La coévolution ainsi que les différentes interactions entre hôte et pathogène contribuent à former la diversité génétique de ces deux organismes. Dans le cadre de cette thèse, nous nous sommes intéressés à l’étude de la variabilité génétique de 1760 gènes immunitaires choisis suivant des critères définis, pour essayer d’expliquer pourquoi il existe une variation individuelle face aux infections. L’objectif principal de ce projet était alors de caractériser et d'analyser de nouveaux gènes et polymorphismes immunitaires pouvant expliquer le contrôle ou la susceptibilité à certaines infections. Deux études pilotes nous ont permis de développer le pipeline de détection de polymorphismes. Pour la première, le polymorphisme des 3 gènes CD28, CTLA4, et ICOS a été caractérisé. Dans la deuxième, nous avons caractérisé le polymorphisme de 10 gènes impliqués dans la réponse immunitaire contre M. tuberculosis. Ces gènes ne sont pas très polymorphes et trois d’entre eux sont très conservés. Ces deux études nous ont aidés à préparer l’analyse à grande échelle avec les mises au point et l’amélioration du pipeline. Nous avons sélectionné 1760 gènes en se basant sur des critères définis. La variabilité génétique a été étudiée dans les populations humaines par une analyse minutieuse in silico de données de séquençage d’exomes générées par différents projets et consortiums pour plus de 700 individus représentant 20 populations à travers le monde. 30 gènes les plus polymorphes ont été ainsi identifiés. Ces gènes pourront être entièrement caractérisés et les données produites pourraient être comparées avec des données de résistance/sensibilité de certaines maladies infectieuses. / Host-pathogen co-evolution and interactions contribute in shaping the genetic diversity of both organisms. The objective of this thesis is to define the genetic basis of variability in disease resistance/susceptibility through the development of large-scale in silico screens to identify novel gene candidates implicated in host-pathogen interactions (such as tuberculosis).A pilot study was conducted on CD28, CTLA4, and ICOS to investigate their polymorphism. As a first step in our study based on data available in the literature, we selected a set of ten genes relevant for the immune response against M. tuberculosis. Seven of these genes were moderately polymorphic, while three of them were highly conserved. This analysis was used to prepare and setup the large scale analysis using the same developed pipeline for polymorphism detection and allele reconstruction. For our in silico, we used sequence data from several projects and consortiums to isolate most polymorphic human genes amongst a list of over 1760 candidates selected based on already established relevance for infections and on evolutionary considerations. A first screen of 64 individuals from eight different populations from several regions of the world was performed and most variable genes were selected for further extensive analyses on a larger panel (715 individuals). 30 most polymorphic genes were thus identified. The extent of polymorphism and the allelic worldwide variants of each of these 30 genes are ready to be fully characterized. The data generated could be compared against infectious disease resistance/susceptibility data to identify potentially relevant gene variation.

Banana is the most important fruit crop in the world but ironically one of the crops least studied. This fruit constitutes a major staple food for millions of people in developing countries and also it is considered the highest selling fruit in the world market making this crop a very important export commodity for the producing countries. At the present time, one of the most significant constraints of banana production that causes significant economical losses are fungal diseases. Among these, Panama disease, also known as Fusarium wilt has been the most catastrophic. Panama disease is caused by the soil-borne fungus Fusarium oxysporum formae specialis (f.sp) cubense (FOC), which infects susceptible bananas through the roots causing a lethal vascular wilt. To date, the race 4 of this pathogen represents the most serious threat to banana production worldwide since most of the commercial cultivars are highly susceptible to this pathogen. Introduction of FOC resistance into commercial cultivars by conventional breeding has been difficult because edible bananas are sterile polyploids without seeds. Genetic transformation of banana, which has already been established in various laboratories around the world has the potential to solve this problem by transferring a FOC race 4 resistance gene into susceptible banana cultivars (eg. Cavendish cultivars). However, a FOC resistant (R) gene has not been isolated. Genes that confer resistance to Fusarium oxysporum have been isolated from tomato and melon using a map-based positional cloning approach. The tomato I2 and melon Fom-2 genes belong to the non-Toll/interleukin like receptors (TIR) subclass of nucleotide-binding site and leucine-rich repeat (NBS-LRR) R genes. These genes confer resistance only to certain races of F. oxysporum in their corresponding plant families limiting their use in other plant families. The fact that these two Fusarium resistance genes share the same basic non-TIR-NBS-LRR structure suggests a similar Fusarium resistance mechanism is shared between the families Solanaceae and Cucurbitaceae. This observation opens the possibility to find similar Fusarium resistance genes in other plant families including the Musaceae. A remarkable discovery of a population of the wild banana Musa acuminata subspecies (ssp.) malaccensis segregating for FOC race 4 resistance was made by Dr. Ivan Buddenhagen (University of California, Davis) in Southeast Asia. Research carried out at Queensland Department of Primary Industries (Australia) using this plant material has demonstrated that a single dominant gene is involved in FOC race 4 resistance (Dr. Mike Smith, unpublished results). Tissue-culture plantlets of this FOC race 4 segregating population were kindly provided to the Plant Biotechnology Program (Queensland University of Technology) by Dr. Mike Smith to be used in our research. This population holds the potential to assist in the isolation of a FOC race 4 resistance gene and other potential Fusarium resistance genes. The overall aims of this research were to isolate and characterise resistance gene candidates of the NBS-type from M. acuminata ssp. malaccensis and to identify and characterise potential Fusarium resistance genes using a combination of bioinformatics and gene expression analysis. Chapter 4 describes the isolation by degenerate PCR of five different classes of NBS sequences from banana (Musa acuminata ssp malaccensis) designated as resistance gene candidates (RGCs). Deduced amino acid sequences of the RGCs revealed the typical motifs present in the majority of known plant NBS-LRR resistance genes. Structural and phylogenetic analyses showed that the banana RGCs are related to non-TIR subclass of NBS sequences. The copy number of each class was estimated by Southern hybridisation and each RGC was found to be in low copy number. The expression of the RGCs was assessed by RT-PCR in leaf and root tissues of plants resistant or susceptible to Fusarium oxysporum f. sp. cubense (FOC) race 4. Four classes showed a constitutive expression profile whereas no expression was detected for one class in either tissue. Interestingly, a transcriptional polymorphism was found for RGC2 whose expression correlated with resistance to FOC race 4 suggesting a possible role of this gene in resistance to this devastating FOC race. Moreover, RGC2 along with RGC5 showed significant sequence similarity to the Fusarium resistance gene I2 from tomato and were chosen for further characterisation. The NBS sequences isolated in this study represent a valuable source of information that could be used to assist the cloning of functional R genes in banana. Chapter 5 describes the isolation and characterisation of the full open reading frame (ORF) of RGC2 and RGC5 cDNAs. The ORFs of these two banana RGCs were predicted to encode proteins that showed the typical structure of non-TIR-NBS-LRR resistance proteins. Homology searches using the entire ORF of RGC2 and RGC5 revealed significant sequence similarity to the Fusarium resistance gene I2 from tomato. Interestingly, the phylogenetic analysis showed that RGC2 and RGC5 were grouped within the same phylogenetic clade, along with the Fusarium resistance genes l2 and Fom-2. These findings suggest that the banana RGC2 and RGC5 are potential resistance gene candidates that could be associated with Fusarium resistance. The case of RGC2 is more remarkable because its expression was correlated to FOC race 4 resistance (Chapter 4). As a first step to test whether RGC2 has a role in FOC race 4 resistance, different expression constructs were made with the ORF of this sequence. One of the constructs contains a RGC2 putative promoter region that was successfully cloned in this work. These constructs will be used to transform susceptible banana plants that can then be challenged with FOC race 4 to assess whether resistance has been acquired by genetic complementation. The results of this thesis provide interesting insights about the structure, expression and phylogeny of two potential Fusarium resistance genes in banana, and provide a rational starting point for their functional characterisation. The information generated in this thesis may lead to the identification of a Fusarium resistance gene in banana in further studies and may also assist the cloning of Fusarium resistance genes in other plant species.