Butyrophilin family genes as putative coeliac disease risk loci

Abstract

Coeliac disease (CeD) is a T-cell mediated autoimmune enteropathy (1). CeD is triggered by the consumption of gluten, the collective name of a group of proteins present in wheat, rye, and barley (1, 2). The prevalence of CeD in the United Kingdom (UK) is around 1% and the number of new CeD diagnoses is rising each year (3, 4). Active CeD, when individuals predisposed to CeD consume dietary gluten, is characterised by an increase in intraepithelial lymphocytes (IELs), and mucosal damage, leading to symptoms such as villous atrophy, abdominal pain, malabsorption and malnutrition (5). Currently, the only treatment for CeD is following a gluten-free diet (6). The genetic risk for CeD is not fully understood, as around 50% of genetic predisposition remains unexplained (7). The HLA loci are estimated to explain around 30% of total genetic risk, specifically the following CeD associated genotypes: HLA-DQ2.5, HLA-DQ8 and HLA-DQ2.2 (8, 9). Recently, the butyrophilin family of immunomodulators have been proposed as putative non-HLA CeD risk loci (10-12). These genes are involved in regulating innate and adaptive immune cells, and in maintaining γδ T-cell subsets (13). Mayassi, et al. (12) have connected the temporary loss of the duodenal butyrophilin heterodimers BTNL3/BTNL8 to the permanent reconfiguration of the local γδ T-cell population in CeD, thus highlighting the importance of butyrophilins in CeD risk. This thesis aimed to explore the interaction of γδ T cells and butyrophilin genes in the context of CeD, in order to identify germline variants in butyrophilin genes and the γδ T cell receptor (TCR) that may confer increased CeD risk. This project confirmed that the risk associated HLA loci were not sufficient to explain CeD risk in the UK Biobank, an open-access database containing the genome-wide genotyping data of 3094 CeD and 29 762 control participants (14). These results underline the need for examining non-HLA risk loci. First, the interaction between butyrophilins and γδ T cells were examined from the γδ T cell side. The germline-encoded HV4 loop of the duodenal Vγ4+ γδ TCRs were examined, that can directly bind the BTNL3 proteins (15, 16). However, no significant differences were found between the HV4 amino acid sequences of 141 CeD and 238 normal samples. Next, the variation in butyrophilin genes were examined in 46 normal and 48 CeD samples. I successfully designed a hybridization probe panel to target butyrophilin genes likely involved in CeD due to their expression patterns in the small intestine and in immune 3 cells, as described by the Human Protein Atlas, a study aiming to characterise the human proteome (17). The samples were successfully sequenced and subjected to gene based burden testing, which quantified the amount of non-synonymous coding variants that were likely to be pathogenic. BTN2A1 had significantly higher gene burden in CeD compared to normal samples. Analysis of the butyrophilin variants were also carried out in the UK Biobank. Single variant testing identified BTN3A1 and BTN3A2 SNPs as significantly associated with CeD in HLA-DQ2.5-matched CeD (n = 1652) and control (n = 6416) participants from the UK Biobank. In summary, this project has shown that the butyrophilin family of genes are promising yet not fully explored immunoregulators connecting innate and adaptive immunity. This thesis provided further evidence for butyrophilins as potential CeD risk loci and identified BTN2A1, BTN3A1, and BTN3A2 SNPs significantly associated with CeD risk. Due to their role in modulating the activity of innate and adaptive immune cells, butyrophilins may be involved in other inflammatory, autoimmune or infectious diseases. However, further studies must be carried out to understand the full extent of the function and role of butyrophilins in CeD risk.