Le développement de nouvelles stratégies d’immunothérapie représente de nos jours un enjeu majeur de santé publique. Dans ce contexte, les lymphocytes T Natural Killer invariant (iNKT) exercent de puissantes activités immuno-modulatrices. Ces cellules ont la particularité de reconnaitre par l’intermédiaire de leur récepteur T (TCR), des (glyco) lipides, présentés par la molécule CD1d exprimée par les cellules présentatrices d’antigènes (APC), notamment les cellules dendritiques (DC). Initialement découvert à partir d’une éponge marine pour ses activités anti-métastatiques, l’alpha-galactosylcéramide (a-GalCer ou KRN) induit rapidement une sécrétion massive, par les cellules iNKT, de cytokines immunomodulatrices telles l’IFN-g, conduisant à la transactivation de nombreuses cellules immunitaires, notamment les cellules Natural Killer (NK), les DC ou encore les lymphocytes Tgd. Cette propriété unique permet aux cellules iNKT de contrôler, chez la souris, le développement des réponses immunes, notamment la réponse anti-tumorale. Face à ces résultats encourageants, des essais cliniques chez l’homme ont été réalisés mais les résultats se sont avérés décevants. Actuellement, il existe deux stratégies pour cibler un antigène à une population cellulaire particulière : 1) le ciblage passif, basé sur la taille des particules ou leur composition et 2) le ciblage actif, basé sur les fortes interactions anticorps/antigène ou ligand/récepteur. Nous avons testé les deux types de ciblages et avons utilisé le même polymère pour constituer nos vecteurs : le PLGA ou poly(lactic coglycolic acid). Dans une première étude (ciblage passif), nous avons comparé l’efficacité de l’encapsulation de l’a-GalCer dans des particules de tailles différentes: des nano (NP) et microparticules (MP). L’a-GalCer vectorisé dans les NP et les MP est endocyté par les DC (voie des clathrines) et active les cellules iNKT in vitro et in vivo mais ne peuvent empêcher leur anergie. Ces résultats décevants nous ont conduits à opter pour le ciblage actif des DC. Dans un premier temps, étant donné les études controversées sur le rôle des DC dans l’anergie des cellules iNKT, nous avons revisité l’implication de ces dernières dans la primo-activation/anergie des cellules iNKT. Nous confirmons ainsi le rôle primordial de DC dans la primo-activation des cellules iNKT mais surtout, nous montrons qu’elles n’induisent pas leur anergie. Représentant une population hétérogène, nous montrons que parmi les DC, les DC CD8a+ sont de puissantes activatrices des cellules iNKT. Nos résultats nous ont ainsi menés à délivrer spécifiquement l’a-GalCer aux DC CD8a+. Pour cela, nous avons greffé sur les NP de PLGA un anticorps anti-DEC205, récepteur lectinique fortement exprimé par les DC CD8a+. In vitro et in vivo, les NP/DEC205/a-GalCer induisent une plus forte activation des cellules iNKT comparativement à l’a-GalCer libre. De même, la co-délivrance d’a-GalCer et d’ovalbumine (OVA) au sein des DC CD8a+ améliore les propriétés adjuvantes de l’a-GalCer en induisant des réponses humorale et cellulaire (lymphocytes T CD8+) spécifiques de l’OVA plus importantes comparativement à la délivrance des deux composés sous leur forme libre. Finalement, de façon intéressante, nous montrons que suite à une primo-activation par les NP/DEC205/a-GalCer, les cellules iNKT sont capables de répondre de nouveau à une seconde stimulation, traduisant l’absence d’anergie des cellules iNKT. En conclusion, nos résultats indiquent que la délivrance spécifique de l’a-GalCer aux DC CD8a+ amplifie la primo-activation des cellules iNKT tout en évitant la mise en place du phénomène d’anergie et ouvrent de nombreuses perspectives dans le cadre de thérapies anti-tumorales et anti-infectieuses. / Nowdays, the development of new immunotherapy strategies represent a major issuein public health. In this context, the invariant Natural Killer T lymphocytes (iNKT) have strongimmunomodulatory properties. This cell population recognizes (glycol)lipid presented by theCD1d molecule expressed by antigen presenting cell (APC) as dendritic cells (DC). Initiallyfound in a marine sponge for its anti-metastatic activities, alpha-galactosylceramide (-GalCer or KRN) induces a massive cytokine production (IFN-, IL-4, IL-17) by iNKT cells.This cytokine burst lead to downstream activation of numerous immune cells like naturalkiller cell (NK), DC or CD8+ T cells. Through this property, iNKT cells regulate numerousimmune responses, including anti-tumoral response. Based on encouraging results in themouse model, -GalCer has been used in anti-tumour therapy in human. Although the drugwas well tolerated, no or moderate clinical responses were observed in patients repeatedlyinoculated with -GalCer. As observed in the mouse system, one potential explanation forthis disappointing observation may lie in the induction of a long-term anergy of human iNKTcells, thus preventing cytokine release upon a recall stimulation. Although controversial,various studies suggest that this phenomonen should be due to a lack of delivery of -GalCer into dendritic cells (DC) and so its presentation by non adequate antigen presentingcell (APC) as B cells. The objective of our work was to optimize -GalCer activity by avoidingiNKT anergy using vectorisation approach. Actually, there are 2 strategies using nanotechnologies to target an antigen to a specific cell population: 1) passive targeting based on the size of the particles, their composition and their surface charge and 2) active targeting based on the strong interactions between an antibody and its antigen or a ligand and its receptor. We have tested the two strategies and therefore we use the same composition of our particles: PLGA or poly(lactic co glycolic acid). This biodegradable and biocompatible molecule is already used in therapy.In a first study (passive targeting), we compared the efficiency of -GalCer encapsulation inparticles with different size: nano (NP) and microparticle (MP). Vectorised -GalCer in NPand MP rapidly activates iNKT cells in vitro. Both type of particles are uptake by DC via aclathrin dependent mechanism. In in vivo approaches, NP/-GalCer and MP/-GalCeractivate iNKT cells but unfortunately could not prevent iNKT cell anergy. These disappointingresults led us to use an active targeting. In first time, because of controversial role of DC iniNKT anergy, we have revisited the role of DC in iNKT primo-activation and anergy. Weconfirm the primordial role of DC in iNKT primo-activation but especially we show that DC donot induce iNKT anergy. DC are heterogeneous and we show that among DC, CD8a+ DCsubpopulation are potent iNKT cells activation. Our results led us to deliver specifically a-GalCer to CD8+ DC. For this, anti-DEC205 antibodies were covalently linked to the surfaceof PLGA NP, DEC205 being highly expressed by CD8+ DC. In vitro and in vivo,NP/DEC205/-GalCer induce a stronger iNKT cell activation relative to free -GalCer (or NPIgG/-GalCer). Moreover, -GalCer and ovalbumin co-delivery in CD8+ DC improve -GalCer adjuvanticity leading to more important humoral and cellular responses. Interestingly,after a primo-activation by NP/DEC205/-GalCer, iNKT cells are able to respond to secondstimulation thus avoiding iNKT cell anergy.In conclusion, our results indicate that specific -GalCer delivery to CD8+ DC improve iNKT cells primo-activation and avoid anergy phenomenon. These findings open several perspectives in anti-tumoral and anti-infectious therapies.

Despite effective antibiotics, Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis, still infects nearly one-third of the world's population. While key immune factors including CD4+ T cells and IFNg production have been identified, there are still many antimicrobial mechanisms yet to be explored. Here we characterized the role of invariant natural killer T (iNKT) cells and GM-CSF during Mtb infection.

In this thesis, I examined the expression of CD1d on respiratory epithelial cells (REC) in human and explored its potential role in mucosal immunity in the lungs. Hitherto, there have been no published reports of CD1d expression on REC though it has been observed on other epithelial surface (notably intestinal epithelial cells). This observation, and work in my supervisor’s laboratory demonstrating CD1d-restricted natural killer T cells (iNKT) cells as early players in the lungs of influenza A virus (IAV)–infected mice prompted my interest in this area. I hypothesized that CD1d is expressed on REC and that it contributes to activation of iNKT cells in the lungs via presentation of endogenous or pathogenic glycolipids. I asked following questions – i) is CD1d expressed on REC ii) can this expression be regulated and iii) does CD1d expression on REC have a function. This thesis provides the first evidence for CD1d expression on human RECs (in cell lines and primary RECs) and also presence of alternatively spliced variants. CD1d expression was inducible by viral-associated signals in vitro and despite being non-professional antigen presenting cells, RECs can present glycolipid (α–GC) to, and activate iNKT cells in a CD1d-dependent process resulting in production of both Th1 and Th2 cytokines. Using whole genome expression profiling, I then showed that iNKT cells expressed a distinct profile of genes while in direct contact with α–GC-bound CD1d on RECs compared to cells separated by transwell membrane. Here early biological pathways were dominated by cytokine and chemokine related genes (JAK-STAT signaling pathways, cytokine-responsive elements and cytokine/chemokine genes) and apoptosis-related genes. This suggested that glycolipid-bound CD1d on REC was capable of inducing a programme of immune activation in iNKT cells. I concluded my work by examining if CD1d expression on RECs influenced its active role in immunity. Using wild type and CD1d-deficient transgenic mice challenged with IAV, I showed that CD1d expression is induced on REC in vivo after viral challenge, and in the absence of CD1d, mice showed worse outcome. RECs isolated from CD1d-deficient mice had a much stronger gene expression profile for pro-inflammatory genes. This suggested that CD1d expression on REC could have a bi-directional effect – on the RECs that expressed CD1d (preventing excessive immune-related genes activation) and on the iNKT cells that it engaged (activation, with pro-immunity effects). The thesis concludes with discussion of the potential implications of these findings and future work to examine hypotheses generated from this work.

The absence of invariant Natural Killer T cells (iNKT cells) in mice infected with Influenza A virus (flu) has previously been shown to augment the expansion of Myeloid-derived suppressor cells (MDSCs), a bone marrow derived population that powerfully suppress the development of viral and tumor immune responses. Moreover, iNKT cell adoptive transfer into flu-infected mice has been shown to abolish the expansion and flu-induced suppressive activity of the MDSCs in a CD1d- and CD40-dependent manner. However, the mechanisms by which this relatively small subset of T cells influence myelopoiesis and MDSC differentiation remain largely unknown. In this manuscript we firstly better define the MDSCs found in flu-infection as IL-10-secreting neutrophils that can suppress T cell proliferation. We then go further to show that the flu-induced ability to suppress T cells is acquired as early as the level of the Granulocyte-Macrophage Progenitors (GMPs) in the bone marrow and that iNKT cells can not only abrogate the suppressive activity of the IL-10-secreting neutrophils in the periphery but also that of the GMPs by a direct CD1d-dependent GCSF-mediated crosstalk. MDSC expansion has previously been shown to be associated with the expression of the myeloid-related protein S100A9, and the mechanism of action of granulocytic-MDSCs shown to be ARG1-dependent. We built upon both these findings to show that iNKT cells influence the expansion and function of the MDSCs in part by regulating S100A9 and ARG1 expression. Following this we then showed for the first time that the acute phase protein Serum Amyloid A (SAA), shown to increase during flu-infection, has a dual reciprocal role: having the ability to up-regulate S100A9 and ARG1 in myeloid cells and differentiate IL-10-secreting suppressive neutrophils, while simultaneously facilitating the ability of the MDSCs to crosstalk with iNKT cells in a CD1d-dependent GCSF-mediated manner to abrogate the SAA-induced suppressive activity. All together the data highlights the complexity of the immune response and the role iNKT cells play in influencing the differentiation of MDSCs during demand-driven myelopoiesis. More importantly however, it further affirms that research into harnessing the immunomodulatory capacity of iNKT cells remains an exciting prospect in bolstering future vaccination strategies and should continue to be pursued.

Global lifespans are longer than ever before and there is an increasing shift towards a more aged global population. Also, the majority of severe disease and deaths in the recent and ongoing COVID-19 pandemic is found in individuals over 60 years of age. Therefore, there is an urgent need to gain insight into how the immune system changes with age; specifically: (1) what are the drivers of chronic, systemic inflammation (‘inflammaging’) that occur in some but not all older individuals and (2) how, in turn, numerical aging and chronic inflammation collide to impact anti-viral immunity and lead to poor infection outcomes. In this body of work, two focused research questions were addressed as part of an overarching goal to determine the links between numerical age, chronic inflammatory status, and immune cell function. First, the role of iNKT cells in the inflammation found with ART-suppressed HIV infection and aging was studied, and second, connections between age and pre-existing immunity to SARS-CoV-2 were investigated. Invariant natural killer T (iNKT) cells are a unique, innate-like T cell subset known to bridge innate and adaptive immune responses and can exert inflammatory and immunosuppressive effector functions. The role of iNKT cells in the chronic inflammation found with ART-suppressed HIV infection and/or normal aging is unclear. Therefore, iNKT cell frequencies and surface phenotypes were measured from a HIV and Aging cohort comprised of ART-suppressed HIV+ subjects and matched uninfected controls stratified by age into younger and older groups and iNKT cell signature were correlated with plasma markers of chronic inflammation. Specific characteristics of iNKT cells were associated with aviremic HIV infection and/or advanced age, and distinct links between iNKT cell signatures and markers of chronic inflammation were found. Further, multivariate analysis (PLS-DA) of the collected dataset revealed that iNKT cell and plasma markers stratified younger from older subjects within both the uninfected and aviremic HIV+ groups. Older age is arguably the strongest predictor of severe clinical outcomes and mortality after SARS-CoV-2 infection. The role of pre-existing, cross-reactive immunity in COVID-19 outcomes is unclear to date. A newly developed, highly sensitive serological assay (the BU ELISA) was used to elucidate links between pre-existing immunity to SARS-CoV-2 and age. We found SARS-CoV-2 receptor binding domain (RBD) and/or nucleocapsid protein (N) reactive antibodies (IgG, IgM, and/or IgA isotypes) in all pre-pandemic subjects tested, with a wide range in antibody levels. SARS-CoV-2 reactive immunoglobulin levels tracked with antibodies specific for analogous viral proteins from endemic coronavirus strains and were lowest in subjects over 70 years of age compared with younger counterparts. In sum, these findings provide evidence of lower pre-existing immunity to SARS-CoV-2 in elderly individuals, and this may account for their poor infection outcomes. In conclusion, the findings in this work provide new insight into the impact of age and chronic inflammation on productive and protective immune responses. These results underscore the need for further investigations into the immune cell mechanisms and inflammaging pathways that subvert healthy aging.

The macrophage receptors that mediate phagocytosis of Borrelia burgdorferi, the Lyme disease spirochete, are unknown despite this cell type's importance in promoting pathogen clearance and inflammation-mediated tissue damage. We now demonstrate that the β2 integrin, Complement Receptor 3 (CR3), mediates the phagocytosis of opsonized and non-opsonized spirochetes by murine macrophages and human monocytes. Although, expression of the surface proteins, CspA and OspE, protects B. burgdorferi from complement-mediated phagocytosis, the versatility of CR3 counteracts the ability of B. burgdorferi to interfere with complement activation and complement-derived opsonins, thus minimizing the bacteria's anti-complement strategy. Interaction of the spirochete with the integrin is not sufficient to internalize B. burgdorferi; however, phagocytosis occurs when the GPI-anchored protein, CD14, is coexpressed in CHO-CR3 cells. CR3-mediated phagocytosis occurs independently of MyD88-induced or inside-out signals but requires the translocation of the integrin to cholesterol rich membrane microdomains. Interestingly, the absence of CR3 leads to marked increases in production of TNF in vitro and in vivo, in spite of reduced spirochetal uptake. Overall, our data establish CR3 as a MyD88-independent phagocytic receptor for B. burgdorferi that also participates in the modulation of the proinflammatory output of macrophages. Macrophages are critical cellular components of the immune response to infectious agents. During infection with B. burgdorferi, macrophages infiltrate the cardiac tissue and induce the activation of invariant NKT cells, leading to the production of the protective cytokine IFNγ. The interaction of macrophages with infectious agents leads to the activation of several signaling cascades, including mitogen activated protein kinases, such as p38 MAP kinase. We now demonstrate that p38 MAP kinase-mediated responses are critical components to the immune response during infection with B. burgdorferi. The inhibition of p38 MAP kinase does not alter the ability of macrophages to phagocytose B. burgdorferi; however, inhibition of p38 during infection with B. burgdorferi results in increased carditis. Through the generation of transgenic mice that express a dominant negative form of p38 MAP kinase specifically in macrophages, we demonstrate that this kinase regulates the production of the iNKT attracting chemokine, MCP-1 and the infiltration of these cells to the cardiac tissue during infection. Overall, the inhibition of p38 MAP kinase during infection with B. burgdorferi specifically in macrophages results in the deficient infiltration of iNKT cells and their diminished production of IFNγ, leading to increased bacterial burdens and inflammation. These results show that p38 MAP kinase provides critical checkpoints for the protective immune response to the spirochete during infection of the heart.

Les cellules iNKT (de l’anglais invariant Natural Killer T) constituent un sous-type particulier de lymphocytes T caractérisé par un profil de type inné. Ces cellules répondent rapidement à des antigènes lipidiques et glycolipidique présentés par le CD1d, une glycoprotéine exprimée par les différentes cellules présentatrices d'antigène. Suite à l’activation, les cellules iNKT sont capables de produire de grandes quantités de cytokines anti-inflammatoires et pro-inflammatoires et elles sont impliquées dans différentes maladies, telles que l'allergie, l'auto-immunité, le cancer et les infections, parmi lesquelles la leishmaniose. Les parasites protozoaires de les espèces Leishmania sont les agents causaux de la leishmaniose, une maladie tropicale négligée dont la manifestation la plus grave affecte les organes viscéraux et qui peut être mortelle si elle n'est pas traitée. Le succès de l'infection dépend de la capacité du parasite à maitriser la réponse immunitaire de l'hôte. Récemment, quelques groupes, y compris le nôtre, ont observé que les parasites Leishmania libèrent des vésicules extracellulaires (VE). Les VE sont formées par une bicouche membranaire lipidique, contenant des lipides, des protéines et du matériel génétique et elles peuvent transmettre des molécules dérivées des pathogènes aux cellules hôte sans contact direct entre les cellules. Les VE produites par les parasites Leishmania et aussi par d’autres protozoaires ont été associés à des effets pro-parasite car elles favorisent un environnement plus permissif à l'établissement de l'infection. Dans cette thèse, nous avons étudié l'effet du matériel extracellulaire (ME), correspondant aux VE et aux molécules non-associées aux VE, libéré par les promastigotes de L. infantum sur les cellules iNKT. Dans le début de ce travail, il a été observé que le ME de L. infantum empêche l'expansion ex-vivo des cellules iNKT humaines à partir de cellules mononucléaires du sang périphérique. Cela a mis en évidence la communication entre les cellules iNKT et le ME de L. infantum, ce qui a été exploré par la suite. Le ME de L. infantum module la capacité très importante des cellules iNKT à produire des cytokines. En effet, le ME de L. infantum empêche la production des différentes cytokines par les cellules iNKT, comme par exemple IL-4 et IFNγ. De plus, nous avons aussi démontré que le ME de L. infantum compète avec l’α-GalCer, un agoniste très puissant des cellules iNKT, pour la liaison à la molécule CD1d, ce qui justifie l’effet inhibiteur dans l'activation des cellules iNKT. Nous avons aussi montré que les lipides qui sont présents dans chaque fraction du ME de L. infantum ont un rôle très important dans l’inhibition de l'activation et l'expansion des cellules iNKT. Ainsi, le ME de L. infantum, par ces lipides, peut participer à l’altération de l’activation des cellules iNKT dépendante du CD1d. Cela ajoute une nouvelle évidence de la contribution du ME de L. infantum dans la subversion de la réponse immunitaire de l’hôte. La communication entre le ME libéré par un pathogène et les cellules iNKT a été étudié pour la première fois, ce qui a suggéré un mécanisme de modulation de ces cellules qui n’avait jamais été décrit. Ce travail ouvre des perspectives pour l'étude de l'interaction de ME libéré par d'autres pathogènes avec des cellules iNKT. De plus, l'analyse des lipides contenus dans le ME de L. infantum pourra aboutir à la découverte de nouvelles molécules spécifiques pour inhiber les cellules iNKT. Cela apporterait des avantages significatifs dans les approches cliniques ciblant la modulation de l'activation des cellules iNKT. / The invariant natural killer T (iNKT) cells constitute a particular subset of T lymphocytes characterized by an innate-like profile. These cells rapidly respond to lipid and glycolipid antigens bound by the glycoprotein CD1d expressed by different antigen presenting cells. Once activated, they release large amounts of anti- and proinflammatory cytokines. Thus, iNKT cells are endowed with a remarkable immunomodulatory potential and they have been implicated in different disorders, such as allergy, autoimmunity, cancer and infection, among which is leishmaniasis. Leishmania spp. are a group of protozoan parasites that includes the causative agents of leishmaniasis. This is a neglected tropical disease in which the most severe form of manifestation affects visceral organs and could be fatal if left untreated. Importantly, the success of Leishmania infection relies on the capacity of the parasite to subvert host immune responses. Recently, a few groups, including ours, observed that Leishmania parasites release extracellular vesicles (EVs). EVs are vesicles formed by a lipid bilayer membrane, containing other lipids, proteins and genetic material on their surface as well as in their lumen. Due to their potential to transmit messages between pathogens and host cells without a direct cell contact, they have been a focus of great interest regarding infection. EVs derived from Leishmania and other protozoan parasites have been associated with pro-parasite effects, by creating a more permissive environment to the establishment of the infection. Herein, we studied the effect of the extracellular material (ExM), which encloses both EVs and vesicle-depleted material, released by L. infantum promastigotes in iNKT cells. In the first steps of this work, it was observed that L. infantum ExM is capable of impairing the expansion of human iNKT cells ex vivo from peripheral blood mononuclear cells. This evidenced the cross-talk between iNKT cells and L. infantum ExM that we explored further. L. infantum ExM also modulates the important capacity of iNKT cells to release cytokines, impairing the production of different cytokines, such as IL-4 and IFNγ by these cells. Furthermore, we also show that L. infantum ExM competes with α-GalCer, a potent iNKT cell agonist, for CD1d binding, which justifies its effect in the impairment of iNKT cell activation. Additionally, we also proved the lipids present in each fraction of L. infantum ExM take an important role in the inhibition of iNKT cell activation and expansion. Thus, L. infantum ExM, through their lipids, is suggested to participate in the impairment of CD1d-mediated activation of iNKT cells, adding a new evidence regarding the contribution of the parasite ExM to subvert host immune responses. To the best of our knowledge, this is the first time that the cross-talk between the ExM released by a microbe and iNKT cells was assessed, shedding light on a mechanism of iNKT cell modulation that remained unexplored so far. This opens new perspectives regarding the study of the interaction of the ExM released by other pathogens with iNKT cells. Moreover, a further analysis of the lipid content of L. infantum ExM might allow the finding of new inhibitory molecules specific to iNKT cells, which can bring significant advantages in clinical approaches targeting the modulation of iNKT cell activation.

Le Virus Humain Herpès 8 (HHV-8) est un Herpèsvirus lymphotrope proche du virus d’Epstein Barr (EBV). Au cours de la lymphoprolifération B qu’est la Maladie de Castleman Multicentrique (MCM) HHV-8 est spécifiquement associé à une profifération de plasmablastes monotypiques IgM/. Ces cellules expriment des facteurs de transcription qui suggèrent que ces cellules sont au stade plasmablastique ou pré-plasmocytaire de la différenciation B.Les cellules invariantes Natural Killer (iNKT) sont des cellules innées qui jouent un rôle dans l’immunité antivirale, en particulier dans le contrôle des Herpèsvirus. Une diminution de ces cellules est associée à l’infection VIH ou l’âge, deux situations associées aux pathologies tumorales associées à HHV-8. Dans la première partie du travail nous avons analysés les iNKT chez des patients MCM et montrés des anomalies de fréquence et de prolifération de ces cellules. Les anomalies des cellules sont associées à des anomalies de répartition des sous populations B mémoires dans le sang circulants et la rate de ces patients. Des expériences de co-cultures montrent que les cellules iNKT pourraient être nécessaires au maintien de ces populations BDans la seconde partie de ce travail, nous avons démontré que la MCM HHV-8 est associée pendant les poussées de a maladie à une circulation dans le sang périphérique de cellules présentant les caractéristiques typiques des plasmablastes décrits jusqu’à présent uniquement dans les tissus lymphoïdes. Nous avons ensuite analysé le profil d’expression génique des cellules B infectées par HHV-8 par rapport à celui de sous populations B normales. Nos résultats montrent clairement que les cellules infectées par HHV-8 présentent un profil d’expression génique très différent de celui des sous populations B normales. Leur profil est par ailleurs caractéristique de plasmablastes. De plus, ces cellules sont en prolifération, modulent négativement l’expression de beaucoup de gènes associés à l’immunité et l’adhésion cellulaire. Enfin, nous confirmons que les cellules B infectées par HHV-8 de la MCM sont bien polyclonales même dans le sang circulant et sans mutations somatiques.Au total, ces résultats nous permettent de proposer un modèle de la physiopathologie de la MCM associée à l’infection HHV-8. / Human Herpesvirus-8 is a B-lymphotropic \γ-herpesvirus closely related to the Epstein-Barr virus (EBV). He is specifically associated with monotypic (IgM/λ) plasmablasts in Multicentric Castleman disease (HHV-8 MCD), which is a B lymphoproliferative disorder. These cells express transcription factors suggesting they are at the plasmablast or pre-pasma cell stage of differentiation. Invariant natural killer T (iNKT) cells are innate-like T cells that play a role in antiviral immunity, specifically in controlling viral replication in EBV-infected B cells. Decline of iNKT cells is associated with age or HIV infection, both situations associated with HHV-8-related diseases. We demonstrated that iNKT cell abnormalities are associated with HHV-8 MCD. These iNKT cell alterations were found to be associated with an imbalance in the frequency of circulating and splenic B cell subsets, and results of Coculturing experiments indicate that iNKT cells may be required for maintaining this cell population. In the second part of our work thesis, we demonstrate that HHV-8 MCD is associated with a unique population of circulating plasmablasts detected during the flare of the disease, with the typical phenotype of the MCD HHV-8-infected plasmablasts in MCD lesions. Then, we used gene expression profile analysis (about 48 000 genes) to further define the phenotype of this MCD HHV-8-infected cells and to investigate the lymphoma relationship to normal B cell subpopulations. The results showed that MCD HHV-8-infected cells displayed a common gene expression profile that is clearly distinct from all the normal B cell subpopulations. The gene expression profile of MCD HHV-8-infected cells was defined as plasmablastic. Moreover, the transcriptomic pattern of MCD HHV-8-infected cells demonstrates that these cells are proliferating and escaping the immune system. Finally, we determined the clonality and the cellular origin of the monotypic circulating plasmablasts by studying the rearranged immunoglobulin heavy genes in LANA+ HHV-8-infected B cells from patients with HHV-8 MCD. Our results show that these cells are polyclonal without somatic mutation. Altogether our results allowed us to elaborate a model of MCD physiopathology.

Les traitements actuels de la leucémie myéloïde chronique (LMC) ne permettent pas d’éliminer la totalité des cellules leucémiques. Dans le but de développer un traitement curatif, il est donc nécessaire de parvenir à une meilleure compréhension des mécanismes sous-jacents des réponses partielles aux traitements, Dans ce travail, nous avons postulé qu’il existe des mécanismes de contrôle extrinsèques de la LMC pouvant influencer l’efficacité des différents traitements. Nous avons choisi d’étudier le rôle potentiel dans la LMC des lymphocytes iNKT, cellules T de type « inné » auxquelles la littérature attribue de nombreuses fonctions antitumorales et des facteurs moléculaires, la cytokine/alarmine IL-33, produite dans la niche hématopoïétique, et son récepteur ST2 à la surface des cellules hématopoïétiques comme cibles de l’IL-33.La première partie de notre travail a ainsi permis de mettre en évidence de profondes altérations fonctionnelles des lymphocytes iNKT chez les patients atteints de LMC ainsi qu’une correction partielle de ces défauts après traitement par l’Imatinib (IM) ou l’IFN-. L’ensemble de ces résultats permet de proposer que l’altération des fonctions des cellules iNKT au cours du développement de la LMC pourrait participer aux mécanismes d’échappement de la tumeur au contrôle par le système immunitaire. La deuxième partie de notre travail a permis de mettre en évidence une expression de la molécule ST2, chaîne spécifique du récepteur à l’IL-33, à la surface des cellules CD34+ de patients atteints de LMC, expression non décelée chez les sujets sains et les patients en rémission après traitement par l’IM. De plus, contrairement aux cellules CD34+ de sujets sains, les cellules progénititrices de patients en phase chronique prolifèrent en réponse à l’IL-33. Enfin, nous avons montré que l’IL-33 est capable de contrecarrer in vitro les effets antiprolifératifs de l’IM. Ainsi nous pouvons émettre l’hypothèse selon laquelle l’IL-33, une cytokine/alarmine, puisse participer aux phénomènes conduisant à la persistance de progéniteurs hématopoïétiques leucémiques chez les patients sous traitement par IM. / To date, treatment of Chronic Myeloid Leukaemia (CML) is not sufficient to completely eradicate leukaemia cells. Hence, in order to develop a curative treatment, it is necessary to have a better understanding of the underlying mechanisms explaining why response to treatment is only partial..We therefore addressed the question whether the extrinsic mechanisms of CML control can affect the effectiveness of different treatments. We first provided evidence of profound functional impairments of iNKT cells in patients with CML. Interestingly these impairments were partially corrected after treatment with Imatinib (IM) or IFN-. Consequently our results suggest that altered functions of iNKT cells during the development of CML could facilitate tumour escape from immune destruction. . The second part of our work revealed that CD34+ progenitors from CML patients upregulate their cell surface expression of the IL-33-specific receptor chain ST2, proliferate and produce cytokines in response to IL-33, conversely to CD34+ cells from healthy individuals. Moreover, ST2 overexpression is normalized following IM therapy, while IL-33 counteracts in vitro IM-induced growth arrest in CML CD34+ progenitors. From these findings, it can be surmised that IL-33, a cytokine/alarmin likely expressed in the hematopoietic niche, facilitates the development of CML and IM resistance.

Made available in DSpace on 2015-07-22T20:50:18Z (GMT). No. of bitstreams: 0 Previous issue date: 2009-11-25 / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Células iNKT restritas a CD1d são protetoras contra o desenvolvimento de melanoma murino, B16F10-Nex2, subcutaneamente em animais singênicos. Essa proteção pode ser deduzida pelo desenvolvimento acelerado do tumor em animais geneticamente deficientes na expressão de CD1d (CD1d-KO), com sobrevida significantemente menor do que a observada com animais WT submetidos ao mesmo desafio de células tumorais. CD1d é uma família de glicoproteínas relacionadas a MHC classe I, envolvidas na apresentação, principalmente em células dendríticas, de antígenos lipídicos para células iNKT. No presente trabalho focalizamos a identificação do lipídeo endógeno expresso em células de melanoma capaz de induzir uma resposta de vigilância imune baseada em células iNKT. Verificamos também a possibilidade de proteger animais contra o desenvolvimento tumoral utilizando células dendríticas primadas com o lipídeo endógeno. A extração dos componentes lipídicos de melanoma foi realizada a partir de tumores crescidos in vivo, evitando-se assim artefatos do cultivo celular in vitro. Foram testados três diferentes protocolos de extração (A, B e C), obtendo-se 14 frações diferentes que foram testadas na ativação do hibridoma DN32.D3, uma linhagem de células NKT murinas imortalizadas. A fração F3 do protocolo A (F3A) ativou o hibridoma DN32.D3 medido pela produção de IL-2. Para uma eficiente apresentação dos lipídeos dessa fração utilizamos com sucesso células dendríticas de medula óssea (BMDC) nos ensaios in vitro e in vivo, para apresentação de F3A e glicolipídeos ativadores de células NKT, agalactosilceramida (a-GalCer) e isoglobotrihexosilceramida (iGb3). Na tentativa de isolar o composto estimulatório presente em F3A de melanoma, essa fração foi analisada por HPTLC revelada com diversos reagentes específicos para resíduos de ácido siálico, açúcares neutros, fosfato e lipídeos totais. A fração também foi submetida a separações em colunas de sílica, ensaio de “immunostaining” quimioluminescente com lectina de Bandeiraea simplicifolia e análises em espectrometria de massa, onde foram identificados gangliosídeos como GM3 bem como glicoesfingolipídeos neutros como iGb3, Gb3, iGb4 e Gb4 por ESI-LIT-MS (electrosptray ionization-linear íon trap-mass spectrometry). Quando iGb3 foi incubado com BMDC e testado com células DN32D3 essas foram ativadas produzindo IL-2. GM3 consistentemente era um inibidor dessa ativação de células iNKT. Ensaios de citotoxicidade foram então realizados e verificamos que células de NKT estimuladas por BMDC incubadas com a-GalCer ou iGb3 circundavam as células tumorais B16F10-Nex2 visualizadas em microscopia de fluorescência e, em ensaio in vitro, as células NKT promoviam lise de até 40% das células tumorais B16F10-Nex2. Realizamos então ensaios in vivo, onde camundongos foram inoculados endovenosamente com células do melanoma murino e tratados ou não com células dendríticas primadas com a-GalCer ou iGb3. Ao excisarmos os pulmões dos animais, notamos que os grupos tratados com lipídeos ativadores de células NKT tinha cerca de 4 vezes menos nódulos pulmonares que o grupo não tratado. Nossos resultados mostram que o melanoma murino B16F10-Nex2 possui a molécula iGb3 e sua precursora iGb4, capaz de ativar células NKT conferindo a essas capacidade citotóxica in vitro contra o melanoma. Esse lipídeo (iGb3) também mostrou um efeito protetor contra metástases pulmonares oriundas do melanoma murino quando apresentado por células dendríticas utilizadas em protocolo de terapia celular. Esse é o primeiro trabalho mostrando que efetivamente um glicolipídeo endógeno ligante de CD1d é capaz de ativar células NKT com efeito protetor antitumoral, através de terapia celular com células dendríticas. Palavras chave: Melanoma B16F10-Nex2, células iNKT, glicoesfingolipídeos iGb3 e iGb4 GM3, células dendríticas, imuno vigilância. / CD1d-restricted iNKT cells are protective against the murine melanoma B16F10- Nex2 growing subcutaneously in syngeneic animals. This is inferred from the fast tumor growth in animals genetically deficient in CD1d (CD1d-KO), which showed a survival time significantly shorter than that of WT animals equally challenged with tumor cells. CD1d belongs to a family of glycoproteins resembling MHC class I, involved in the presentation, chiefly in dendritic cells, of lipidic antigens to iNKT cells. In the present work we focus on the identification of an endogenous lipid component expressed in melanoma cells able to induce an immunosurveillance response based on iNKT. We also investigated the possibility of animal protection against tumor development by using dendritic cells primed with the endogenous lipid. The extraction of membrane lipid components was carried out from in vivo grown tumors, thus avoiding artifacts from the in vitro grown cultures. Three different extraction protocols were tested (A, B, C), and 14 different fractions were obtained and tested for the activation of hybridoma DN32.D3, a cell line of immortalized murine NKT cells. Fraction F3 of protocol A (F3A) activated hybridoma DN32.D3 to produce IL-2. For an efficient presentation of lipids from this fraction we successfully used bone marrow dendritic cells (BMDC) on in vitro and in vivo assays. F3A and NKT-cell activating glycolipids, agalactosylceramide (a-GalCer) and isoglobohexosylceramide (iGb3), were tested. In the attempt to isolate the stimulatory component present in the melanoma F3A fraction, HPTLC was used and revealed with several specific reagents for sialic acid residues, neutral sugars, phosphate and total lipids. The fraction was also separated in silica columns, immunostained with Bandeiraea simplicifolia BS-1 lectin and analyzed by mass spectrometry. Ganglioside GM3 and neutral glycosphingolipids iGb3, Gb3, iGb4 and Gb4 were identified by ESI-LIT-MS (electrospray ionization- linear ion trap- mass spectrometry). By incubation of iGb3 with BMDC and these with DN32.D3 cells, the latter were activated to produce IL-2. GM3 consistently inhibited the activation of iNKT cells. Cytotoxicity assays were then carried out, in which we found that NKT cells stimulated by BMDC, primed with a-GalCer or iGb3, encircled the B16F10-Nex2 tumor cells as visualized by fluorescent microscopy. NKT cells promoted lysis of up to 40% of tumor cells. In vivo tests showed that mice injected endovenously with murine melanoma cells and treated with dendritic cells primed with a-GalCer or iGb3, had lungs with 4-fold fewer nodules than an equally challenged but untreated group. The present results show that the murine melanoma B16F10-Nex2 expresses iGb3 and its precursor iGb4, being able to activate NKT cells and conferring them a cytotoxic activity in vitro against melanoma. Such lipid (iGb3) was also protective in vivo reducing the melanoma pulmonary metastases when presented by dendritic cells used in cellular therapy protocol. This is the fist work showing effectively that an endogenous CD1d-restricted glycolipid able to activate iNKT cells display a protective antitumor effect, using cellular therapy with dendritic cells. / TEDE / BV UNIFESP: Teses e dissertações