Tacrolimus (Tacro) is a potent immunosuppressant and a central agent in the prevention of posttransplantation rejection. Tacro is characterized by a narrow therapeutic index and wide interindividual pharmacokinetic fluctuation. The contribution of human UDP-glucuronosyltransferase (UGT) in its metabolism has not been extensively studied. In vitro metabolism studies support that the liver produced Tacro-glucuronide (Tacro-G) while its formation was minimal or undetectable in the presence of intestine and kidney microsomes. Among 16 human UGTs tested, UGT1A4 was the sole enzyme involved in Tacro-G formation. This conclusion is supported by the finding of inhibition with a specific substrate of UGT1A4 lamotrigine with Ki values similar for both human liver and UGT1A4 microsomes and the correlation with trifluoperazine-glucuronide formation by liver microsomes (rs = 0.551; p = 0.02). Formation of Tacro-G by liver samples varied among individuals (6.4-fold variation; n = 16), and common nonsynonymous variants may contribute to this variability. In the human embryonic kidney 293 cellular model, no significant differences in enzyme kinetics could be revealed for UGT1A4*2 (P24T) and *3 (L48V), whereas the allozyme *4 (R11W) displayed a 2-fold higher velocity (p < 0.01) compared with the UGT1A4*1 enzyme preparation. In human liver samples, carriers of the UGT1A4 variants did not display statistically different efficiency in Tacro-G formation compared with homozygote for the reference genotype UGT1A4*1/*1. We conclude that UGT1A4 is the major isoform involved in Tacro glucuronidation, whereas additional studies are required to assess the contribution of UGT1A4 genetic factors in tacrolimus glucuronidation variability.
Among factors involved in patient and graft survival after transplantation, immunosuppressant exposure and efficiency is critical to optimize clinical outcome. Immunosuppressive regimens in posttransplantation typically include a calcineurin inhibitor, cyclosporine, or tacrolimus (Tacro), as well as other steroid and immune modulators such as mycophenolate mofetil and azathioprine, to name only a few. Tacro suppresses lymphocyte proliferation and interleukin synthesis, possibly through a calcineurin-dependent inhibition of nuclear translocation and subsequent activation of nuclear factor of activated T-cell transcription factors (Halloran, 2004).
Because of the complex pharmacokinetics, pharmacodynamics, and pharmacogenetics of Tacro, the potential exists for adverse reactions and a high incidence of drug interactions. Tacro is characterized by a narrow therapeutic index and a large interindividual variation that would be explained, at least in part, by its metabolism. Tacro is metabolized by cytochromes P450 (CYP3A4 and CYP3A5) in both liver and small intestine and transported by P-glycoprotein (ABCB1) efflux transporter (Wallemacq et al., 2009). Drugs that are substrates of these pathways, as well as inhibitors and inducers, can cause significant interactions with Tacro (Mertz et al., 2009). In addition, genetic variations in genes encoding for these proteins have been shown to interfere with Tacro pharmacokinetics, efficacy, and toxicity (Kuypers et al., 2010; Staatz et al., 2010). Tacro is also conjugated to glucuronide (G) by the UDP-glucuronosyltransferase enzymes (UGTs) (Firdaous et al., 1997; Strassburg et al., 2001). In support of this, Tacro-G is measured in human bile (Firdaous et al., 1997). However, limited data are available regarding the involvement of specific enzymes. In this study, we compared the glucuronidation activity for Tacro of major human metabolizing tissues, identified main UGT isozymes involved in the formation of Tacro-G, and tested common genetic variants in UGT1A4-coding regions.
Materials and Methods
Chemical and Reagents.
Tacro was obtained from New England Biolabs Ltd. (Pickering, ON, Canada). Lamotrigine (LTG), 3′-azido-3′-detocythymidine (AZT), and trifluoperazine (TFP) were purchased from Sigma-Aldrich Ltd. (Oakville, ON, Canada); deuterated cyclosporine was obtained from Toronto Research Chemical (Downsview, ON, Canada); and anti-calnexin antibody was from CEDARLANE (Burlington, ON, Canada). All chemicals and reagents were of the highest grade commercially available. Liver tissues from 16 individuals were kindly provided by Dr. Ted T. Inhaba (Department of Pharmacology, University of Toronto, Toronto, ON, Canada) (Trottier et al., 2010). Human pooled colon and kidney microsomes were purchased from BD Gentest (Woburn, MA) and Celsis In Vitro Technologies (Chicago, IL), respectively.
Microsomal protein extracts from UGT1A- and UGT2B-overexpressing human embryonic kidney (HEK) cells and tissues were prepared by differential centrifugation and quantified by Western blot, and assays were conducted as described previously (Benoit-Biancamano et al., 2009). UGT1A4 first exon was amplified by polymerase chain reaction from each liver sample and sequenced as described earlier (Benoit-Biancamano et al., 2009). The initial screening was performed for 16 h with 200 μM Tacro at 37°C for liver, kidney, and intestine microsomes and nine human UGT1A and seven UGT2B functionally active enzymes characterized to date (Fig. 1). Kinetic parameters were then assessed in liver samples from individual donors and in one human intestine pool, as well as positive UGT, namely UGT1A4 reference protein *1 (R11P24L48) as well as for its common variant alloenzymes *2 (Thr24), *3 (Val48), and *4 (Trp11). Assays were performed using Tacro concentrations ranging from 25 to 750 μM for 1 h, whereas TFP assays were done according to the procedure described earlier (Court, 2005; Benoit-Biancamano et al., 2009). Inhibition constant (Ki) values were determined by using Tacro (25–250 μM), LTG (0–3 mM), and AZT (0.1–2.5 mM). The enzyme kinetic model was selected as described previously (Benoit-Biancamano et al., 2009), using Sigma Plot 11.2 assisted by Enzyme Kinetics 1.3 (SSI, San Jose, CA). Values are expressed as mean ± S.D. of triplicate determinations of three independents experiments. Enzymatic activities using UGT-overexpressing cell lines were considered statistically significant for p values <0.05, according to the Student's t test variance analysis. The relationship between UGT1A4 genotype and UGT activity in liver samples was evaluated by performing an analysis of variance; assuming an unequal variance between groups, we used Kruskal-Wallis nonparametric analysis of variance by ranks. Correlations were evaluated by Spearman rank correlation coefficient.
Tacro-G Detection by Mass Spectrometry.
Detection of Tacro-G, expressed as peak area ratio of Tacro-G over cyclosporine d4, was performed by liquid chromatography (LC) coupled to a triple-quadrupole mass spectrometer (API3200; Applied Biosystems, ON, Canada) operating with a turbo ion-spray source in single ion monitoring mode. The high-performance liquid chromatography system consisted of the Agilent 1200 LC (Agilent Technology, Ville Saint-Laurent, QC, Canada), coupled to a Luna C8 column (50 × 4.6 mm, 5 μm; Phenomenex, Torrance, CA). The mobile phase is composed of binary solution, 1 mM ammonium acetate, and 0.01% acetic acid (solvent A) and methanol with 1 mM ammonium acetate and 0.01% acetic acid (solvent B). The analytes were eluted at 0.9 ml/min flow rate, using the following protocol: initial condition, 50% B; 0 to 0.5 min, linear gradient 50 to 97% B; 0.5 to 3 min, isocratic 97% B; 3 to 3.1 min, linear gradient 97 to 50% B; 3.1 to 6 min, isocratic 50% B. Electrospray was set to positive-ion mode with an ionization and declustering potential energy of 5000 and 100V, respectively, and the ion source temperature was set to 500°C. The following mass ions (m/z) were used for detection: 997.8 [M+NH4]+ for Tacro-G and 1224 [M+NH4]+ for cyclosporine d4 (internal standard).
Results and Discussion
Glucuronidation of Tacro Is Exclusively Performed by UGT1A4.
Data reveal that Tacro glucuronidation is formed by liver microsomes and to a lesser extent by intestine proteins, consistent with a previous report (Strassburg et al., 2001). Enzymatic assays performed with recombinant UGT1A and UGT2B proteins showed that Tacro-G, detected by mass spectrometry, was found exclusively for UGT1A4. This predominant contribution of UGT1A4 in Tacro glucuronidation pathway contrasts with a previous study that identified UGT2B7 as the major UGT enzyme involved in Tacro-G formation based on enzymatic assays using microsomal protein extracts from Sf9 cells infected with UGT-recombinant baclovirus and detection of the glucuronide product by thin-layer chromatography (Strassburg et al., 2001).
The evidence against a role for UGT2B7 in Tacro-G formation include the absence of its glucuronidation by UGT2B7-HEK293 microsomes and by a commercial preparation of UGT2B7 recombinant proteins, both of which are positive for the glucuronidation of a UGT2B7 substrate, AZT (data not shown). In addition, Tacro-G remained undetectable in the presence of human kidney microsomes that abundantly express UGT2B7 but not UGT1A4 (Fig. 1C) (Ohno and Nakajin, 2009). A role for UGT2B7 in Tacro-G formation is not significant, and this reaction would be performed by UGT1A4 exclusively. Data from competitive assays performed with liver and UGT1A4-HEK293 microsomes using specific substrates LTG for UGT1A4 and AZT for UGT2B7 (Court, 2005) also sustain this conclusion. Tacro-G formation was influenced in a concentration-dependent manner by LTG for both liver and UGT1A4-derived proteins but not in the presence of AZT. Furthermore, the apparent Ki values estimated for inhibition of Tacro glucuronidation by LTG were almost identical in liver and UGT1A4 (0.43 ± 0.26 and 0.47 ± 0.28 mM, respectively), and this result supports the involvement of this sole enzyme (Fig. 2). Another observation is the affinity of UGT1A4, which appears to be higher for Tacro than for LTG, as indicated by the Ki > Km (Ki = 0.43 ± 0.26 and 0.47 ± 0.28 mM for liver and UGT1A4 microsomes, respectively). Lastly, a significant correlation was observed between formation of Tacro-G and TFP-G by liver microsomes (rs = 0.551; p = 0.02). Therefore, we conclude that UGT1A4 is responsible for Tacro-G formation and that UGT2B7 is not involved.
Tacro-G Formation May Be Altered by Common UGT1A4 Variant Allozymes.
Experimental evidence demonstrated that genetic polymorphisms in the UGT1A4 gene significantly influence glucuronidation activity in vitro (Ehmer et al., 2004; Sun et al., 2006; Benoit-Biancamano et al., 2009). We sought to examine whether common polymorphisms in the coding region of UGT1A4 (frequency ≥1%) influence Tacro-G formation (Fig. 3). Kinetics for Tacro revealed that variant allozymes UGT1A4*1 (R11P24L48), UGT1A4*2 (Thr24), UGT1A4*3 (Val48), and UGT1A4*4 (Trp11) have similar apparent Km values (Table 1). UGT1A4*4 (Trp11) significantly altered Tacro-G formation rates compared with the reference UGT1A4*1 protein (3.14 ± 0.51 for *4 versus 1.65 ± 0.34 for *1; p = 0.007), resulting in a 67% increase in the Vmax/Km ratio for UGT1A4*4. It remains to be determined whether UGT1A4*4 could possibly result in a lower tacrolimus exposure in vivo. No significant changes of Tacro-G formation were revealed for UGT1A4*2 and UGT1A4*3 allozymes, whereas significant alteration of enzyme kinetics for other substrates has been reported previously, which suggests a substrate-specific effect of these coding variations (Ehmer et al., 2004; Benoit-Biancamano et al., 2009). Formation of Tacro-G by 16 liver samples further supports variability among individuals (0.16–1.03 area · min−1 · mg−1; 6.4-fold variation). Kinetic parameters of homozygote carriers of UGT1A4*1 (n = 2) and heterozygotes UGT1A4*1/*2 (n = 2) or UGT1A4*1/*3 (n = 2) were also assessed; however, no individual with the UGT1A4*4 allele could be tested because of its low frequency (Table 1). Rates of Tacro-G formation among *1/*1 carriers was highly similar (Vmax = 0.32 ± 0.03 area · min−1 · mg−1). Compared with livers with the *1/*1 genotype, mean values of Tacro-G formation (Vmax) for UGT1A4*1/*2 and UGT1A4*1/*3 carriers were not statistically different. However, these observations are based on a very limited number of samples and should be confirmed in a larger liver bank. We cannot exclude either the potential effect of common variations in regulatory or the noncoding regions of the UGT1A4 gene, which previously exhibited strong linkage with these cSNPs (Benoit-Biancamano et al., 2009).
In conclusion, data indicate that Tacro-G is mainly produced by the liver and is exclusively dependent on the UGT1A4 pathway. Structural information on the position of glucuronidation of Tacro requires additional investigations. In addition, common genetic variations in the UGT1A4 gene are associated with an altered in vitro formation of Tacro-G and may have a potential effect in vivo.
Participated in research design: Guillemette and Lévesque.
Conducted experiments: Laverdière and Caron.
Contributed new reagents or analytic tools: Laverdière, Caron, and Guillemette.
Performed data analysis: Laverdière, Caron, and Guillemette.
Wrote or contributed to the writing of the manuscript: Laverdière, Caron, Harvey, Lévesque, and Guillemette.
Other: Lévesque and Guillemette acquired funding for this research.
This work was supported by the Canadian Institutes of Health Research [Grant MOP-89954] (to C.G. and É.L.); and the Canada Research Chair Program (C.G.). I.L. was supported by the Fonds de l'Enseignement et de la Recherche of the Faculty of Pharmacy (Laval University); É.L. is the recipient of a Canadian Institutes of Health Research Clinician Scientist Salary Award (Phase 2); and C.G. is the holder of a Canada Research Chair in Pharmacogenomics.
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
- human embryonic kidney
- liquid chromatography.
- Received February 21, 2011.
- Accepted April 12, 2011.
- Copyright © 2011 by The American Society for Pharmacology and Experimental Therapeutics