Associations Of Maternal Urinary Levels Of Neonicotinoid Insecticides During Early Pregnancy And Gene Polymorphism With Infant Reproductive Hormones

The massive global use of neonicotinoid insecticides(NNIs)has resulted in widespread population exposure.Animal studies have shown that exposure to NNIs during pregnancy can hinder reproductive development in offspring;however,no relevant population studies are available.Infancy urinary reproductive hormone levels are important biomarkers for assessing reproductive development and function.Commonly used indicators of reproductive hormones include the follicle-stimulating hormone(FSH)and luteinizing hormone(LH),as well as estrogens and androgens.In this study,898mother-infant pairs with a mean infant age of 1.16 months were included based on a prospective birth cohort to measure infancy urinary reproductive hormone levels,establish reference range of urinary reproductive hormones in healthy infants at 1 month of age,and investigate the potential predictors of infancy urinary reproductive hormones.Since early pregnancy is a critical window for fetal reproductive development,this study analyzed the association between the levels of maternal urinary NNIs and their metabolites during early pregnancy and infancy urinary reproductive hormones.To further provide clues to the potential mechanisms by which NNIs exposure affects reproductive hormone levels,given the important roles of FSH and LH in germ cell genesis,development and maturation,and regulation of estrogen and androgen,and that FSH and LH levels are associated with polymorphisms of genes regulating FSH or LH.In this study,we further analyzed the interaction of maternal urinary NNIs and their metabolites and polymorphisms of genes regulating FSH or LH in association with infancy urinary level of FSH or LH.Part 1:Associations between Maternal Urinary Levels of Neonicotinoid Insecticides during Early Pregnancy and Infancy Reproductive HormonesObjective:(1)To assess the levels and construct reference ranges of infancy urinary reproductive hormones and further explore the predictors of infancy urinary reproductive hormones.(2)To investigate the association of single maternal urinary level of NNIs and their metabolites as well as the mixtures during the early pregnancy with infancy urinary reproductive hormones,and to identify the NNIs that contribute most to the associations.Methods:Based on a birth cohort in Wuhan,898 mother-child pairs with a mean infancy age of 1.16 months recruited between 2015 and 2019 who donated both maternal urine samples during the early pregnancy and infancy urine samples around 1-month-old were included.(1)This study measured urinary levels of estrogens and androgens,including estrone(E1),Estradiol(E2),testosterone(Testo),dihydrotestosterone(DHT),dehydroepiandrosterone(DHEA),and androstenedione(A4)as well as FSH and LH of the infants.The non-parametric rank-sum test was used to evaluate the differences in infancy urinary reproductive hormones of male and female infants.After excluding infants whose parents with a history of infertility treatment and pregnancy complications,or infants of preterm birth,low birth weight or in vitro fertilization,the 2.5th and 97.5th percentile of infancy urinary reproductive hormones of male or female infants was used to construct the empirical reference ranges of infancy urinary reproductive hormones among healthy 1-month-old infants.In addition,the multiple linear regression model was used to identify the potential predictors of infancy urinary reproductive hormones.(2)Four common maternal urinary NNIs and six metabolites in the early pregnancy,including acetamiprid(ACE),clothianidin(CLO),imidacloprid(IMI),thiamethoxam(THM),clothianidin(CLO),imidacloprid(IMI),thiamethoxam(THM),desmethyl-acetamiprid(DM-ACE),desmethyl-clothianidin(DM-CLO),5-hydroxy-imidacloprid(5-OH-IMI),imidacloprid-olefin(IMI-olefin),desnitro-imidacloprid(DN-IMI),and 6-chloronicotinic acid(6-Cl NA)were measured.NNIs and their metabolites with a detection rate greater than 60%were used as independent variables,infancy urinary reproductive hormones with a detection rate greater than 50%were used as dependent variables.Multiple linear regression models and weighted quantile sum(WQS)regression models were used to explore the associations of single maternal urinary NNIs and their metabolites or their mixtures during the early pregnancy with infancy urinary reproductive hormones,respectively,and to derive the weight of contribution of each substance to the association.NNIs and their metabolites as well as urinary reproductive hormones were included into the model after being adjusted with urinary specific gravity and natural logarithmic transformed.The stratified analysis by infant sex.And the sensitive analysises were also performed after excluding infants whose parents with a history of infertility treatment and pregnancy complications,or infants of preterm birth,low birth weight or in vitro fertilization,or randomly selected 288 male infants to 1:1 matched with female infants.Results:(1)The detection rates of all urinary reproductive hormones of both male and female infants were higher than 50%,except for E1 and E2,with a detection rate lower than 5%.Urinary FSH of male infants were significantly lower than female infants(P<0.05),and urinary LH,LH/FSH ratio,Testo,DHT,and A4 of male infants were significantly higher than female infants(P<0.05).The reference range of urinary FSH,LH,LH/FSH ratio,Testo,DHT,DHEA and A4 of male infants was 0.46～2.07 m IU/m L,0.48～1.25 m IU/m L,0.47～1.56,0.02～7.05 ng/m L,<LOQ～0.30 ng/m L,4.91～190.83ng/m L and 0.03～0.37,respectively.The reference range of urinary FSH,LH,LH/FSH ratio,Testo,DHT,DHEA and A4 of female infants was 0.46～7.68 m IU/m L,0.40～0.85m IU/m L,0.08～1.22,<LOQ～3.33 ng/m L,<LOQ～0.14 ng/m L,1.77～187.98 ng/m L and0.01～0.34 ng/m L,respectively.The predictors of urinary reproductive hormones were not identical among male infants and female infants;however,there was more consistency in that lower infancy urinary reproductive hormone levels except for higher levels of the ratio of LH to FSH and A4 were observed among artificially fed or mixed-fed infants compared with breastfeeding;infancy urinary reproductive hormone levels were generally lower in infants of multiparous mothers compared with those of primiparous mothers,and infancy urinary reproductive hormones were generally lower in infants of mothers had a history of infertility treatment relative to those whose mothers had no history of infertility treatment.(2)Multiple linear regression analysis in the total population revealed that maternal urinary 5-OH-IMI was significantly and negatively associated with infancy urinary FSH(P<0.05),maternal urinary 5-OH-IMI,IMI,and IMI-olefin were all significantly and negatively associated with infancy urinary LH(P<0.05),and maternal urinary 5-OH-IMI and IMI-olefin were both significantly and negatively associated with infancy urinary A4(P<0.05).WQS regression model analysis in the total population revealed that the WQS index of maternal urinary NNIs and their metabolite in early pregnancy was significantly and negatively associated with infancy urinary FSH,LH,Testo,DHT,and A4,with 5-OH-IMI contributing the most to the above associations.In the stratification analysis by infant sex,significant negative associations(P<0.05)between maternal urinary 5-OH-IMI,IMI,and IMI-olefin in the early pregnancy and infancy urinary A4 levels were observed in male infants in the multiple linear regression analysis.WQS regression model analysis in male infants revealed that the WQS index of maternal urinary NNIs and their metabolite in early pregnancy was significantly and negatively associated with infancy urinary A4 of male infants,with 5-OH-IMI contributing the most.In female infants,maternal urinary DN-IMI was significantly and negatively associated with infancy urinary DHT(P<0.05).Maternal urinary CLO in early pregnancy was significantly and negatively associated with LH/FSH ratio in female infants(P<0.05).WQS regression model analysis in female infants revealed that the WQS index of maternal urinary NNIs and their metabolite in early pregnancy was significantly and negatively associated with infancy urinary LH in female infants,with 5-OH-IMI contributing the most.The WQS index of maternal urinary NNIs and their metabolite in early pregnancy was significantly and negatively associated with urinary DHT in female infants,with DN-IMI and 5-OH-IMI contributed the most.The results of sensitivity analysis were generally consistent with the above analysis.Conclusion:(1)Urinary FSH was significantly lower in male infants than in female infants,and urinary LH,LH/FSH ratio,Testo,DHT and A4 were significantly higher in male infants than in female infants.Artificial or mixed feeding,multiparous and maternal history of infertility treatment were associated with lower infancy urinary reproductive hormone levels in male and female infants.The reference range of urinary reproductive hormones explored in this study for healthy 1-month-old male and female infants may provide reference for future studies based on urinary reproductive hormones in infants.(2)Exposure to NNIs and their metabolites in early pregnancy was associated with lower levels of the infancy urinary FSH,LH and Testo,DHT,and A4,with 5-OH-IMI contributing the greatest weight,suggesting an important role for 5-OH-IMI in influencing reproductive hormone levels.Part 2.Interaction between Maternal Urinary Levels of Neonicotinoid Insecticides during Early Pregnancy and Polymorphisms of Genes Regulating Gonadotropin in Associations with Infant Levels of GonadotropinObjective:To investigate the interaction effects between maternal urinary levels of NNIs and their metabolites during pregnancy and polymorphisms of genes regulating FSH or LH on the associations with infancy urinary levels of FSH or LH,and further assess the sex difference in the gene-environment interaction.Methods:A total of 210 out of the 898 mother-child pairs with sufficient cord blood samples were included in this part.Illumina Human Asian Screening Array(ASA)gene chip was used for the genome-wide genotyping of the cord blood.Genes in the pathways regulating FSH and LH levels were collected from the Gene set enrichment analysis(GSEA),with 85 and 10 genes collected,respectively.A total of 215 and 6 single nucleotide polymorphisms(SNPs)loci in the related genes of FSH and LH that met the quality control criteria and with the linkage disequilibrium coefficient r~2<0.80 were extracted from the typing results,respectively.Based on the results of the first part of the study,the interaction between 5-OH-IMI,which contributed most to the association of maternal urinary NNIs and their metabolites with infancy urinary FSH and LH,and the identified SNPs,which significantly associated with infancy urinary levels of FSH and LH in the co-dominant,dominant,recessive genetic models,was analyzed in the total population,male,and female infants,respectively.The population was divided into low and high concentration groups based on the 66.7th percentile of the maternal urinary level of 5-OH-IMI.And we predict the function of SNPs with significant interactions in Haplo Reg database,and analyze the histone modifications and chromatin accessibility of the genomic regions where these SNPs are located in the hypothalamus,pituitary,ovary,or testis in the Cistrome database.We further performed the expression quantitative trait locus(e QTL)analysis to predict the association of SNPs with their gene expression levels in the hypothalamus,pituitary,ovary,or testis with the Genotype-tissue expression(GTEx)database.Results:After adjusting for all the potential confounders,a total of 19 SNPs were identified with P<0.05 in the association with infancy urinary FSH in co-dominant,dominant,or recessive model analysis.These 19 SNPs were located in seven genes,including the Cytochrome P450 oxidoreductase(POR),Smoothened(SMO),T-Box transcription factor 3(TBX3),Estrogen receptor 2(ESR2),AKT serine/threonine kinase 1(AKT1),WW domain containing oxidoreductase(WWOX),and protein kinase C alpha(PRKCA).The ESR2 rs1256031,WWOX rs11150123,WWOX rs8064012,POR rs41301394,and PRKCA rs58162388 significantly interacted with 5-OH-IMI in the association with FSH in female infants.Bioinformatic analysis suggested that all of the above five loci are potentially functional genetic variants.The e QTL analysis revealed that ESR2 rs1256031 G allele was significantly positively associated with expression level of ESR2 in the pituitary(P<0.05).WWOX rs11150123 G allele was significantly positively associated with the expression level of WWOX in the ovary(P<0.05).POR rs41301394 T allele was significantly positively associated with expression level of POR in the pituitary,ovary,and testis(P<0.05).Conclusion:In this study,ESR2 rs1256031,WWOX rs11150123,WWOX rs8064012,POR rs41301394,and PRKCA rs58162388 significantly interacted with maternal urinary5-OH-IMI in association with infancy urinary FSH in female infants.It is suggested that genes such as ESR2,WWOX,POR,and PRKCA may be involved in the process of NNIs and their metabolites disrupting the level of FSH in infants,especially in female infants.The results of this study may provide clues to the mechanisms of NNIs and their metabolites in influencing reproductive hormones.However,future validation in a larger sample size population is still needed.