Loss-of-Function Variants Influence the Human Metabolome. B. Yu1, AH. Li1, G. Metcalf2, DM. Muzny2, AC. Morrison1, TH. Mosley3, RA. Gibbs2, E. Boerwinkle1,2 1) Human Genetics Center, University of Texas Health Science Center at Houston, Houston, TX; 2) Human Genome Sequencing Center, Baylor College of Medicine, Houston, TX; 3) The Memory Impairment and Neurodegenerative Dementia Research Center, University of Mississippi, Jackson, MS.

   Loss-of-function (LoF) variants are more frequent in individuals of African descent and are powerful instruments for identifying disease susceptibility genes. The metabolome is a collection of small molecules resulting from a multitude of biologic processes and can act as biomarkers of disease. We sequenced and annotated the exomes and measured 308 serum metabolites in a sample of 1,361 African-Americans (AAs) from the Atherosclerosis Risk in Communities (ARIC) Study. On average, each individual had 112 LoF variants and were homozygous for 7. Single SNP tests (for MAF > 5%) and gene burden tests (for cMAF 5%) were performed for each metabolite. We identified 12 novel genes (p < 5.0E-7) harboring more than 50 LoF variants affecting the metabolome. Depending on the metabolite, these loci were associated with 19-51% of the difference in metabolite levels, with an average effect of 38%. For example, six LoF mutations in SLCO1B1 were consistently associated with high levels of hexadecanedioate (cMAF = 2.7%, p = 9.3E-10), a C16 dicarboxylic acid. SLCO1B1 is an organic ion transporter and follow-up studies showed pleiotropic effects on other dicarboxylic acids (e.g. tetradecanedioate, p = 9.0E-5). These associations were replicated in an independent sample of 616 ARIC AAs (p = 4.6E-5 and p = 1.6E-5, respectively). Reflecting an important role of fatty acid oxidation in myocardial energy metabolism, these two metabolites were significant predictors of incident heart failure beyond the traditional risk factors, whereby higher levels of hexadecanedioate and tetradecanedioate were associated with increased risk (p = 3.0E-7 and p = 4.3E-3, respectively). In a second example, a LoF mutation in CD36 (Y325X, MAF = 8.4%), which encodes a membrane-bound fatty acid transporter, was associated with reduced levels of octanoylcarnitine and decanoylcarnitine (p = 3.9E-8 and p = 2.7E-7, respectively), and the associations were replicated in an independent sample of ARIC AAs (p = 0.02 and p = 0.01, respectively). These two fatty acylcarnitines compounds are biomarkers for medium-chain acyl-CoA dehydrogenase (MCAD) deficiency (MIM: 201450), and CD36 is an important component of platelet and monocyte biology. Our findings suggest a role of CD36 in regulating acylcarnitine levels. Taken together, these results provide new insights into gene function and the understanding of disease etiology by integrating -omic technologies in a deeply phenotyped population.

You may contact the first author (during and after the meeting) at