Genetic analysis of the anticoagulant proteins Antithrombin, Protein C and Protein S

Abstract

Cardiovascular diseases (CVDs) are the number one case of dead globally according to the World Health Organisation (WHO) ahead of cancer and respiratory diseases. Most of CVDs are caused by a combination of genetic, environmental and lifestyle factors and therefore are considered complex diseases. Alterations in the coagulation cascade are strongly related to these diseases and can cause severe problems in the circulatory system that might lead to death, through the formation of thrombus that prevent blood from flowing. We focused this study on the identification of genetic markers in general population that are associated with some proteins of the coagulation cascade. To do this, in association with the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium, we meta-analysed summary statistics from multiple Genome Wide Association Studies (GWAS). GWAS test for association in millions of genetic positions against a phenotypic trait in thousands of individuals. In our case, three of the most important anticoagulant blood proteins were measured, Antithrombin (AT), Protein C (PC) and Protein S (PS). These three proteins have been studied before in smaller samples sizes. Our meta-analysis involved more participants from different ancestries, increasing the power from previous studies. AT was analysed in 27,783 individuals (25,095 from European (EU) ancestry and 2,688 from African-American (AA) ancestry), PC in 19,285 individuals (16,597 from EU and 2,688 from AA) and PS was divided in two smaller phenotypes, PST and PSF, analysed in 6,257 and 4,006 individuals, respectively. Multiple independent loci that have been identified previously were detected in our meta-analyses, including the coding genes for our three proteins, SERPINC1 (AT), PROC (PC) and PROS1 (PS), but we identified some novel associations, too. For AT, a genomic region containing SNX17 and GCKR genes in EU population (best SNP rs4665972, trans-ethnic p-value = 4.41·10-16, EU p-value = 1.26·10-16), BAZ1B also in EU population (best SNP rs13244268, trans-ethnic p-value = 4.006·10-9, EU p-value = 6.00·10-9) and another genomic region containing TXNL4B and HP in AA population (best SNP rs5471, trans-ethnic p-value = 4.374·10-26, AA p-value = 7.76·10-25) were novel associations; for PC, most of the identified variants were already known or in high Linkage Disequilibrium with known variants, but we identified a novel variant in EU population, rs150070344, at SPG11 gene with a p-value of 4.07·10-8. For PS, we found a novel association at ORM1 gene for both phenotypes, PST and PSF (best SNP rs150611042, PST p-value = 1.03·10-15, PSF p-value = 1.16·10-19). In conclusion, the use of larger sample sizes has allowed the detection of new associated variants. Furthermore, we have been able to observe the already known differences that exist between the linkage disequilibrium blocks of two different populations, in our case EU and AA. The application of conditional analysis on these results will allow to know more details about the causal variants. This study has been done in the Genomics of Complex Diseases Group, led by José Manuel Soria, within the Research Institute of Hospital Sant Pau, and using international genetic data from the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium.