Linkage analysis of hypertriglyceridemia in a single large family identifies 3 novel potentially pathogenic variants. E. A. Rosenthal1, J. Ranchalis1, J. D. Brunzell2, A. G. Motulsky3, D. A. Nickerson3,4, E. M. Wijsman1,5, G. P. Jarvik1,3 1) Dept Med Gen, Univ Washington School of Medicine, Seattle, WA; 2) Div. of Metabolism, Endocrinology, and Nutrition, Univ Washington School of Medicine, Seattle, WA; 3) Dept Genome Sci, Univ Washington, Seattle, WA; 4) Dept of Bioengineering, Univ Washington, Seattle, WA; 5) Dept of Biostatistics, Univ Washington, Seattle, WA.
Hypertriglyceridemia (HTG) is a risk factor for cardiovascular disease, a leading cause of death in the U. S. There are few pharmacological treatments, with non-compliance rates >5% due to side effects. Investigating the genetics of HTG may lead to new drug targets. There are ~35 known SNPs that explain only ~10% of variation in triglycerides (TG). Due to the genetic heterogeneity of HTG, a family study design is optimal for identification of novel loci with large effect size as the same mutation can be seen in many relatives and co-segregation with TG observed. We considered HTG in a large family of European American descent (N=121, 85 with TG, 82 with marker genotypes on ~5cM map), ascertained for familial combined hyperlipidemia, a diagnosis defined by variable atherogenic lipoprotein levels in multiple relatives. This family has 4 relatives with HTG > top ½% of U.S. adults, suggesting a strong genetic component. We log transformed TG after adjusting for sex and age using population data (log-adjTG). Known pathogenic SNPs do not explain log-adjTG in this family. Bayesian segregation analysis supports 1 quantitative trait locus (QTL) (posterior probability = 0.71). We analyzed the data with Bayesian Markov chain Monte Carlo joint oligogenic linkage and association using Loki2.4.7, as it does not need a fixed trait model, handles large pedigrees, and imputes missing genotypes. To evaluate linkage evidence, we calculated the Bayes factor (BF) for 2 cM intervals, which compares the posterior and prior odds that a QTL is located in a region. We detected linkage to chrs. 7, 14, 17 and 18 (maxBF=25, 44, 183 and 29, respectively). Whole exome data on 16 individuals revealed 3 highly conserved novel missense single nucleotide variants (SNVs) at SLC25A40 on chr. 7, PRKCH on chr.14, and PLD2 on chr. 17. Jointly, these SNVs explained 49% of the genetic variance (Vg) in log-adjTG; only the SLC25A40 SNV was significantly associated with log-adjTG (p=0.0005), adjusting for pedigree structure, and explained 28% Vg. The SLC25A40 protein is involved in inner mitochondrial membrane transport. The SNV causes a TYR125CYS substitution just outside the 2nd helical transmembrane region. Cysteines stabilize folded proteins through disulfide bonds; an extra cysteine may allow for a protein configuration change that disrupts protein function. Whole gene testing in subjects from the Exome Sequencing Project confirmed the association between TG and SLC25A40 (p=0.03).
You may contact the first author (during and after the meeting) at