Erythropoiesis is the process by which the hematopoietic tissue of the bone marrow produces red blood cells (erythrocytes). The principal function of erythrocytes is to deliver oxygen from the lungs to the other tissues of the body. To perform their duty as oxygen carriers, erythrocytes require iron. Failure to incorporate adequate iron into heme results in impaired erythrocyte maturation, leading to microcytic, hypochromic anemia. The erythroblasts mature in organized niches called erythroblastic island consisting of a central macrophage that extends cytoplasmic protrusion to a ring of surrounding erythroblasts. Although the feature of erythroblast islands is now recognized as an important contributor to normal erythroid development, as well as altered erythropoiesis in diverse diseases, as anemia of inflammation and chronic disease, myelodysplasia, thalassemia, and malarial anemia, how macrophages support erythroid maturation is poorly understood. The most important systemic factors that influence iron availability is hepcidin, a circulating peptide that maintains iron homeostasis. Hepcidin is an hyposideremic hormone made predominantly by hepatocytes acting to downregulate iron absorption by the duodenal cells and iron release by the macrophages. Elevated levels of hepcidin in the bloodstream shut off iron absorption, and, conversely, low levels of circulating hepcidin allow increased iron export into the bloodstream. Aberrations in hepcidin expression or responsiveness to hepcidin result in disorders of iron deficiency and iron overload, which represent an important public health issue worldwide. As befits an iron regulatory hormone, hepcidin gene expression is highly regulated by the iron status of the body. While the molecular mechanisms of hepcidin induction by iron begin to be relatively well understood, not much is known concerning the absolute requirement of hepcidin repression in conditions of anemia, accelerated erythropoiesis and hypoxia to stimulate iron release by duodenal and macrophagic cells. Several line of evidence indicate that erythroid precursors most likely communicate directly their iron needs to the liver to influence the production of hepcidin and thus the amount of iron available for use. However, the mechanism by which erythroid cells modulate hepcidin production of liver cells has not yet been elucidated. The major goal of this application is to identify soluble molecules released by erythroid cells during their differentiation and to study their role in iron metabolism both at the central level, through the regulation of hepcidin production by liver cells, and at the local level, through their effects to the bone marrow macrophages. The ability of these molecules to regulate hepcidin production by liver cells will be tested by Partner 1 and Partner 2 will determine whether these molecules can also modify the physiology of the erythroblastic island macrophages. The question of whether the liver serine protease matriptase 2, a recently characterized hepcidin gene repressor, is the link between these soluble factors and hepcidin gene expression will be specifically addressed through ex vivo and in vivo studies through the generation and analysis of different mouse models. Improved understanding of how soluble erythroid factors impact on iron homeostasis locally, through maintaining the physiological integrity of the erythroid island, and systematically, through regulation of hepcidin synthesis, will be important to limit iron-mediated pathology, in particular anemia, a condition that affected one-quarter of the world’s population and constitutes a public health problem in both developing and industrialized countries.

Summary Acquisition of immune tolerance to gene transfer products is a major issue for the success of novel therapeutic modalities in the field of gene therapy of rare monogenic disorders. This is of particularly high importance for therapeutic transgenes delivered in non-lymphoid tissues and detailed knowledge of the molecular and cellular pathways that control the initiation of cytotoxic immune responses (CD8+ T cell responses) is unknown. I wish here to dissect the molecular components which control the presentation of transgene product delivered in muscle tissue as well as the mechanisms leading to recruitment of regulatory T cell activities and sustained immune tolerance. Using defined cell-associated antigens delivered with clinically relevant adeno-associated viral vector (AAV) systems, I will assess the contribution of the direct and indirect MHC class I presentation pathways leading to cross-tolerisation of CD8 T cell responses. For this, I will use engineered AAV vectors in which was inserted in 3’ of the transgene the target of microRNA142.3p (miRNA strongly expressed in the hematopoietic system) to restrict transgene expression in muscle tissues, avoiding direct expression in antigen presenting cells. This system allows us to characterize for the first time the contributions of the direct and indirect pathway of transgene product presentation in vivo. Using 1) mice genetically invalidated for different components of the direct and indirect MHC class I presentation pathways, 2) cellular ablation or transfer of key players such as regulatory T cells, and 3) novel immune-modulation approaches such as the anti-T cell mAb (anti-CD3) used in the clinics to reset tolerance in autoimmune type 1 diabetes, we will delineate the rules that govern the induction of tolerance to foreign transgenes delivered in peripheral tissues. I believe that the fundamental knowledge acquired through this program will be important to develop novel tolerance induction modalities for both gene therapy and autoimmunity applications. Publications of the team Chappert P, et al. Antigen-specific Tregs impair CD8 T cell priming by blocking early T cell expansion. Eur J Immunol. (2010) 40:339-50 Saveanu L, et al. IRAP identifies an endosomal compartment required for MHC class I cross-presentation. Science. (2009). 325:213-7 Chappert P, et al. Antigen-driven interactions with dendritic cells and expansion of foxp3+ regulatory T cells occur in the absence of inflammatory signals. J Immunol. (2008) 180:327-34 Lorain S, Gross DA, et al. Transient immunomodulation allows repeated injections of AAV1 and correction of muscular dystrophy in multiple muscles. Mol Ther. (2008) 16:541-7 Firat E, et al. The role of endoplasmic reticulum-associated aminopeptidase 1 in immunity to infection and in cross-presentation. J Immunol. (2007) 178:2241 Gross DA, et al. Simple conditioning with mono-specific CD4+CD25+ regulatory T cells for bone marrow engraftment and tolerance to multiple gene products. Blood. (2006) 108:1841-8 Saveanu L, et al. Concerted peptide trimming by human ERAP1 and ERAP2 aminopeptidase complexes in the endoplasmic reticulum. Nat Immunol. (2005) 6:689 Saveanu L, et al. Dendritic cells: open for presentation business Nat Immunol. (2005) 6:7 Review Gross DA, et al. High vaccination efficiency of low-affinity epitopes in antitumor immunotherapy. J Clin Invest. (2004) 113:425-33 Gross DA, et al. CD4+CD25+ regulatory T cells inhibit immune-mediated transgene rejection. Blood. (2003) 102:4326-8

The origin and stability of cooperation is a long-standing and challenging question in evolutionary biology. Emergence, maintenance, and dynamics of interactions between individuals, are difficult to explain in a natural world, typically perceived as driven by ruthless, “survival of the fittest” competition. However, instances of cooperation are abundant, from multicellularity and resource sharing, to quorum sensing and virulence triggered by collective behaviour in bacteria. Currently, we still do not know how cheater individuals, which benefit from cooperation without paying its costs, are prevented from invading populations of cooperating individuals. Leading theories, such as reputation, kin selection, reciprocity, or “green beards”, have shown promising results, but scarcely tested, even in microbial communities, most successful experimental systems for the study of cooperation. Here we test a novel hypothesis based on two striking features of social interactions in microorganisms (1) cooperation mechanisms are often based on secretion of public good (a compound beneficial not just to organisms paying the metabolic cost to produce it, but to all the organisms in the population), (2) genes encoding for public good molecules are preferentially located on mobile genetic elements. In particular, plasmids carrying public good genes may enforce cooperation by actively transforming cheaters into public good producing, cooperating individuals. Reversely, if cooperation is necessary for survival in a certain environment, it may in turn promote conjugation that helps it be maintained. In the last part of our study, we will also address complex ecological dynamics, by studying population extinction due to loss of cooperation, a process akin to aging of multicellulars. To test the theories on cooperation and information transfer, we will use a combination of two powerful research techniques, experimental and digital evolution. Experimental evolution is a growing and successful field as exemplified by the E. coli Long Term Experimental Evolution project by Dr. Richard Lenski and Drosophila laboratory evolutionary radiation by Dr. Michael Rose. However, as done presently, experimental evolution remains time/energy-consuming and limited in terms of statistical power and number of different experimental conditions that can be tested simultaneously. To continue improving this effective methodology, we propose to develop a new research platform, the Robot Assisted Competition and Evolution (RACE). Coupling directed evolution and synthetic biology approaches to the robotic automation would replace most of the repetitive operations of every day bench-work and enable long term and high throughput experimental evolution. RACE will be paired with the in silico models developed using the state-of-the-art simulation Aevol. We will evolve thousands of populations of digital organisms and examine billions of computer programs that reproduce, conjugate, mutate, and compete with each other. Aevol organisms have been implemented specifically with bacteria in mind but generally represent yet another instance of evolution, this time on a computer chip. We will conduct parallel in vitro and in silico experiments, directly compare the results to test our theoretical predictions and use any differences to understand the role of system-specific parameters in the evolution and maintenance of cooperation and information transfer. In the past decade, the knowledge of E. coli systems biology and genomics as well as the complexity of digital organisms in simulations like Aevol have grown and matured so we can now successfully bridge the gap between the two. The potential of RACE and Aevol goes beyond just understanding the evolution of cooperation and plasmid conjugation; once fully developed, our approach will provide a versatile toolbox for testing general problems in ecology and evolution dealing with complex, multi-agent dynamics.

Immune tolerance is key to the maintenance of the integrity of organisms against foreign invaders with respect to self-constituents. Deregulation of this mechanism promotes the occurrence of life-threatening autoimmune diseases that affect 5-10% of the general population. Previously, the product of the autoimmune regulator (Aire) gene was shown to play a key role in immune tolerance. Indeed, Aire induces medullary epithelial cells in the thymus (MECs) to synthesize and present a large repertoire of peripheral self-antigens (a “self-shadow”), leading to the clonal deletion of self-reactive maturing T cells and thereby protecting against autoimmune manifestations. Recently, important advances have enlightened the intriguing mode of action of Aire in showing that Aire recruits the transcriptional machinery at silenced genes and activates transcription elongation. In addition to activating transcription elongation, preliminary results indicate that Aire induces 3’UTR shortening of its sensitive transcripts in MECs, and that these cells show an accumulation of miRNAs. The overall goal of my project is to describe a post-transcriptional control of the Aire-driven expression in the thymus that leads to higher levels of Aire-dependent self-antigens, and to identify key molecular players involved in this mechanism. I propose as a first task to establish a widespread increase of miRNAs in MECs by quantifying the expression of a comprehensive panel of miRNAs. As miRNAs bind to 3’UTRs, and as 3’UTRs are subjected to dynamic regulation, we will assay the extent of 3’UTR length regulation in MECs by using Affymetrix whole-transcript microarrays and by analyzing the data at the single probe level (task 2). In order to characterize the escape of the Aire-induced genes from a repressive effect of miRNAs, we will assay the Aire-specific 3’UTR shortening in WT vs Aire-KO MECs by mRNA high-throughput sequencing (task 3). Subsequently, we will perform reporter assays in a MEC cell line in order to evaluate the increase of protein levels of short 3’UTR isoforms compared to long 3’UTR isoforms. We will also assay the impact of miRNAs on the post-transcriptional repression, by mutating the complementary sites of miRNAs in long 3’UTR reporters. The following objective (task 4) will be to identify the factors involved in the 3’UTR shortening triggered by Aire. We will first set up an in vitro model of Aire-triggered 3’UTR shortening, based on transient transfection of a MEC cell line with an Aire expression vector and a dual-luciferase reporter construct containing a prototypic 3’UTR whose transcript undergoes Aire-specific 3’UTR shortening as a read-out. Subsequently, an extended set of RNA-binding factors potentially involved in Aire’s mode of action will be screened by short-hairpin RNA (shRNA)-containing lentivirus infection of the in vitro model that we set up. Finally, to validate in vivo the effect of some RNA-binding proteins identified by the shRNA screen (the best candidates), we will generate knock-down mice using a high-speed lentigenic approach, which is based on oocyte infection with very-high titer shRNA-containing lentiviruses (task 5). Altogether the expected results should uncover an important layer of control of self-antigen expression in the thymus, at the post-transcriptional level. Completion of this project should shed a new light on our understanding of the processes that mediate the establishment of immune tolerance in the thymus, and provide new potential targets for therapeutic intervention in autoimmune diseases.