Les cellules souches de l’épiderme interfolliculaire humain, appelées cellules souches kératinocytaires (CSK), assurent l’homéostasie et le renouvellement du tissu durant toute la vie d’un individu grâce à leur importante capacité d’autorenouvellement. Ma thèse de doctorat a porté sur l’étude des effecteurs moléculaires impliqués dans la balance entre prolifération et quiescence dans un modèle in vitro de CSK, isolées de manière clonale et définies par le terme holoclone. Je me suis tout d’abord intéressée à la réponse des holoclones à l’effet d’un régulateur important de la prolifération et de la quiescence des cellules souches adultes : le facteur de croissance TGF β1. Mon travail s’est ensuite focalisé sur l’étude d’un gène agissant en aval de la voie de signalisation TGF β, le facteur de transcription Klf4, et dont le rôle dans la biologie des cellules souches adultes reste largement méconnu. Klf4 est en effet surtout décrit pour son rôle dans la reprogrammation des cellules somatiques en cellules iPS. Le maintien de la sensibilité des holoclones aux signaux inhibiteurs de la croissance constitue un gage de la normalité des CSK. Notre étude de la réponse des holoclones à l’effet antiprolifératif du TGF β1 montre que les holoclones, dotés d’un fort potentiel de croissance, caractéristique des CSK, conservent leur sensibilité à l’effet du TGF β1. Ces résultats nous ont permis de valider les holoclones comme constituant un modèle pertinent pour caractériser la biologie normale des CSK et décrypter les contrôles moléculaires de l’état souche. Le modèle des holoclones a été exploité dans le cadre d’une approche de génomique fonctionnelle visant à déterminer le rôle du facteur de transcription Klf4 dans les CSK. L’utilisation de vecteurs lentiviraux exprimant un shARN dirigé contre l’ARNm de Klf4 nous a permis d’étudier l’impact d’une modulation fine du niveau d’expression de Klf4 sur les propriétés des holoclones. La répression transcriptionelle de Klf4, d’environ un facteur 2, favorise de manière significative l’expansion du compartiment clonogénique au sein des holoclones. Ce gain de fonction concerne à la fois les potentiels de croissance et de reconstruction épidermique des holoclones. Un aspect important de ce travail a concerné la recherche des réseaux moléculaires régulés par Klf4 dans les holoclones. Une analyse du transcriptome nous a permis de montrer que Klf4 participe au contrôle des mécanismes de cycle cellulaire et de différenciation. Klf4 interviendrait également dans la régulation des voies de signalisation TGF β/BMP et Wnt, connues pour exercer des rôles clés dans la biologie des cellules souches. Klf4 constituerait donc un censeur de l’activité du compartiment immature dans l’épiderme interfolliculaire. Il participerait aux mécanismes de régulation du cycle cellulaire et serait susceptible d’intervenir dans le contrôle de l’autorenouvellement du compartiment souche. / Stem cells present within the human interfollicular epidermis, which are defined as keratinocytes stem cells (KSC), ensure the homeostasis and renewal of the tissue throughout the whole individual life. These functions are related to their important self-renewal capacity. My PhD project was focused on the knowledge of the molecular effectors involved in the control of the balance between proliferation and quiescence in KSC. This scientific question was investigated in an in vitro model of KSC which were clonally derived and characterized as holoclones. Holoclones are controlled by mitogenic growth factors and also by antiproliferative signals. One of these regulators is the growth factor TGF β1 which plays an important role in the control of quiescence and cell proliferation within several adult stem cell systems. In the context of growth inhibition by TGF β1, I have studied the role of a downstream gene of the TGF β pathway, the transcription factor Klf4, whose role in adult stem cell biology remains unclear. In fact, Klf4 is mostly described for its involvement in the reprogramming process of somatic cells into iPS cells. The maintenance of holoclone sensitivity to cell growth inhibitors is a critical parameter of KSC normal physiology. Holoclones possess an extensive growth capacity, which is characteristic of KSC. Despite this high proliferation rate, holoclones are still responsive to the antiproliferative effect of TGF β1. These results allowed us to validate the use of holoclone as a relevant model of non-transformed KSC suitable for the characterization of the role of candidate stemness genes in KSC biology, such as Klf4. The holoclone model was exploited to perform a functional genomic approach to investigate the role of Klf4 in KSC. We have developed a shRNA-based gene knock-down method using lentiviral vectors to assess the impact of Klf4 down-modulation on holoclone functional properties. Our results show that Klf4 down modulation controls the expansion of the clonogenic compartment present within holoclone progeny. This gain-of-function, which is maintained at the long term level, leads to an increase in holoclone 3D epidermis reconstruction capacity. A major point of this project was to elucidate the molecular networks controlled by Klf4 in holoclones. Microarray data show that Klf4 regulates the expression of several genes related to pathways involved in the control of stem cell fate. In particular, we identified many transcripts related to TGF β/BMP and Wnt signallings. Interestingly, the majority of the modulated transcripts are involved in the regulation of cell cycle and in keratinocyte differentiation process. All together these results suggest a critical role Klf4 as a stemness censor of the most immature compartment activity. Klf4 is likely to be involved in cell cycle regulation of KSC compartment and in the control of KSC self-renewal process.

Queratinócitos são um veículo bastante atrativo para a transferência gênica ex vivo e liberação sistêmica uma vez que as proteínas secretadas por estas células podem atingir a circulação via um mecanismo similar ao processo natural. Um eficiente vetor retroviral (LXSN) contendo o gene do hormônio de crescimento de camundongo (mGH) foi utilizado para transduzir queratinócitos humanos primários. Os queratinócitos transduzidos apresentaram um nível de secreção in vitro alto e estável atingindo até 11 g mGH/106 células/dia. Os epitélios formados por estes queratinócitos geneticamente modificados apresentaram, porém, uma queda na taxa de secreção > 80 % quando foram retirados da placa de cultura utilizando um procedimento clássico. A substituição desta metodologia clássica por uma cultura organotípica resolveu completamente este problema. Camundongos anões imunodeficientes (lit/scid) implantados com estes enxertos organotípicos foram acompanhados durante 4 meses, e apresentaram um aumento de peso significativo (P<0,05) nos primeiros 40 dias. Níveis circulatórios de mGH atingiram um pico de 21 ng/mL 1 h após o implante, mas estes níveis rapidamente atingiram níveis basais (~2 ng/mL). Os queratinócitos humanos primários apresentaram portanto altos níveis de expressão in vitro e os maiores níveis circulatórios, porém por um breve período de tempo, reportados até o momento para GH neste tipo de células. Em conjunto com resultados que mostraram uma recuperação considerável da eficiência de secreção de mGH em cultura por enxertos organotípicos retirados dos animais, foram discutidos os fatores que ainda impedem a utilização clínica deste modelo promissor de terapia gênica cutânea. / Keratinocytes are a very attractive vehicle for ex vivo gene transfer and systemic delivery, since proteins secreted by these cells may reach the circulation via a mechanism which mimics the natural process. An efficient retroviral vector (LXSN) encoding the mouse growth hormone gene (mGH) was used to transduce primary human keratinocytes. Organotypic raft cultures were prepared with these genetically modified keratinocytes and were grafted onto immunodeficient dwarf mice (lit/scid). Transduced keratinocytes presented a high and stable in vitro secretion level of up to 11 g mGH/106cells/day. Conventional epidermal sheets made with these genetically modified keratinocytes, however, showed a drop in secretion rates of > 80% simply due to detachment of the epithelium from its substratum. Substitution of conventional grafting methodologies with organotypic raft cultures completely overcame this problem. The stable long-term grafting of such cultures onto immunodeficient dwarf (lit/scid) mice could be followed for more than 4 months, and a significant weight increase (P<0.05) over the control group was observed in the first 40 days. Circulating mGH levels revealed a peak of 21 ng/mL just 1h after grafting, but unfortunately these levels rapidly fell to baseline values (~ 2 ng/mL). mGH-secreting primary human keratinocytes presented the highest in vitro expression and peak circulatory levels reported to date for a form of GH with this type of cells. Together with data showing that excised implants can recover in culture a remarkable fraction of their original in vitro mGH secretion efficiency, the factors that might still hamper the success of this promising model of cutaneous gene therapy are discussed. SUMÁRIO

Autosomal dominant epidermolytic ichthyosis (EI) is a rare disease characterized by intra-epidermal blistering due to mutations in either of two keratin genes, KRT1 and KRT10, expressed by suprabasal keratinocytes. Autosomal recessive congenital ichthyosis (ARCI) is a non-blistering, hyperkeratotic disease caused by mutations in one of the following genes: ABCA12, ALOX12B, ALOXE3, TGM1, CYP4F22, NIPAL4 and SLC27A4, which are all essential for skin barrier homeostasis. ARCI and EI often respond well to treatment with retinoids, but the mechanism of action is unclear. The aim of this thesis was to increase the knowledge of pathogenic pathways in ichthyosis and to find new explanations to the effect of retinoids. In vitro studies of immortalized keratinocytes from EI patients showed an abnormal keratin aggregation after heat stress, that could be partially inhibited by pre-treatment with all-trans retinoic acid (ATRA) or retinoic acid receptor α-agonists. ATRA treatment also reduced the relative expression of mutated vs wildtype KRT10. The clearance of ATRA in human keratinocytes was found to be mediated by CYP26B1. In skin biopsies from ARCI patients, immunofluorescence analysis of 12R-LOX, eLOX-3, TGM1, ichthyin and FATP4 showed altered expression, not only of the mutated protein, but also of the other proteins. These observations are consistent with a feedback regulatory mechanism by which the loss of one protein results in an up-regulation of other proteins. Furthermore, 12R-LOX, eLOX-3 and TGM1 were intimately co-localized in stratum corneum, as were ichthyin and FATP4, suggesting that the proteins are linked to the same metabolic pathway. When treated with a CYP26 inhibitor known to raise the endogenous ATRA level of the skin, two patients with NIPAL4 mutations, initially exhibiting increased co-localization signals for 12R-LOX and eLOX-3, displayed normalized lipoxygenase expressions and showed clinical improvement. In conclusion, mechanisms are proposed by which pathogenic keratin aggregations in EI and epidermal protein deficiencies in ARCI patients may be mitigated by retinoids. Furthermore, the vivid crosstalk between proteins incriminated in ARCI suggests that these enzymes operate along a common metabolic pathway essential for producing barrier lipids in stratum corneum. Any abrogation of this production may cause barrier failure, hence resulting in a compensatory hyperkeratosis characteristic of congenital ichthyosis.

Ions in the soil solution traverse many layers (epidermis, exodermis, central cortex, and endodermis) within the root to reach the stele. The endodermis is present in almost all vascular plants while the exodermis is found only in majority of angiosperm roots tested. The maturation of the exodermis and the death of epidermis alter the plasma membrane surface areas (PMSA) potentially available for ion uptake. Do these changes reduce the ion uptake in proportion to the loss of absorptive surface areas? To answer this question onion (Allium cepa L cv. Wolf) adventitious root segments representing above features: Immature Exodermis Live Epidermis (IEXLEP), Mature Exodermis Live Epidermis (MEXLEP), Mature Exodermis Dead Epidermis (MEXDEP) were excised. Using a compartmental elution technique, radioactive sulphate and phosphate present in various internal compartments were quantified. Quantities of ions moved across the plasma membrane, a summation of quantities in the cytoplasm, ‘vacuole’, and ‘bound’ compartments, indicated that the maturation of the exodermis reduces the uptake of sulphate but not phosphate. In contrast, epidermal death reduced the movement of both ions across the plasma membranes. Although there is a reduction in the available PMSA with the maturation of the exodermis and death of the epidermis, these events do not necessarily reduce the ion movement into the plasma symplast. The endodermal cells of onion roots deposit suberin lamellae as secondary walls. As seen in cross-sections some cells remain without these lamellae and are known as ‘passage cells’. What is the pattern of suberin lamella deposition along the root? Is the suberin lamella a continuous layer? To answer these questions, endodermal layers isolated from onion adventitious roots were used in the present study. These layers were observed using four stains (Sudan Red 7B, Fluorol yellow 088 [Fy], berberine, and Nile red) and three microscopes (compound-white light, compound-epifluorescence and confocal scanning). In differentiating cells with and without suberin lamellae in endodermal layers Sudan Red 7B served the best results for compound-white light microscope, Fy for compound-epifluorescence microscope and Nile for confocal laser scanning microscope (CLSM). Suberin lamellae deposition initiated almost in a random manner; they continued to be deposited resulting in the production of longitudinal files alternating with files with passage cells, and were ultimately deposited in almost all cells at a distance of 255 mm from the tip. The suberin lamellae are perforated with pores, a consistent feature even as far as 285 mm from the tip. These pores may serve as portals for water, ions, and pathogen movement.

Ions in the soil solution traverse many layers (epidermis, exodermis, central cortex, and endodermis) within the root to reach the stele. The endodermis is present in almost all vascular plants while the exodermis is found only in majority of angiosperm roots tested. The maturation of the exodermis and the death of epidermis alter the plasma membrane surface areas (PMSA) potentially available for ion uptake. Do these changes reduce the ion uptake in proportion to the loss of absorptive surface areas? To answer this question onion (Allium cepa L cv. Wolf) adventitious root segments representing above features: Immature Exodermis Live Epidermis (IEXLEP), Mature Exodermis Live Epidermis (MEXLEP), Mature Exodermis Dead Epidermis (MEXDEP) were excised. Using a compartmental elution technique, radioactive sulphate and phosphate present in various internal compartments were quantified. Quantities of ions moved across the plasma membrane, a summation of quantities in the cytoplasm, ‘vacuole’, and ‘bound’ compartments, indicated that the maturation of the exodermis reduces the uptake of sulphate but not phosphate. In contrast, epidermal death reduced the movement of both ions across the plasma membranes. Although there is a reduction in the available PMSA with the maturation of the exodermis and death of the epidermis, these events do not necessarily reduce the ion movement into the plasma symplast. The endodermal cells of onion roots deposit suberin lamellae as secondary walls. As seen in cross-sections some cells remain without these lamellae and are known as ‘passage cells’. What is the pattern of suberin lamella deposition along the root? Is the suberin lamella a continuous layer? To answer these questions, endodermal layers isolated from onion adventitious roots were used in the present study. These layers were observed using four stains (Sudan Red 7B, Fluorol yellow 088 [Fy], berberine, and Nile red) and three microscopes (compound-white light, compound-epifluorescence and confocal scanning). In differentiating cells with and without suberin lamellae in endodermal layers Sudan Red 7B served the best results for compound-white light microscope, Fy for compound-epifluorescence microscope and Nile for confocal laser scanning microscope (CLSM). Suberin lamellae deposition initiated almost in a random manner; they continued to be deposited resulting in the production of longitudinal files alternating with files with passage cells, and were ultimately deposited in almost all cells at a distance of 255 mm from the tip. The suberin lamellae are perforated with pores, a consistent feature even as far as 285 mm from the tip. These pores may serve as portals for water, ions, and pathogen movement.

L'allergie de contact est une pathologie en constante augmentation et particulièrement répandue dans les pays industrialisés. Aucune thérapie n'existe actuellement et seule l'éviction totale de l'allergène permet d'éviter toute nouvelle réaction d'allergie. La base de l'allergie de contact est la formation d'une liaison entre l'allergène et les protéines épidermiques. C'est cette étape chimique clé qu' il est important de comprendre afin de développer de nouvelles méthodes dites« alternatives» dans le cadre de l'actuelle législation sur les cosmétiques. Le but de ce travail de thèse a été d'étudier la réactivité de différents allergènes connus face aux acides aminés et protéines d'épidermes reconstruits de type SkinEthic® par la technique RMN HRMAS dérivée de la RMN du solide. Six allergènes ont ainsi été étudiés et leur réactivité a été comparée a celle observée en solution face à une protéine modèle. / Contact dermatitis is one of the most common health problem and highly prevalent in industrialized countries. No therapy currently exists and only the total eviction of the allergen can prevent further allergie reaction.The key molecular event in skin sensitization is the formation of a bond between the allergen and the epidermal proteins. Due to the recent legislation on cosmetic and to help avoid the inappropriate use of new allergens, the understanding of this key step has to be expanded in order to develop new alternative metbods.The aim of this PhD work is to study the reactivity of some allergens towards amino acids and proteins presents in reconstructed human epidermis like SkinEthic® using the HRMAS NMR technique. Six allergens have been studied and their reactivity was compared to that observed ln solution with a model protein.

Les microARNs sont de petits ARN non codants régulant négativement l'expression génique au niveau post-transcriptionnel. Ils interviennent dans de nombreux processus biologiques et leur rôle dans la régulation de l'homéostasie cutanée est clairement démontrée. Cependant, leur fonction durant le vieillissement épidermique n'a jamais été étudiée. Nous avons donc réalisé une analyse exhaustive du miRnome épidermique durant son vieillissement afin d'identifier les microARNs différentiellement exprimés avec l'âge dans ce tissu. Plusieurs microARNs significativement modulés dans des kératinocytes âgés, nous ont permis d'établir une signature du vieillissement épidermique. Parmi eux, les deux brins du microARN miR-30a sont induits dans les épidermes âgés. La construction d'un lentivirus permettant la surexpression stable et inductible de ce microARN a facilité son étude fonctionnelle dans un modèle organotypique d'épiderme reconstruit. Nous avons observé que la surexpression de ce microARN dans un modèle de culture tridimensionnelle induit un phénotype épidermique présentant des similitudes avec celui observé durant son vieillissement chronologique et caractérisé par une forte altération de la différenciation des kératinocytes, par une perturbation de sa fonction barrière et par une augmentation de l'abondance des cellules apoptotiques. Ce projet de thèse a permis l'identification de 3 cibles directes de miR-30a dans les kératinocytes. Il s'agit de LOX, codant pour la lysyl oxydase qui intervient dans la balance prolifération/différenciation des kératinocytes, d'AVEN, un inhibiteur de caspase et d'IDH1, codant pour l'isocitrate déshydrogénase, enzyme du métabolisme énergétique. Ainsi, ce projet de thèse a révélé un nouveau microARN acteur du vieillissement épidermique et a permis de de mettre à jour de nouveaux mécanismes moléculaires expliquant certaines altérations phénotypiques observées dans l'épiderme avec l'âge / MicroRNAs are small non-coding RNA that negatively regulate gene expression at the post-transcriptional level. There are involved in many biological processes and play a key role in the regulation of skin homeostasis. However, their function during epidermal aging has never been studied. We performed an exhaustive analysis of the epidermal miRnome during its aging in order to identify microRNAs differentially expressed with age in this tissue. Several microRNAs significantly modulated in elderly keratinocytes, allowed us to establish a signature of epidermal aging. Among them, the two strands of the microRNA miR-30a are induced in aged epidermis. The construction of a lentivirus allowing inducible and stable overexpression of this microRNA facilitated its functional study in an organotypic model of reconstructed epidermis. We observed that the overexpression of this microRNA in a three-dimensional culture model induces an epidermal phenotype similar of those observed during its chronological aging characterized by a strong alteration of keratinocyte differentiation, by a disturbance of its barrier function and by an increase in the abundance of apoptotic cells. This thesis project allowed the identification of three miR-30a targets in keratinocytes : LOX encoding lysyl oxidase, which plays a role in proliferation/differentiation balance of keratinocytes, AVEN encoding a caspase inhibitor and IDH1 encoding isocitrate deshydrogenase, a key enzyme of cellular metabolism.Our work revealed a new miRNA actor and deciphered new molecular mechanisms to explain some alterations observed in epidermis during aging, especially those concerning keratinocytes differentiation and apoptotic death

Le facteur de transcription HIF-1 est un hétérodimère composé d’une sous-unité α et d’une sous-unité ß. HIF-1 est capable de reconnaître une séquence consensus appelée HRE (HIF Response Element) et de réguler l’expression de plus de 200 gènes cibles impliqués dans divers mécanismes cellulaires. Nous nous intéressons à étudier le rôle de HIF-1α dans la peau, d’une part dans la régulation des enzymes de la réparation de l’ADN suite à l’irradiation UVB, d’autre part dans la physiologie cutanée.Nos résultats montrent bien que HIF-1α régule l’expression des gènes participant à la réparation de l’ADN (XPC et XPD). Ces gènes contiennent dans leurs régions promotrices des HRE de HIF-1α. La quantification de l’immunoprécipitation de chromatine révèle des HRE putatifs dans les gènes codant pour d'autres protéines de la réparation de l'ADN (XPB, XPG, CSA et CSB), ce qui suggère que HIF-1α est un régulateur clé de la machinerie de réparation de l'ADN. Nous avons prouvé que HIF-1α est indispensable à l’adhésion des kératinocytes par sa régulation exercée sur la laminine-332 et les intégrines (α6 et ß1). L’absence de l’expression de HIF-1α empêche aussi la reconstruction des épidermes à partir des kératinocytes humains. Nos résultats ont montré que les souris invalidées pour HIF-1α développent avec l’âge un phénotype d’inflammation dans plusieurs régions. Ces souris sont très sensibles au moindre stress consécutif à une blessure et une irradiation UVB. L’induction de l’inhibition de HIF-1α dans des souris inductibles avec le tamoxifène indique un détachement de l’épiderme au niveau des couches supra-basales. Ces souris meurent deux semaines après injection du tamoxifène / The transcription factor HIF-1 is a heterodimer composed of an α and ß subunit. HIF-1 is capable of recognizing a consensus sequence called HRE (hypoxia Response Element) and regulate the expression of more than 200 target genes involved in various cellular mechanisms. We are interested in studying the role of HIF-1α in the skin physiology.Our results show that HIF-1α regulates the expression of two main factors (XPC and XPD) involved in nucleotide excision repair through binding on HRE in their promoter regions. Quantitative chromatin immunoprecipitation assays further revealed putative HREs in the genes encoding other DNA repair proteins (XPB, XPG, CSA and CSB), suggesting that HIF-1α is a key regulator of the DNA repair machinery. We proved that HIF-1α is essential for keratinocyte adhesion through its regulation exerted on laminin-332 and integrins (α6, ß1). The lack of HIF-1α expression also prevents the reconstruction of epidermis by human keratinocytes. Our results showed that mice constitutively depleted for HIF-1α in their epidermis develop with age a phenotype of inflammation in several regions. These mice are very sensitive to the stress resulting from wound injury and UVB irradiation. HIF-1α depletion in the epidermis of inducible mice using tamoxifen results in a detachment of the epidermis in suprabasal layers. These mice die within two weeks after injection of tamoxifen

La peau est un organe très exposé à l’environnement et fournit la première ligne de défense contre de nombreux pathogènes. Cette fonction est remplie dans l’épiderme murin par les cellules de Langerhans (LCs) et les cellules T dendritiques de l’épiderme (DETCs). Alors que le développement de ces cellules a bien été étudié, peu d’expériences ont été effectuées sur leur renouvellement en condition homéostatique chez des animaux adultes sans manipulations. Nous avons alors développé un système de traçage cellulaire par fluorescence multicolore pour étudier l’homéostasie des LCs et des DETCs. Cette approche de «fate mapping» m’a permis de mettre en évidence un modèle dans lequel le réseau adulte des LCs est formé d’unités prolifératives adjacentes composées de LCs en division et leurs cellules filles. Nous avons identifié que les cellules en division étaient majoritairement représentées par la fraction la plus immature des LCs, suggérant que ces LCs peuvent régénérer leur réseau grâce à une capacité de prolifération limitée. Lors d’une inflammation importante, les LCs sont renouvelées par des progéniteurs issus de la moelle osseuse et s’organisent également en unités de prolifération. Je me suis ensuite intéressé à l’homéostasie des DETCs. Ce réseau est formé de la même manière par des unités prolifératives de DETCs. Un modèle de greffe de peau nous a permis de montrer que les DETCs semblent renouveler les cellules disparues dans une zone restreinte. En conclusion, mes travaux de thèse ont permis de révéler les dynamiques cellulaires qui régissent l’homéostasie des cellules immunitaires de l’épiderme. / The skin is an organ very much exposed to the environment and supplies the primary line of defence against several pathogens. In the mouse model epidermis, this function is fulfilled by Langerhans’ cells (LCs) and dendritic T cells (DETCs). While LCs and DETCs development have thoroughly been studied, few experiences have been carried out concerning the renewal of these cells through homeostatic conditions in adult “nonmanipulated” animals. Then we have designed a new system of fate mapping, by way of multi-coloured fluorescence to study the LCs and DETCs homeostasis. This method of fate mapping allowed me to highlight a model in which the adult network of LCs is made up of adjacent proliferating units, made of dividing LCs and of their daughter cells. We have identified that the dividing cells were mainly represented by the most immature fraction of LCs, suggesting that these LCs can renew their network thanks to a limited ability to proliferate. During significant inflammation, LCs are renewed by progenitors coming from the bone marrow and organize themselves in proliferation units as well. I also took an interest in the homeostasis of DETCs. In the same way as for the LCs, this network seems to be made up of DETCs proliferating units. A model of skin graft led us to show that the DETCs seem to renew the missing cells in a restricted area. As a conclusion, my research work allowed me to reveal the cellular dynamism which governs the homeostasis of the epidermis’ immune cells.

Exocyst is protein complex evolutionary conserved in yeasts, animals and plants, which plays a role in control of cell morphogenesis and polarity. It is a tethering complex whose function is to attach secretory vesicles to specifi c foci on plasma membrane. Complex exocyst is formed by eight subunits. Subunit EXO70 is encoded by 23 paralogue genes in genome of Arabidopsis thaliana. Mutation in paralogue subunit EX070H4 causes defect in trichome maturation. Mutant trichomes have thin, not reinforced cell wall, making them soft and elastic. Transcription of EXO70H4 gene is induced by UV radiation, therefore observations of plants cultivated on UV-B radiation were done. Analysis of mutants cultivated on UV-B radiation revealed hyperaccumulation of vesicules in cytoplasm, which were visible by light microscope. Hyperaccumulation was not observed in control plants cultivated on UV-B radiation, but thickening of cell wall was induced. Th is reaction to UV in trichomes hasn't been described yet. Analysis of cellular localization made with YFP tagged constructs revealed that EXO70H4 localizes into mobile corpuscules associating with Golgi apparatus. It was found with yeast two hybrid system that EXO70H4 interacts with TRS120, subunit of tethering complex TRAPPII which is active in Golgi apparatus....