Genetic Architecture of Left Ventricular Hypertrophy
Biology and Biomedical Sciences: Human and Statistical Genetics
Date of Award
Doctor of Philosophy (PhD)
Chair and Committee
C. Charles Gu
LVH is a strong predictor of cardiovascular morbidity and mortality, with a significant genetic component, and as such, poses a major medical and financial challenge to the general community. Therefore understanding the genetic basis of LVH is important to developing better diagnosis, treatment and prevention of the disease. Traditional candidate gene studies and more recent genome wide association studies have identified genetic loci spanning multiple chromosomes associated with LVH, providing some early insight into the remarkably complex genetics of LVH. However much remains to be known about the functional roles of these loci and the relationships amongst themselves, as well as with their environmental counterparts, that determine one's risk of developing the disease and are essential for making effective prediction and prevention strategies. In the present project, I set out to systematically study the genetic architecture of LVH by investigating the different layers of complexity, from single-variant effects to pairwise GxG/GxE interactions, to the roles of biological pathways and clinical phenotyping modalities to gain deeper understanding. The idea stems off a working model that highlights previous work, particularly candidate genes and interplay between metabolism and inflammation in the development of LVH. Our work takes advantage of a recently completed GWAS study of LVH for discovery, and existing resource of a previous study of LVH, with an older GWAS platform, for replication and validation.
At the single-variants level, we took the GWAS approach and investigated, beyond conventional analysis, the variable genetic associations to LVH status defined by two different echocardiographic imaging modalities for measuring LV mass. We identified both shared genetic factors (CDH13) as well as the unshared ones (MYOM1, MYOCD), reflecting that phenotype measurement is an important factor in overall LVH complexity. This is a notable progress in improved understanding of the remarkable architecture of this disease. At the layer of pairwise interactions, I continue to use GWAS approach for studying gene-environment and gene-mitochondrial interactions but chose an extended candidate genes approach for studying chromosomal/nuclear GxG interactions to curtail false positives. The analysis of interactions among variants in 347 candidate genes identified a large number of potentially relevant pairwise interactions including a statistically significant one between MEF2B and IL1RAP, two key genes in hypertrophic signaling pathways and IL1 receptor signaling respectively. The genome-wide examination of interactions between nuclear and mitochondrial genomes is a relatively unexplored area and a novel part of this project; we found many interesting trans-nuclear mitochondrial interactions including on between TMOD3 and MT-COX1 significant at the genome-wide level (p=1.68E-09). Furthermore, genome-wide analyses of gene-environment interaction were carried out for hypertensive medications and serum triglyceride levels. The gene-medication analyses identified 21 significant interactions at p≤10^-6 including suggestive evidence of interaction between Angiotensin receptor blockers and bile acid transporter SLC10A4. The gene-triglyceride analyses identified 4 significant interactions at p≤10^-5, many corroborated the energy metabolism hypothesis of LVH, including one between triglyceride and PGM5, a key enzyme in cellular glucose utilization and energy homeostasis. Finally, at the layer of concerted effects of many factors, an extensive search for relevant biological pathways important to LVH was performed by extensive examination of 1410 known biological pathways compiled from existing databases for significant enrichment of association signal hits. This confirmed many important gene-sets or pathways that are known to be associated with LVH (e.g., "regulation of fat cell differentiation" and "basic mechanism of PPAR on gene expression"), and also revealed multiple novel gene-sets that were not associated with LVH before but corroborate well with our working-model (e.g.,"IL13 biosynthesis" and "succinate dehydrogenase activity").
Taken together, the present research goes beyond single-variant analysis and attempts to gain valuable insights into the complex genetic architecture of LVH by systematically studying collective multi-variable effects of gene-gene, gene-environment interactions, and of organic gene-sets in biological pathways, on LVH phenotypes in real GWAS studies. The analyses of the different layers rediscovered some of what we already know; but more importantly, they also led us to exciting new findings about LVH genetics such as the potential of trans-nuclear mitochondrial interactions. The results painted a complex yet seemingly organized architecture of the LVH genetics that has many components waiting for us to further characterize their functional importance to the development and modulation of LVH. Due to the modest sample sizes of the real studies, the lower density of genotyping platform in the replication sample, and due to the lack of independent studies with ECHO phenotypes in general, further replication of the findings are in order. Nonetheless, our work demonstrates the importance of analyzing interaction effects in order to illustrate the complex architecture of LVH. The findings provide a rich collection of new hypotheses for further study of functional roles of these collective effects, and for continual development of improved prevention and intervention strategies to cure and manage the disease.
Barve, Ruteja A., "Genetic Architecture of Left Ventricular Hypertrophy" (2014). All Theses and Dissertations (ETDs). 1284.