The majority of human proteins are post-translationally modified by covalent addition of one or more complex oligosaccharides (glycans). Alterations in glycosylation processing are associated with numerous diseases and glycans are attracting increasing attention both as disease biomarkers and as targets for novel therapeutic approaches. Using a recently developed high performance liquid chromatography (HPLC) method for high-throughput glycan analysis, genome-wide association studies (GWAS) of 33 directly measured and 13 derived N-glycan features were performed in 3533 individuals from four European isolated populations. Polymorphisms at six loci were found to show genome-wide significant association with plasma concentrations of N-glycans. Several of these gene products have well characterised roles in glycosylation, however, SLC9A9 and HNF1A were two of the novel findings. Subsequent work performed by collaborators found HNF1A to be a “master regulator” of genes involved in the fucosylation of plasma N-glycans. Additionally, this work led to the discovery that N-glycans could act as biomarkers to discriminate HNF1A-MODY from type 1 and type 2 diabetes mellitus (T1D, T2D) patients. After the success of the total plasma N-glycan GWAS, it was thought that stronger and more biologically interpretable associations may be found from the investigation of N-glycans isolated from a single protein. Glycosylation of immunoglobulin G (IgG) influences IgG effector function by modulating binding to Fc receptors. To identify genetic networks that govern IgG glycosylation, N-linked IgG glycans were quantitated using ultra performance liquid chromatography (UPLC) in 2247 individuals from the same four European populations from the previous study. GWAS of the 77 N-glycan measures identified 15 loci with a p-value < 5x10-08. Four loci contained genes encoding glycosyltransferases, while the remaining loci contained genes that have not previously been implicated in protein glycosylation. However, most have been associated with autoimmune and inflammatory conditions and/or hematological cancers. Several high-throughput methods for the analysis of N-glycans have been developed in the past few years but thorough validation and standardization of these methods is required before significant resources are invested in large-scale studies. To this end, four of these methods were compared, UPLC, multiplexed capillary gel electrophoresis (xCGE), and two mass spectrometric (MS) methods, for quantitative profiling of N-glycosylation of plasma IgG in a subset of 1201 individuals recruited from two of the cohorts used in the previous GWAS studies. A “minimal” dataset was compiled of N-glycan structures able to be measured by all four methods. To evaluate their accuracy, correlations were calculated for each structure in the minimal dataset. Additionally, GWAS was performed to test if the same associations would be observed across methodologies. Chromatographic methods with either fluorescent or MS-detection yielded slightly stronger associations than MS-only and xCGE, but at the expense of lower levels of throughput. Advantages and disadvantages of each method were identified, which should aid in the selection of the most appropriate method for future studies. This work shows that it is possible to identify new loci that control glycosylation of plasma proteins using GWAS and the potential of N-glycans for biomarker development. It also provides some guidelines for methodology selection for future studies of N-glycans.

Les variations en nombre de copies (CNV) sont des gains ou pertes d’une séquence d’ADN et peuvent avoir un rôle dans la susceptibilité à certaines maladies. Les CNVs peuvent être détectés par les puces de SNPs de haute résolution en analysant les intensités des allèles avec des algorithmes de détection des CNVs tels que CNV partition, PennCNV et QuantiSNP. Dans cette thèse, nous avons évalué les performances de ces outils pour la détection des CNVs au niveau pangénomique et pour les tests d'association. Nous avons également étudié des stratégies d'association combinant les informations de l'allèle et du nombre de copies pour des SNP situés dans des CNV. Nous avons appliqué ces outils pour mener une étude d’association pan-génomique avec les CNV en utilisant les données de l'étude espagnole du cancer de lavessie (SBC)/EPICURO générées par la puce Illumina 1M.Nos résultats montrent une faible fiabilité et une faible sensibilité des algorithmes de détection des CNV. Dans la région du gène GSTM1 où un CNV très fréquent existe qui est associé au risque de cancer de la vessie, nous avons constaté que les algorithmes de détection des CNV ont de faibles performances. Néanmoins, l’utilisation de la mesure d'intensité des allèles dans les tests d'association peut alors être une alternative intéressante car cela nous a permis de détecter cette association connue. Pour les SNPs situés dans des CNVs, nous avons étudié plusieurs stratégies de tests d'association et nous avons montré que la plus puissante était d’utiliser un modèle avec deux termes correspondant respectivement à la somme et à la différence du nombre de copies des deux allèles. Finalement, en appliquant ces stratégies à l'étude (SBC)/EPICURO, nous avons identifié des CNVs potentiellement associés au risque de cancer de la vessie, ainsi que des SNP dont l'allèle et le nombre de copies pourraient être impliqués dans le risque de cancer de la vessie. / Copy number variations (CNVs) are losses or gains of DNA sequences that may play a role in specific disease susceptibility. CNVs can be detected by high-resolution SNP-arrays through the analysis of allele intensities with CNV calling algorithms such as CNVpartition, PennCNV and QuantiSNP. In this thesis, we identified and assessed the performances of available tools for CNV calling and for association testing, at the genome-wide level. We also investigatedassociation strategies that combine information on both the allele and the number of copies for SNPs located in CNV regions. We applied these tools to conduct a genome-wide association study with CNV using data from the Spanish Bladder Cancer (SBC)/EPICURO Study generated by the Illumina 1M SNP-array. Our results showed a low reliability and a low sensitivity of the investigated CNV calling algorithms applied to SNP-array data. The GSTM1 locus shows a very frequent CNV that is associated with bladder cancer (BC) risk. We reported that the calling algorithms performed very poorly in identifying this CNV. We proposed using allele intensity measures (LRR) as a screening step to assess association as it allowed the detection of the GSTM1 CNV association with BC. To combine the allele and the number of copies for SNPs located in CNV regions, we investigated several strategies of association testing and we showed that the more powerfulone used a two-term model with the sum and the difference of the number of copies of both alleles. Finally, by applying these strategies to the (SBC)/EPICURO Study, we identified CNV regions potentially associated with BC risk, as well as SNPs for which both the allele and the number of copies could be involved in BC risk.

Ankylosing spondylitis (AS) is a common inflammatory arthritis of the spine and other affected joints, which is highly heritable, being strongly influenced by the HLA-B27 status, as well as hundreds of mostly unknown genetic variants of smaller effect. The aim of my research was to confirm some of the previously observed genetic associations and to identify new associations, many of which are in biological pathways relevant to AS pathogenesis, most notably the IL-23/T_H17 axis (IL23R) and antigen presentation (ERAP1 and ERAP2). Studies presented in this thesis include replication and refinement of several potential associations initially identified by earlier GWAS (WTCCC-TASC, 2007 and TASC, 2010). I conducted an extended study of IL23R association with AS and undertook a meta-analysis, confirming the association between AS and IL23R (non-synonymous SNP rs11209026, p=1.5 x 10-9, OR=0.61). An extensive re-sequencing and fine mapping project, including a meta-analysis, to replicate and refine the association of TNFRSF1A with AS was also undertaken; a novel variant in intron 6 was identified and a weak association with a low frequency variant, rs4149584 (p=0.01, OR=1.58), was detected. Somewhat stronger associations were seen with rs4149577 (p=0.002, OR=0.91) and rs4149578 (p=0.015, OR=1.14) in the meta-analysis. Associations at several additional loci had been identified by a more recent GWAS (WTCCC2-TASC, 2011). I used in silico techniques, including imputation using a denser panel of variants from the 1000 Genomes Project, conditional analysis and rare/low frequency variant analysis, to refine these associations. Imputation analysis (1782 cases/5167 controls) revealed novel associations with ERAP2 (rs4869313, p=7.3 x 10-8, OR=0.79) and several additional candidate loci including IL6R, UBE2L3 and 2p16.3. Ten SNPs were then directly typed in an independent sample (1804 cases/1848 controls) to replicate selected associations and to determine the imputation accuracy. I established that imputation using the 1000 Genomes Project pilot data was largely reliable, specifically for common variants (genotype concordence~97%). However, more accurate imputation of low frequency variants may require larger reference populations, like the most recent 1000 Genomes reference panels. The results of my research provide a better understanding of the complex genetics of AS, and help identify future targets for genetic and functional studies.