Profiles Of Urinary Polycyclic Aromatic Hydrocarbons Metabolites-Associated DNA Methylation And Their Associations With Heart Rate Variability

Polycyclic aromatic hydrocarbons(PAHs)are persistent organic pollutants that widespread in the environment.Inhalation and dietary intake are considered to be the most common routes for PAHs exposure.After entering human body,PAHs can be distributed by the systemic circulation to the whole body.PAHs are metabolized by several enzymes and excreted through urine and feces mainly in the form of mono-hydroxylated PAHs(OH-PAHs).In addition to causing a variety of cancers including lung cancer,PAHs are also associated with subclinical cardiovascular disease.Heart rate variability(HRV)reflects the autonomic nervous modulation of the heart,whose reduction is suggested as an early indication of cardiovascular autonomic dysfunction.Substantial epidemiological studies have suggested the associations between PAHs exposure and HRV reduction,but the underlying mechanisms remain to be elucidated.DNA methylation is an important epigenetic modification in which a methyl group is added to the base cytosine in the dinucleotide 5’-Cp G-3’.Recent studies have reported that PAHs exposure can induce DNA methylation change,however,previous studies were based on the global DNA methylation or the candidate gene-based approaches.Study to identify Cp G sites in association with PAHs exposure on a genome-wide scale is scarce.Whether PAHs exposure-associated DNA methylation may play a role in the relationship between PAHs exposure and HRV reductions remain to be clarified.Furthermore,little is known about the potential roles in gene regulation and relevant pathways for the PAHs exposure-associated DNA methylation.Epidemiological studies have reported that smoking is linked to the reduction of HRV.However,except for PAHs,tobacco smoke contains a large number of other environmental pollutants that may serve as potential confounders in the relationship between PAHs exposure and HRV alternations.Therefore,it is important to explore the profiles of DNA methylation associated with PAHs exposure and their associations with HRV in non-smokers,which can provide a relatively accurate estimation of their relationships.In this cross-sectional study,we aim to explore profiles of DNA methylation associated with urinary PAHs metabolites,to evaluate their associations with HRV alternation among non-smokers,and further to investigate the association of PAHs exposure-related DNA methylation with the transcription levels of their corresponding genes and to explore the related pathways of the annotated genes.We firstly performed a genome-wide DNA methylation association analysis of 10 different urinary mono-hydroxylated PAHs(OH-PAHs)and total OH-PAHs among 304 Chinese non-smokers to identify the profiles of urinary OH-PAHs metabolites-associated DNA methylation.For the cytosine-phosphate-guanine(Cp G)sites significantly associated with urinary OH-PAHs metabolites,we investigated their associations with HRV indices.We further explored the correlation between DNA methylation levels and the m RNA levels of their corresponding genes,and investigated the related pathways of the annotated genes.Part I Genome-wide analysis of DNA methylation with urinary polycyclic aromatic hydrocarbons metabolitesObjective: This study aims to investigate the profiles of urinary PAHs metabolites associated DNA methylation among adult non-smokers.Methods: This cross-sectional study included a total of 304 non-smoking adults,including 90 community residents of Wuhan(Wuhan panel)and 62 community residents of Zhuhai(Zhuhai panel)in the baseline survey of Wuhan-Zhuhai(WHZH)cohort,99 healthy participants recruited in Shiyan(Shiyan panel),and 53 healthy coke oven workers(COW panel).We measured 12 different OH-PAHs in urine using gas chromatography-mass spectrometry(GC-MS)system.Two of the 12 different OH-PAHs were excluded in the current study given that the urinary levels of two OH-PAHs were generally below their corresponding limits of detection.We measured DNA methylation levels at > 485,000 Cp G sites in whole blood using Illumina Infinium Human Methylation450 K Bead Chip among those participants.We performed the genome-wide DNA methylation association analysis of 10 different urinary OH-PAHs and total OH-PAHs(the sum of ten OH-PAHs)in Wuhan panel,Zhuhai panel,COW panel,and Shiyan panel separately.Multiple linear regression models were used to test the associations between DNA methylation and urinary OH-PAHs in each panel,where the methylation value at each Cp G site was included as the dependent variable and each urinary OH-PAH was included as the independent variable.All models were adjusted for age,gender,body mass index,drinking status,blood cell counts,and surrogate variables.We performed a fixed-effect meta-analysis with a sample-size weighted approach to obtain P values and an inverse-variance weighted method to obtain estimates of effect size.Genome-wide significance was defined as Benjamin-Hochberg corrected false discovery rate(FDR)< 0.05.Result: We identified three Cp G sites whose methylation levels were significantly positively associated with urinary OH-PAHs.Two Cp G sites,cg18223625 located in COL20A1 and cg07805771 near SLC16A1,were significantly or marginal significantly associated with a higher urinary level of 2-hydroxynaphthalene(2-OH-Nap)after Benjamini-Hochberg correction(β = 0.431 for cg18223625 and 0.354 for cg07805771,FDR = 0.016 and 0.056,respectively).While cg09235308,mapped to PLEC1/MIR661,was significantly associated with total OH-PAHs(β = 0.478,FDR = 0.004).However,we did not observe any Cp G sites significantly associated with the other 9 OH-PAHs within the genome-wide significance threshold(FDR < 0.05).Conclusions: We identified three Cp G sites significantly associated with urinary OH-PAHs levels among adult non-smokers,among which cg18223625 and cg07805771 were significantly positively associated with urinary 2-OH-Nap while cg09235308 was significantly positively associated with urinary total OH-PAHs.No Cp G was found to be associated with the other 9 OH-PAHs at the genome-wide significance.Our findings provided new evidence that PAHs exposure was linked to differential DNA methylation.Further study with a larger sample size is needed to validate our findings and identify more Cp G sites associated with PAHs exposure.Part II Associations of urinary polycyclic aromatic hydrocarbons metabolites-associated DNA methylation with heart rate variabilityObjective: We aim to evaluate the associations of urinary 2-OH-Nap and total OH-PAHs associated DNA methylation with HRV alternations among adult non-smokers,and to investigate their potential relationships with gene transcription levels and the relevant pathways of the identified Cp G sites.Methods: In this cross-sectional study,the HRV indices were measured in a total of 205 subjects from the abovementioned Wuhan panel(n = 90),Zhuhai panel(n = 62),and COW panel(n = 53)panels.HRV was measured using a 3-channel digital Holter monitor for 10 minutes.Data from each Holter assessment was recorded automatically and processed using the Impresario and Cardio-navigator Plus software.Two-time domain parameters and three frequency domain parameters were generated.Time domain parameters include standard deviation of all R-R intervals(SDNN),root mean square of successive differences in adjacent R-R intervals(r-MSSD),while frequency domain parameters include low frequency power(LF),high frequency(HF)power,and total power(TP).The Human HT-12 version 4 Expression Bead Chip was used to detect the m RNA profiles in subjects of Shiyan panel(n = 99).We firstly used multiple linear regression models to test the relationships between urinary 2-OH-Nap as well as total OH-PAHs and HRV indices in Wuhan panel,Zhuhai panel and COW panel separately.Models were adjusted for age,gender,BMI,drinking,working years(for COW panel only),and regular exercise.Secondly,for the Cp G sites identified to be significantly associated with urinary 2-OH-Nap and total OH-PAHs,we explored their associations with HRV measurements by using multiple linear regression models in the Wuhan panel,Zhuhai panel,and COW panel separately with adjustment of age,gender,BMI,drinking,working years(for COW panel only),regular exercise,and blood cell counts.Inverse-variance meta-analysis was performed to combine the results of associations between DNA methylation and HRV of Wuhan panel,Zhuhai panel,and COW panel.For the significant Cp G sites associated with urinary 2-OH-Nap and total OH-PAHs,we further applied multiple linear regression models with adjustment for age and gender to test the correlations of DNA methylation levels with the m RNA levels of the corresponding genes in the 99 participants of Shiyan panel.To identify plausible pathways associated with urinary OH-PAHs associated Cp G sites,we performed Gene Ontology(GO)pathway analysis associated with urinary 2-OH-Nap and total OH-PAHs.Result: In the present study,compared with individuals in the lowest group of urinary 2-OH-Nap,participants in the highest group had lower TP(P for trend = 0.036),lower SDNN(P for trend = 0.105)and lower LF(P for trend = 0.169)among participants from COW.However,we did not observe significant association of urinary2-OH-Nap or total OH-PAHs with HRV indices in participants from Wuhan panel and Zhuhai panel(all P > 0.05).The methylation levels of urinary 2-OH-Nap associated cg18223625,cg07805771 and total OH-PAHs associated cg09235308 were all negatively associated with HRV.We found each 1-unit increase in inverse-normal transformed methylation level of cg18223625 was associated with 5.77% lower SDNN[β(95% CI)=-5.77%(-11.29%,-0.25%)],7.58% lower r-MSSD [β(95% CI)=-7.58%(-14.12%,-1.03%)],17.09% lower HF [β(95% CI)=-17.09%(-34.17%,-0.02%)],and 13.51% lower TP [β(95% CI)=-13.51%(-25.45%,-1.57%)](all P <0.05).In addition,cg18223625 was also negatively associated with LF [β(95% CI)=-7.90%(-20.89%,5.10%),P = 0.234].Each 1-unit increase in inverse-normal transformed methylation level of cg07805771 was associated with 7.51% lower SDNN[β(95% CI)=-7.51%(-13.17%,-1.86%)],17.78% lower HF [β(95% CI)=-17.78%(-35.33%,-0.22%),and 17.70% lower TP [β(95% CI)=-17.70%(-29.92%,-5.47%)(all P < 0.05).Besides,cg07805771 was also negatively associated with r-MSSD [β(95% CI)=-6.39%(-13.11%,0.33%),P = 0.063] and LF [β(95% CI)=-8.45%(-22.03%,5.14%),P = 0.223].Furthermore,we found each 1-unit increase in inverse-normal transformed methylation level of cg09235308 was associated with3.31% lower SDNN [β(95% CI)=-3.31%(-9.17%,2.55%),P = 0.268],4.33% lower r-MSSD [β(95% CI)=-4.33%(-11.33%,2.68%),P = 0.226],1.54% lower LF [β(95% CI)=-1.54%(-15.33%,12.26%),P = 0.827],13.61% lower HF [β(95% CI)=-13.61%(-31.70%,4.49%),P = 0.141],and 13.49% lower TP [β(95% CI)=-13.49%(-25.99%,-0.99%),P = 0.034].When investigating the correlation of DNA methylation with m RNA levels,no statistically significant correlation was found between the methylation levels of the identified Cp G sites and the m RNA levels of their annotated or adjacent genes(all P > 0.05).Gene Ontology pathway analysis identified that 2-OH-Nap associated top Cp G sites annotated to the genes that were involved in the inflammatory processes,sinoatrial node cell action potential,and sinoatrial node cell to atrial cardiac muscle cell signaling(P < 0.001),while the top Cp G sites associated with total OH-PAHs were enriched for transmission of nerve impulse,and inflammatory response(P < 0.001).Conclusions: The methylation levels of cg18223625 and cg07805771 that associated with urinary 2-OH-Nap,and cg09235308 associated with total OH-Nap were inversely associated with HRV indices among adult non-smokers.Despite no significant association was found between the methylation levels of the identified Cp G sites and the expression levels of their annotated or adjacent genes,we found the top Cp G sites associated with urinary 2-OH-Nap and total OH-PAHs annotated to the genes that were involved in the inflammatory processes,sinoatrial node cell action potential,sinoatrial node cell to atrial cardiac muscle cell signaling,and nerve impulse.These findings suggested that DNA methylation changes associated with PAHs exposure may play a potential role in HRV alternation,which providing mechanism evidence for the relationship of PAHs exposure with subclinical cardiovascular disease.