Introduction

As a member of the calcineurin inhibitor (CNI) family, cyclosporine (CsA) is a first-line immunosuppressant widely used to prevent allograft rejection after solid organ transplantation. However, the bioavailability of cyclosporine ranges from less than 5% to 89% in transplant patients1. Owing to its narrow therapeutic index and marked inter-individual pharmacokinetic variation, therapeutic drug monitoring (TDM) is necessary to adjust the dosage and reduce toxicity2. Nevertheless, TDM is hysteretic for optimizing efficacy and limiting toxicity of cyclosporine. Therefore, identifying factors that affect the pharmacokinetics of cyclosporine has great potential for improving the safety and efficacy profile.

Cyclosporine is a substrate of cytochrome 3A4 and 3A5 (CYP3A4 and CYP3A5) and P-glycoprotein (P-gp/MDR1, encoded by ABCB1). Most previous studies have focused on the influence of genetic variants in the genes encoding CYP3A4, CYP3A5, and P-gp on CsA pharmacokinetics but have yielded conflicting results3. In addition to the differences in study populations, particularly in sample size and ethnicities, there might be other, as-yet undiscovered, genetic factors that influence the expression or function of CYP3A4, CYP3A5 or P-gp to be discovered.

The pregnane X receptor (PXR, encoded by NR1I2) is reported to be the key nuclear receptor regulating the expression of CYP3A4, CYP3A5, and ABCB14. Factors affecting the expression or function of PXR, such as single nucleotide polymorphisms (SNPs), may influence the expression of downstream target genes. T25385C, G24113A, C6994T, C4356T, and G7635A have been reported to be associated with CYP3A4 phenotype, activity and content, whereas A11156C has been associated with variable P-gp levels5,6,7. A24622T and C24446A in exon1 have been found to be associated with PXR levels in previous studies from our laboratory8, suggesting that these two SNPs may also be correlated with the expression of CYP3A4, CYP3A5, or ABCB1.

NF-κB, a protein complex found in almost all animal cell types, is a transcription factor critical for inflammatory responses. It has long been observed that inflammatory responses and infections decrease drug metabolism capacity in humans9,10. The possible effects of NF-κB on enzymes and transporters related to drug metabolism have been studied because NF-κB is a key regulator of inflammation. Gu et al revealed that NF-κB competitively binds to the retinoid X receptor (RXR), thus preventing the PXR-RXR complex from binding to consensus DNA sequences in the regulatory regions of downstream genes, including CYP3A4 and ABCB111. The NFKB1 gene encodes the p50 subunit of NF-κB, which complexes with p65 to produce the major form of NF-κB, and possesses a functional common insertion/deletion (−94 ins/del ATTG) mutation in its promoter region. Deletion of the ATTG alleles may lead to the loss of binding to nuclear proteins and reduced promoter activity and hence decreased production of the NF-κB p50 isoform12.

The aim of this study was to comprehensively evaluate the influence of SNPs in CYP3A4, CYP3A5, ABCB1, NR1I2, and NFKB1 on cyclosporine concentration in a group of Chinese renal transplant recipients during the early stage after transplantation.

Materials and methods

Ethics statement

The study was performed in accordance with the Declaration of Helsinki, and ethical approval for this study was obtained from the ethics committee of the First Affiliated Hospital of Sun Yat-Sen University. Written informed consent was obtained from all subjects.

Patients and therapy

A total of 101 renal transplant recipients (70 men and 31 women) who received a renal transplant between 2005 and 2011 at the Department of Organ Transplant, the First Affiliated Hospital, Sun Yat-sen University were enrolled in this study. The average age was 44.5±13.0 years (range 18–75 years), and the average body weight was 60.4±10.6 kg.

All patients were maintained on a triple immunosuppressive regimen consisting of cyclosporine (Neoral, Novartis Pharma, Basel, Switzerland) as a calcineurin inhibitor, mycophenolate mofetil (Cellcept; Roche, Basel, Switzerland) as a purine inhibitor and steroids (prednisolone, Guangdong Huanan Pharmacy Ltd, Dongguan, China). CsA was administered at a starting dose of 4–6 mg/kg per day following the surgery and administered twice a day in equal amounts. The daily dosage was then adjusted according to trough concentration (C0), to achieve a target concentration of 200 ng/mL. The steroid regimen was 1 g/d intravenously (IV) administered methylprednisolone at the time of surgery followed by 30 mg/d of oral prednisolone, which was progressively tapered to 15–20 mg/d by the end of the first month. Mycophenolate mofetil was given at a dose of 1 g twice daily (bid) during the first month following surgery. Furthermore, other medications known to affect CsA pharmacokinetics, except for prednisolone, were not administered; these included CYP3A or P-gp inducers and inhibitors, such as calcium channel blockers (diltiazem, nicardipine and verapamil), antiepileptics (phenytoin and carbamazepine), antimycotics (fluconazole and ketoconazole), macrolide antibiotics (erythromycin and clarithromycin) and St John's wort. Patients with abnormal hepatic function and combined organ transplants were also excluded.

Data collection

Body weight, CsA dosages and whole blood CsA concentrations were recorded on days 7–18 after transplantation, when CsA concentrations were stable or dosages had been unchanged for a period of time. Two-milliliter blood samples for the CsA assay were collected before 8:00 AM, immediately before the morning dose. The samples were assayed using the commercially available CsA whole blood monoclonal antibody fluorescence polarization assay13. The weight-adjusted CsA dosage (mg/kg per d) and the dose-adjusted trough concentration [C0/D, (ng/mL)/(mg/kg) per d] were calculated.

DNA extraction and genotyping

Total genomic DNA was extracted from peripheral leukocytes according to a previously described method14. CYP3A5*3(A6986G), CYP3A4*1G (G20230A), ABCB1 C1236T, G2677T/A, and C3435T polymorphisms were detected by using previously reported polymerase chain reaction restriction-fragment length polymorphism (PCR-RFLP) methods15,16,17. The NFKB1 −94 ins/del ATTG polymorphism was detected by TaqMan® SNP Genotyping Assay. Polymorphisms of NR1I2, including T25385C, A24622T, C24446A, G24113A, C4356T, A601G, G7635A, A11156C, and C6994T, were genotyped by direct sequencing5,6,7,8.

Statistical analysis

The pair-wise linkage disequilibrium (LD) for SNPs was estimated by SHEsis (http://analysis2.bio-x.cn/myAnalysis.php). ABCB1 1236-2677-3435 haplotype analysis was performed by PHASE 2.1. Groups were compared using nonparametric tests. For the analysis of continuous pharmacologic variables, we used patient genotypes as categorical independent variables. The Hardy-Weinberg equilibrium test was performed using an appropriate χ2 test. The Mann-Whitney U test was used for comparisons between two groups, and the Kruskal-Wallis H test was used for comparisons among several groups. Statistical analysis was performed using SPSS version 17.0 for Windows (SPSS Inc, Chicago, IL, USA). The results are expressed as the mean±standard deviation (SD). A P-value less than 0.05 was considered statistically significant.

Results

Genotypes frequencies

A total of 101 renal transplant recipients were genotyped for polymorphisms in CYP3A5, CYP3A4, ABCB1, NFKB1, and NR1I2. All mutant allele and genotype frequencies were in agreement with previous reports in Han Chinese populations. No differences were observed between allograft recipients and the healthy population, and no deviations from the Hardy-Weinberg equilibrium were observed.

CYP3A4*1G was in moderate linkage disequilibrium (LD) with CYP3A5*3 (D'=0.63). A moderately high degree of LD between ABCB1 C1236T, G2677T/A, and C3435T was also observed. For C1236T-G2677T/A, G2677T/A-C3435T, and C3435T-C1236T, D' was 0.56, 0.79 and 0.72, respectively, which agreed with previous reports. NR1I2 T25385C and G24113A were in complete linkage disequilibrium (D'=1), whereas NR1I2 C4356T was strongly linked to A-601G SNP (D'=0.785). No significant linkage was found between other individual combinations of SNPs.

According to the haplotype analysis, the frequencies of the three major haplotypes (1236-2677-3435) of ABCB1 were 32.3% for TTT, 25.7% for TGC and 20.4% for CGC. The remaining haplotypes constituted 21.6% of the patients' haplotypes.

Association between ABCB1 genotype and CsA dose-adjusted trough blood concentrations

A significantly higher CsA C0/D was observed in ABCB1 2677 TT carriers than in the other 2677 genotype carriers [90.4±24.5 vs 67.8±26.8 (ng/mL)/(mg/kg), P=0.001] (Figure 1A).

Figure 1
figure 1

Correlation of the ABCB1 (A) 2677TT genotype, (B) 3435TT genotype and (C) 1236TT-2677TT-3435TT haplotype with CsA C0/D on d 7–18 after transplantation in renal transplant recipients.

PowerPoint slide

Similarly, carriers of the ABCB1 3435 TT genotype also had a significantly higher CsA C0/D than carriers of the 3435 CT together with CC genotypes [92.0±24.0 vs 68.4±26.5 (ng/mL)/(mg/kg), P=0.002] (Figure 1B).

Although patients with the ABCB1 1236TT genotype tended to have higher CsA C0/D than those with 1236CC or 1236CT genotypes, no significant association was observed between the CsA C0/D and ABCB1 C1236T genotype (P=0.097).

When combining the effects of ABCB1 C1236T, G2677T/A and C3435T, carriers of the 1236TT-2677TT-3435TT haplotype had a considerably higher CsA C0/D than carriers of other genotypes [97.2±21.8 vs 68.7±26.9 (ng/mL)/(mg/kg), P=0.001] (Figure 1C).

Association between NFKB1 genotypes and CsA dose-adjusted trough blood concentrations

Patients with the NFKB1 −94 ATTG ins/ins genotype typically had a higher CsA C0/D than those with the NFKB1 −94 ATTG del/del genotype [78.5±32.8 vs 61.1±19.4 (ng/mL)/(mg/kg)], but the difference was not statistically significant (P=0.069) (Figure 2A).

Figure 2
figure 2

Correlation of NFKB1 genotypes with the dose-adjusted trough concentration of cyclosporine. (A) Influence of the NFKB1 −94 ins/del ATTG genotype on CsA C0/D on d 7–18 after transplantation in renal transplant recipients. (B) Influence of the NFKB1 −94 ins/del ATTG genotype on CsA C0/D on d 7–18 after transplantation in non-carriers of the ABCB1 2677 TT and 3435 TT genotypes.

PowerPoint slide

Association between CYP3A5, CYP3A4, and NR1I2 genotype and CsA dose-adjusted trough blood concentrations

Carriers of the CYP3A5*1/*1 and *1/*3 genotypes were combined as CYP3A5 expressers, and carriers of the *3/*3 genotype were defined as non-expressers. No significant difference in CsA C0/D was observed between CYP3A5 expressers and non-expressers. CYP3A4*1G SNP was also found not to influence CsA C0/D. When the effects of CYP3A5*3 and CYP3A4*1G were combined, we were still unable to observe any effect of the CYP3A4-CYP3A5 haplotype on CsA C0/D.

Although numerous SNPs within the NR1I2 gene were included in this study, no association was observed between CsA C0/D and any of the NR1I2 SNPs (data not shown).

Stratification analysis

Stratification analysis was performed to eliminate any confounding effects of the genes that were included in this study. This study was designed to explore an independent effect of CYP3A4, CYP3A5, NFKB1, and NR1I2 genotype on CsA C0/D after 'standardization' for P-gp activity.

Among non-carriers of the ABCB1 2677 TT genotype and 3435 TT genotype, carriers of the NFKB1 −94 ATTG ins/ins genotype had a significantly higher CsA C0/D than carriers of the −94 ATTG del/del genotype [75.9±32.9 vs 55.1±15.1 (ng/mL)/(mg/kg), P=0.026] (Figure 2B).

Discussion

The current study is a comprehensive study on polymorphisms of drug metabolic enzymes (CYP3A4 and CYP3A5), a transporter (P-gp) and upstream regulators of them (PXR and NF-κB), to explore the cause of the large inter-individual variation in cyclosporine concentration observed in renal transplant recipients. Moreover, the current study is also directly evaluate the potential effect of NF-κB-related polymorphisms on drug metabolism.

ABCB1 C1236T, G2677T/A, and C3435T were the most commonly and extensively studied SNPs of ABCB1, which can form different haplotypes. Carriers of the TT genotype of ABCB1 C3435T or G2677T/A have been reported to have significantly minimized P-glycoprotein activity compared to wild type activity18,19, suggesting a higher CsA C0/D in carriers of the ABCB1 3435 TT and 2677 TT genotypes.

Separate studies on 44 Caucasian liver transplant recipients20 and 88 Middle Eastern renal transplant recipients21 up to 1 month after transplantation both found carriers of the ABCB1 3435 TT genotype had a higher CsA C0/D than carriers of 3435 CC or CT genotype. By contrast, in 103 Chinese renal transplant recipients, Qiu et al reported a lower C0/D in carriers of the ABCB1 2677GG wild-type genotype than in carriers of the ABCB1 2677TT variant genotype in the first 8 to 30 d after transplantation22,23. Consistent with these studies, we found that the ABCB1 2677TT and 3435TT genotypes and 1236TT-2677TT-3435TT haplotype were closely correlated with a higher cyclosporine dose-adjusted C0. Moreover, a more obvious influence was found in the 1236TT-2677TT-3435TT haplotype than in each SNP independently; the 1236TT-2677TT-3435TT haplotype contributes more to the variation in CsA concentrations, as determined by multiple linear regression analysis. These findings, which are in agreement with previous observations, suggest that the use of the ABCB1 haplotype is superior to SNP analysis for predicting concentrations of cyclosporine24.

As a nuclear factor that takes part in the control of as many as 150 target genes, including many inflammatory genes, NF-κB is a focal point and its related pathways and relationships with many autoimmune diseases and cancers have been widely studied. The majority of studies on NF-κB-related polymorphisms, such as the NFKB1 −94 ins/del ATTG mutation, have been focused on their association with the incidence of tumors and inflammatory diseases25. Although it has long been observed that inflammatory responses and infections decrease drug metabolism capacity in human, a correlation between NF-κB-related polymorphisms and drug metabolism has never been reported. In a previous in vitro study, Gu et al reported that the structural component of NF-κB, p65, can competitively bind to RXR, which is necessary to form PXR-RXR heterodimers. The competitive binding of p65 to RXR, therefore, disrupts the interaction of PXR-RXR heterodimers with the consensus DNA sequences in the regulatory regions of target genes, thus significantly suppressing gene expression11. It is possible that NF-κB may have an effect on CsA concentrations by suppressing PXR-mediated regulation of CYP3A4/ABCB1. Hence, we speculated that the NFKB1 polymorphism might result in altered NF-κB expression, leading to reduced CYP3A4/ABCB1 expression and increased CsA concentrations.

As reported previously, the del ATTG allele may result in decreased NFKB1 transcript levels and therefore decreased p50 protein production12. Thus, the ins/ins ATTG genotype is speculated to be associated with a higher NF-κB protein content, and consequently lower CYP3A4 expression and ultimately higher CsA concentrations than the levels in carriers of the del/del ATTG genotype. This finding is consistent with our results that carriers of the −94 ATTG ins/ins genotype had a significantly higher CsA C0/D than carriers of the −94 ATTG del/del genotype in non-carriers of the ABCB1 2677 TT and 3435 TT genotypes. These observations suggest that immunoregulation has a notable effect on CsA concentrations. Further study is needed to assess our speculations and reveal the underlying mechanisms.

Although the importance of CYP3A5*3 in tacrolimus concentrations has been fairly obvious, its influence on CsA concentrations is still in dispute. Several researchers26,27,28 have observed a significant association between CYP3A5*3 and CsA trough concentrations. However, no association between CYP3A5*3 and CsA concentrations was found in this study, which is in consistent with some previous reports29,30,31. CYP3A4*1G is a CYP3A4 SNP with the highest occurrence in Chinese populations, but a definitive function has not been reported. A series of studies examined the relationship between the CYP3A4*1G polymorphism and drug metabolism but resulted in inconsistent findings32,33,34. In agreement with our observation, CYP3A4*1G was not associated with cyclosporine concentrations in 126 renal recipients27; however, Qiu et al22 and Hu et al35 found that it was associated with a lower CsA concentration. Differences in study populations, sample size, ethnicities and particularly in the duration after transplantation might be possible confounding factors. During the early stage after transplantation, steroids were used at a high dose, which might have induced CYP3A.

Although they were reported to be functional in previous studies, none of the NR1I2 SNPs were found to be correlated with CsA concentrations in our study5,6,8. However, most of these studies were in vitro, without the consideration of many possible influences caused by other factors in vivo. Another caveat to be taken into account is the induction of PXR by steroids36 because the recipients in this study, who were at an early stage after transplantation, routinely received a high dose of steroids. Therefore, the impact of NR1I2 SNPs on cyclosporine concentrations at a stable stage after transplantation in patients who received a lower dose of steroids needs to be studied.

In summary, this study reports the potential effect of NF-κB-related polymorphisms on clinical drug metabolism. We also found that the ABCB1 1236TT-2677TT-3435TT haplotype was most significantly correlated with CsA concentrations in Chinese renal transplant recipients. Patients who carry the ABCB1 1236TT-2677TT-3435TT haplotype may be at a greater risk of high CsA concentrations that could lead to hepatotoxicity and nephrotoxicity. NFKB1 polymorphisms were identified as a minor significant factor associated with CsA concentrations in this study, and patients with the NFKB1 ins/ins ATTG genotype tended to have higher CsA concentrations. We also propose that pre-transplant genotyping of the ABCB1 C1236T, G2677T/A, C3435T, and NFKB1 ins/ins ATTG genotypes may be used as an index for CsA dosing in clinical practice and that the initial CsA dose for carriers of the ABCB1 1236TT-2677TT-3435TT haplotype and NFKB1 −94 ins/ins ATTG genotype be lower to prevent possible toxicities during the early stage of transplantation.

Author contribution

Yu ZHANG, Jia-li LI, Min HUANG, and Chang-xi WANG designed the research. Yu ZHANG, Jia-li LI, Xue-ding WANG, Qian FU, and Long-shan LIU performed the research. Yu ZHANG, Jia-li LI, and Wen-ying SHU contributed new reagents and analytic tools. Yu ZHANG and Zhuo-jia CHEN analyzed the data. Yu ZHANG, Jia-li LI, and Wen XIE wrote the paper.