Functional characterisation of genetic variants influencing human food preferences using bioinformatics and animal models

背景：营养是影响人类健康的重要因素之一。不健康的食物选择可能会导致几种疾病，如肥胖、代谢综合征、心血管疾病和一些癌症。虽然许多研究都集中在这些疾病的遗传学上，但遗传学家最近才开始关注食物摄入量，尽管结果非常有限。在食物供应不成问题的国家，食物偏好是推动食物选择和消费的最重要因素之一。各种研究表明，与自我报告的食物频率问卷相比，食物喜好指标更能预测实际的长期消费。我们最近的全基因组关联研究强调了许多与特定食物偏好差异相关的基因。尽管取得了这些成功，但没有一个基因编码味觉或嗅觉受体，因此它们在决定食物偏好方面的作用仍然知之甚少。该项目的目的是确定哪些已知的味觉偏好基因有可能导致代谢状态的改变，并在细胞和动物模型中描述这些基因的功能。该项目有三个要素：1)定义与食物偏好相关的遗传变异对报告食物消费的影响-使用英国生物库所有500,000名参与者的食物频率问卷，以及他们的基因数据(2017年初可用)，我们将评估之前描述的16个食物偏好基因变异和味觉受体基因功能变异对报告食物消费的影响。虽然众所周知，这些类型的问卷容易出现报告偏差，但数量非常大的问卷提供了如此强大的统计能力，甚至可以检测到非常小的影响。这一步对于选择哪些遗传变异被发扬光大非常重要。2)所选遗传变异的生物信息学特征--将使用生物信息学工具来描述在前面步骤中选择的变异的特征，以制定关于它们影响哪些基因、它们如何影响基因以及在哪些组织中的假设。这一点很重要，因为在许多情况下，遗传变异并不调节最接近的基因，而是具有更广泛的影响。识别实际的靶点，其组织表达，以及与已知的食欲和新陈代谢相关的通路相互作用的潜力，对于形成机械假说是至关重要的。3)利用细胞培养和动物模型研究遗传变异的功能特征--最相关的遗传变异和基因将在动物模型和细胞培养中进行功能研究，分3个主要阶段进行。(A)绘制与食物偏好有关的大脑回路中候选基因的表达图谱(例如，涉及黑素皮质素MC4受体的下丘脑通路或与脂肪和糖类偏好有已知联系的催产素信号通路)。(B)在体外相关神经细胞模型(例如人类SH-SY5Y)中，模拟CRISPR介导的基因敲除对神经元功能和典型信号的影响。(C)CRISPR--以胚胎中最有希望和/或最易处理的基因为目标，迅速产生一种新的转基因模型，用于在体内全面描述基因缺失对行为食物偏好(Leng和Menzies，CIP)和体内代谢(Morton，CVS)的影响。遗传变异的综合生物信息学特征对大量公开可用资源的体外和体内数据挖掘动物模型的新基因/生物学模型，包括细胞培养、基于细胞的分析、共聚焦成像、尖端转基因工具、体内行为和代谢分析

Background: Nutrition is one of the most important factors influencing human health. Unhealthy food choices can lead to several disorders such as obesity, metabolic syndrome, cardiovascular disease and some cancers. Although many studies have focused on the genetics of these disorders, geneticists have only recently started looking at food intake, although with very limited results. In countries where food availability is not an issue, food preference is one of the most important factors driving food choice and consumption. Various studies suggest that food-liking measures are better predictors of actual long-term consumption compared to self-reported food frequency questionnaires. Our recent genome-wide association studies have highlighted numerous genes associated with differences in specific food preferences. Despite these successes, none of the genes encode taste or olfactory receptors and so their role in determining food preferences is still poorly understood. The aim of the project is to identify which of the known taste preference genes have the potential to lead to altered metabolic states and to characterise these genes functionally in cell and animal models. The project has three elements: 1) Definition of the impact of the genetic variants associated with food preferences on reported food consumption - Using the food frequency questionnaires for all 500,000 participants to UK Biobank, and their genotype data (available in early 2017), we shall assess the impact of the 16 previously described food preference genetic variants and functional variants in taste receptor genes on reported food consumption. Although it is known that these types of questionnaires are subject to reporting biases, the very large numbers provide such statistical power that even very small effects can be detected. This step will be important to select which genetic variants are carried forward. 2) Bioinformatic characterisation of the selected genetic variants - The variants selected in the previous steps will be characterized using bioinformatics tools to formulate hypotheses regarding which genes they affect, how they influence the genes and in which tissues. This is important because in many cases genetic variants do not regulate the closest gene but rather have longer range effects. The identification of the actual target, its tissue expression, and potential for interaction with known appetite- and metabolism-related pathways is of fundamental importance for formulating mechanistic hypotheses. 3) Functional characterisation of the genetic variants through the use of cell culture and animal models - The most relevant genetic variants and genes will be studied functionally in animal models and cell cultures, in 3 main stages. (a) Mapping of expression of the candidate genes in brain circuits involved in food preference (e.g. hypothalamic pathways involving the melanocortin MC4 receptor or oxytocin signalling with known links to fat and sugar preference).(b) Modelling the effects of CRISPR-mediated gene knockouts on neuronal function and canonical signalling after CRISPR-targeted knockout or variant-associated regulatory deletion/modification of the gene in a relevant neuronal cell model in vitro (e.g. human SH-SY5Y). (c) CRISPR-targeting of the most promising and/or tractable gene in embryos to rapidly generate a novel transgenic model for full in vivo characterization of the impact of gene deletion on behavioural food preferences (Leng and Menzies, CIP) and on in vivo metabolism (Morton, CVS).Training Outcomes General epidemiology Handling and analysis of very large-scale food frequency questionnaire data. Integrative bioinformatics characterisation of genetic variants Data mining of large publicly available resources In vitro and in vivo modelling of novel genes/biology of animal model, including cell culture, cell-based assays, confocal imaging, cutting-edge transgenic tools, in vivo behavioural and metabolic assays