Association and interaction effect of UCP2 gene polymorphisms and dietary factors with congenital heart diseases in Chinese Han population

Congenital heart diseases (CHDs) are the most common birth defects and the leading cause of non-infectious deaths in infants, with an unknown etiology. We aimed to assess the association of genetic variations in UCP2 gene, dietary factors, and their interactions with the risk of CHDs in offspring. The hospital-based case–control study included 464 mothers of children with CHDs and 504 mothers of healthy children. The exposures of interest were maternal dietary factors in early pregnancy and UCP2 genetic variants. Logistic regression analyses were used to assess the association and interaction of UCP2 gene and dietary factors with CHDs. Our results found that the polymorphisms of UCP2 gene at rs659366 and rs660339, together with maternal dietary factors including excessive intake of pickled vegetables and smoked foods were associated with increased risks of CHDs in offspring. Regular intake of fresh meat, fish and shrimp, and milk products were associated with lower risks of CHDs in offspring. Besides, positive interaction between the dominant model of rs659366 and excessive intake of pickled vegetables was found in the additive interaction model (RERI = 1.19, P = 0.044). These findings provide the theoretical basis for gene screening and a new clue for the prevention of CHDs in offspring.


Results
Basic characteristics in case and control groups. The basic characteristics of 504 controls and 464 cases are shown in Table 1. The mean gestational age was 28.56 ± 4.70 in the case group, which was significantly higher than that in the control group 27.77 ± 5.40 (P = 0.016). In addition, we found that maternal education level, average annual income, history of diabetes, smoking history, second-hand smoke exposure history, history of drinking alcohol, and folic acid supplement in the case group were significantly different from the control group (P < 0.05). These factors which were significantly different across groups were adjusted when estimating the association of UCP2 gene polymorphisms and maternal dietary factors with the risk of CHDs.

Association of UCP2 genetic variants and maternal dietary factors with the risk of CHDs. All
SNPs of the UCP2 gene in the controls were conformed to Hardy-Weinberg equilibrium shown in Supplement Table S1 (P > 0.05). Meanwhile, the associations of UCP2 gene polymorphisms and maternal dietary factors with the risk of CHDs based on univariate analysis are shown in Table 2. After adjusting for basic characteristics that were significantly different across groups, the associations of UCP2 gene polymorphisms and maternal dietary factors with the risk of developing CHDs are examined under each gene model based on multivariate logistic regression (Fig. 1).
Besides, our results showed that mothers with excessive intake of pickled vegetables (aOR = 1.67, 95% CI 1.17-2.37, P = 0.004) and smoked foods (aOR = 1.95, 95% CI 1.36-2.79, P < 0.001) were more likely to increase the risk of developing CHDs compared to those with never intake. In contrary, regular intake of fresh meat (aOR = 0.53, 95% CI 0.29-0.97, P = 0.038), fish and shrimp (aOR = 0.43, 95% CI 0.24-0.78, P = 0.005), and milk products (aOR = 0.26, 95% CI 0.18-0.38, P < 0.001) were lower risk of developing CHDs than in those who did not consume. Table 3 shows the effect of the interaction of the dominant model of rs659366 and maternal dietary factors on the risk of CHDs. After adjusting for basic characteristics that were significantly different across groups, our results suggested that there was statistically significant positive interaction between the dominant model of rs659366 and excessive pickled vegetables (RERI = 1.19, 95%CI 0.03-2.35, P = 0.044). Compared with mother carrying the CC genotype and having no pickles vegetable intake, those carrying the CT + TT genotype with excessive pickled vegetables intake have a higher risk of CHDs (OR = 2.96, 95% CI 1.90-4.59, P < 0.001). Besides, we further analyzed the independent effect of maternal UCP2 genetic polymorphisms on the risk of CHDs by stratification of maternal Linkage disequilibrium (LD) test and haplotype analysis. The r-square values of the linkage disequilibrium test among SNPs of maternal UCP2 gene are summarized in Table 4. Our results showed that there were not strong correlations (all r 2 < 0.8) among these three SNPs of UCP2 gene (r 2 = 0.630 between rs659366 and rs660339; r 2 = 0.588 between rs659366 and rs591758; r 2 = 0.623 between rs660339 and rs591758). Linkage disequilibrium (LD) analysis of the UCP2 SNPs between cases and controls is shown in Supplement Fig. S1. The r-square values and log-odds scores indicated that there were no potential linkage disequilibrium blocks between these SNPs. Therefore, we did not perform a haplotype analysis of these SNPs.

Discussion
CHDs are the most common birth defects in the world and the major cause of infant non-infection mortality 30 . Although most studies have found that genetic and environmental factors were closely associated with CHDs, the etiology of most cases remains unknown 5,6 . In the present study, we first assess the association of the the polymorphisms of UCP2 gene and the risk of CHDs. Our results suggested that the polymorphisms of UCP2 gene at rs659366 and rs660339 were associated with increased risks of CHDs after adjustment for confounding factors, which indicated that UCP2 gene could play an important role in the development of CHDs. Over the past decades, there were no studies for whether the UCP2 gene could affect the risk of CHDs in offspring. However, www.nature.com/scientificreports/ some evidence has suggested that the mutations of UCP2 genes can affect the activity or expression levels of UCP2 by increasing the coupling of oxidative phosphorylation, which might cause the reduction of energy expenditure and subsequently contribute to the development of obesity and diabetes 14. At present, most previous studies have found that two common polymorphisms of UCP2 gene including rs659366 (located in the promoter region) and rs660339 (a missense variant in exon 4) were closely associated with obesity and diabetes 12,31,32 . Furthermore, maternal obesity and diabetes have been identified as risk factors of CHDs, which indirectly indicated that genetic variants of UCP2 may be associated with the risk of CHDs in offspring. Besides, the polymorphisms of UCP2 gene also are associated with the excessive accumulation of ROS 33,34 . Previous studies have suggested that excessive accumulation of ROS may lead to the abnormal development of the embryonic heart through irreversible damage to cell membranes, DNA, and other cellular structures 13,21,35 . Meanwhile, excessive accumulation of ROS also may exert cytotoxic effects in cardiomyocytes, thereby resulting in the abnormal development of the embryonic heart 21,36 . In a word, findings from the present study support our results. This implies that UCP2 gene might become the susceptibility gene for CHDs in offspring, and helps to provide a new thread for finding candidate genes for CHD in the future.
In the present study, we also analyzed the association of maternal dietary factors with the risk of developing CHDs in offspring. Our results suggested that excessive intake of pickled vegetables and smoked food were significantly associated with increased risks of CHDs in offspring. Pickled vegetables contain large amounts of nitrite. The previous study has strongly indicated that excessive intake of nitrite can directly affect atrioventricular valve formation by yielding too much NO signaling, thereby causing abnormal development of zebrafish heart 37 . Besides, smoked food usually contained a high concentration of polycyclic aromatic hydrocarbons (PAH). After ingestion, PAH compounds can be transferred through the placenta to the fetus 38 . Their metabolites may bind to DNA to form PAH-DNA adducts, which might exert reproductive and developmental toxicity effects 39 . Previous animal studies and population research [40][41][42][43] have found that the exposure of PAH during embryonic life may derail the concerted expression of genes critical to normal cardiovascular system development and increase the risk of developing CHDs in offspring. On the contrary, our results showed that the regular intake of fresh meat, fish and shrimp, and milk products may decrease the risk of developing CHDs in offspring, which is consistent with most previous studies [44][45][46] . These three foods contain large amounts of proteins, lipids, unsaturated fatty acids, essential amino acids, and minerals, which are beneficial to the heart development of the embryonic.
Besides, we also detected the interactions between the UCP2 gene and maternal dietary factors on the risk of developing CHDs in offspring. Our study firstly reported a positive interaction between pickled vegetable intake and the dominant model of rs659366. Our results suggested that the estimated joint effect on the additive model of rs659366 and excessive intake of pickled vegetables together was greater than the sum of the estimated effects of genotype alone and pickled vegetable intake alone, which indicated that genetic and dietary factors jointly cause the occurrence of CHDs in offspring. Although previous studies 28, 29 have found a significant interaction between the UCP2 gene and dietary factors on the development of metabolic diseases, the involved pathways and the regulatory mechanism are still unclear and need further research. The observed interaction provides a new idea for further finding the etiology of CHDs in offspring and providing prevention strategies.
This study had some limitations. Firstly, because many complicated severe CHDs during pregnancy may have been induced by therapeutic induction or the fetus has died in utero, we only selected CHDs treated with www.nature.com/scientificreports/ surgery after birth as the study participant. It was impossible to select samples by random sampling, which inevitably led to selection bias and affected the representativeness of the results to a certain extent. Secondly, maternal diet during pregnancy was retrospectively reported by the mother with CHDs waiting for surgery in the hospital, therefore, recall bias was inevitable. Although a previous study 47 has suggested that maternal diet during pregnancy could be recalled well after birth, we cannot overlook the possible limitation of recall bias. Thirdly, although some previous studies 28, 29 have found a significant interaction between the UCP2 gene and dietary intake on the development of metabolic diseases, the involved pathways and the regulatory mechanism of interaction between genetic and dietary factors is unclear, which emphasizes the necessity of further research. Fourth, the specific inclusion and exclusion criteria, as well as a very local population in our study may limit the extrapolation of our findings and make it difficult to compare directly with other studies. Fifth, the limitation that only three SNPs of UCP2 gene were selected in our study cannot be ignored, because UCP2 gene has more than ten SNPs. Although our study suggested the significant statistical association between UCP2 gene and the risk of CHDs in offspring, more SNPs should be detected in the future to obtain more genetic information and provide more clues for the study of genetic susceptibility sites of CHDs. Sixth, residual confounders are always of www.nature.com/scientificreports/ concern in observational studies. Although the present study adjusted for a wide range of potential confounders for CHDs, we still could not exclude the possibility that other unmeasured or inadequately measured factors have confounded the true association. Besides, due to the sample size is still relatively small, we only focus on the risk of total CHDs when assessing the effect of maternal dietary factors, the polymorphisms of UCP2 gene, and their interactions on CHDs in offspring. These limitations highlight the urgent need for large samples, multicenter, prospective, and different ethnic populations studies to further confirm our findings.

Conclusion
The polymorphisms of rs659366 and rs660339 in UCP2 gene and maternal dietary factors were found to be associated with the risk of CHDs in offspring. These results provide the theoretical basis for gene screening and a new clue for the prevention of CHDs in offspring. However, the limitations of our study still should be carefully considered. Whether the study findings can be applied to other groups remains to be explored.

Materials and methods
Ethics statement. The study was approved by the Ethics Committee of Xiangya School of Public Health Central South University (No. XYGW-2018-36) and performed according to the Declaration of Helsinki. The design and purpose of the current study were clearly described in the research protocol; the protocol was registered at the Chinese Clinical Trial Registry with registration number ChiCTR1800018492 and is available at http:// www. chictr. org. cn/ listb ycrea ter. aspx. All participants provided written informed consent before completing an enrollment questionnaire as well as providing biological samples.

Study design and data sources.
We conducted a hospital-based case-control study in Changsha City, Hunan province, China, from March 2018 to August 2019. A total of 803 children with CHD and 889 children without any congenital malformations were recruited in Hunan Children's Hospital. During recruitment, every mother of children was interviewed face to face by using a structured questionnaire. A total of 803 mothers who had a fetus with CHDs was recruited to the case group; of these, 124 mothers who were non-Han Chinese, 56 not completing the questionnaire, 26 belonging to multiple pregnancies, and 133 not providing the blood sample were excluded. Finally, a total of 464 cases were involved in our analysis. During the same period, total 889 mothers were recruited in the control group who had a baby without any birth defects; of these, 98 mothers who were non-Han Chinese, 101 not having a complete record of the questionnaire, 19 belonging to multiple pregnancies, and 167 not providing the blood sample were excluded. Not all the selected controls were available, and only 504 controls were included.
Inclusion criteria. In the present study, the exposures of interest were maternal dietary habits in early pregnancy and the polymorphisms of UCP2 gene. The outcome of interest was CHDs that included the following specific phenotypes: atrial septal defect (ASD), ventricular septal defect (VSD), patent ductus arteriosus (PDA), atrioventricular septal defect (AVSD), tetralogy of Fallot (TOF), and complete transposition of great arteries (TGA).
To minimize recall bias of exposure by mothers to the greatest extent, all cases and controls recruited were younger than 1 year old. All cases were selected from the same hospital but different clinics during the same study period as the cases, which were diagnosed by using ultrasonography and confirmed by surgery. And the control group was recruited from the Department of Child Healthcare after health counseling or medical examination. During recruitment, every mother of children was interviewed face to face by using a structured questionnaire by some professionally trained investigators and provided the blood sample.
Epidemiological investigation. The field epidemiological investigation mainly included basic demographic criteria gestational age, maternal education level, average annual income, body mass index (BMI), history of diabetes, smoking history, second-hand smoke exposure history, history of drinking alcohol, folic acid supplement, and dietary habits. Among them, a self-administered food frequency questionnaire was developed based on a semi-quantitative food frequency questionnaire (SQFFQ) which have been used to investigate maternal dites during the pregnancy in a previous study 27 and local eating habits, which was considered validated (The Crohnbach's α coefficient was 0.81 and test-retest reliability was 0.92). The frequency of consumption was defined as follows: the average weekly intake of food less than twice was defined as no food intake, and the average weekly intake of food more than twice was defined as excessive food intake (i.e., pickled vegetables, smoked foods, barbecued foods, and fried foods) or regular food intake (i.e., fresh meat, fish and shrimp, fresh eggs, fresh vegetables, soy foods, and milk products). After completing the epidemiological investigation, for each mother, three to five milliliters of blood samples were collected in EDTA by venipuncture for genotyping. Plasma and blood cells were separated from the blood by centrifugation and finally stored at − 80 °C for later laboratory analysis.
SNP selection and genotyping. The genomic DNA was extracted from peripheral blood cells by using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA, USA) based on the manufacturer's standard protocol and dissolved in sterile TBE buffer. Presently, considering this fact that there few studies on the association between UCP2 gene variants and risk of CHDs in offspring, we selected candidate loci of UCP2 gene mainly based on previous similar studies that assessed the association of UCP2 gene variants with risk of diabetes mellitus and obesity. As a result, these genetic loci including rs659366, rs660339, and rs591758 in UCP2 genes were selected as candidate loci for this study. The polymorphism of UCP2 gene was genotyped using the matrix-assisted laser www.nature.com/scientificreports/ desorption and ionization time-of-flight mass spectrometry Mass Array system (Agena iPLEXassay, San Diego, CA, USA). The primer sequences of rs659366 were ACG TTG GAT GAA ACG CAC GTG TTT GTC CCG (forward) and ACG TTG GAT GTT TAA TTG GCT GAC CGTCC (reverse). The primer sequences of rs660339 were ACG TTG GAT GGA TCC AAG GAG AAA GTC AGG (forward) and ACG TTG GAT GTG GTC AGA ATG GTG CCC ATC (reverse). The primer sequences of rs591758 were ACG TTG GAT GGG GAA AGC ACT GTA AAA CCA (forward) and ACG TTG GAT GTC TCC CCA ACT TCT TAG AGC (reverse). The laboratory technician, who performed the genotyping, retyped and double-checked each sample, and recorded the genotype data, was blinded to whether the samples were from cases or controls. The error rate of genotyping was lower than 5%.
Statistical analysis. Statistical analysis was performed using SPSS 24.0 software (SPSS Inc., Chicago, IL, USA). Demographic characteristics and the SNP genotypes of UCP2 were evaluated using the Pearson chisquared test, Fisher's exact test (for categorical variables), Student's t-test, and Wilcoxon's rank-sum test (for continuous variables). Hardy-Weinberg equilibrium (HWE) was tested for every group. Linkage disequilibrium (LD) and haplotype analysis were performed using HaploView 4.2. The associations of UCP2 genetic variants and maternal dietary factors with the risk of CHDs were estimated by calculating odds ratios (ORs) and their 95% confidence intervals (CIs) using logistic regression models. The forest map was performed using SPSS 24.0 software (SPSS Inc., Chicago, IL, USA) and Review Manager Version 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration). Besides, we also examine additive interactive effects of the UCP2 gene and maternal dietary factors on the risk of developing CHDs in offspring. The relative excess risk due to interaction (RERI), OR, and 95% CIs were calculated using Microsoft Excel according to Knol et al. 48 . The tests were performed significantly for a two-sided P value not exceeding 0.05, except where otherwise specified. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.