From noncoding variant to phenotype via SORT1 at the 1p13 cholesterol locus

Abstract

Recent genome-wide association studies (GWASs) have identified a locus on chromosome 1p13 strongly associated with both plasma low-density lipoprotein cholesterol (LDL-C) and myocardial infarction (MI) in humans. Here we show through a series of studies in human cohorts and human-derived hepatocytes that a common noncoding polymorphism at the 1p13 locus, rs12740374, creates a C/EBP (CCAAT/enhancer binding protein) transcription factor binding site and alters the hepatic expression of the SORT1 gene. With small interfering RNA (siRNA) knockdown and viral overexpression in mouse liver, we demonstrate that Sort1 alters plasma LDL-C and very low-density lipoprotein (VLDL) particle levels by modulating hepatic VLDL secretion. Thus, we provide functional evidence for a novel regulatory pathway for lipoprotein metabolism and suggest that modulation of this pathway may alter risk for MI in humans. We also demonstrate that common noncoding DNA variants identified by GWASs can directly contribute to clinical phenotypes.

Figure 1. The human chromosome 1p13 locus is preferentially associated with very small LDL and liver gene expression

a, Mean serum lipid and lipoprotein particle levels in homozygotes for the minor haplotype of the 1p13 locus (minor allele of rs646776) vs. homozygotes for the major haplotype (major allele of rs646776), normalized to the mean level in minor haplotype homozygotes, in the MDC-CC cohort (measured by ion mobility) and the PARC cohort (measured by gradient gel electrophoresis). LDL-L = large LDL; LDL-M = medium LDL; LDL-S = small LDL; LDL-VS = very small LDL. b, Relative gene positions in and around the 1p13 locus; * indicates position of rs646776. c, Mean expression of local genes in homozygotes for the major 1p13 haplotype (major allele of rs646776) vs. heterozygotes vs. homozygotes for the minor 1p13 haplotype (minor allele of rs646776), normalized to the mean level in major haplotype homozygotes, in samples of human liver, human subcutaneous adipose, and human omental adipose. d, Mean expression of PSRC1, CELSR2, SORT1, and TCF7L2 (negative control) mRNA, standardized to B2M expression, and sortilin protein, standardized to α-tubulin, in samples of human liver from homozygotes for the major 1p13 haplotype (major allele of rs12740374) vs. heterozygotes vs. homozygotes for the minor 1p13 haplotype (minor allele of rs12740374) if available, normalized to the mean level in major haplotype homozygotes. P values derived from linear regression analyses or unpaired t test. Error bars show s.e.m.

Figure 2. rs12740374 is responsible for haplotype-specific difference in transcriptional activity

a, Map of 1p13 SNPs genotyped in ~20,000 individuals of European descent relative to CELSR2 and PSRC1 genes. The six SNPs with strongest association with LDL-C (indicated with boxes), comprising a single haplotype, define the 6.1 kb region between the stop codons of the two genes. b, Firefly luciferase expression from constructs transfected into Hep3B human hepatoma cells. Both the major (darker colors) and minor (lighter colors) haplotypes of the 6.1 kb region were subcloned in forward and reverse orientations into a basal firefly luciferase construct with the SV40 promoter. Shown are ratios of firefly luciferase expression to Renilla luciferase expression (expressed from cotransfected plasmid), measured 48 hours after transfection, normalized to the mean ratio from the major haplotype, forward orientation construct. Error bars show s.e.m., N = 2. c, Both the major and minor haplotypes of a minimal 2.1 kb region were subcloned into the basal construct. Single nucleotide alterations were introduced individually into the minor haplotype, changing minor alleles of SNPs into major alleles. Shown are ratios of firefly luciferase expression to Renilla luciferase expression normalized to the mean ratio from the major haplotype construct. Error bars show s.e.m., N = 4.

Figure 3. rs12740374 alters a C/EBP transcription factor binding site

a, The human DNA sequence surrounding rs12740374, major and minor alleles, and orthologous DNA sequence in mouse. The major allele of rs12740374 disrupts one of two core elements (position 2-3, 8-9) in the predicted consensus binding site on which a C/EBP dimer binds. b, Electrophoretic mobility shift assays (EMSA) with labeled probes matching the C/EBP consensus binding site, the rs12740374 minor allele (T) sequence, and the rs12740374 major allele (G) sequence. Competition assays were performed with 100-fold excess of cold probe. Either of two C/EBPα antibodies was used to compete for binding and/or shift the protein-DNA complex. c, Relative firefly luciferase expression from constructs with haplotypes of 2.1 kb region transfected into Hep3B cells. Single nucleotide alterations were introduced into constructs as indicated, altering rs12740374 and the three other core recognition nucleotides in the predicted C/EBP binding site. d, e, Relative firefly luciferase expression from constructs with haplotypes of 6.1 kb region transfected into (d) Hep3B human hepatoma cells with or without concomitant transduction with A-C/EBP (dominant negative C/EBP) cDNA via lentivirus and (e) NIH 3T3 fibroblasts with or without concomitant transduction with C/EBP-α cDNA via lentivirus. f, Relative SORT1 expression, determined as a ratio with B2M expression by qRT-PCR, in Hep3B cells [homozygous major (GG) at rs12740374] or SK-HEP-1 human hepatoma cells [heterozygous (GT) at rs12740374] with or without concomitant transduction with A-C/EBP cDNA via lentivirus. Error bars show s.e.m., N = 3 for each experiment.

Figure 4. Overexpression or knockdown of Sort1 in mouse liver alters plasma lipids and lipoproteins

Adeno-associated virus 8 (AAV8) vectors either containing no gene, murine Sort1 cDNA, or murine Psrc1 cDNA were administered via intraperitoneal injection; phosphate-buffered saline or siRNA duplex targeting firefly luciferase or mouse Sort1 and prepared in lipidoid formulation was administered weekly via tail vein injection at 2.0 mg/kg. Plasma samples were collected prior to injection and at various time points after injection, and were subjected: individually to analytical chemistry (Mira autoanalyzer) to measure total cholesterol (a, e, g); as pooled samples to FPLC (d, f, h), from which fractions 10 to 26 were used to calculate LDL-C levels (a, e, g); individually to NMR to measure LDL particle concentrations (b). P values calculated with unpaired t test, shown if P < 0.05. Error bars show s.e.m. a-d, Apobec1−/−; APOB Tg mice (five mice per group). b, NMR measurements at six weeks. LDL-L = large LDL; LDL-M/S = medium small LDL; LDL-VS = very small LDL. c, The mice were intraperitoneally injected with Pluronic F-127 detergent to block VLDL triglyceride lipolysis and permit assessment of the rate of VLDL secretion. Plasma samples were collected at baseline, one hour, two hours, and four hours after injection. VLDL particle concentrations were measured from pooled samples with NMR. e, f, Apobec1−/−; APOB Tg mice (five mice per group). g, h, Ldlr−/− mice (five mice per group).