Genetic variation shapes the chromatin accessibility landscape and transcriptional responses in mouse adipose tissue

TL;DR

This study shows how genetic differences between two mouse strains affect chromatin accessibility and gene regulation in adipose tissue, offering insights into how non-coding genetic variants influence obesity-related diseases.

Contribution

The study reveals that genetic variation directly alters chromatin accessibility and transcription factor binding in mouse adipose tissue, linking non-coding variants to gene regulation.

Findings

Chromatin accessibility differences in adipose tissue are strain-specific and driven by genetic variation.

Genetic variants at transcription factor binding sites correlate with changes in TF occupancy and chromatin accessibility.

Differentially expressed genes are located near regions of altered chromatin accessibility caused by genetic variation.

Abstract

Most of the disease associated genetic variants identified in genome wide association studies have been mapped to the non-coding regions of the genome. One of the leading mechanisms by which these variants are thought to affect disease susceptibility is by altering transcription factor (TF) binding. Even though inbred mouse strains have been commonly used to investigate polygenic diseases, less is known on how their genetic differences translate to the level of gene regulation and chromatin landscape. Here, we investigated how genetic variation affects chromatin accessibility in the epididymal white adipose tissue (eWAT) of C57BL/6J and 129S1/SvImJ mice, which are commonly used to study diet-induced obesity, fed either chow or high-fat diet. We show that differences in chromatin accessibility are almost exclusively strain-specific and driven by genetic variation. In addition, we integrate ATAC-seq (chromatin accessibility) and H3K27ac ChIP-seq (active regulatory regions) data to show that tissue-specific TF binding sites are commonly found in the active regulatory regions hosting TF motif altering variants in eWAT. Using footprint analysis, we also show that TF occupancy is consistent with TF binding motif scores at the genetically altered loci. In addition, we validate these findings by extending the analysis to ATAC-seq and H3K27ac ChIP-seq data obtained from the liver. We employ RNA-seq to show that differentially expressed genes are co-located with differentially accessible regions hosting genetic variants. Overall, our findings highlight the connection between differential chromatin accessibility and genetic variation across metabolically central tissues of a mouse model for polygenic obesity.

Most of the variants associated with polygenic diseases are outside the protein coding regions of the genome. In this study, we took benefit of the known genetic differences of two inbred mouse strains that are commonly used to study polygenic obesity to investigate the impact of the non-coding genetic variation on the chromatin landscape and the regulation of gene expression in the adipose tissue. Using several complementary sequencing and bioinformatic methods, we show that chromatin accessibility at genetically different transcription factor (TF) binding sites is associated with the strength of the altered TF binding motifs. We also identify the key TFs bound to the regions of genetically determined chromatin accessibility. We further show how the genetic effects propagate to the level of gene expression. Together, these findings give valuable insight into the mechanisms by which genetic variation affects gene regulation and provide future targets for research.

Figures