Associations Of Plasma Long Non-coding RNAs With Incident Acute Coronary Syndrome And Its Influencing Factors

Coronary heart disease（CHD）is the major cause of death worldwide,while acute coronary syndrome（ACS）is the most severe and life-threatening subtype of CHD.In recent years,the incidence and mortality of ACS continue to increase.The steady increase subsequently leads to an ever-increasing disease burden and a serious public health concern that need to be addressed.However,the development of ACS involves multiple genetic and environmental factors,and the underlying mechanism remains to be explored.Emerging evidence suggested that long non-coding RNAs（lnc RNAs）,a type of important epigenetic regulator,might link the environmental factors to disease development.Lnc RNAs may be involved in the pathological processes of ACS through multiple molecular mechanisms.Lnc RNAs are relatively stable in peripheral blood,and their expression levels differ at different stages of disease development.However,most of the evidence was derived from case-control studies.These studies were limited by small sample sizes,without fully control of confounding factors.Moreover,it is lack of prospective study investigating the associations of circulating lnc RNAs with incident ACS.Therefore,it is urgent to prospectively investigate the associations between circulating lnc RNAs and incident ACS.In addition,environmental pollutants including metals/metalloids;unfavorable lifestyles including smoking,drinking,physical inactivity,and prolonged or insufficient sleep duration;individuals’health conditions including obesity,hypertension,hyperlipidemia,and diabetes;and occupational exposures including shift work were linked with adverse cardiovascular outcomes.However,limited studies have evaluated the associations of these cardiovascular risk factors with circulating lnc RNAs.Furthermore,it is lack of studies to explore the mediation role of circulating lnc RNAs on the associations between cardiovascular risk factors and incident ACS.Therefore,in this prospective nested case-control study,we aimed to discover and validate plasma lnc RNAs that were significantly related with incident ACS,to estimate the associations of cardiovascular risk factors with the identified lnc RNAs,and further to explore the roles of plasma lnc RNAs in the associations of cardiovascular risk factors with incident ACS.This dissertation includes the following two parts:PartⅠAssociations of plasma lnc RNAs with incident acute coronary syndromeObjective:To discover and validate plasma lnc RNAs associated with incident ACS in a prospective nested case-control study,and predict the possible functions of the identified plasma lnc RNAs.Methods:The nested case-control study was based on the Dongfeng-Tongji cohort,where 38295 subjects were enrolled in 2013.A total of 595 incident ACS cases were identified until the end of 2016 during the follow-up period.Each ACS case was randomly matched to a control on sex,age（within 3 years）,as well as the date of blood collection（within 1 month）.A total of 23 incident ACS and 23 matched controls were randomly selected as discovery population.We profiled plasma levels of lnc RNAs and m RNAs by microarray using Agilent human lnc RNA+m RNA Array V4.0.Differentially expressed lnc RNAs and m RNAs were selected as candidate lnc RNAs and m RNAs through Wilcoxon signed-rank test.We performed absolute quantification of the candidate lnc RNAs using droplet digital polymerase chain reaction（dd PCR）in the plasma samples of other 572 incident ACS and 572 matched controls.After excluding 14 pairs of incident ACS case-control samples that failed quality control,a total of 558 incident ACS and 558matched controls were included in the final analyses.The associations between candidate lnc RNAs and incident ACS were modeled using multivariable-adjusted conditional logistic regression analyses.To predict the possible functions of the validated lnc RNA,we identified candidate m RNAs which were related to the validated lnc RNA and included them in the Gene Ontology（GO）enrichment analysis.Results:In the discovery stage,based on the screening criteria including fold change>2or<0.5,and Benjamini-Hochberg corrected false discovery rate（FDR）<0.05,we identified five significantly down-regulated lnc RNAs as candidate lnc RNAs（MIR4435-2HG,AL731557,NEAT1,AC008014,and TSPOAP1-AS1）.In the validation stage,among the five candidate lnc RNAs,we observed that higher levels of plasma MIR4435-2HG were associated with a lower risk of incident ACS.Compared with those in the lowest quartile of MIR4435-2HG levels(<8.89×10^-2),the odds ratio（OR）and 95%confidence interval（CI）for subjects in the highest quartile(>13.76×10^-2)was 0.64(95%CI:0.45,0.92;P_trend=0.02).No significant associations were observed between the other four lnc RNAs and incident ACS(all P_trend>0.05).Results of GO enrichment analysis suggested that m RNAs correlated with MIR4435-2HG were primarily enriched in platelet activation and coagulation processes.Conclusions:Plasma levels of MIR4435-2HG,which primarily enriched in platelet activation and coagulation processes,were negatively associated with incident ACS.Further studies are warranted to explore the underlying functions of the association between MIR4435-2HG and incident ACS.Part II Mediation analysis of plasma MIR4435-2HG on the associations between cardiovascular risk factors and incident acute coronary syndromeObjective:To investigate the associations of cardiovascular risk factors with plasma MIR4435-2HG levels,and explore whether MIR4435-2HG would mediate the associations between cardiovascular risk factors and incident ACS.Methods:This study was based on 558 pairs of incident ACS case-control samples in the validation stage of Part I.We evaluated the associations of age,gender,lifestyles including smoking,drinking,metabolic equivalent（MET）,sleep duration,midday napping,and tea consumption;individuals’health conditions including body mass index（BMI）,family history of coronary heart disease,hypertension,hyperlipidemia,diabetes,blood pressures,blood lipids,and blood glucose;and occupational exposures including shift work with plasma MIR4435-2HG levels using generalized linear regression models.The associations between plasma metals/metalloids and MIR4435-2HG were performed among the 460 pairs of incident ACS case-control,who had sufficient plasma samples for metals/metalloids measurement.Plasma levels of 18 metals/metalloids were measured by inductively coupled plasma mass spectrometry.We performed generalized linear regression models and the LASSO penalty regression model to assess the associations between plasma metals/metalloids levels and MIR4435-2HG levels.In addition,based on the results of the association analyses between these cardiovascular risk factors and incident ACS,we further estimated whether plasma MIR4435-2HG levels would mediate the associations between cardiovascular risk factors and incident ACS using mediation analyses.Results:The findings suggested that age,BMI,shift work,molybdenum,and rubidium levels were associated with MIR4435-2HG levels.We observed a significant reduction in MIR4435-2HG levels with increasing age（β=-0.004,95%CI:-0.007,-0.001;P=0.01）.Compared with the individuals with normal BMI（18.5-23.9 kg/m²）,MIR4435-2HG levels were lower among overweight individuals with BMI between 24.0-27.9 kg/m²（β=-0.057,95%CI:-0.106,-0.008;P=0.02）and obese individuals with BMI≥28.0 kg/m²（β=-0.109,95%CI:-0.183,-0.034;P=0.004）.Result from the mediation analysis showed that MIR4435-2HG levels mediated 11.30%(P_mediation=0.04)of the association between BMI and incident ACS.Compared with individuals who were free of shift work history,those who experienced more than 10 years of shift work had lower MIR4435-2HG levels（β=-0.058,95%CI:-0.113,-0.004;P=0.04）.In the analyses of the associations between plasma metals/metalloids levels and MIR4435-2HG levels.We found five plasma metals/metalloids（aluminum,arsenic,cobalt,molybdenum,rubidium）were significantly associated with MIR4435-2HG levels in the single metal models.Results from the polymetallic models suggested that molybdenum levels were negatively associated with MIR4435-2HG levels,with theβ（95%CI）per interquartile range（IQR）increase in plasma molybdenum levels being-0.069（-0.098,-0.039）.Meanwhile,rubidium levels were positively associated with MIR4435-2HG levels,with theβ（95%CI）per IQR increase in plasma rubidium levels being 0.042（0.009,0.074）.The associations of molybdenum and rubidium levels with MIR4435-2HG levels in polymetallic models were further validated in the LASSO penalty regression model.In addition,higher levels of molybdenum might attenuate the positive association between rubidium levels and MIR4435-2HG levels(P_interaction=0.002).In further mediation analysis,we did not find the mediation role of plasma MIR4435-2HG levels on the association between rubidium levels and incident ACS(P_mediation>0.05).Conclusions:Age,BMI,and shift work were negatively associated with plasma MIR4435-2HG levels,and MIR4435-2HG partly mediated the association between BMI and incident ACS.Moreover,molybdenum and rubidium levels were negatively and positively associated with MIR4435-2HG levels,respectively.Higher levels of molybdenum might attenuate the positive association between rubidium levels and MIR4435-2HG levels.No significant mediation role of MIR4435-2HG was observed on the association between rubidium levels and incident ACS.These findings provided epidemiological evidence for further elucidating the underlying mechanism between BMI and incident ACS.