved by utilizing allele-specific primers. Each primer was designed to amplify fragments from a specific allele. Primer Vin was specific for mutations conferring the ben^^"^ phenotype. It was expected to amplify a 171 bp. DNA fragment from the ben^"^ alleles only. Primers BenHR and BenMR were specific for mutations conferring the ben"" and ben'^'' phenotypes, respectively. They were expected to amplify 172 bp. and 165 bp. DNA fragments from the ben"" and ben"^" alleles, respectively. Of the 953 isolates tested, 414 (69.9%) were benomyl sensitive (ben^) and 179 (30.1%) were benomyl resistant. All the benomyl resistant alleles were ben^"", since neither the ben"" nor the ben"" alleles were detected. Frequencies of benomyl resistance were 23%, 24%, and 23% for the 1997 collections, and were 46%, 26% and 38% for the 1998 collections for benomyl, Azindoyle and water treatments, respectively. Growth assay was performed to evaluate the applicability of using PCR in monitoring benomyl resistance in fungal field populations. Tests were performed on 14 isolates representing the two phenotypes (ben^ and ben^"'' alleles) characterized by PCR. Results of those tests were in agreement with PCR results. Enzyme digestion was also used to evaluate the accuracy and reliability of PCR products. The mutation associated with the ben^"'' phenotype creates a unique site for the endonuclease enzyme Bsh^236^ allowing the use of enzyme digestion. Isolates characterized by PCR as ben^'^'^ alleles had this restriction site for the SsA7l2361 enzyme. The most time consuming aspect of this study was growing fungal isolates on culture media for DNA extraction. In addition, the risk of contamination or losing the fungus during growth processes was relatively high. A technique for extracting DNA directly from lesions on leaves has been used (Luck and Gillings 1 995). In order to apply this technique in experiments designed to monitor fungicide resistance, a lesion has to be homogeneous for fungicide sensitivity. For this purpose, PCR protocol was used to determine lesion homogeneity. One hundred monoconidial isolates of V. inaequalis from 10 lesions (10-conidia/ lesion) were tested for their phenotypes with respect to benomyl sensitivity. Conidia of six lesions were homogeneous, while conidia of the remaining lesions were mixtures of ben^ and ben^ phenotypes. Neither the ben" nor the ben' phenotype was detected.

Feature Selection and Classification Using Age Layered Population Structure Genetic Programming

The curse of dimensionality is a major problem in the fields of machine learning, data mining and knowledge discovery. Exhaustive search for the most optimal subset of relevant features from a high dimensional dataset is NP hard. Sub–optimal population based stochastic algorithms such as GP and GA are good choices for searching through large search spaces, and are usually more feasible than exhaustive and deterministic search algorithms. On the other hand, population based stochastic algorithms often suffer from premature convergence on mediocre sub–optimal solutions. The Age Layered Population Structure (ALPS) is a novel metaheuristic for overcoming the problem of premature convergence in evolutionary algorithms, and for improving search in the fitness landscape. The ALPS paradigm uses an age–measure to control breeding and competition between individuals in the population. This thesis uses a modification of the ALPS GP strategy called Feature Selection ALPS (FSALPS) for feature subset selection and classification of varied supervised learning tasks. FSALPS uses a novel frequency count system to rank features in the GP population based on evolved feature frequencies. The ranked features are translated into probabilities, which are used to control evolutionary processes such as terminal–symbol selection for the construction of GP trees/sub-trees. The FSALPS metaheuristic continuously refines the feature subset selection process whiles simultaneously evolving efficient classifiers through a non–converging evolutionary process that favors selection of features with high discrimination of class labels. We investigated and compared the performance of canonical GP, ALPS and FSALPS on high–dimensional benchmark classification datasets, including a hyperspectral image. Using Tukey’s HSD ANOVA test at a 95% confidence interval, ALPS and FSALPS dominated canonical GP in evolving smaller but efficient trees with less bloat expressions. FSALPS significantly outperformed canonical GP and ALPS and some reported feature selection strategies in related literature on dimensionality reduction.

Feature Selection and Classification Using Age Layered Population Structure Genetic Programming

The curse of dimensionality is a major problem in the fields of machine learning, data mining and knowledge discovery. Exhaustive search for the most optimal subset of relevant features from a high dimensional dataset is NP hard. Sub–optimal population based stochastic algorithms such as GP and GA are good choices for searching through large search spaces, and are usually more feasible than exhaustive and determinis- tic search algorithms. On the other hand, population based stochastic algorithms often suffer from premature convergence on mediocre sub–optimal solutions. The Age Layered Population Structure (ALPS) is a novel meta–heuristic for overcoming the problem of premature convergence in evolutionary algorithms, and for improving search in the fitness landscape. The ALPS paradigm uses an age–measure to control breeding and competition between individuals in the population. This thesis uses a modification of the ALPS GP strategy called Feature Selection ALPS (FSALPS) for feature subset selection and classification of varied supervised learning tasks. FSALPS uses a novel frequency count system to rank features in the GP population based on evolved feature frequencies. The ranked features are translated into probabilities, which are used to control evolutionary processes such as terminal–symbol selection for the construction of GP trees/sub-trees. The FSALPS meta–heuristic continuously refines the feature subset selection process whiles simultaneously evolving efficient classifiers through a non–converging evolutionary process that favors selection of features with high discrimination of class labels. We investigated and compared the performance of canonical GP, ALPS and FSALPS on high–dimensional benchmark classification datasets, including a hyperspectral image. Using Tukey’s HSD ANOVA test at a 95% confidence interval, ALPS and FSALPS dominated canonical GP in evolving smaller but efficient trees with less bloat expressions. FSALPS significantly outperformed canonical GP and ALPS and some reported feature selection strategies in related literature on dimensionality reduction.

Markovian Processes, Two-Sided Autoregressions and Finite-Sample Inference for Stationary and Nonstationary Autoregressive Processes.

In this paper, we develop finite-sample inference procedures for stationary and nonstationary autoregressive (AR) models. The method is based on special properties of Markov processes and a split-sample technique. The results on Markovian processes (intercalary independence and truncation) only require the existence of conditional densities. They are proved for possibly nonstationary and/or non-Gaussian multivariate Markov processes. In the context of a linear regression model with AR(1) errors, we show how these results can be used to simplify the distributional properties of the model by conditioning a subset of the data on the remaining observations. This transformation leads to a new model which has the form of a two-sided autoregression to which standard classical linear regression inference techniques can be applied. We show how to derive tests and confidence sets for the mean and/or autoregressive parameters of the model. We also develop a test on the order of an autoregression. We show that a combination of subsample-based inferences can improve the performance of the procedure. An application to U.S. domestic investment data illustrates the method.

Assymétries cérébrales lors de traitement de l’information visuelle rapide : investigations chez une population clinique et neurologiquement saine

Le phénomène de Clignement Attentionnel (Attentional Blink, AB), fait référence à une diminution transitoire du rapport exact d’une deuxième cible (C2) si celle-ci est présentée trop tôt après une première cible (C1) lors d’une présentation visuelle sérielle rapide (rapid serial visual presentation, RSVP), et ce, quand les deux cibles doivent être rapportées. Cette étude a examiné l’existence possible d’asymétries hémisphèriques dans le traitement attentionnel ainsi que l’éventualité que la présentation de cibles à deux hémisphères différents puisse diminuer le AB chez des participants neurologiquement sains et l’abolir dans le cas d’un patient callosotomisé. Pour ce faire, nous avons employé un paradigme modifié du AB dans lequel les cibles pouvaient apparaître dans n’importe quelle de quatre RSVP, une dans chaque quadrant du champ visuel, pour permettre des essais dans lesquels les deux cibles puissent être présentées au même hémisphère et d’autres où chaque cible était présentée à un hémisphère différent. Bien que nous n’ayons trouvé aucune diminution de l’effet AB lors de présentation inter-hémisphérique, dans les deux populations à l’étude, le taux de bonnes réponses globales à la deuxième cible était plus élevé quand les cibles étaient présentées à des hémisphères différents. Nous avons également trouvé un avantage de l’hémisphère gauche chez le patient callosotomisé.

Étude fonctionnelle et génétique d'une population de lymphocytes T CD4-CD8- impliquée dans la résistance au diabète auto-immun chez la souris

Le diabète auto-immun résulte de la destruction des cellules bêta pancréatiques sécrétrices d’insuline par les lymphocytes T du système immunitaire. Il s’ensuit une déficience hormonale qui peut être comblée par des injections quotidiennes d’insuline d’origine exogène, toutefois il demeure à ce jour impossible de guérir les patients atteints de la maladie. De façon générale, un système immunitaire sain reconnaît une multitude d’antigènes différents et assure ainsi notre défense à l’égard de différents pathogènes ou encore de cellules tumorales. Il arrive cependant que, pour des raisons génétiques et/ou environnementales, les lymphocytes T puissent s’activer de façon aberrante suite à la reconnaissance d’antigènes provenant du soi. C’est ce bris de tolérance qui mène au développement de pathologies auto-immunes telles que le diabète auto-immun. Afin de limiter l’auto-immunité, des mécanismes de sélection stricts permettent d’éliminer la majorité des lymphocytes T présentant une forte affinité envers des antigènes du soi lors de leur développement dans le thymus. Certains de ces lymphocytes réussissent toutefois à échapper à l’apoptose et migrent en périphérie afin d’y circuler en quête d’un antigène spécifiquement reconnu. Il est alors primordial que des mécanismes périphériques assurent le maintien de la tolérance immunitaire en faisant obstacle à l’activation et à la prolifération des lymphocytes T auto-réactifs. L’une des avenues afin d’inhiber le développement de réponses immunitaires aberrantes est la génération de lymphocytes T régulateurs. Ces cellules, d’origine thymique ou périphérique, peuvent arborer différents phénotypes et agissent via de multiples mécanismes afin d’inactiver et/ou éliminer les cellules impliquées dans l’apparition de pathologies auto-immunes. L’utilisation de modèles murins transgéniques a permis la mise en évidence d’une population peu caractérisée de lymphocytes T au potentiel régulateur. En effet, la proportion de ces cellules T n’exprimant pas les corécepteurs CD4 et CD8 (double négatives, DN) a été inversement corrélée à la prédisposition à l’auto-immunité chez ces ii souris. L’objectif principal de cette thèse est de démontrer la fonction immuno-régulatrice des lymphocytes T DN, tout en investiguant les facteurs génétiques responsables du maintien de cette population cellulaire. Nous avons observé que les lymphocytes T DN exercent une activité cytotoxique à l’égard des lymphocytes B de façon spécifique à l’antigène, via la libération de granules cytolytiques contenant du granzyme B et de la perforine. Par ailleurs, nous avons établi qu’un unique transfert adoptif de ces cellules est suffisant afin d’inhiber le développement du diabète auto-immun chez des hôtes transgéniques prédisposés à la maladie. Le recours à des souris déficientes pour l’expression du gène CD47 a permis de constater que la voie de signalisation CD47-Sirp est essentielle dans le maintien de la proportion des lymphocytes T DN. De plus, le locus murin de prédisposition au diabète auto-immun Idd13, qui contient le gène Sirp, a été identifié pour son rôle dans la régulation de la proportion de ces cellules. Finalement, une analyse génétique a révélé que d’autres intervalles génétiques sont impliqués dans le contrôle de la population des lymphocytes T DN. Parmi ceux-ci, un locus situé en région proximale du chromosome 12 a été validé grâce à la création de souris congéniques. Grâce aux résultats présentés dans cette thèse, notre compréhension de la biologie ainsi que de la régulation des lymphocytes T DN est approfondie. Ces connaissances constituent un pas important vers la création de thérapies cellulaires novatrices permettant de prévenir et de guérir diverses pathologies auto-immunes.