Long Noncoding RNA TRIBAL Links the 8q24.13 Locus to Hepatic Lipid Metabolism and Coronary Artery Disease

Abstract

Background: Genome-wide association studies identified a 20-Kb region of chromosome 8 (8q24.13) associated with plasma lipids, hepatic steatosis, and risk for coronary artery disease. The region is proximal to TRIB1, and given its well-established role in lipid regulation in animal models, TRIB1 has been proposed to mediate the contribution of the 8q24.13 locus to these traits. This region overlaps a gene encoding the primate-specific long noncoding RNA transcript TRIBAL/TRIB1AL (TRIB1-associated locus), but the contribution of TRIBAL to coronary artery disease risk remains untested.

Methods: Using recently available expression quantitative trait loci data and hepatocyte models, we further investigated this locus by Mendelian randomization analysis. Following antisense oligonucleotide targeting of TRIBAL, transcription array, quantitative reverse transcription polymerase chain reaction, and enrichment analyses were performed and effects on apoB and triglyceride secretion were determined.

Results: Mendelian randomization analysis supports a causal relationship between genetically determined hepatic TRIBAL expression and markers of hepatic steatosis and coronary artery disease risk. By contrast, expression data sets did not support expression quantitative trait loci relationships between coronary artery disease-associated variants and TRIB1. TRIBAL suppression reduced the expression of key regulators of triglyceride metabolism and bile acid synthesis. Enrichment analyses identified patterns consistent with impaired metabolic functions, including reduced triglyceride and cholesterol handling ability. Furthermore, TRIBAL suppression was associated with reduced hepatocyte secretion of triglycerides.

Conclusions: This work identifies TRIBAL as a gene bridging the genotype-phenotype relationship at the 8q24.13 locus with effects on genes regulating hepatocyte lipid metabolism and triglyceride secretion.

Keywords: bile acids and salts; cholesterol; coronary artery disease; genotype; hepatocytes; lipids.

Figure 1.

Impact of top UK Biobank coronary artery disease (CAD) single-nucleotide polymorphism (SNP) on lipid traits and TRIBAL expression. A, rs6982502_ T allele (alternate/protective) linked to reduced CAD (odds ratio, <1), low-density lipoprotein (LDL-C), and triglyceride (TG; negative β-values). B, rs6982502_C allele (reference/risk) linked to increased TRIBAL expression. Stockholm-Tartu Atherosclerosis Reverse Network Engineering Task effect based on reference allele. C, Violin plot of liver TRIBAL expression as a function of rs6982502 genotype according to Genotype-Tissue Expression; P=0.008. adj. indicates adjusted; and LIV, liver.

Figure 2.

P-P plots of coronary artery disease (CAD) vs expression quantitative trait loci (eQTL) values for TRIBAL and TRIB1 in the liver and blood. ENSG00000253111 expression is not present in eQTLGen (whole blood), while a single significant eQTL within a 500-Kb window centered on TRIB1 was nominally (P<0.05) linked to TRIB1 expression in the liver. TRIBAL (liver), but not TRIB1 (whole blood), eQTLs are enriched in genome-wide association study (GWAS) significant CAD hits (P=5.27×10⁻⁶⁰ vs P=0.299). Data were analyzed via eQTpLot. SNP indicates single-nucleotide polymorphism.

Figure 3.

Impact of TRIBAL suppression by antisense oligonucleotides (ASOs) on local gene expression. A, Schema of ASO cognate positions relative to the TRIBAL1 transcript drawn approximately to scale. B and C, RNA expression in TRIBAL ASO-treated hepatocytes. RNA expression was measured in primary hepatocytes from 3 distinct donors targeted for 72 hours with TRIBAL ASO1, ASO2, or ASO4, as indicated. B, Two distinct donors, each treated with either ASO1 or ASO2. C, Third donor, treated with 3 distinct ASOs. Values shown were normalized internally to PPIA (ΔCt method) and to the matching control (CTL) ASO values. Error bars represent the standard deviation (STD). P values (paired t test with matching NT values): *P<0.05, **P<0.01, ***P<0.001. NT indicates non-target.

Figure 4.

Suppression of major regulators of liver metabolism by TRIBAL antisense oligonucleotides (ASOs). Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis of samples submitted to transcriptome array (A), or replication set (B), using a third donor and 3 TRIBAL cognate ASOs. Transcript abundance, expressed relative to the NT value as determined by qRT-PCR of hepatocytes treated with ASO1 or the NT ASO for 72 hours. Each point represents a biological replicate. Mean±SD is shown. Statistical significance was tested using 2-tailed Student t tests. *P<0.05, **P<0.01, ***P<0.001. NT indicates non-target.

Figure 5.

TRIBAL suppression reduces hepatocyte triglyceride (TG) secretion without affecting apoB secretion. Levels of TG (A) and apoB (B) in cell media of 72-hour antisense oligonucleotide (ASO)–treated hepatocytes (donor HU8412) were measured using a fluorescence-based lipase assay and ELISA, respectively. In C, the ratio of TG to APOB was calculated for each biological replicate and is expressed relative to the corresponding CTLASO value, indicating that TRIBAL suppression reduced lipidation of apoB lipoproteins. Each point represents a distinct biological replicate. Control apoB and TG values were 1.7 µg/mL (SD, ±0.34) and 224 ng/mL (SD, ±83), respectively. Statistical significance was tested using a 1-way Student t test (vs CTLASO), under the hypothesis that TG and apoB secretion would be inhibited. *P<0.05, **P<0.01, ***P<0.001. CTLASO indicates control antisense oligonucleotide.