Polymorphic variants in Sweet and Umami taste receptor genes and birthweight

The first thousand days of life from conception have a significant impact on the health status with short, and long-term effects. Among several anthropometric and maternal lifestyle parameters birth weight plays a crucial role on the growth and neurological development of infants. Recent genome wide association studies (GWAS) have demonstrated a robust foetal and maternal genetic background of birth weight, however only a small proportion of the genetic hereditability has been already identified. Considering the extensive number of phenotypes on which they are involved, we focused on identifying the possible effect of genetic variants belonging to taste receptor genes and birthweight. In the human genome there are two taste receptors family the bitter receptors (TAS2Rs) and the sweet and umami receptors (TAS1Rs). In particular sweet perception is due to a heterodimeric receptor encoded by the TAS1R2 and the TAS1R3 gene, while the umami taste receptor is encoded by the TAS1R1 and the TAS1R3 genes. We observed that carriers of the T allele of the TAS1R1-rs4908932 SNPs showed an increase in birthweight compared to GG homozygotes Coeff: 87.40 (35.13–139.68) p-value = 0.001. The association remained significant after correction for multiple testing. TAS1R1-rs4908932 is a potentially functional SNP and is in linkage disequilibrium with another polymorphism that has been associated with BMI in adults showing the importance of this variant from the early stages of conception through all the adult life.

Polymorphisms in the TAS1Rs family have been associated to food intake and overweight in children 32 and adults 33 , sweet taste and sucrose detection threshold and sensitivity 34,35 and with food intake and gastric cancer 36 . We investigated the effect of polymorphic variants in the TAS1R1, TAS1R2 and TAS1R3 genes on birth weight on 1077 newborns of Caucasian origin.

Results
Among the 1077 full term-newborns recruited, 25 (2.3%) were discarded because they had a genotyping call rate lower than 75%. The average SNP call rate of the remaining samples was 98.70%, with a minimum of 95.64% for TAS1R3-rs111615792 and a maximum of 99.53% for TAS1R1-rs17029626 and for TAS1R1-rs4908563. The QC analysis showed a concordance among duplicates greater than 99%. All the SNPs allelic and genotyping frequencies resulted in Hardy-Weinberg equilibrium (p-value > 0.01). The average weight measured for males was 3376 g and 3251 g for females. The average gestational time was 39 + 4 weeks (39 + 3 for males and 39 + 4 for females). The relevant characteristics of the population are given in Table 1.

Results of the association analysis between epidemiologic variables and genotypes on birthweight.
We performed a linear regression analysis to evaluate the association with the genetic and nongenetic variables and birthweight. We observed that birthweight is significantly associated with sex with males been heavier than females, Coeff = 124.5; 95% CI (72.17-176.83); p-value = 3.11 × 10 -6 . Gestational age showed a very strong proportional effect in birthweight increase with longer gestational time (expressed in weeks) Coeff = 160.77; 95% CI (140.96-180.58); p-value = 5.72 × 10 -57 . We also observed an inverse association between maternal smoking and birthweight with a coeff = − 202.95; 95% CI (− 328.09 to 77.81); p-value = 0.001). The results are shown in Table 2.
All the analysis conducted to establish the effect of genetic variables were corrected for these three variables. We observed four associations between the genotypes and birthweight, however only one resulted statistically significant after correction for multiple testing. In specific, heterozygous carriers of the T allele of the TAS1R1-rs4908932 SNP showed an increase in birthweight compared to GG homozygotes Coeff: 87.40 (35.13-139.68) p-value = 0.001. We observed an additional association in the TAS1R1 gene and birthweight, namely the positive effect of the G allele of TAS1R1-rs4908930 Coeff rs4908930 = 50.60; 95% CI (0.70-100.49); p-value = 0.047, compared to the common A homozygotes. Finally, we observed two associations in the TAS1R2 gene, an increase in birthweight for the G allele homozygotes of TAS1R2-rs4920566 SNP (Coeff = 71.65; 95% CI (2.40-140.91); p-value = 0.043) and for the carriers of the G allele of the TAS1R2-rs9701796 (Coeff = 136.98; 95% CI (7.80-266.16); p-value = 0.038). The SNPs selected for the TAS1R3 gene did not show any statistically significant association with birth weight. The results of these analysis are shown in Table 3.
Functional relevance of the SNPs. In the GTEx database there are no eQTLs for TAS1R2-rs9701796 and TAS1R2-rs4920566, while for TAS1R1-rs4908932 there is only one eQTL in the aorta artery. The database shows ten eQTLs for TAS1R1-rs4908930, but none in the gastro-enteric tract. RegulomeDB assigns a rank of 4 to TAS1R1-rs4908932 and TAS1R1-rs4908930 and a rank of 5 for TAS1R2-rs9701796 and TAS1R2-rs4920566. Table 1. Characteristics of the population in study. a This information was available for 1046 subjects. The measurement unit is "weeks". b This information was available for 1041 subjects. The measurement unit is "g". c This information was available for 733 subjects. The unit of measure is "cm". d This information was available for 1029 subjects. The unit of measure is "years". e This information was available for 731 subjects. The measurement unit is "kg/m 2 ". f This information was available for 736 subjects. The unit of measure is "g. g This information was available for 1037 subjects; h-This information was available for 1043 subjects. www.nature.com/scientificreports/ Haploreg, shows that all the SNPs are situated in potentially methylated regions and that they alter a potential TF binding site, although with a null or modest effect.

Discussion
Birthweight can have a dramatic impact in the development and health of the newborn in the short-period and in the long-term period. There are several well-known factors that have a strong effect on birthweight and genetic variability has been investigated in this regard. GWAS have identified a relatively small number of risk loci and the hereditability explained is still limited. We have taken into consideration the genetic variability in taste receptors since it has been repeatedly associated with BMI and metabolic traits, dietary behaviours and human pathologies [15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31] . This study has been conducted on more than 1000 new-borns of Italian origin collected by the Santa Chiara Hospital of Pisa with the aim to further improve our knowledge on the epidemiologic and genetic factors that influence birthweight.  www.nature.com/scientificreports/ We observed a very strong effect of sex, gestational age and maternal smoking on the birthweight as expected and reported in the literature 37,38 .
The most novel findings of this study are represented by several associations between the SNPs and birthweight, in particular, we found two hits in the TAS1R1 (rs4908930, rs4908932) and TAS1R2 (rs4920566, rs9701796) genes. None of these SNPs has been previously reported with body weight in adults or children. The best finding, both considering the strength of the association (i.e. coefficient) and the statistical significance (p value) that we observed was the average increase of 87 g for the carriers of the T allele of TAS1R1-rs4908932 (p = 0.001). This finding is the only one that remains significant after multiple testing correction according to Bonferroni.
According to Haploreg TAS1R1-rs4908932 is situated 29 bp in the 3′-UTR of the TAS1R1 gene and could modify the methylation status of the gene in 24 tissues among which several belonging to the GI tract including the colon, the liver, the stomach and the pancreas. It is interesting to note that the SNP possibly modifies the methylation status also in several foetal organs, including the intestine. In addition, Haploreg also suggests that the SNP could modify the binding of several transcription factors including HNF4A which is involved in the development of the intestines and mutation of which have been associated with metabolic diseases [39][40][41][42] . These indications suggest a potential key role of the SNP in regulating the gene expression that could translate in the differential birthweight observed. TAS1R1-rs4908932 is in moderate LD with several SNPs that have been found to be associated with human traits. Of particular relevance is rs6577584 (r 2 = 0.114, D' = 0.9402 in the European individuals of the 1000G project, according to LDlink) since it has been found to be associated, at genome wide level, with BMI in adults in a study conducted using the UK biobank repository 43 , highlighting the importance of the locus from foetal life to adulthood.
In addition to BMI variants in LD with TAS1R1-rs4908932 have also been reported to be associated with cardiovascular disease, age at menopause, and medication use as reported in the GWAS catalogue 43 , making this a potential pleiotropic locus of the genome.
The effect size on birthweight associated TAS1R1-rs4908932 observed in our study is rather large (87 g) compared with what reported by others 10,12,44 for several SNPs associated with birthweight. For example, Beaumont and colleagues, in a study that included UK biobank data, reported a maximum effect size of around 50 g 10 . Even though, the SNPs are not the same and therefore not directly comparable, this difference could be at least partially explained by the size of the two studies with ours being smaller. It would be therefore important to replicate our finding in a larger cohort of individuals to better compute the effect of the variant.
A clear strength of this study is the fact that the individuals collected are consecutive and therefore they are an unbiased representation of the individuals born in a high-volume centre such as the Santa Chiara university hospital in Pisa. Moreover, population stratification is not an issue considering that all the subjects of the study have been collected in the same center.
In conclusion the association that we propose here between TAS1R1-rs4908932 and birthweight is corroborated by a study-wise significance and reflects the association of a locus on chromosome 1 that has been already observed in adults showing the importance of this variant from the early stages of conception through all the adult life.

Materials and methods
Study population. This study was carried out on 1077 new-born recruited at the Division of Neonatology of the Santa Chiara Hospital (Pisa, Italy) from 2015 to 2019. Inclusion criteria in the recruitment were term birth defined as a gestational age ≥ 37 weeks and five minutes Apgar score ≥ 7 that indicates that the newborn is in good health and does not require additional interventions or treatments 45,46 . Exclusion criteria consisted of an Apgar score < 7, suspicion of genetic syndrome or metabolic diseases. For each individual 5 ml of blood were collected from the cord at birth and anthropometric measures at birth (birth weight, length, head circumference) were retrospectively collected. In addition, mother's age, maternal smoking behavior, pre-pregnancy BMI, maternal gestational diabetes, maternal pre-gravidic diabetes, maternal weight increase during pregnancy were also collected. The parents of all subjects signed a written informed consent form and the study was approved by the ethical committee of the Meyer Children Hospital of Florence which is the appointed IRB for all the pediatric studies in the Tuscany region, all methods were performed in accordance with the relevant guidelines and regulations. Additional information has been given elsewhere 47 .

Selection of polymorphisms.
The selection of polymorphisms included in the study focused on tagging (tSNP) and functional SNPs. The choice of the tSNPs was based on the Linkage Disequilibrium (LD) in the Caucasian population. To select the tSNP of TAS1R1, TAS1R2 and TAS1R3, the chromosomal position of these genes was inserted in Ensemble genome browser's VCF to PED conversion tool (online version http://grch3 7.ensem bl.org/Homo_sapie ns/Tools /Vcfto Ped), which returns a linkage pedigree file and a marker information file with European population genotype data. These files were uploaded to Haploview, a bioinformatic software (https :// www.broad insti tute.org/haplo view/haplo view version 4.2) used to choose tagging SNPs 48 . The search criteria for tSNP set on the Haploview software were: r 2 > 0. 8

Statistical analysis.
For each SNP Hardy-Weinberg equilibrium was assessed using Pearson chi-square test. To estimate how the anthropometric and epidemiologic variables and genetic variability affects birth weight a linear regression model was used, calculating the regression coefficient, its confidence interval (95% CI) and the p-value. If the regression coefficient takes positive values, it indicates that the average weight increases for carriers of the rare allele, on the contrary, in the case of negative values, the regression coefficient represents how much the birth weight is reduced on average in the presence of rare allele. For the genetic variables we used a codominant model of inheritance putting the most common allele as reference category in each analysis. Applying the Bonferroni correction (dividing the threshold value of 0.05 by the number of SNPs) we considered the p-value = 0.0024 as threshold for statistical significance.
Bioinformatic tools. To test the functional relevance of the SNPs associated with birthweight we used three bioinformatic tools. The Genotype-Tissue Expression (GTEx-version v7release, GRCh37/hg19 assembly, data access on 04/01/2019) project to identify if the variants where associated with gene expression. Such variants are called expression quantitative trait loci (eQTLs). RegulomeDB 2.0 (https ://www.regul omedb .org/regul omesearc h/) was used, instead, to assess whether the variant is associated with regulatory potential. RegulomeDB assigns to each SNP a rank, going from 1a to 7 in descending order of the accumulated evidences of functional relevance of the SNP. The accumulated evidence consists in several items such as whether the SNPs binds to one or more transcription factor, if it lies in a DNAse sensitive region (i.e. a region that is accessible to transcription factors), and it is an eQTL. Finally, we used Haploreg v4.1 (https ://pubs.broad insti tute.org/mamma ls/haplo reg/ haplo reg.php) to assess whether the variants belong to transcription factor binding sites or if they are in methylated DNA regions.

Data availability
The data for this work will be made available to researchers who submit a reasonable and detailed request to the corresponding author, conditional to approval of the Ethics Commission of the Meyer Children Hospital of Florence which is the appointed IRB for all the pediatric study in the Tuscany region. Data will be stripped from all information allowing identification of study participants.