Studies On The Multimodal Associations Between Low-Density Lipoprotein Cholesterol And Stomach Cancer Risk And The Mechanisms Of Anti-Stomach Cancer Effects Through Targeting Susceptibility Gene LPCAT3

Background:The morbidities and mortalities of stomach cancer are at the forefront of malignancies.In spite of amazing advancements in treatments,delayed diagnosis and high recurrence rate of stomach cancer prompt poor prognoses.Previous epidemiological studies have shown that lipidomic biomarkers such as high-density lipoprotein cholesterol（HDL-C）,low-density lipoprotein cholesterol（LDL-C）,and triglyceride（TG）are associated with the risk of digestive system cancers,and that targeting their related transcriptional susceptibility genes to inhibit tumor growth.Because of reverse causality and confounding bias,phenotypic association results from traditional observational studies are often uncertain.Thus,genetic association evidences make it possible to overcomes the limitations of traditional observational studies.In this study,the following two questions are primarily addressed:1.Systematical explorations of phenotypic and genetic association patterns between signature lipidomic biomarkers（HDL-C,LDL-C,and TG）and stomach cancer risk,and identification of the significant genetic susceptibility gene related to signature lipidomic biomarkers with causal effects on the risk of stomach cancer risk.2.Exploration of the mechanisms of targeting related susceptibility genes exerting anti-stomach cancer effect.Part 1 Signature lipidomic biomarkers and the risk of stomach cancer risk:a prospective cohort study and multi-omics Mendelian randomization analysis Objective:To explore the phenotypic and genetic associations between signature lipidomic biomarkers（HDL-C,LDL-C and TG）and stomach cancer risk,to further clarify the indicative roles of signature lipidomic biomarkers on stomach cancer risk in the given population,and to distinguish the significant genetic susceptibility gene related to signature lipidomic biomarkers with causal effects on the risk of stomach cancer risk using multi-omics summary-data-based Mendelian randomization（SMR）analyses.Methods:1.Phenotypic association pattern analyses:In this study,based on the large-scale prospective cohort study of UKB,participants were conducted rigorously screened,and 319,568 participants were identified for this study.Hazard ratios（HRs）and 95%confidence intervals（CIs）of signature lipidomic biomarkers on the risk of stomach cancer derived from the multivariable-adjusted Cox proportional hazards regression model.Restricted cubic splines（RCSs）were further represented to estimate nonlinear associations between signature lipidomic biomarkers and the risk of stomach cancer.In addition,whether estimates of signature lipidomic biomarkers on the risk of stomach cancer were intervened by age and sex was examined with age-and sex-specific subgroup analyses by subdivisions of signature lipidomic biomarkers（HDL-C:<1.0 mmol/L,1.0-1.6 mmol/L,≥1.6 mmol/L;LDL-C:<3.4 mmol/L,3.4-4.1 mmol/L,≥4.1 mmol/L;TG:<1.7 mmol/L,1.7-2.2 mmol/L,≥2.2 mmol/L）with reference of Singapore MOH Clinical Practice Guidelines 12/2016 and the results of RCSs.2.Calculation of polygenic risk score of signature lipidomic biomarkers:Summary-level European-ancestry meta-analysis GWAS datasets of HDL-C,LDL-C,and TG without UKB samples retrieved from The Global Lipids Genetics Consortium were defined as based data,and rigorous quality control procedures were applied to extract independent single nucleotide polymorphisms（SNPs）to be used to calculate polygenic risk scores（PRSs）of signature lipidomic biomarkers.Individual-level phenotypic and genotype data of signature lipidomic biomarkers based on European-ancestry UKB samples was used as target data in order to avoid substantial inflation of PRS-phenotype relationship due to sample overlap between target data and base data.PRSs of signature lipidomic biomarkers were calculated with the average method in the PRSice-2 software（Linux version）.3.Genetic association pattern analyses:With the evidence of linear or nonlinear phenotypic associations between signature lipidomic biomarkers and the risk of stomach cancer,we supposed that nonlinear MR analyses causally assessed the nonlinear phenotypic associations,otherwise,linear MR analyses were primarily used to assess the potential linear causal associations for linear or nonlinear phenotypic associations.For nonlinear MR analyses,participants were stratified into three subgroups by residual of signature lipidomic biomarkers.Then,piecewise linear method was utilized to estimate causal associations between signature lipidomic biomarkers and the risk of DSCs in each stratum（local average causal effects,LACE）,whose nonlinearity was assessed with Quadratic test and the Cochran Q test.Linear MR analyses using ratio of coefficients method were conducted for two-stage MR analyses.At the same time,several sensitivity analyses were used to verify the robustness of nonlinear and linear causal association results.In addition,age-and sex-specific stratified MR analyses were conducted to quantify LACEs within the residuals of above three categories of signature lipidomic biomarkers.4.Determination of the related susceptibility genes using multi-omics summary-data-based Mendelian randomization analyses:Base on summarization above MR results that genetically predicated LDL-C concentration linearly causally associated with the risk of stomach cancer,multi-omics SMR analyses based on using LDL-related cis-e QTL and cis-m QTL datasets were utilized to further investigate underlying molecular pleiotropic associations between LDL-C characterized by genetic susceptibility in LDL-C related genes and the risk of stomach cancer,and Bayesian inference-based colocalization was used to estimate the posterior probability of shared SNPs.Above results determined the potential target for the treatment and prevention of stomach cancer.Results:1.Phenotypic association patterns between signature lipidomic biomarkers and stomach cancer risk:An independent linear negative phenotypic association was demonstrated between HDL-C and stomach cancer risk(P_nonlinearity>0.05,P_overall<0.05),and an independent nonlinear phenotypic association was observed between LDL-C concentration and stomach cancer risk(P_nonlinearity<0.05,P_overall<0.05).Sensitivity analyses still showed robust results after excluding participants within the first two-year follow-up time,additionally adjusting for additional medication use,and adjusting for other signature lipidomic biomarkers concentrations.Furthermore,in age-and sex-specific subgroup analyses,there were no interactions between signature lipidomic biomarkers and age or sex on the risk of stomach cancer(P_interaction>0.05).For participants aged 60 years or older,the categories of HDL-C<1.0 and≥1.6 mmol/L were associated with higher and lower risks of stomach cancer（P<0.05）.Meanwhile,the category of LDL-C≥4.1 mmol/L showed a negative correlation with the risk of stomach cancer（P<0.05）.For female participants,the categories of HDL-C<1.6 and LDL-C≥4.1 mmol/L were respectively positively and negatively correlated with the risk of stomach cancer（P<0.05）.For male participants,the category of HDL-C≥1.6mmol/L was associated with a lower risk of stomach cancer（P<0.05）.2.Genetic association patterns between signature lipidomic biomarkers and stomach cancer risk:Results of nonlinear and linear MR analyses overall suggested that genetically predicted LDL-C concentration had a linearly negatively causal effect on the risk of stomach cancer（Quadratic P=0.901,Cochran Q P=0.434;HR:0.340（0.137-0.843）,P=0.020）,and there were no causal associations between other signature lipidomic biomarkers and the risk of stomach cancer.Even though in sensitivity analyses with additional adjustments of potential confounders and other MR methods,the shapes of causal associations and causal estimates of signature lipidomic biomarkers on the risk of stomach cancer remained consistent with the findings of primary nonlinear MR analyses and linear MR analyses.In age-and sex-specific stratified MR analyses,integrating linear and nonlinear MR analyses indicated that there were no nonlinear causal patterns between signature lipidomic biomarkers and the risk of stomach cancer in all age and sex groups（Quadratic P>0.05,Cochran Q P>0.05）,but linear negative causal associations were observed between LDL-C and the risk of stomach cancer in participants aged 60 years or older and female participants.Results of age-and sex-stratified MR analyses showed that the category of LDL-C≥4.1 mmol/L had inverse causal associations with the risk of stomach cancer in participants aged 60 years or older and female participants.3.LPCAT3,as the significant LDL-C-related susceptibility gene,is positively causally associated with the risk of stomach cancer:Using 104 090 SNPs(P_SNP-LDL-C<5×10^-8)related to the expression of 246 LDL-C-related gene transcripts,and 149575 SNPs(P_SNP-LDL-C<5×10^-8)related to 1113 LDL-C-related methylated Cp G sites as GIVs,a key LDL-C-related susceptibility gene,LPCAT3,was identified that was causally associated with the risk of stomach cancer.Both transcriptome and DNA methylation levels confirmed that LPCAT3 expression was causally positively associated with risk of stomach cancer,and the results of colocalization analysis suggested that LPCAT3 expression shared causal variations with stomach cancer.Conclusions:1.Genetic association evidence showed that there was a negative linear causal correlation between LDL-C and the risk of stomach cancer.Although phenotypic association evidence supported a linear negative association between HDL-C and stomach cancer risk,no genetic association evidences suggested a linear or nonlinear causal association between genetically predicted HDL-C and stomach cancer risk.Phenotypic and genetic association evidences both showed that LDL-C≥4.1mmol/L decreased the risk of stomach cancer in female participants and participants aged 60years or older2.The multi-omics SMR analyses based on transcriptome and epigenetics suggested that LPCAT3 could be regarded as a significant LDL-C-related susceptibility gene that was positively causally associated with the risk of stomach cancer,and its expression shared causal variation with stomach cancer.Part 2 The effects of targeting LDL-C-related susceptibility gene LPCAT3 on the biological behaviors of stomach cancerObjective:To explore the feasibility of LDL-C-related significant susceptibility gene LPCAT3 as a direct target of indole-3-carbinol（I3C）,and to explore the mechanisms of anti-stomach cancer effects through I3C targeting susceptibility gene LPCAT3 with in vivo and in vitro experiments.Methods:1.Bioinformatics analyses of LPCAT3:Firstly,based on TCGA database and GTEx database（414 stomach cancer tissues vs.210 normal or paracancerous stomach tissues）,the differences of LPCAT3 expression between stomach cancer and normal or paracancerous stomach tissues were compared from m RNA level,and the diagnostic value of LPCAT3 expression was determined with diagnostic receiver operating characteristic（ROC）curve.The distribution differences in clinical characteristics（age,sex,pathological stage,pathological TNM-staging,and histological grade）and the differences in overall survival（OS）,disease specific survival（DSS）,and progression free interval（PFI）were also analyzed between high and low expression of LPCAT3.2.The effects of indole-3-carbinol on the biological behavior of stomach cancer:In vitro experiments（CCK-8 test,clone formation test,scratch test,transwell test,Annexin V/PI double staining method and western blotting）were used to investigate the effects of different indole-3-carbinol（I3C）concentrations（0,100,200,300,and 400μM）on the proliferation,migration,invasion,and apoptosis of stomach cancer cell lines;In vivo experiment（stomach cancer xenograft model in nude mouse）was utilized to explore the effect of I3C on the growth of stomach cancer.3.Effects of targeting regulation of the significant LDL-C-related susceptibility gene LPCAT3 by indole-3-carbinol on the biological behaviors of stomach cancer:The transcriptome sequencing analysis was used to identify the expression of LPCAT3 in I3C-treated stomach cancer cells.The q RT-PCR experiment was used to detect the expression of LPCAT3 expression level in stomach cancer cells treated with different I3C concentrations（0,100,200,300,and 400μM）.Moreover,rescue experiments were performed in overexpressed LPCAT3 stomach cancer cells,and CCK-8 test,Annexin V/PI double staining,and western blotting were used to detect the proliferation,apoptosis,and the expression of apoptosis-related proteins Bax and Bcl-2 in stomach cancer cells among the following 4 groups:0μM I3C+NC group,200μM I3C+NC group,0μM I3C+LPCAT3 overexpression group,and 200μM I3C+LPCAT3 overexpression group.Results:1.LPCAT3 was abnormally highly expressed in stomach cancer tissue,and had a favorable diagnostic value and was related to pathological T-staging.Results of bioinformatic analyses showed that LPCAT3 expression in stomach cancer was significantly higher than that in normal or paracancerous tissues,and LPCAT3 m RNA and protein expressions were relatively highly expressed in stomach cell lines AGS and HGC-27 compared with that in the normal stomach mucosal epithelial cell line GES-1.Meanwhile,the diagnostic ROC curve showed that the higher LPCAT3 had a good diagnostic value for stomach cancer.Although there were no significant differences among OS,DSI,and PFI between high and low LPCAT3 expression,results of clinical correlation analysis suggested that the high expression of LPCAT3 was related to the advanced pathological T-staging.2.Indole-3-carbinol inhibited the proliferation,migration,invasion,and apoptosis of stomach cancer cells,and inhibited the growth of stomach cancer xenograft:The results of CCK-8 test,clone formation test,scratch test,transwell test,Annexin V/PI double staining method,and western blotting of treatments with increasing I3C concentrations for stomach cancer cells showed that the proliferations,migrations,and invasion abilities of stomach cancer cells decreased,and the apoptosis rates increased,and the expression of pro-apoptotic protein Bax increased,and apoptosis inhibitor protein Bcl-2 decreased in concentration-dependent manners.The result of stomach cancer xenograft model in nude mouse suggested that I3C significantly inhibited the growth of stomach cancer xenograft in nude mice.3.Targeted regulation of LPCAT3 by indole-3-carbinol inhibited proliferation and induced apoptosis of stomach cancer cells:The results of the transcriptome sequencing analysis showed that the expression of LPCAT3 was significantly down-regulated in the I3C-treated stomach cancer cell.Meanwhile,results of q RT-PCR showed that LPCAT3 expression was down-regulated in the I3C-treated stomach cancer cells with concentration-dependent manners.Moreover,results of the rescue experiment suggested that LPCAT3 overexpression significantly reversed the effects of I3C on proliferation inhibition and apoptosis induction in stomach cancer cells.Conclusions:1.Higher LPCAT3 expression was observed in stomach cancer.Although there was no significant difference of prognosis between samples with high and low LPCAT3expression,the LPCAT3 expression was abnormally higher in advanced pathological T-staging,and exhibited a good diagnostic accuray in stomach cancer.2.I3C inhibited the growth of stomach cancer xenografts,and significantly inhibited the proliferation,migration,invasion,and induction of apoptosis in stomach cancer cells,indicating that it might directly bind LPCAT3 and inhibit LPCAT3 activity to act an anti-stomach cancer agent.Overall conclusions:1.There was a linear negative causal correlation between LDL-C and the risk of stomach cancer,and a higher LDL-C concentration reduced the risk of stomach cancer,the effect that is more noticeable in female participants and participants aged 60 or older.2.The causal association between LDL-C characterized by genetic susceptibility in all LDL-C-related genes and stomach cancer risk was concluded by multi-omics SMR analyses,and LPCAT3 was identified as the significant LDL-C related susceptibility gene,whose expression was positively causally associated with the risk of stomach cancer.3.LPCAT3 was abnormally highly expressed in stomach cancer and associated with the advanced pathological T-staging.Furthermore,I3C could directly bind LPCAT3 and inhibit LPCAT3 activity to exert an anti-stomach cancer effect.