Trans-ancestry Fine Mapping and Molecular Assays Identify Regulatory Variants at the ANGPTL8 HDL-C GWAS Locus

Maren E. Cannon, Qing Duan, Ying Wu, Monica Zeynalzadeh, Zheng Xu, Antti J. Kangas, Pasi Soininen, Mika Ala‐Korpela, Mete Civelek, Aldons J. Lusis, Johanna Kuusisto, Francis S. Collins, Michael Boehnke, Hua Tang, Markku Laakso, Yun Li, Karen L. MohlkeG3 2017 · DOI

TL;DR

This study pinpoints regulatory variants near the ANGPTL8 gene that influence high-density lipoprotein cholesterol (HDL-C) levels. By combining fine-mapping techniques with molecular assays, the researchers identified specific genetic variants that alter gene expression in liver and adipose tissue, shedding light on the complex genetic underpinnings of HDL-C regulation.

What Problem Does This Paper Address?

Genome-wide association studies (GWAS) have identified numerous genetic loci associated with HDL-C, a crucial marker for cardiovascular health. However, pinpointing the exact causal variants and their mechanisms of action within these loci remains a significant challenge. This is particularly true for loci like the one near the ANGPTL8 gene, where multiple GWAS have identified associated variants across different ancestral populations. The problem is to move beyond statistical association to understand the functional genetic basis of HDL-C variation, specifically by identifying regulatory variants that explain these associations and their tissue-specific effects.

Methods

The researchers employed a multi-pronged approach:

  • Trans-ancestry Fine-Mapping: They used the MANTRA algorithm to perform meta-analyses of GWAS data from European (METSIM) and African American (WHI) populations. This method accounts for shared ancestry and potential differences in effect sizes across populations to narrow down candidate variants.
  • Prioritization Algorithms: CAVIAR and PAINTOR were used to further prioritize candidate variants by integrating association statistics with functional genomic annotation data (e.g., histone marks, transcription factor binding sites from ENCODE and Roadmap Epigenomics projects).
  • Reporter Assays: Dual luciferase assays were conducted in various human and mouse cell lines (HepG2, Huh-7, SGBS, SW872, 3T3L1, MIN6, 293T) to test the regulatory activity of candidate variants and their surrounding regions. This allowed them to assess tissue-specific effects on gene expression.
  • Electrophoretic Mobility Shift Assays (EMSAs): EMSAs were used to investigate allele-specific protein binding to DNA regions containing candidate variants, providing evidence for differential transcription factor interactions.
  • Expression Quantitative Trait Loci (eQTL) Analysis: They analyzed RNA expression data from subcutaneous adipose tissue to identify variants associated with the expression levels of nearby genes, specifically ANGPTL8 and DOCK6.

Datasets

  • METSIM Study: A cohort of 10,197 men from Eastern Finland, used for lipid measurements and genotyping.
  • Women's Health Initiative (WHI) SNP Health Association Resource (SHARe) study: A cohort of 8421 self-identified African American participants, used for lipid measurements and genotyping.
  • ENCODE and Roadmap Epigenomics Projects: Publicly available datasets providing functional genomic annotations (histone marks, transcription factor binding, DNase hypersensitivity sites) across various cell types.
  • 1000 Genomes Project: Used for LD calculations and imputation.

Key Insights and Ideas

  • Tissue-Specific Regulation: The ANGPTL8 gene exhibits tissue-specific expression, primarily in liver and adipose tissue. A 400-bp promoter region appears to be a key driver of this liver specificity, while enhancer regions within 5 kb upstream contribute to regulation in both liver and adipose.
  • Shared Locus, Complex Regulation: The ANGPTL8 locus associated with HDL-C shows evidence of shared regulatory mechanisms across different ancestries, but the genetic architecture is complex, potentially involving multiple functional variants.
  • ANGPTL8 as the Target Gene: While both ANGPTL8 and its neighboring gene DOCK6 showed eQTL associations, only ANGPTL8 expression levels were directly associated with HDL-C levels, strongly implicating ANGPTL8 as the primary target gene.
  • Functional Variants Identified: Specific variants, particularly rs12463177, demonstrated allelic differences in transcriptional activity and protein binding, suggesting a direct role in regulating ANGPTL8 expression.
  • Haplotype Effects: The study highlights that the combined effect of multiple variants within a haplotype can contribute significantly to the overall association signal, even if individual variants show modest effects.

Key Results

  • A 400-bp promoter region increased transcriptional activity fivefold in HepG2 (liver) cells, while a 5-kb region showed enhancer activity in liver and adipose tissues.
  • Fine-mapping and functional assays identified 13 candidate variants.
  • Variant rs12463177 showed a significant 1.5-fold increase in transcriptional activity (P = 0.004) and differential protein binding, with the HDL-C-increasing allele (G) showing enhanced activity.
  • Six additional variants (rs17699089, rs200788077, rs56322906, rs3760782, rs737337, and rs3745683) also showed evidence of allelic differences in transcriptional activity and/or protein binding.
  • eQTL analysis revealed that variants associated with HDL-C were also associated with ANGPTL8 and DOCK6 mRNA levels in adipose tissue, but only ANGPTL8 expression was linked to HDL-C.

Limitations and Future Work

  • In Vitro vs. In Vivo: The functional effects observed in cell lines, particularly the modest effect of rs12463177, may not fully recapitulate the physiological impact in vivo.
  • Complex Interactions: The study suggests a complex regulatory mechanism involving multiple variants and potentially enhancer-repressor interactions. Elucidating these precise roles and interactions will require further investigation, including in vivo studies.
  • Transcription Factor Identification: While some protein binding was observed, the specific transcription factors involved in allele-specific binding for variants like rs737337 remain to be identified.
  • DOCK6 Role: While ANGPTL8 is the primary suspect, the possibility of a role for DOCK6 cannot be entirely ruled out, though evidence points more strongly to ANGPTL8.

Future work should focus on in vivo validation of the identified functional variants and their regulatory mechanisms, as well as a deeper understanding of the transcription factors that mediate these effects. Further investigation into the complex interplay of multiple variants at this locus is also warranted.

Summary updated

Abstract

Recent genome-wide association studies (GWAS) have identified variants associated with high-density lipoprotein cholesterol (HDL-C) located in or near the <i>ANGPTL8</i> gene. Given the extensive sharing of GWAS loci across populations, we hypothesized that at least one shared variant at this locus affects HDL-C. The HDL-C-associated variants are coincident with expression quantitative trait loci for <i>ANGPTL8</i> and <i>DOCK6</i> in subcutaneous adipose tissue; however, only <i>ANGPTL8</i> expression levels are associated with HDL-C levels. We identified a 400-bp promoter region of <i>ANGPTL8</i> and enhancer regions within 5 kb that contribute to regulating expression in liver and adipose. To identify variants functionally responsible for the HDL-C association, we performed fine-mapping analyses and selected 13 candidate variants that overlap putative regulatory regions to test for allelic differences in regulatory function. Of these variants, rs12463177-G increased transcriptional activity (1.5-fold, <i>P</i> = 0.004) and showed differential protein binding. Six additional variants (rs17699089, rs200788077, rs56322906, rs3760782, rs737337, and rs3745683) showed evidence of allelic differences in transcriptional activity and/or protein binding. Taken together, these data suggest a regulatory mechanism at the <i>ANGPTL8</i> HDL-C GWAS locus involving tissue-selective expression and at least one functional variant.
BibTeX
@inproceedings{cannon2017i,
  author = {Maren E. Cannon and Qing Duan and Ying Wu and Monica Zeynalzadeh and Zheng Xu and Antti J. Kangas and Pasi Soininen and Mika Ala‐Korpela and Mete Civelek and Aldons J. Lusis and Johanna Kuusisto and Francis S. Collins and Michael Boehnke and Hua Tang and Markku Laakso and Yun Li and Karen L. Mohlke},
  title = {{<i>Trans</i>-ancestry Fine Mapping and Molecular Assays Identify Regulatory Variants at the <i>ANGPTL8</i> HDL-C GWAS Locus}},
  booktitle = {G3},
  year = {2017},
  doi = {10.1534/g3.117.300088},
}
Figure