Plasma Metal Concentrations,Serum Metabolite Levels,and Their Associations With Risk Of Incident Acute Coronary Syndrome

Acute coronary syndrome(ACS)refers to a group of clinical syndromes caused by acute myocardial ischemia and is considered as the most severe clinical type of coronary heart disease.The occurrence and development of ACS are the results of interactions between genetic factors and environmental factors,but the association between metal exposure and ACS risk and its potential mechanisms still need to be explored.Metabolomics technology can comprehensively detect hundreds to thousands of small molecular metabolites in organisms.Therefore,exploring metabolic changes caused by metal exposure using metabolomics technology can help to elucidate the potential mechanisms underlying the health effects caused by metal exposure.However,current metabolomics studies related to metal exposure mostly explored the associations of exposure to individual metal with metabolites among special populations such as pregnant women,and the associations between exposure to multiple metals and metabolites in the general population remain to be further investigated.In addition,the application of metabolomics technology to comprehensively analyze the metabolic features related to diseases can help to elucidate the etiology and mechanisms of diseases.However,no prospective metabolomics study has explored the associations of circulating metabolites with ACS risk.Mendelian randomization(MR)approach can be used to examine causality,but the scarcity of genomewide association studies(GWAS)on circulating metabolites has limited the application of MR method to explore the causality of metabolites with ACS risk.Moreover,the combination of metabolomics technology and mediation analysis methods is helpful to investigate the mediation effects of metabolites on the associations between environmental exposure and health outcomes,which will further facilitate the understanding of the pathogenesis of complex environment-related diseases.However,no research has been conducted to explore the mediating role of metabolite in the association between metal exposure and ACS risk.Given the above background,this study is based on the prospective Dongfeng-Tongji cohort.Firstly,we conducted a cross-sectional study to investigate the associations of plasma metal concentrations with serum metabolite levels and discover metal-associated metabolic pathways.Then we conducted a nested case-control study to investigate the associations between serum metabolite levels and risk of incident ACS,identify incident ACS-associated metabolites and further examine the causality of the identified associations between metabolites and ACS risk.Finally,we conducted a nested case-control study to investigate the associations between plasma metal concentrations and risk of incident ACS,and further explore the mediating roles of incident ACS-associated metabolites in the associations between plasma metal concentrations and risk of incident ACS.This study consists of the following three parts:Part 1.Associations of plasma metal concentrations with serum metabolite levelsObjectives:We aim to investigate the associations of multiple plasma metals with serum metabolite levels and discover metal-associated metabolic pathways,in order to provide a scientific basis for further elucidating the mechanisms underlying the associations between metal exposure and metabolic diseases.Methods:We conducted a cross-sectional study among 1992 participants from the baseline population of the Dongfeng-Tongji cohort who were recruited in 2008.We measured plasma concentrations of 23 metals using inductively coupled plasma-mass spectrometry(ICP-MS),and excluded three metals with detection rates less than 50%and three other metals which were not well-represented by plasma level for exposure assessment.Finally,17 plasma metals were included in subsequent analysis.We performed serum untargeted metabolomic profiling using ultra performance liquid chromatography-mass spectrometry(UPLC-MS).After matching to authentic chemical standards database and quality control,189 annotated metabolites were retained for subsequent analysis.We used multivariable linear regression models to investigate the associations between individual metal exposure and metabolite levels,and then performed metabolic pathway enrichment analysis.We also applied twoway orthogonal partial least squares(O2PLS)and Bayesian kernel machine regression(BKMR)to explore the associations between mixed metal exposure and metabolites.Results:When false discovery rate(FDR)<0.05 was considered statistically significant,we characterized the metabolic profiling of each metal,and we found that seven metals were associated with the biosynthesis of unsaturated fatty acids pathway(As,Ba,Co,Pb,Ti,V,and Zn),two metals with the linoleic acid metabolism pathway(As and Pb),two metals with the alpha-linolenic acid metabolism pathway(Ba and Pb),and one metal with the aminoacyl-tRNA biosynthesis pathway(Ni).Moreover,O2PLS analysis identified that five metabolites,including aspartylphenylalanine,tetradecenoic acid(free fatty acid[FFA]14:1),uridine,Carnitine C14:2 and lysophosphatidylcholine(LPC)18:2,contributed most to the joint covariation between metal and metabolite data matrices(SSjoint=12.3%,8.3%,8.0%,7.4%,and 7.3%,respectively).Two-stage BKMR analysis found that the combined exposure of aluminum,arsenic,barium,and zinc was significantly positively associated with aspartylphenylalanine level,with a synergistic interaction between arsenic and barium(P for interaction=0.048);the combined exposure of barium,cobalt,manganese,and lead was significantly positively associated with Carnitine C14:2 level,with a synergistic interaction between barium and lead(P for interaction<0.001).Conclusions:Metal exposure is closely associated with changes in serum metabolic profiling,mainly involving metabolic pathways including biosynthesis of unsaturated fatty acids,linoleic acid metabolism,alpha-linolenic acid metabolism,and aminoacyl-tRNA biosynthesis.Part 2.Associations between serum metabolite levels and risk of incident acute coronary syndromeObjectives:To identify metabolites associated with risk of incident ACS and explore the causality of the associations.Methods:We performed serum non-targeted metabolomics using UPLC-MS in a nested case-control study within the Dongfeng-Tongji cohort,including 500 incident ACS cases during the follow-up from the baseline survey in 2008 to the end of 2013,and 500 controls matched 1:1 by age(±3 years)and sex.Firstly,we applied penalized logistic regression model with elastic-net regularization to select metabolites associated with risk of incident ACS.Subsequently,we performed genome-wide association studies(GWAS)on the incident ACS-associated metabolites in the Dongfeng-Tongji cohort to verify single nucleotide polymorphism(SNP)sites previously reported in the literature and determined the genetic instruments.Finally,based on the selected incident ACS-associated metabolites,we carried out one-sample Mendelian randomization(MR)analysis to examine the causality between these metabolite levels and ACS risk,as well as two-sample MR analysis to examine the potential causal associations of these metabolite levels with cardiovascular risk factors.Results:After adjustments for age,sex,BMI,smoking,drinking,physical activity and family history of CHD,we observed that aspartylphenylalanine and tetracosanoic acid levels were significantly positively associated with risk of incident ACS,while 1,5-anhydroD-glucitol level was significantly negatively associated with risk of incident ACS(all Ptrend<0.05).Compared with the lowest quartile,the odds ratios(OR)and 95%confidence intervals(CIs)of the highest quartile were 2.02(1.38,2.93)for aspartylphenylalanine,0.57(0.38,0.87)for 1,5-anhydro-D-glucitol,and 1.82(1.22,2.71)for tetracosanoic acid,respectively.After further adjustments for conventional cardiovascular risk factors such as blood pressure,blood lipids,and fasting glucose as well as use of antihypertensive,anti diabetic and lipid-lowering medications,we still observed significant positive associations of aspartylphenylalanine and tetracosanoic acid levels with risk of incident ACS(P-trend=0.015 and 0.034,respectively),whereas the negative association of 1,5anhydro-D-glucitol level with risk of incident ACS was attenuated to the null and not independent of fasting glucose.Besides,we found that the negative association between 1,5-anhydro-D-glucitol level and risk of incident ACS was causal(P<0.05)before adjusting for fasting glucose,and the positive associations of aspartylphenylalanine level with hypertension(P<0.05)and hypertriglyceridemia(P=0.077)were causal in two-sample MR analysis.Conclusions:Our study identified aspartylphenylalanine as positively associated with risk of incident ACS,and replicated 1,5-anhydro-D-glucitol and tetracosanoic acid which had been previously reported to be negatively and positively associated with cardiovascular disease(CVD)risk respectively.These findings may provide new clues to the etiological study of ACS.Part 3.The mediation roles of serum metabolite levels in the associations between plasma metal concentrations and risk of incident acute coronary syndromeObjectives:To investigate the associations of plasma metal concentrations with risk of incident ACS,and further explore the mediation roles of incident ACS-associated metabolites in the associations between plasma metal concentrations and risk of incident ACS.Methods:The study population of this study is based on the nested case-control study in Part 2,and the detection methods of plasma metal concentrations and serum metabolome are the same as those in Part 1.We first used multivariable conditional logistic regression models to investigate the associations between plasma metal concentrations and risk of incident ACS.Then we applied generalized linear models to explore the relationships between incident ACS-associated plasma metals and the three metabolites identified to be associated with risk of incident ACS in Part 2,including aspartylphenylalanine,1,5anhydro-D-glucitol,and tetracosanoic acid.Finally,we performed mediation analysis to evaluate the mediation roles of serum metabolite levels in the associations between plasma metal concentrations and risk of incident ACS.Results:Plasma concentrations of aluminum,arsenic,barium,lead,manganese,and titanium were significantly positively associated with risk of incident ACS(all Ptrend<0.05).Compared with the lowest quartile,the ORs and 95%CIs corresponding to the highest quartile were 2.13(1.37,3.30)for aluminum,1.65(1.08,2.52)for arsenic,1.77(1.16,2.71)for barium,2.06(1.34,3.19)for lead,1.72(1.11,2.66)for manganese,and 1.60(1.01,2.53)for titanium,respectively.In addition,plasma concentrations of aluminum,arsenic,barium,lead,manganese,and titanium were significantly positively associated with serum aspartylphenylalanine level(all P<0.001),and plasma concentration of titanium was significantly positively associated with serum tetracosanoic acid level(P<0.05).Mediation analysis showed that aspartylphenylalanine mediated 24.90%,21.40%,8.98%,and 19.43%of the associations of plasma barium,lead,manganese,and titanium with ACS risk respectively(all P<0.05).Conclusions:Plasma concentrations of aluminum,arsenic,barium,lead,manganese,and titanium were significantly positively associated with risk of incident ACS,and aspartylphenylalanine level might partially mediate these associations.