CHARACTERIZATION OF THE ONTOGENY AND INTER-INDIVIDUAL VARIATION OF GENES IN CHOLESTEROL AND STATIN PATHWAYS

Abstract

Introduction: Previous studies have shown that gene expression can change throughout development,1 and therefore genotype-phenotype associations found in adults might not be observed in children of all ages. The purpose of this study was to 1) characterize the ontogeny of 30 genes in pathways related to cholesterol synthesis and/or statin action or toxicity in pediatric liver samples, and 2) assess the in vitro and in vivo consequences in children of genetic variation in 3-hydroxy-3-methyl-glutaryl-Coenzyme A reductase (HMGCR) and heterogeneous nuclear ribonucleoprotein A1 (HNRNPA1), two genes implicated in altered cholesterol levels and/or statin response.2,3 Methods: RNA and DNA were isolated from pediatric liver samples (n=62), and DNA was isolated from patients in the Cardiology Pharmacogenomic Repository (CPR) (n=195). For Aim 1, the ontogeny of mRNA expression from 30 genes related to cholesterol and statin metabolism as measured by RNA-seq in the liver samples was assessed. For Aim 2, all postnatal samples were genotyped for rs1920045 (HNRNPA1) and rs3846662 (HMGCR), and genotypes were tested for association with either alternative splicing in vitro (liver samples) or plasma lipid levels in vivo (CPR). Statistical analyses on liver sample data were conducted with Kruskal-Wallis or Wilcoxon tests with Bonferroni correction. Analysis of CPR samples was completed with ANOVA, Kruskal-Wallis or Wilcoxon tests with Bonferroni correction. Samples were stratified by race and analyses were repeated. Liver sample use was deemed non-human subjects research and the CPR was approved by the Pediatric IRB. Results: Analysis of postnatal liver samples revealed age-related changes in Ensembl-based total and primary mRNA transcript expression of ABCB1 (p<0.0008). Inclusion of prenatal samples revealed 13 additional genes with age-related expression changes during development in either the Ensembl or UCSC based data. The ratios of alternative to canonical transcripts of HMGCR trended towards significance in the HMGCR and HNRNPA1 dominant genotype models (p=0.0465, 0.0470 respectively). CPR analysis suggested a relationship between HMGCR genotype and low-density lipoprotein cholesterol (LDL-C) or total cholesterol (TC) in African Americans, or HNRNPA1 genotype and TC in Caucasians, but these relationships did not achieve statistical significance. Conclusion: Although trends in age-related changes in gene expression, and genotype-phenotype associations, were observed for several genes of interest, the number of statistically significant associations was limited by use of a stringent criterion for multiple testing as well as intra-group variability and relatively small sample sizes given the amount of variability observed. Until these issues are resolved in a larger number of samples, it is premature to conclude genotype-phenotype associations observed in adults will also be present in children at all ages.