Impact Of UGT1A9,UGT1A8,UGT2B7,ABCC2,ABCG2 And SLCO1B3 Genetic Polymorphisms On Mycophenolic Acid Metabolism In Chinese Renal Transplant Recipients

BackgroundMycophenolate mofetil(MMF)is an immunosuppressant that is widely used after organ transplantation.In general,MMF combined with calcineurin inhibitors can prevent allograft rejection in transplant patients.After oral administration,MMF is hydrolyzed to form mycophenolic acid(MPA).MPA is a reversible and uncompetitive inhibitor of inosine 5’-monophosphate dehydrogenase(IMDH)that inhibits synthesis of guanosine nucleotides,which is essential for lymphocytes.By inhibiting proliferative responses of T-and B-lymphocytes,MPA reduces the incidence of renal graft loss due to rejection.MPA usually reaches maximal concentration within the first hour MMF oral administration and achieves second maximal concentration from the sixth to eighth hour after oral MMF administration through enterohepatic circulation.MPA clearance is dominated by glucuronidation and enterohepatic circulation.In the liver,MPA is mainly glucuronized by uridine diphosphate glucuronosyltransferases(UGTs)to>90%inactive mycophenolic acid glucuronide(MPAG)and<10%active mycophenolic acid acyl glucuronide(AcMPAG),which inhibits lymphocyte proliferation.Of the UGTs,UDP-glucuronosyltransferase-2B7(UGT2B7,encoded by the UGT2B7 gene)is mainly responsible for generating AcMPAG.The UDP-glucuronosyltransferase family 1 member A8(UGT1A8,encoded by the UGT1A8 gene)and UDP-glucuronosyltransferase family 1 member A9(UGT1A9,encoded by the UGT1A9 gene)are mainly responsible for generating MPAG.Part of the MPAG is reabsorbed by the organic anion-transporting polypeptide 1B3(OATP1B3,encoded by the SLCO1B3 gene)before excretion into bile.Part of MPAG is secreted in bile via the multidrug resistance-associated protein 2(MRP2,encoded by ABCC2)and the breast cancer resistance protein(BCRP,encoded by ABCG2)and gut bacteria convert it back to MPA with subsequent reabsorption.This process is called enterohepatic circulation of MPA.MPA has a narrow therapeutic range and its pharmacokinetics show large,interindividual variation that is affected by many factors including genetic and clinical.Investigating the factors contributing to interindividual variability in MPA pharmacokinetics is highly desirable for improving therapeutic efficacy and reducing side effects for patients.UGTs,OATP1B3,MRP2 and BCRP are important enzymes and transporters involved in MPA metabolism.Single nucleotide polymorphisms(SNPs)of these genes may be key factors affecting MPA pharmacokinetics.However,most of studies didn’t analyze the SNP-SNP interaction for MPA pharmacokinetics,which is also important for analyzing the effect of SNP combinations.The development of bioinformatic tools and databases gave us the opportunity to search and prioritize SNPs associated with gene expression.In this study,we searched new SNPs associated with variation in MPA pharmacokinetics using bioinformatic databases.The newly found SNPs and SNPs previously reported in the literature were validated in 408 Chinese renal transplant recipients.In addition,the SNP-SNP interactions for MPA pharmacokinetics and the impact of clinical factors on MPA pharmacokinetics were evaluated and validated.Participants and methodsIn this study,recipients of renal transplant between December 2009 and July 2017 at Nanfang Hospital were selected for this retrospective study.All kidney transplant recipients were treated with identical immunosuppressive regimen including MMF,calcineurin inhibitor(tacrolimus or cyclosporine),and steroids.Trough blood concentration(C0)of MPA was adjusted by dosage(D)and weight.Dose-adjusted MPA trough concentration(C0/D)was the ratio of measured MPA trough concentration C0 divided by corresponding daily MPA dose(D)expressed as mg/kg body weight.C0/D on days 3-8 after transplantation was selected as representative ratio parameters of the early phase after renal transplantation.Candidate SNPs in UGT1A9,UGT1A8,UGT2B7,ABCC2,ABCG2,and SLCO1B3 were selected two ways.First,we selected SNPs reported in the literature as possibly affecting MPA pharmacokinetics or corresponding gene activity with a minor allele frequency(MAF)>1%in Han Chinese in Beijing(CHB)population(data from 1000 Genomes Project phase 3 release).Second,significant SNPs or expression quantitative trait loci(eQTLs)assigned to these genes were collected from bioinformatic tools and databases.Genotyping was performed by high-resolution melting(HRM)assays.In addition to analyzing associations between single SNP and MPA C0/D individually,gene-gene(SNP-SNP)interactions for MPA C0/D among the 11 SNPs were analyzed using dominant,recessive,and additive genetic models.Pearson correlation was used to analyze associations between C0/D and clinical factors.Results1.Ten SNPs of UGT1A9,UGT1A8,UGT2B7,ABCC2,ABCG2,and SLCO1B3,reported to possibly affect MPA pharmacokinetics or corresponding gene activity,were selected as candidate SNPs.In addition,a SNP with a significant p-value was collected from bioinformatic tools and databases.Finally,these 11 SNPs were selected as candidate SNPs.2.The impact of the 11 SNPs on MPA C0/D was analyzed using dominant,recessive,and additive genetic models.Of the 11 SNPs,UGT2B7 rs7662029 was significantly associated with MPA C0/D using the dominant model(p=0.024)and additive model(p=0.028).The UGT2B7 rs7662029 GG group had a higher MPA C0/D than other genotypic groups.ABCC2 rs717620 was significantly associated with MPA C0/D using the recessive model(p=0.043).Compared with other genotypic groups,the ABCC2 rs717620 TT group had a higher MPA C0/D.No significant differences in C0/D were observed for the other 9 SNPs using the three genetic models.3.SNP-SNP interactions for MPA C0/D were detected among the 11 SNPs using dominant,recessive and additive genetic models.Using the additive genetic model,significant SNP-SNP interactions for MPA C0/D were detected between ABCC2 rs717620 and UGT1A9 rs2741049(p=0.002).Patients carrying more ABCC2 rs717620 T allele and UGT1A9 rs2741049 C allele had higher MPA C0/D.Similar SNP-SNP interactions were found between ABCC2 rs717620 and UGT1A8 rs1042597(p=0.002).Patients carrying more ABCC2 rs717620 T allele and UGT1A8 rs1042597 C allele had higher MPA C0/D.4.Correlation between MPA C0/D and clinical factors were analyzed using Pearson correlation analysis.MPA C0/D was significantly positively correlated with age,weight,hematocrit,hemoglobin,albumin and direct bilirubin.Conclusion1.UGT2B7 rs7662029 and ABCC2 rs717620 may affect MPA metabolism.2.Gene-gene(SNP-SNP)interactions have an effect on MPA C0/D,UGT1A9 rs2741049 and UGT1A8 rs1042597 had independent interactions with ABCC2 rs717620,respectively.3.Clinical factors can also affect the pharmacokinetics of MPA.