Detecting epistasis within chromatin regulatory circuitry reveals CAND2 as a novel susceptibility gene for obesity.

Genome-wide association studies have identified many susceptibility loci for obesity. However, missing heritability problem is still challenging and ignorance of genetic interactions is believed to be an important cause. Current methods for detecting interactions usually do not consider regulatory elements in non-coding regions. Interaction analyses within chromatin regulatory circuitry may identify new susceptibility loci. We developed a pipeline named interaction analyses within chromatin regulatory circuitry (IACRC), to identify genetic interactions impacting body mass index (BMI). Potential interacting SNP pairs were obtained based on Hi-C datasets, PreSTIGE (Predicting Specific Tissue Interactions of Genes and Enhancers) algorithm, and super enhancer regions. SNP × SNP analyses were next performed in three GWAS datasets, including 2286 unrelated Caucasians from Kansas City, 3062 healthy Caucasians from the Gene Environment Association Studies initiative, and 3164 Hispanic subjects from the Women’s Health Initiative. A total of 16,643,227 SNP × SNP analyses were performed. Meta-analyses showed that two SNP pairs, rs6808450–rs9813534 (combined P = 2.39 × 10−9) and rs6808450–rs3773306 (combined P = 2.89 × 10−9) were associated with BMI after multiple testing corrections. Single-SNP analyses did not detect significant association signals for these three SNPs. In obesity relevant cells, rs6808450 is located in intergenic enhancers, while rs9813534 and rs3773306 are located in the region of strong transcription regions of CAND2 and RPL32, respectively. The expression of CAND2 was significantly downregulated after the differentiation of human Simpson–Golabi–Behmel syndrome (SGBS) preadipocyte cells (P = 0.0241). Functional validation in the International Mouse Phenotyping Consortium database showed that CAND2 was associated with increased lean body mass and decreased total body fat amount. Detecting epistasis within chromatin regulatory circuitry identified CAND2 as a novel obesity susceptibility gene. We hope IACRC could facilitate the interaction analyses for complex diseases and offer new insights into solving the missing heritability problem.

全基因组关联研究已经确定了许多肥胖的易感位点。然而，遗传力缺失问题仍然具有挑战性，并且忽视基因相互作用被认为是一个重要原因。当前检测相互作用的方法通常不考虑非编码区域的调控元件。染色质调控回路内的相互作用分析可能会确定新的易感位点。 我们开发了一种名为染色质调控回路内相互作用分析（IACRC）的流程，以识别影响体重指数（BMI）的基因相互作用。基于Hi - C数据集、PreSTIGE（预测基因和增强子的特定组织相互作用）算法以及超级增强子区域获得了潜在的相互作用单核苷酸多态性（SNP）对。接下来在三个全基因组关联研究（GWAS）数据集中进行了SNP×SNP分析，包括来自堪萨斯城的2286名无亲缘关系的白种人、来自基因 - 环境关联研究倡议的3062名健康白种人以及来自妇女健康倡议的3164名西班牙裔受试者。 总共进行了16,643,227次SNP×SNP分析。荟萃分析表明，经过多重检验校正后，两个SNP对rs6808450 - rs9813534（合并P = 2.39×10⁻⁹）和rs6808450 - rs3773306（合并P = 2.89×10⁻⁹）与BMI相关。单SNP分析未检测到这三个SNP的显著关联信号。在与肥胖相关的细胞中，rs6808450位于基因间增强子区域，而rs9813534和rs3773306分别位于CAND2和RPL32的强转录区域。人辛普森 - 戈拉比 - 贝梅尔综合征（SGBS）前脂肪细胞分化后，CAND2的表达显著下调（P = 0.0241）。在国际小鼠表型联盟数据库中的功能验证表明，CAND2与瘦体重增加和总体脂肪量减少相关。 在染色质调控回路内检测上位性将CAND2确定为一种新的肥胖易感基因。我们希望IACRC能够促进复杂疾病的相互作用分析，并为解决遗传力缺失问题提供新的见解。