Multilocus Genetic Association Study On Candidate Genes Of Non-alcoholic Fatty Liver Disease And Its Spleen-yang Deficiency Syndrome

Objective: The aim of this study was to explore the genes polymorphisms and genetic biomarkers susceptible for NAFLD and the spleen-Yang deficiency pattern(SYDP,the basic and core pattern of NAFLD in TCM),and to develop a risk prediction model for NAFLD,based on a candidate gene study using multiple SNPs.Methods: A community-based case-control study was conducted among Chinese Han subjects since April 2017 to July 2018 in Zhangjiang community of Pudong New District in Shanghai,China.This study included 341 NAFLD subjects(163 SYDP subjects and 178 non-SYDP subjects),341 healthy control subjects and 332 subjects for validation set.A total of 22 SNPs in 11 genes reported by GWAS or candidate gene studies were genotyped using the Sequenom Mass ARRAY system.The statistical analyses included genetic association study,LD and haplotype analyses,logistic regression model,ROC analyses and trend analyses among others.All statistical analyses were performed using SPSS 21.0,SHEsis,Haploview 4.2,Med Calc 17.1.0 and related bioinformatics databases.Results: For all SNPs,the allele frequency distribution in our study were comparable to the overall frequency in East Asians reported in The 1000 Genomes Project and there was no significant deviation from HWE in control group except for the SNP failed to be designed.The subjects enrolled in this study were representative.1.The results of the candidate gene study of NAFLD showed:(1)The rs1260326,rs780094 and rs780093 in GCKR,and rs626283 and rs641738 in MBOAT7 demonstrated significant differences in allele frequency distribution(P <0.05);The rs2294918 in PNPLA3,and rs1260326,rs780094 and rs780093 in GCKR demonstrated significant differences in genotype frequency distribution(P < 0.05);(2)Genotype association analyses showed rs1260326,rs780094 and rs780093 in GCKR,and rs641738 in MBOAT7 all increased NAFLD risk and all the risk alleles were T.After adjustment for gender,age,BMI and current smoking,rs1260326,rs780094 and rs780093 all increased NAFLD risk in codominant,dominant,recessive and additive genetic models(P < 0.05);and rs641738 increased NAFLD risk in additive genetic model(P < 0.05);(3)The LD analyses showed rs1260326-rs780094-rs780093 was identified as a strong LD block.Haplotype analyses showed the haplotype C C C decreased NAFLD risk(P <0.05),and the haplotype T T T increased NAFLD risk(P < 0.05).2.The results of the risk prediction model for NAFLD combining SNPs and metabolic risk factors showed:(1)The additive effects of the sum of risk alleles of rs780094 in GCKR and rs641738 in MBOAT7 suggested that the prevalence of NAFLD showed an increasing trend with the increase in the number of risk alleles(P for trend< 0.05);After adjustment for confounding factors,the risk and OR for NAFLD also showed an increasing trend with the increase in the number of risk alleles(P for trend< 0.05);(2)A total of four metabolic risk factors for NAFLD were selected using univariate and multivariate analyses(stepwise regression),including BMI,WC,FPG and TG.Then a“NCI” formula calculated to predict NAFLD risk was established using logistic regression model;(3)ROC analyses showed the “NCI + SNP”,combined of additive effects of SNPs and metabolic risk factors,had the largest AUCs to predict NAFLD in both training set and validation set.In validation set,the AUC of NCI + SNP was 0.757,and the statistical significance of the AUCs differences showed NCI + SNP was larger than NCI,HSI,Ty G and LAP(P < 0.05).3.The results of the genetic biomarkers of SYDP of NAFLD showed:(1)Between SYDP group and control group: the rs1260326,rs780094 and rs780093 in GCKR demonstrated significant differences in both allele and genotype frequency distribution(P < 0.05).Genotype association analyses showed the three SNPs all increased SYDP risk and all the risk alleles were T;After adjustment for confounding factors,the three SNPs all increased SYDP risk in codominant,dominant,recessive and additive genetic models(P < 0.05);(2)Between non-SYDP group and control group: the three SNPs demonstrated no deviation in either allele or genotype frequency distribution;After adjustment for confounding factors,there was still no deviation in all the five genetic models(P >0.05).Between SYDP group and non-SYDP group: all the three SNPs demonstrated significant differences in both allele and genotype frequency distribution;After adjustment for confounding factors,the three SNPs were associated with SYDP in related genetic models(P < 0.05);(3)The LD analyses showed the three SNPs was identified as a strong LD block.Haplotype analyses showed the haplotype C C C decreased SYDP risk(P < 0.001),and the haplotype T T T increased SYDP risk(P < 0.001);(4)The association of the three SNPs and the degree of SYDP suggested that the allele C showed a decreasing trend with the increase of the degree of SYDP(P for trend<0.05),and the allele T showed no trend(P for trend> 0.05).The NAFLD-SYDP subjects with the mutant type CT-TT of rs1260326 and rs780094 had higher SYDP scores than the subjects with the wild type CC(P < 0.05),but there was no deviation in rs780093(P > 0.05).Conclusions: 1.The rs1260326,rs780094 and rs780093 in GCKR,and rs641738 in MBOAT7 were associated with NAFLD in Chinese Han subjects.2.The risk prediction model combining of additive effects of rs780094 in GCKR and rs641738 in MBOAT7 with metabolic risk factors had a high predictive value for predicting NAFLD.3.The rs1260326,rs780094 and rs780093 in GCKR were identified as the genetic biomarkers of SYDP of NAFLD,and the three SNPs were associated with the degree of SYDP.