Search

The connection between genotype and phenotype is not straightforward due to the multiple mechanisms through which DNA mutations can affect phenotypic manifestations and the numerous ways polymorphisms can interact to produce a final outcome. This is particularly true for the immune system which is characterized by a complex interplay between multiple cell subsets. For this project we adopted a knowledge driven approach for the identification of association and epistasis events in the context of immune cells and immune diseases. The aim of our work was two-fold: to characterize the influence of genetic variants on transcriptional regulation in an immune cell subset; and to study the role played by association and epistasis in a complex immune disease characterized by the interaction of multiple cell types. To evaluate the impact of polymorphisms on expression in an immune cell subset we focused on neutrophils and performed a genome-wide expression Quantitative Trait Loci (eQTL) study. We found significant eQTLs for 832 distinct HUGO genes with a strong overlap both with published whole blood eQTLs and with GWAS signals. To study the role played by association and epistasis on a complex immune phenotype we used a published allergic rhinitis (AR) cohort of individuals of Chinese ethnicity and limited our search to a gene known for its role in immune suppression by T regulatory cells (Treg), CD39. Differences in CD39 expression on Treg cells have already been described for multiple immune diseases and CD39 polymorphisms have been associated with various complex phenotypes. In our study we showed that a CD39 polymorphism (rs7071836) was associated with the surface expression of this molecule on Treg cells but not on other CD39-expressing leukocyte subsets. Together with another polymorphism in the promoter region of FAM134B (rs257174), SNP rs7071836 was found to impact susceptibility to AR. Polymorphism rs257174 was shown to affect expression of its cis gene in monocytes but notably not in Treg cells. We were also able to show that the expression of CD39 and FAM134B was inversely correlated in whole blood implying that the genetic interaction had an effect in vivo. In addition we were able to suggest a mechanism for the interaction of CD39 in Treg cells and FAM134B in monocytes by showing that extracellular ATP levels, which inversely correlate with CD39 expression, regulated FAM134B expression in monocytes. Throughout this thesis we did not limit our analysis to the identification of statistical association or epistasis but, whenever possible, we tried to characterize the biological impact of statistical findings. To do this we developed a tool for the identification of potential functional SNPs from disease/trait-associated polymorphisms. This novel tool (ArchiLD) has the capability of overlapping linkage disequilibrium plots with the biological knowledge provided by the UCSC Genome Browser and can be used to try and bridge the gap between statistical association and biological mechanisms. In this thesis we propose a biologically-oriented statistical approach for the identification of association and epistasis in immune cells and apply it both in the context of gene regulation and in the context of complex diseases. Due to the large number of tests performed for both analyses and the relatively small size of our cohorts, whenever possible, we tried to validate our statistical results using replication cohorts, published studies or additional biological experiments to reduce the possibility of false positives.