Les infections fongiques invasives, sont des infections sévères grevées d’une lourde mortalité. Elles sont actuellement un problème majeur de santé publique et leur incidence augmente. Les candidémies représentent la quatrième cause d’infection hématogène nosocomiale aux Etats-Unis, les cryptococcoses sont responsables de 600 000 décès chaque année en Afrique et l’aspergillose invasive infecte 10% des patients transplantés de cellules souches. La mortalité de ces infections reste élevée avec des taux de mortalité de 50, 15 et 40% respectivement. L’augmentation de leur incidence est due à l’accroissement des populations immunodéprimées à risque de développer des infections fongiques en raison de l’augmentation des thérapeutiques immunosuppressives et de l’allongement de la durée de vie des patients immunodéprimés. Les infections fongiques invasives surviennent chez des patients immunodéprimés, majoritairement dans un contexte d’immunodépression acquise (neutropénie, chimiothérapie, greffe de cellules souches périphériques ou transplantation d’organe solide, diabète, infection par le virus de l’immunodéficience humaine), mais également secondairement à un déficit immunitaire héréditaire (granulomatose septique chronique, déficit immunitaire combiné, neutropénie congénitale, défaut de l’axe IFNγ-IL12). Cependant certains patients développent des infections fongiques invasives sans immunodépression ou facteurs de risques identifiés. Nous avons donc émis l’hypothèse que ces infections avaient possiblement une origine génétique non identifiée. Au cours de ma thèse, j’ai étudié une cohorte de patients présentant des infections fongiques invasives sans facteur favorisant identifié afin de rechercher une étiologie génétique à ces infections. Le premier groupe de patients que j’ai étudié présentait une dermatophytose invasive ou dermatophytose profonde sans immunodépression. Contrairement à la dermatophytose superficielle, en général bénigne et fréquente dans la population générale; la dermatophytose profonde est une infection rare, invasive et sévère, dans laquelle les dermatophytes (qui sont des champignons filamenteux) envahissent les tissus dermiques et hypodermiques, les ganglions et parfois les organes profonds. Les patients que j’ai étudié étaient tous originaires d’Afrique du Nord, pour la plupart issus de familles consanguines, dont certains avec des cas multiples. Ces observations suggéraient une origine génétique de la dermatophytose profonde avec une hérédité probablement récessive. Au cours de ma thèse, j’ai étudié les caractéristiques cliniques, immunologiques et génétiques de 18 patients atteints d’une dermatophytose profonde, issus de neuf familles Marocaines, Algériennes, Tunisiennes ou Egyptiennes. En parallèle, j’ai étudié des patients ayant présenté des infections fongiques avec localisations cérébrales. L’une de ces patients a présenté des abcès cérébraux suite à une infection disséminée à Exophiala dermatitidis et trois patients ont développé des infections du système nerveux central à Candida spp.. Les infections invasives à Exophiala dermatitidis sont des infections rares, avec de fréquentes atteintes du système nerveux central, survenant majoritairement chez des patients sans déficit immunitaire identifié suggérant l’existence d’une origine génétique probable méconnue chez ces patients. Les candidoses invasives surviennent habituellement chez des patients neutropéniques, ayant récemment subi une intervention chirurgicale ou étant porteurs d’un cathéter intraveineux. Parmi les candidoses invasives, les localisations au système nerveux central sont rares, et classiquement rapportées chez des nouveau-nés prématurés ou suite à une intervention neurochirurgicale. J’ai par ailleurs étudié une patiente ayant développé des infections invasives des tissus sous-cutanés et des adénopathies dues à un champignon filamenteux. (...) / Invasive fungal diseases are a major health problem as they are severe infections complicated with high mortality rates and with rising incidence. Invasive fungal diseases occur mainly in patients with acquired immunodeficiencies, but also with primary immunodeficiencies (chronic granulomatous disease, defect in IFN-ϒ/IL-12 axis, congenital neutropenia). However, few patients develop invasive fungal disease without known risk factor. We therefore hypothesized that these infections probably have an unidentified genetic etiology. I studied a cohort of patients who developed invasive fungal diseases without risk factors and searched for a genetic etiology to their infections. The first group of patients presented with deep dermatophytosis without known immunodeficiency. Deep dermatophytosis is a rare, invasive and severe infection where dermatophytes invade dermis, hypodermis, lymph nodes and sometimes deep organs. I could study clinical, immunological and genetic characteristics of 18 patients from nine families who presented deep dermatophytosis. I also studied patients who developed central nervous system (CNS) fungal infections; one patient with CNS Exophiala dermatitidis infection and three patients with CNS Candida spp. infection. Invasive E. dermatitidis infections are rare, with frequent CNS location, mainly reported in patients without known immunodeficiencies, suggesting a potential unknown genetic etiology in these patients. CNS candidiasis are also rare infections usually occuring in preterm neonates or following neurosurgery. Based on literature data previously reporting a large consanguineous Iranian family with CARD9 deficiency that developed chronic mucocutaneous and central nervous system candidiasis; according to candidate gene approach, I sequenced CARD9 in all patients. CARD9 is an adaptor protein expressed by myeloid cells that signals downstream Dectin-1 and Dectin2 that are the main Pattern Recognotion Receptor implicated in antifungal immunity. I identified in all studied patients homozygous CARD9 mutations. Among 18 patients with deep dermatophytosis, 16 had homozygous nonsense Q289X and two homozygous missense R101C mutation in CARD9. I identified R18W, R35Q and R70W homozygous missense mutations in the patients who developed E. dermatitidis and two patients who developed CNS candidiasis, respectively. Transmission was autosomal recessive for all patients, except for the one with E. dermatitidis infection who had an uniparental disomy. In contrast with controls, CARD9 expression is abolished in Q289X, reduced in R70W and normal in R18W patients’ myeloid cells. CARD9 deficient patients whole blood and dendritic cells display a selective response defect to Candida albicans and Saccharomyces cerevisiae ; with IL-6 and TNF-α production impairment after Candida albicans and Saccharomyces cerevisiae stimulation. This defect can explain elective fungal susceptibility of CARD9 deficient patients to invasive fungal infections. This work evidenced that CARD9 deficiency was the main genetic etiology of deep dermatophytosis. It also could evidence that CARD9 deficiency is associated with Exophiala dermatitidis and Candida spp. CNS infections. This susceptibility is associated with proinflammatory cytokines defect by dendritic cells and whole blood to fungal agents. Various fungal clinical phenotypes in CARD9 deficient patients assess CARD9 central role in skin and central nervous system antifungal immunity.

The innate immune system is the first line of defence against infection. It is comprised of physicochemical barriers and a variety of cell types including macrophages and dendritic cells. Pathogens express specific pathogen associated molecular patterns (PAMP) which are recognised by pattern recognition receptors (PRR) on macrophages to initiate an innate immune response. Gram-negative bacteria such as Salmonella enterica serovar Typhimurium express a range of bacterial PAMPs recognised by Toll-like receptors (TLRs) including lipopolysaccharides (LPS) recognised by TLR-4 and lipoproteins by TLR-2. The activation of TLRs results in activation of nuclear factor κB (NF-κB) to drive transcription of mRNA coding for pro-inflammatory proteins such as tumor necrosis factor α (TNF-α) and pro-interleukin (IL) 1β. Myeloid cells also possess intracellular PRRs including the nucleotide-binding domain and leucine-rich repeat (NLR) family. NLR family CARD domain- containing protein 4 (NLRC4) and NLR family pyrin domain-containing protein 3 (NLRP3) are the main NLRs engaged in recognising S. Typhimurium infection, leading to formation of the inflammasome. The inflammasome is a macromolecular complex assembled in the cytoplasm, and usually contains a NLR, the structural protein apoptosis-associated speck-like protein containing a CARD (ASC) and effector enzymes such as cysteine-dependent aspartate-directed protease (caspase) -1 and caspase-8. This structure is responsible for processing the cytokines pro- IL-1β and pro-IL-18 to their mature form and is involved in triggering a pro-inflammatory process of cell death termed pyroptosis. The formation of the inflammasome therefore results in cell death and secretion of proinflammatory cytokines which play important roles in controlling infections. Inflammasome activity must be tightly coordinated, as its dysregulation is associated with a variety of auto-inflammatory and auto-immune diseases. The signalling events leading to inflammasome assembly are poorly understood and the molecules involved in fine-tuning its activity are only beginning to be discovered. The aim of this thesis was to discover new molecules involved in inflammasome activation and/or in keeping its activity in check. To achieve this goal, I performed S. Typhimurium infection assays in primary bone marrow derived macrophages (BMDM) derived from C57BL/6 mice wild type (WT) and compared the resulting cellular viability, intracellular bacteria counts and IL-1β production to that of BMDMs derived from C57BL/6 mice lacking proteins involved with, or suspected to be involved with, innate immune activity. Amongst the proteins I studied, caspase recruitment domain 9 (CARD9) inhibited inflammasome-mediated IL-1β production. Multiple independent genome-wide association studies link this protein to inflammatory pathologies such as Crohn's disease, but its role in canonical inflammasomes was largely unexplored. To investigate how CARD9 inhibits inflammasome-mediated IL-1β production I have conducted assays in WT and Card9-/- BMDMs, including stimulation of specific NLRs with their purified ligands, infection with bacterial strains deficient in NLRC4 activation, and infection assays in presence of pharmacological inhibitors. By employing these approaches, I observed that CARD9 has a negative role on NLRP3-dependent IL-1β production. Specifically, in response to activation of the NLRP3 by Salmonella infection, CARD9 negatively regulates pro-IL-1β transcription, and decreases IL-1β processing by inhibiting spleen tyrosine kinase (SYK)-mediated NLRP3 activation and represses caspase-8 activity in the inflammasome. CARD9 expression is suppressed in the course of S. Typhimurium infection which may act as a mechanism to increase IL-1β production during the infection. In conclusion, I have established a connection between CARD9 and IL-1β production by the canonical NLRP3 inflammasome and elucidated some of the mechanisms involved in this process. I have also found evidence that other proteins are likely to be involved in inflammasome regulation and the elucidation of their roles will be addressed in future studies.