Individual differences in taste and their association with genes, dietary behaviour, and brain structure

Abstract

Taste perception plays a key role influencing human dietary behaviour and further health consequences. It has been shown that genetic and neurological factors contribute to variation in taste, but their underlying mechanisms remain largely underexplored. For example, it is unclear how much of the variance in sweet taste is due to genetics, whether the association between sweet and bitter tastes is due to genetic covariance, and whether variation in brain structure is associated with taste. The goal of this work is to extend current knowledge in individual differences in human taste perception of sweetness and bitterness by showing their relationships with genes, dietary behaviour, and brain morphology. We perform quantitative and statistical genetic analyses using an extensively phenotyped and genotyped twin sample of Australian adolescents (n = 1999), with replication and extension making use of two publically available datasets from the Human Connectome Project (HCP; n = 1101) and the UK Biobank (N = 438,870).In Chapter 1, we employed structural equation modelling (variance components analysis) to provide the first evidence that approximately 30% of variation in the perceived intensity of sweet compounds, including sugars (i.e. glucose and fructose) and high-potency sweeteners (i.e. aspartame and neohesperidin dihydrochalcone) is due to genetics. Furthermore, we identified a common genetic factor accounting for more than 75% of the genetic variance in the perception for each of these sweet compounds, suggesting that the perception of both sugars and high-potency sweeteners was regulated by a common set of genes.In Chapter 2, we demonstrated that a quarter of the genetic variance in perceived sweetness (i.e. a weighted mean score of the four sweet tastes from Chapter 1) is shared with at least half of the genetic variance in the perceived bitterness of quinine, sucrose octaacetate (SOA), and caffeine. The genetic association between sweetness and the bitterness of propylthiouracil (PROP) becomes evident after adjusting for the TAS2R38 genotype. These results reveal shared genetic pathways for the human perception of sweetness and bitterness.To pinpoint the source of genetic variation in bitter taste, in Chapter 3, we performed a genome-wide association analysis (GWAS). As previous work was underpowered to detect variants with small effects, we used a bivariate approach to boost power. We identified two putative novel variants with small effects (< 2%) on chromosomes 7 and 12 for the perceived intensity of denatonium benzoate (DB) and SOA, respectively. We provided the first independent replication for the caffeine bitterness on chromosome 12 and confirmed the previously identified variants on chromosomes 7 and 12 for PROP and quinine, respectively. Building on the common source of genetic variances identified in Chapter 2, we showed evidence for pleiotropy that each of the three variants (for quinine, caffeine, and SOA) on chromosome 12 is associated with more than one of the bitter tastes (quinine, SOA, caffeine, and DB). These findings offer a useful starting point for determining the biological pathways linking perception of bitter substances.We investigated the effect of taste perception on diet-related outcomes. Although previous findings of the association between bitter taste perception and bitter beverage intake were inconsistent, we used two-sample Mendelian randomization (Chapter 4) to demonstrate a causal relationship between the perceived bitterness of PROP, quinine, and caffeine and the consumption of coffee, tea, and alcohol among UK Biobank participants. In Chapter 5, our longitudinal analyses showed that the perceived sweetness in adolescence is a predictor for body mass index (BMI) in early adulthood. We also showed that this association is partly due to genetics using structural equation modelling and polygenic risk prediction approaches.In Chapter 6, we conducted an exploratory analysis to examine the associations between the volumes of 84 brain regions of interest (ROI) and the perceived sweetness and bitterness (PROP, quinine, caffeine). The volumes of 5 ROIs (right cuneus gyrus, right inferior temporal gyrus, left transverse temporal gyrus, and left and right caudates) were nominally associated with both sweet and bitter tastes. Additionally, we replicated an association between quinine bitterness and the volume of left entorhinal gyrus using data from HCP. This study provides the first evidence for an association between brain morphology and taste intensity ratings.In conclusion, we used structural equation modelling to show that sweet taste is heritable and the association between sweet and bitter tastes is largely due to genetic covariance. Additionally, bivariate GWAS identified variants with small effects on bitter tastes and revealed their pleiotropy. Furthermore, results from Mendelian randomization and longitudinal analyses evidence the potential causal impact of taste on dietary behaviour and BMI. These findings enhance our understanding in the genetic architecture of taste and shed light on the personalized nutrition and medicine. Lastly, we showed that the volume of specific brain regions is associated with taste perception, which provides new insights into the gustatory network and suggests a potential role of brain structure in taste.