Important Role And Molecular Mechanism Of Gene-gene Interaction In Coronary Artery Disease

Coronary artery disease（CAD）is the most common cardiovascular disease and the leading cause of death worldwide.CAD is caused by the formation of coronary artery wall plaque（atherosclerosis,AS）in coronary arteries that causes vascular stenosis,resulting in insufficient blood,oxygen,and nutrient supply to the heart,causing chest pain（angina）,shortness of breath,myocardial infarction（MI）,and sudden cardiac death.CAD is a complex disease that is caused by both genetic factors and environment factors as well as their interactions.In recent years,genome-wide association studies（GWAS）have become the most important genetic method for studying complex genetic diseases.GWAS is effective in discovering the susceptibility loci or susceptibility gene for complex diseases and explaining the genetic basis and pathogenic mechanisms of complex diseases,which may have significance in clinical prevention,diagnosis and treatment of diseases.GWAS have identified almost 200susceptibility gene loci for CAD,which represent breakthrough advances in genetics of CAD,thus laying a solid foundation for identify molecular mechanisms underlying the pathogenesis of CAD.Despite these important advances on genetics of CAD,the known GWAS loci cannot explain the full heritability of CAD,and much remains to be done to identify the rest of heritability of CAD.Gene-gene interaction has been proposed as one genetic mechanism that can explain some portions of heritability for common complex human diseases.Gene-gene interaction,also referred to as epistasis significantly affects phenotypes,and may have a profound impact on the development of human diseases such as CAD.The major goal of this study was to identify gene-gene interactions in CAD using an easily generalizable multi-stage approach we developed.Our easily generalizable multi-stage approach was used to overcome the difficulty of identifying gene-gene interactions in CAD.Our approach consists of multiple steps that combine statistical and functional approaches to functionally characterize GWAS variants.As proof of principle,we combine information from global gene expression profiling,functional interactions,and genetic interactions to robustly identify gene-gene interactions involved in CAD.In this thesis research project,we started with the ANRIL gene that encodes a long noncoding RNA（lncRNA）and has been studied in our laboratory for many years.In earlier studies,we performed whole genome expression profiling microarrays to identify downstream genes regulated by ANRIL after it was knocked down.We identified 8 downregulated genes,including AHNAK2,CLIP1,CXCL11,ENC1,EZR,LYVE1,WASL,and TNFSF10,and two upregulated genes,including TMEM100 and TMEM106B.Further studies showed that ANRIL regulated CAD-relevant endothelial cell functions through CLIP1,EZR,and LYVE1.However,two upregulated genes were not characterized in detail.Therefore,this project focused on ANRIL regulation of TMEM100 encoding a Transmembrane Protein 100 and TMEM106B encoding Transmembrane Protein 106B.We transfected two types of endothelial cells,including HCAECs or EA.hy926 cells with plasmids or si RNAs and used them for various assays for assessment of endothelial cell functions,including monocyte adhesion to endothelial cells,and transendothelial migration of monocytes,the two key cellular processes critical to the initiation of CAD.In order to determine whether the identified gene-gene interactions at the molecular and cellular level is relevant to gene-gene interactions at the genetic level,we performed genetic analysis using three large-scale study populations,including CADRDIo GRAMplus C4D,Gene ATLAS,and the UK Biobank.We analyzed genomic variants in 100 kb flanking regions of TMEM100 and TMEM106B,and tested whether they were associated with risk of CAD.A meta-analysis of 636,539 samples in total revealed that one variant rs11509880 located in intron 4 of TMEM106B was significantly associated with risk of CAD.No variant in TMEM100 was found to be associated with risk of CAD.Most importantly,we found significant genetic interaction between ANRIL variant rs2383207and TMEM106B variant rs3807865 with genotyping data available in 343,145 unrelated samples in UK Biobank.Similarly,we identified significant genetic interaction between two other CAD GWAS variants,genetic variant rs6903956 in ADTRP encoding Androgen-Dependent TFPI Regulatory Protein and variant rs17465637 in MIA3encoding Melanoma Inhibitory Activity Family SH3 Domain ER Export Factor 3.The main results are as follows:1.Real-time RT-PCR analysis confirmed the results from global gene expression profiling,and showed that knockdown of ANRIL at 9p21.3 GWAS CAD locus upregulated two downstream genes,TMEM100 and TMEM106B.Molecular and cellular characterization of endothelial cell functions indicated that the increased monocyte adhesion to endothelial cells（ECs）and transendothelial migration of monocytes（TEM）,two critical processes in the initiation of CAD,by ANRIL knockdown were reversed by knockdown of TMEM106B,but not of TMEM100.The data suggest that ANRIL initiates CAD-related endothelial processes by regulating the expression of TMEM106B,but not TMEM100.2.The decreased monocyte adhesion to ECs and TEM induced by ANRIL overexpression was reversed by overexpressing TMEM106B,but not TMEM100.The data further show that ANRIL initiates CAD-related endothelial processes by regulating the expression of TMEM106B,but not TMEM100.3.Countrary to ANRIL transcript DQ485454,TMEM106B expression was>2-fold higher in coronary artery tissue samples from CAD patients than from control coronary artery tissues.However,there was no significant difference on the expression level of TMEM100 between coronary artery tissue samples from CAD patients and that from normal tissues.Thus,similar to ANRIL,TMEM106B,but not TMEM100,is associated with CAD.4.We found highly significant association between genomic variants in TMEM106B（but not in TMEM100）and CAD(p=1.9×10^-8).Statistical analysis further revealed significant gene-gene interaction between ANRIL variant rs2383207 and TMEM106B rs3807865（p=0.009）.5.A similar approach involving global gene expression profiling,functional studies,and statistical analysis also identified significant interaction between rs6903956 in ADTRP and rs17465637 in MIA3（p=0.005）.In summary,our data provide new data to support the notion that ANRIL is the causal gene for CAD at the 9p21.3 GWAS locus,and TMEM106B is the causal gene at the 7q21.11 CAD locus.The causal genes for CAD,ANRIL and TMEM106B,may serve as targets for development of drugs and other therapeutic agents to prevent and treat CAD.Furthermore,we identified two pairs of epistatic gene-gene interactions between GWAS loci for CAD.These results offer important insights into the genetic architecture and molecular mechanisms for the pathogenesis of CAD.Identification of epistasis or gene-gene interactions is paramount to our understanding of the genetic architecture of a human disease,and may have an important implication on future precision diagnosis and risk stratification of common human diseases based on polygenic risk scores and other methods.Importantly,we developed a simple but robust new method to detect gene-gene interactions between GWAS loci of CAD,which involves global gene expression profiling,functional studies,and statistical analysis.Our strategy could be extended to understand other complex diseases.Altogether,these studies start to provide important mechanistic insights of the molecular mechanism of the interaction between these susceptibility genes to affect the risk of CAD.In addition,our study provides novel insights into the function of CAD susceptibility genes and biological pathways,which may be involved in the pathogenesis of CAD.Further studies may establish gene-gene interaction networks for complex diseases as important targets for developing prevention and treatment strategies for coronary artery disease.