Mrp3), and Abcg2 (Bcrp1) Are the Main Determinants for Rapid Elimination of Methotrexate and Its Toxic Metabolite 7-Hydroxymethotrexate in Vivo

Published OnlineFirst December 8, 2009; DOI: 10.1158/1535-7163.MCT-09-0668

3350

Abcc2 (Mrp2), Abcc3 (Mrp3), and Abcg2 (Bcrp1) are the main determinants for rapid elimination of methotrexate and its toxic metabolite 7-hydroxymethotrexate in vivo

Maria L. H. Vlaming,1 Anita van Esch,1 Zeliha Pala,3 and Abcg2 together mediate the rapid elimination of Els Wagenaar,1 Koen van de Wetering,1 MTX and 7OH-MTX after i.v. administration and can Olaf van Tellingen,2 and Alfred H. Schinkel1 to a large extent compensate for each other's absence. This may explain why it is still comparatively safe to Divisions of 1Molecular Biology and 2Clinical Chemistry, The use a toxic drug such as MTX in the clinic, as the risk Netherlands Cancer Institute, Amsterdam, the Netherlands; of highly increased toxicity due to dysfunctioning of and 3Department of Pharmacology, Faculty of Pharmacy, Istanbul University, Istanbul, Turkey ABCC2, ABCC3, or ABCG2 alone is limited. Neverthe- less, cotreatment with possible inhibitors of ABCC2, ABCC3, and ABCG2 should be done with utmost cau- Abstract tion when treating patients with methotrexate. [Mol The multidrug transporters ABCC2, ABCC3, and ABCG2 Cancer Ther 2009;8(12):3350–9] can eliminate potentially toxic compounds from the body and have overlapping substrate specificities. To in- Introduction vestigate the overlapping functions of Abcc2, Abcc3, and Abcg2 in vivo, we generated and characterized The ATP-binding cassette (ABC) transporters ABCC2, Abcc3;Abcg2−/− and Abcc2;Abcc3;Abcg2−/− mice. We ABCC3, and ABCG2 are transmembrane proteins that are subsequently analyzed the relative impact of these involved in the export of potentially toxic endogenous transport proteins on the pharmacokinetics of the anti- and exogenous compounds from the cell. ABCC2 and cancer drug methotrexate (MTX) and its main, toxic, ABCG2 are expressed in the apical cell membrane of hepa- metabolite 7-hydroxymethotrexate (7OH-MTX) after i.v. tocytes and epithelial cells of small intestine and kidney, administration of MTX (50 mg/kg). Whereas in single pumping their substrates into bile, feces, and urine. ABCC3 is expressed basolaterally in hepatocytes and intestinal epi- and double knockout mice, the plasma and liver concen- – trations of MTX and 7OH-MTX decreased rapidly after thelial cells and pumps its substrates toward the blood (1 MTX administration, in the Abcc2;Abcc3;Abcg2−/− 5). Due to their sites of expression, these ABC transporters mice, they remained very high. One hour after adminis- are involved in the elimination of potential toxins from the tration, 67% of the MTX dose was still present in livers body, thereby protecting the organism from these toxins. of Abcc2;Abcc3;Abcg2−/− mice as MTX or 7OH-MTX ABCC2, ABCC3, and ABCG2 have very broad substrate versus 7% in wild-type, showing dramatic liver accumu- specificities. They can influence the pharmacokinetics of a lation of these toxic compounds when Abcc2, Abcc3, wide range of (anticancer) drugs and carcinogens, as well as potentially toxic endogenous compounds such as bile and Abcg2 were all absent. Furthermore, the urinary – and fecal excretion of the nephrotoxic metabolite 7OH- salts, bilirubin, or porphyrins (1 6). The overlap in substrate MTX were increased 27- and 7-fold, respectively, in specificity of these three proteins is relatively large. They Abcc2;Abcc3;Abcg2−/− mice. Thus, Abcc2, Abcc3, can, for example, all transport the anticancer drugs etopo- side and methotrexate (MTX), as well as a range of glucuronide conjugates of drugs and endogenous compounds (1–9). The widely used anticancer and antirheumatic drug MTX Received 7/21/09; revised 10/15/09; accepted 10/27/09; published is transported by many ABC transporters such as ABCB1, OnlineFirst 12/8/09. ABCC1-5, and ABCG2 in vitro (1,6,7,10,11).Using − − Grant support: NKI 2003-2940 from the Dutch Cancer Society. Z. Pala Abcc2;Abcg2 / mice, we have shown recently that Abcc2 was supported by a grant of the Scientific and Technological Research Council of Turkey. K. van de Wetering was supported by ZonMw TOP and Abcg2 are the main determinants for the biliary excre- grant 40-00812-98-07-028. tion of MTX and its main toxic metabolite 7-hydroxymetho- The costs of publication of this article were defrayed in part by the trexate (7OH-MTX) after i.v. administration of MTX (50 payment of page charges. This article must therefore be hereby marked −/− advertisement in accordance with 18 U.S.C. Section 1734 solely to mg/kg; ref. 12). Furthermore, using Abcc2 and Abcc2; −/− indicate this fact. Abcc3 mice, we found that when Abcc2 was absent (up- Note: Supplementary material for this article is available at Molecular regulated), Abcc3 in the liver caused increased sinusoidal Cancer Therapeutics Online (http://mct.aacrjournals.org/). efflux of MTX and 7OH-MTX from liver (where 7OH- Requests for reprints: Alfred H. Schinkel, Division of Molecular Biology, MTX is primarily formed) to blood, leading to increased uri- The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX −/− Amsterdam, the Netherlands. Phone: 31-20-5122046; Fax: 011-31-20- nary excretion of both compounds (13). Using Abcc3 6691383. E-mail: [email protected] mice, Kitamura et al. (14) recently showed that Abcc3 in- Copyright © 2009 American Association for Cancer Research. creases the systemic exposure to [3H]MTX after oral admin- doi:10.1158/1535-7163.MCT-09-0668 istration or continuous infusion in mice by mediating

Mol Cancer Ther 2009;8(12). December 2009

Molecular Cancer Therapeutics 3351

basolateral efflux in liver and duodenum. They found no ments Australia Pty. Ltd. MRPr1 and M4I-80 were kind gifts − − difference between wild-type and Abcc4 / mice in the oral of Dr. G.L. Scheffer (Free University Hospital, Amsterdam, plasma pharmacokinetics of [3H]MTX. Also, in ABCC2- the Netherlands). A polyclonal antibody against mouse deficient rats, it was shown that the biliary excretion of Abcc2 (25) was kindly provided by Prof. Dr. J.-M. Fritschy MTX was only 10% of the biliary MTX excretion in wild- (University of Zürich, Zürich, Switzerland). Real-time PCR type rats (15). Combining these literature data, there seem primers were from QIAGEN. to be important and overlapping functions of especially Western Analysis Abcc2, Abcc3, and Abcg2 in the pharmacokinetics of Crude membrane fractions from tissues were prepared as MTX and 7OH-MTX. Also, in humans, it was shown that described (23, 26). Western blotting was done as described specific mutations or polymorphisms in ABCC2 or ABCG2 (27). Equal protein loading was confirmed by Ponceau S could influence the pharmacokinetics of MTX, which was staining of the membranes after transfer. Abcc1 (Mrp1), in some cases associated with increased toxicity or efficacy Abcc2 (Mrp2), and Abcc4 (Mrp4) were detected with mono- in patients with loss-of-function mutations (16–20). No clonal antibodies MRPr1 (dilution, 1:1,000), a polyclonal anti- correlations between ABCC3 polymorphisms and MTX murine Abcc2 antibody (dilution, 1:1,000), and M4I-80 (di- pharmacokinetics or toxicity have been reported yet. lution, 1:400), respectively. Bound primary antibodies were Although quite extensive in vivo studies on the influence detected by incubating the blot with horseradish peroxi- of MTX (and 7OH-MTX) have been done already, the rela- dase–labeledrabbitanti-ratIgG(1:1,000;DAKO)orgoat tive effect of the different ABC transporters is unclear. When anti-rabbit IgG (1:1,000; DAKO). Densitometric analysis using single (and even double) knockout mice, the effects of was done using the TINA 2.09 software program (Raytest). the deleted gene(s) might be underestimated because other Real-time PCR Analysis transporters may compensate for their loss. For example, we RNA isolation, cDNA synthesis, and real-time PCR anal- did not find an effect of Abcc3 on MTX pharmacokinetics ysis on livers of female mice (n = 3) were done as described − − after i.v. bolus administration when we analyzed Abcc3 / (28). mice, but when Abcc2 was additionally deleted, a clear ef- Histologic, Clinical-Chemical, and Hematologic fect of Abcc3 expression was found (13). The same was true Analysis for the effect of Abcg2 on 7OH-MTX pharmacokinetics, Histologic analysis of tissues (n = 5, male and female), which only became apparent in the absence of Abcc2 (12). standard clinical chemistry analyses on serum (twice within In the present study, we describe the generation and char- atimespanof1.5y,n =5–6), and standard hematologic − − − − acterization of Abcc3;Abcg2 / and Abcc2;Abcc3;Abcg2 / analysis of male and female mice (n = 6) were done as de- mice. These mice were subsequently used to further unravel scribed (21). the overlapping and possibly compensatory functions of Plasma and Tissue Pharmacokinetic Experiments − − Abcc2, Abcc3, and Abcg2 in determining the pharmacoki- MTX was administered to female wild-type, Abcc3;Abcg2 / − − netics of MTX and 7OH-MTX in vivo.Weshowherethat and Abcc2;Abcc3;Abcg2 / mice (n =3–13) by injecting 5 μL/g these three transporters are the main determinants of the body weight of a 10 mg/mL MTX solution in saline into the rapid elimination of MTX and 7OH-MTX from the liver tail vein. Animals were killed by terminal bleeding through and plasma. Other ABC transporters seem less important cardiac puncture under methoxyflurane anesthesia at 7.5, in these processes. 15, 30, 60, or 120 min after MTX administration, and organs were removed at each time point. Fecal and Urinary MTX Excretion Experiment − − − − Materials and Methods Female wild-type, Abcc3;Abcg2 / and Abcc2;Abcc3;Abcg2 / Animals mice (n =5–9) were individually housed in Ruco Type M/1 Mice were housed and handled according to institution- stainless steel metabolic cages. They were allowed 24 h to al guidelines complying with Dutch legislation. The gener- adapt before MTX was injected at 50 mg/kg into the tail − − − − − − ation of Abcc2 / (21), Abcc3 / (22), Abcg2 / (23), Abcc2; vein as described above, and feces and urine were collected − − − − Abcc3 / (13, 24), and Abcc2;Abcg2 / mice (12) was de- for 24 h. − − scribed before. Abcc3;Abcg2 / mice were generated by High Performance Liquid Chromatography Analysis of crossbreeding the two single knockout strains. Abcc2; MTX and 7OH-MTX − − Abcc3;Abcg2 / mice were generated by crossbreeding Collected organs and feces were homogenized in an ice- − − − − Abcc2;Abcc3 / and Abcc2;Abcg2 / mice. All animals were cold 4% bovine serum albumin solution, and plasma was of >99% FVB background and between 9 to 14 wk of age. diluted in human plasma before high performance liquid Animals were kept in a temperature-controlled environ- chromatography analysis. Urine was diluted in water. ment with a 12-h light/12-h dark cycle. They received a MTX and 7OH-MTX concentrations in the different matrices standard diet (AM-II, Hope Farms) and acidified water were determined as described (29). ad libitum. Statistical Analysis Chemicals Unless otherwise indicated, the two-sided unpaired Stu- MTX (Emthexate PF, 25 mg/mL) was from Pharmachem- dent's t test was used to assess the statistical significance of ie and 7OH-MTX was from Toronto Research Chemicals, differences between wild-type and knockout mice. When Inc. Methoxyflurane (Metofane) was from Medical Develop- statistical differences between more than two groups were

Mol Cancer Ther 2009;8(12). December 2009

3352 Methotrexate in Abcc2;Abcc3;Abcg2−/− Mice

analyzed, one-way ANOVA followed by Tukey's multiple we determined mRNA expression levels of other genes that comparison test was done, as indicated. Results are pre- may be involved in MTX metabolism and transport in livers − − sented as means ± SDs. Differences were considered to of female knockout mice. We found that, similar to Abcc2 / − − be statistically significant when P < 0.05. Averaged concen- and Abcc2;Abcg2 / mice (12), mRNA levels of the enzyme trations for each time point were used to calculate the area aldehyde oxidase 1, which is implicated in the conversion of under the plasma concentration versus time curve (AUC) MTX to 7OH-MTX (32), was increased 2.5-fold compared from t = 0 to the last sampling point by the linear trape- with wild-type mice (n =3,P = 0.008; data not shown). Al- zoidal rule; SEMs were calculated by the law of propaga- dehyde oxidase 3 mRNA expression in liver was not signif- tion of errors (30). icantly different from wild-type (data not shown). Furthermore, whereas we previously found that mRNA of the uptake transporter Slco1b2 was increased ∼2-fold in − − − − Results Abcc2 / and Abcc2;Abcg2 / mice (12), mRNA levels of Macroscopic and Microscopic Analysis of Abcc3; Slco1a4 and Slco1b2 were not significantly different from −/− −/− − − − − Abcg2 and Abcc2;Abcc3;Abcg2 Mice wild-type in Abcc3;Abcg2 / and Abcc2;Abcc3;Abcg2 / mice − − − − Abcc3;Abcg2 / and Abcc2;Abcc3;Abcg2 / mice were via- (data not shown). ble, fertile, and had normal life spans and body weights. Serum Clinical Chemistry and Hematologic Analysis Macroscopic and microscopic histologic and pathologic of Abcc3;Abcg2−/− and Abcc2;Abcc3;Abcg2−/− Mice analysis did not reveal obvious specific aberrations in tis- Table 1 shows clinical chemistry parameters of Abcc3; − − − − sues of both knockout strains, including the liver. However, Abcg2 / and Abcc2;Abcc3;Abcg2 / mice that were different −/− −/− in Abcc2;Abcc3;Abcg2 mice (but not in Abcc3;Abcg2 from wild-type mice. The results of all measured parameters mice), the liver weight was increased 41% to 67% compared are shown in Supplementary Table S1. Conjugated bilirubin with wild-type (Fig. 1A). Increased liver weight was previ- levels were below the detection limit of the analyzer − − ously shown in other Abcc2-deficient strains (Fig. 1A; (2 μmol/L). In the Abcc2;Abcc3;Abcg2 / mice, the serum to- refs.12,13,21,31).However,theliverweightinAbcc2; tal bilirubin concentrations were mildly increased compared − − − − Abcc3;Abcg2 / mice (both male and female) was also signif- with wild-type, as was previously shown for Abcg2 / and − − icantly higher than in most Abcc2, Abcc3, and Abcg2 single Abcc2;Abcg2 / mice (12, 23). In contrast to what was previ- − − − − and double knockout mice (ANOVA). ously shown for Abcc2 / and Abcc2;Abcg2 / mice (12, 21, Protein and mRNA Expression of other ABC 33), we did not detect increased conjugated bilirubin levels − − − − Transporters in Tissues of Abcc3;Abcg2 / and Abcc2; in Abcc2;Abcc3;Abcg2 / mice. This is in line with the earlier − − Abcc3;Abcg2 / Mice observation that Abcc3 is responsible for transport of conju- Protein expression of Abcc1 in liver of male wild-type, gated bilirubin from liver to blood when Abcc2 is absent (13). − − − − Abcc3;Abcg2 / ,andAbcc2;Abcc3;Abcg2 / mice was low Besides bilirubin, triglyceride serum concentrations were al- − − and therefore hard to quantify (data not shown). Abcc2 pro- so increased ∼2-fold in Abcc2;Abcc3;Abcg2 / mice (but not in − − − − tein in liver of male and female Abcc3;Abcg2 / mice was not Abcc3;Abcg2 / mice; Table 1). Furthermore, total protein le- − − different from wild-type mice (dat not shown). Abcc2 pro- vels were mildly increased in serum of Abcc2;Abcc3;Abcg2 / − − − − tein in kidney of female Abcc3; Abcg2 / mice was not differ- mice (Table 1). In male (but not female) Abcc2;Abcc3;Abcg2 / ent from wild-type either. mice, alanine aminotransferase, creatinine, and urea serum Abcc4 protein expression in kidney of female Abcc2; concentrations were also mildly but significantly increased − − Abcc3;Abcg2 / mice was increased 1.6-fold compared with (Table 1). − − − − wild-type mice and similar to that of Abcc2;Abcg2 / mice Hematologic analysis of Abcc3;Abcg2 / mice did not − − (Fig. 1B). Abcc4 protein in livers of male Abcc3;Abcg2 / show consistent differences with wild-type mice (data not − − − − and Abcc2;Abcc3;Abcg2 / mice was low and therefore hard shown). Female Abcc2;Abcc3;Abcg2 / mice had, similar to − − to quantify (data not shown). In female Abcc3;Abcg2 / mice, other Abcc2-deficient mice (12, 21), mildly reduced hemo- liver Abcc4 protein was not different from wild-type. In fe- globin levels compared with wild-type (7.2 ± 0.2 mmol/L − − − male Abcc2;Abcc3;Abcg2 / mice, Abcc4 liver expression versus 7.8 ± 0.3 mmol/L, n =6,P = 2.3 × 10 3). In males, seemed to increase about 2-fold compared with wild-type, this was not found. − − comparable with that in Abcc2;Abcg2 / mice (12), although Effect of Abcc2, Abcc3, and Abcg2 on MTX and 7OH- this was hard to quantify due to low expression levels in all MTX Pharmacokinetics strains (data not shown). Quantitative real-time PCR analy- We have previously shown that Abcc2 and Abcg2 are the − − sis confirmed that in female Abcc2;Abcg2 / and Abcc2; main determinants for the biliary excretion of MTX and − − Abcc3;Abcg2 / mice, Abcc4 mRNA expression in liver was 7OH-MTX after i.v. administration of MTX, (50 mg/kg; significantly (14- to 19-fold) increased compared with wild- ref. 12). Furthermore, when Abcc2 is absent, (upregulated) − − type mice (Fig. 1C). Abcc4 mRNA levels in Abcc3;Abcg2 / Abcc3 in the liver causes increased efflux of MTX and 7OH- mice were not different from wild-type mice (Fig. 1C). MTX from liver to blood, leading to increased urinary excre- Expression Levels of Genes Involved in Methotrexate tion of these compounds (13). To further investigate the Disposition functional overlap and importance of these ABC transpor- Because we aimed to analyze the pharmacokinetics of ters, we administered MTX i.v. at a dose of 50 mg/kg to − − − − − − − − MTX using Abcc3;Abcg2 / and Abcc2;Abcc3;Abcg2 / mice, wild-type, Abcc3;Abcg2 / ,andAbcc2;Abcc3;Abcg2 / mice

Mol Cancer Ther 2009;8(12). December 2009

Molecular Cancer Therapeutics 3353

Figure 1. Physiologic characterization of Abcc3;Abcg2−/− and Abcc2;Abcc3;Abcg2−/− mice. A, liver weight (as percentage of body weight) of male (left) and female (right) wild-type, Abcc2−/−, Abcc2;Abcg2−/−, Abcc2;Abcc3−/−, Abcc3;Abcg2−/−, and Abcc2;Abcc3;Abcg2−/− mice [columns, means (n = 3–12); bars, SD; **, P < 0.01; ***, P < 0.001, relative to wild-type, ANOVA). Abcc3−/− and Abcg2−/− mice had no aberrant liver weight (12, 13). B, protein levels of Abcc4 (molecular weight, 150 kDa) in crude membrane fractions of kidney samples from two independent female wild-type, Abcc2; Abcc3;Abcg2−/−, and (for comparison) Abcc2;Abcg2−/− mice (dilution experiment). Densitometric analysis of the blots showed that on average Abcc4 expression was increased 1.6-fold in Abcc2;Abcc3;Abcg2−/− mice compared with wild-type. The lane with the positive control (Sf-9 vesicles containing ABCC4; ref. 40), is indicated with “+”. Underglycosylation of ABCC4 in the Sf-9 cells causes the faster migration compared with the murine Abcc4. The amount of protein loaded in each lane (in micrograms) is noted above the respective lanes. C, mRNA expression of Abcc4 in liver of female wild-type, Abcc2;Abcg2−/−, Abcc3;Abcg2−/−, and Abcc2;Abcc3;Abcg2−/− mice. Results are expressed as the fold change in liver of knockout mice compared with wild-type mice. Data were normalized against the endogenous control glyceraldehyde-3-phosphate dehydrogenase (n = 3; ***, P < 0.05; ns, not significant; ANOVA, statistical significance was determined by comparison of ΔCt values of different strains). Columns, means; bars, SD. and analyzed the pharmacokinetics of MTX and 7OH-MTX. mentary Table S2). This suggests a markedly delayed over- The results for all compound and single Abcc2, Abcc3, and all elimination of MTX in these mice (Fig. 2A). Abcg2 knockout mice (12, 13) are given in Supplementary We have shown before (12) that after an initial high hepat- Tables S2 (for MTX) and S3 (for 7OH-MTX). Figure 2 shows ic accumulation shortly after administration, MTX was rap- − − plasma and liver levels of MTX and 7OH-MTX in wild-type, idly eliminated from the liver in wild-type, Abcc2 / ,and − − − − − − Abcc3;Abcg2 / ,andAbcc2;Abcc3;Abcg2 / mice. Further- Abcc2;Abcg2 / mice. This is likely due to (upregulated) more, the combined liver levels of MTX and 7OH-MTX Abcc3 in these strains, leading to increased sinusoidal efflux − − − − are shown in Supplementary Fig. S1. For comparison, the (13). In Abcc3;Abcg2 / mice, as in Abcg2 / mice (Supple- − − pharmacokinetics of MTX and 7OH-MTX in Abcc2;Abcg2 / mentary Table S2; ref. 12), MTX elimination from the liver mice (12) are also shown in each figure. was significantly delayed (Fig. 2B). Still, after 120 minutes, −/− Whereas the AUCplasma for MTX was increased 1.6-fold in the MTX liver levels in Abcc3;Abcg2 mice were back to − − − − − − Abcg2 / mice (12), the plasma levels of MTX in Abcc3;Abcg2 / wild-type levels. In contrast, in Abcc2;Abcc3;Abcg2 / mice mice were similar to wild-type (Supplementary Table S2; 120 minutes after administration, the liver levels of MTX Fig. 2A), suggesting that Abcc3 is necessary for the in- were increased up to 7-fold compared with wild-type con- − − creased MTX plasma levels in Abcg2 / mice. The role of trol liver (Fig. 2B). Abcc3 was also illustrated by the reduced MTX plasma le- We also determined plasma and tissue levels of the toxic − − − − vels in Abcc2;Abcc3;Abcg2 / mice compared with Abcc2; metabolite 7OH-MTX (Fig. 2C and D). In the Abcc2;Abcg2 / − − Abcg2 / mice up to 60 minutes after administration mice, the plasma levels of 7OH-MTX increased rapidly, prob- (Fig. 2A). Whereas in all other strains, the MTX plasma ably due to reduced biliary excretion of MTX, 7OH-MTX, or − − levels rapidly decreased 30 minutes after administration, both (12). In Abcc2;Abcc3;Abcg2 / mice, however, this − − in Abcc2;Abcc3;Abcg2 / mice, they stayed relatively high increase was less pronounced, showing that part of the − − up to 120 minutes after administration, leading to a 1.4-fold increased plasma 7OH-MTX concentrations in Abcc2;Abcg2 / increased AUCplasma compared with wild-type (Supple- mice was dependent on the presence of (overexpressed)

Mol Cancer Ther 2009;8(12). December 2009

3354 Methotrexate in Abcc2;Abcc3;Abcg2−/− Mice

hepatic Abcc3 (Fig. 2C, see also below). It should be noted Table S2; Fig. 3A; ref. 12). The 7OH-MTX kidney levels also − − that in the Abcc2;Abcc3;Abcg2 / mice, substantial amounts followed the plasma levels, resulting in substantial kidney − − of 7OH-MTX still reached the circulation, leading to in- accumulation in both Abcc2;Abcg2 / and Abcc2;Abcc3; −/− creased plasma levels and a 9.6-fold increased AUCplasma Abcg2 mice at 60 minutes after administration (Fig. 3C). compared with wild-type mice over the first 120 minutes We have shown before that the presence of MTX and (Supplementary Table S3; Fig. 2C). This indicates that in 7OH-MTX in the small intestine after i.v. administration − − the Abcc2;Abcc3;Abcg2 / mice, substantial amounts of (up to about 50% of the dose in 60 minutes) is mainly due 7OH-MTX can still leave the liver over the basolateral to their biliary excretion by Abcc2 and Abcg2, whereas di- membrane, even without Abcc3. rect excretion across the intestinal wall is relatively low (12, − − Analysis of hepatic 7OH-MTX concentrations clearly 13). Figure 3B shows that in Abcc2;Abcc3;Abcg2 / mice, the showed the importance of Abcc2, Abcc3, and Abcg2 for excretion of MTX into the small intestine was clearly re- the elimination of 7OH-MTX from the liver. Liver levels of duced to only 15% of wild-type levels at 60 minutes after 7OH-MTX were dramatically (up to 90-fold at t = 120 min- administration (Supplementary Table S2). Despite markedly − − utes) increased in the Abcc2;Abcc3;Abcg2 / mice (Fig. 2D), increased hepatic levels in the triple knockout mice also when compared with all double knockout strains, in- (Fig. 2B), relatively little MTX was excreted into the small − − cluding Abcc2;Abcc3 / mice (Supplementary Table S3). At intestine, once more showing that Abcc2 and Abcg2 are 60 minutes, the cumulative MTX and 7OH-MTX levels in the predominant factors involved in the hepatobiliary excre- wild-type as well as single and double knockout strains tion of MTX into the small intestine. The somewhat higher − − did not exceed 22% of the given dose (ranging from 7.0 ± small intestinal MTX levels in Abcc2;Abcc3;Abcg2 / com- − − − 0.9% of the dose in wild-type to 22.0 ± 7.2% of the dose in pared with Abcc2;Abcg2 / mice (P = 4.8 × 10 3 by Student's − − − − Abcc2;Abcc3 / mice). However, in Abcc2;Abcc3;Abcg2 / t test at 60 minutes) probably reflected the higher liver MTX mice, the combined MTX and 7OH-MTX liver levels concentrations (Fig. 3B). Clearly, also in the absence of amounted to 67.1 ± 3.5% of the dose (Supplementary Abcc2 and Abcg2, there is some remaining biliary excretion Fig. S1), illustrating a drastically reduced liver elimination of MTX, as we still found low levels of MTX in the small − − − − of both compounds in these mice. In the triple knockout intestine of Abcc2;Abcg2 / and Abcc2;Abcc3;Abcg2 / mice. mice, accumulated 7OH-MTX was slowly released from Figure 3D shows that, despite the increased liver levels − − the liver, primarily toward plasma, resulting in relatively of 7OH-MTX in the Abcc2;Abcc3;Abcg2 / mice, the excre- high 7OH-MTX plasma levels over a prolonged time frame tion into the small intestine was significantly lower than (Fig. 2C and D). in wild-type mice. This confirms that Abcc2 and Abcg2 MTX levels in the kidney in general reflected the plasma are the main transporters responsible for (biliary) excretion of MTX levels, leading to 7- and 2-fold increased kidney levels 7OH-MTX into the small intestine. However, like for MTX − − − − in Abcc2;Abcg2 / and Abcc2;Abcc3;Abcg2 / mice, respec- (see above), the small intestinal levels of 7OH-MTX were sig- tively, at 60 minutes after administration (Supplementary nificantly higher at 60 minutes after administration in Abcc2;

Table 1. Clinical chemical analysis of serum from wild-type, Abcc3;Abcg2−/−, and Abcc2;Abcc3;Abcg2−/− mice

Parameter Sex Strain − − − − Wild-type Abcc3;Abcg2 / Abcc2;Abcc3;Abcg2 /

Total Bilirubin (μmol/L) Male 2.2 ± 0.4 2.7 ± 0.8 3.7 ± 0.5* † Female 3.0 ± 1.0 3.7 ± 1.9 4.2 ± 1.3 ALAT (U/l) Male 34 ± 5 36 ± 7 55 ± 14* Female 27 ± 8 22 ± 9 33 ± 6 ‡ Creatinine (μmol/L) Male 11 ± 2 19 ± 4 25 ± 4* Female 17 ± 4 17 ± 3 20 ± 3 ‡ Urea (mmol/L) Male 10 ± 1 13 ± 2 12 ± 1§ Female 10 ± 2 10 ± 2 11 ± 1 ‡ Total protein (g/l) Male 48 ± 2 50 ± 2 52 ± 2 ‡ Female 49 ± 2 50 ± 3 53 ± 2 ‡ Triglyceride (mmol/L) Male 1.7 ± 0.4 1.8 ± 0.4 3.4 ± 1.0 ∥ Female 2.0 ± 0.7 2.0 ± 0.6 3.3 ± 0.8

NOTE: Results are presented as means ± SD (n =5–6). Abbreviation: ALAT, alanine aminotransferase. *P < 0.001 compared with wild-type mice (Student's t test was used for statistical analysis). † Was not significant in this measurement (P = 0.13), but was significantly higher in a previous data set. All other measured parameters are shown in Supple- mentary Table S1. ‡ P < 0.01 compared with wild-type mice (Student's t test was used for statistical analysis). §P < 0.05 compared with wild-type mice (Student's t test was used for statistical analysis). ∥ P < 0.05.

Mol Cancer Ther 2009;8(12). December 2009

Molecular Cancer Therapeutics 3355

Figure 2. Plasma and liver pharmacokinetics of MTX and 7OH-MTX after i.v. administration of MTX (50 mg/kg) to female wild-type, Abcc2;Abcg2−/−, Abcc3; Abcg2−/−,andAbcc2;Abcc3;Abcg2−/− mice. A, MTX plasma concentration versus time curves of the different strains, semi-log plot [points, mean (n =3–13); bars, SD]. MTX plasma AUC values for the depicted strains (min·μg/mL): wild- type, 444 ± 44; Abcc2;Abcg2−/−, 1,446 ± 229; Abcc3;Abcg2−/−, 451 ± 26; Abcc2; Abcc3;Abcg2−/−, 603 ± 56. B, MTX liver level versus time curves of the different strains [points, mean (n =3–9); bars, SD]. MTX liver AUC values for the depicted strains (min·mg/g liver): wild-type, 12.8 ± 1.1; Abcc2;Abcg2−/−, 12.3 ± 1.5; Abcc3; Abcg2−/−, 20.5 ± 1.5; Abcc2;Abcc3; Abcg2−/−, 24.3 ± 1.3. C, 7OH-MTX plasma concentration versus time curves of the different strains [points, mean (n =3–13); bars, SD]. 7OH-MTX plasma AUC values for the depicted strains (min·μg/mL): wild- type, 17 ± 3; Abcc2;Abcg2−/−, 210 ± 36; Abcc3;Abcg2−/−,20±2;Abcc2;Abcc3; Abcg2−/−,163±17.D, 7OH-MTX liver level versus time curves of the different strains [points, means (n =3–9); bars, SD]. 7OH- MTX liver AUC values for the depicted strains (min·mg/g liver): wild-type, 2.3 ± 0.3; Abcc2;Abcg2−/−, 5.8 ± 0.6; Abcc3; Abcg2−/−, 3.1 ± 0.3; Abcc2;Abcc3; Abcg2−/−, 28.0 ± 1.5.

− − − − Abcc3;Abcg2 / compared with Abcc2;Abcg2 / mice (P < 0.01, low plasma and kidney 7OH-MTX concentrations in ANOVA), suggesting that the increased liver and plasma these strains. The fecal excretion of 7OH-MTX was, like − − concentrations in these mice led to somewhat increased that of MTX, somewhat increased in Abcc3;Abcg2 / mice. − − (hepatobiliary and/or direct) excretion into the small in- In contrast, in Abcc2;Abcc3;Abcg2 / mice, both the uri- testine of 7OH-MTX, not mediated by Abcc2 or Abcg2. nary and fecal excretion of 7OH-MTX were markedly in- Fecal and Urinary Excretion of MTX and 7OH-MTX creased (27.0- and 7.3-fold, respectively) compared with − − The cumulative excretion of MTX and 7OH-MTX over wild-type (Fig. 4B). In fact, in Abcc2;Abcc3;Abcg2 / mice, − − 24 hours into urine and feces in wild-type, Abcc3;Abcg2 / , the total amount of 7OH-MTX recovered in urine and − − and Abcc2;Abcc3;Abcg2 / miceisshowninFig.4.Despite feces represented ∼43% of the total MTX dose, as op- the delayed hepatic elimination of MTX and 7OH-MTX early posed to <8% in wild-type and double knockout strains after administration (see above), especially in the triple (Fig. 4C and D). knockout mice, virtually all of the administered MTX was excreted as MTX or 7OH-MTX within 24 hours (Fig. 4C and D). The total urinary excretion of MTX was not different Discussion from wild-type in the compound knockout strains, show- To investigate the possibly overlapping functions of ing that even if Abcc2, Abcc3, and Abcg2 are absent, Abcc2, Abcc3, and Abcg2, we generated and character- − − − − MTX can still leave the liver and be excreted via the ized Abcc3;Abcg2 / and Abcc2;Abcc3;Abcg2 / mice. De- urine. The fecal excretion of MTX was not significantly spite the absence of up to three ABC transporters with − − different from wild-type in the Abcc2;Abcc3;Abcg2 / mice overlapping or complementary functions in the protection − − either. Interestingly, in Abcc3;Abcg2 / mice, the fecal ex- from toxic compounds (1–4), both strains are viable, fer- cretion of MTX was significantly increased compared tile, and have a normal life span. They also do not dis- with wild-type (Fig. 4A). play obvious physiologic or pathologic aberrations under The urinary excretion of 7OH-MTX in wild-type and standard housing conditions, except for an increased liver − − Abcc3;Abcg2 / mice was low, in line with the relatively size and some mild changes in serum clinical chemistry.

Mol Cancer Ther 2009;8(12). December 2009

3356 Methotrexate in Abcc2;Abcc3;Abcg2−/− Mice

− − We therefore consider these double and triple knockout Abcg2 / mice, showing that at least in the protective envi- mice quite suitable for pharmacokinetic, toxicologic, and ronment of our animal facility, Abcc2, Abcc3, and Abcg2 do physiologic research on the functions of Abcc2, Abcc3, not have crucial overlapping physiologic functions. and Abcg2. We used the mice to evaluate the overlapping We found that after i.v. administration of MTX, Abcc2, or complementary functions of Abcc2, Abcc3, and Abcg2 in Abcc3, and Abcg2 play important and clearly overlapping the pharmacokinetics of the anticancer and antirheumatic or complementary roles in the elimination of MTX and drug MTX and its main, toxic, metabolite 7OH-MTX in vivo. 7OH-MTX from the liver. Whereas in single and double We show here that Abcc2, Abcc3, and Abcg2 together are knockout mice there was only relatively mild accumulation the primary ABC transporters responsible for the fast elim- of both compounds in the liver (Supplementary Tables S2 − − ination of both MTX and 7OH-MTX from the liver and and S3), in the Abcc2;Abcc3;Abcg2 / mice this was much body, and that they can largely compensate for the loss of more dramatic, especially for 7OH-MTX, and it lasted lon- each other. Absence of all three ABC transporters leads to a ger as well (Supplementary Fig. S1; Fig. 2B and D). Further- dramatic accumulation of MTX and 7OH-MTX in the liver, more, whereas in all other strains the plasma levels of MTX as well as prolonged systemic exposure to both compounds and 7OH-MTX at 120 minutes after i.v. administration of after MTX administration. MTX were quite low and comparable with wild-type (Sup- − − − − In Abcc2;Abcc3;Abcg2 / mice, some changes in clinical plementary Tables S2 and S3), in Abcc2;Abcc3;Abcg2 / mice, chemistry parameters of serum were found (Table 1). How- they were still relatively high, suggesting prolonged sys- ever, although the differences were consistent in two sepa- temic exposure to these toxic compounds in the combined − − rate measurements, they were relatively small, and absence of these transporters. In Abcc2;Abcc3;Abcg2 / mice, extensive pathologic analysis of liver and kidney of these the total amount of 7OH-MTX recovered in urine and feces mice did not reveal any signs of pathologic lesions. Further- represented ∼43% of the total MTX dose, as opposed to <8% more, the mice had normal life spans. This suggests that in wild-type and double knockout strains (Fig. 4C and D). It there are no serious effects on the health of Abcc2;Abcc3; is likely that the absence of Abcc2, Abcc3, and Abcg2 leads

Figure 3. Kidney and small intestinal tissue + contents pharmacokinetics of MTX and 7OH-MTX after i.v. administration of MTX (50 mg/kg) to female wild- type, Abcc2;Abcg2−/−, Abcc3;Abcg2−/−,andAbcc2; Abcc3;Abcg2−/− mice. A, MTX kidney level versus time curves of the different strains [points, mean (n =3–9); bars, SD]. B, MTX small intestinal tissue and contents level versus time curves of the different strains [points, mean (n =3–9); bars, SD]. C, 7OH-MTX kidney level versus time curves of the different strains [points, mean (n =3–9); bars, SD]. D, 7OH-MTX small intestinal tissue and contents level versus time curves of the different strains [points, mean (n =3–9); bars, SD]. Note the differences in axis scales.

Mol Cancer Ther 2009;8(12). December 2009

Molecular Cancer Therapeutics 3357

Figure 4. Urinary and fecal excretion of MTX and 7OH-MTX 24 h after i.v. administration of MTX (50 mg/kg) to female wild-type, Abcc2;Abcg2− / − , Abcc3;Abcg2−/−,andAbcc2;Abcc3; Abcg2−/− mice. A, urinary (left) and fecal (right) excretion of MTX (as % of the dose). B, urinary (left)andfecal(right) excretion of 7OH-MTX (as % of the dose). C, cumulative urinary and fecal excretion of MTX (as % of the dose). D, cumulative urinary and fecal excretion of 7OH-MTX (as % of the dose). Columns, means (n =5–9; *, P < 0.05; **, P < 0.01; ***, P < 0.001); bars, SD. Levels of 7OH-MTX in feces of wild-typemicewereclosetoback- ground and therefore hard to quantify (background ~3% of the dose). Levels of 7OH-MTX in feces of all strains may be slightly overestimated due to high background peaks in this matrix. Note the differences in axis scales.

to increased retention of MTX in the liver and subsequent 7OH-MTX in patients treated with MTX. This may cause extensive 7OH-MTX formation and accumulation in the increased risk of hepatotoxicity when patients with poly- liver. This is subsequently eliminated over a relatively pro- morphisms or mutations in one or more of these genes are longed period via both urine and feces. Combined, these treated with MTX, as was previously shown for MTX-treated results show that Abcc2, Abcc3, and Abcg2 can to a large African-American patients carrying an ABCC2 mutation extent compensate for absence of one another in the liver. (19). However, as it may be unlikely for patients to have Other ABC transporters, which have been shown to trans- polymorphisms or mutations in all three transporters, the port MTX in vitro, such as Abcc1, Abcc4, Abcc5, and P-gp risk of increased MTX-related hepatotoxicity could be rela- (10, 11, 34, 35), do not seem to play a significant role in the tively small. rapid elimination of MTX and 7OH-MTX in vivo,atleast In fact, it is worth noting that the functional overlap or not after bolus i.v. administration of MTX at 50 mg/kg. complementarity between no less than three different pro- Abcc2, Abcc3, and Abcg2 clearly are important for the teins that we observed here guarantees a considerable ro- disposal of MTX and 7OH-MTX early after administra- bustness in the protection from MTX, a xenobiotic toxin. tion. However, also in the absence of all three proteins, This makes sense from both biological and clinical-therapeutic virtually all of the administered MTX was excreted as perspectives. Although individual deficiencies in Abcc2, MTX or 7OH-MTX within the first 24 hours after admin- Abcc3, or Abcg2 can affect the clearance and systemic expo- istration (Fig. 4). This shows that other (low capacity) sure of MTX to a certain extent, from our study, it is clear elimination systems are apparently still able to eliminate that the consequences would be much more severe in the these compounds from the body. Which mechanisms are absence of functions of all these proteins. Thus, overall pro- responsible for this is not clear and may be investigated tection against MTX toxicity is not entirely dependent on in the future. the function of one single gene and protein, but it is deter- The combined absence of Abcc2, Abcc3, and Abcg2 led to mined by at least three different genes. This means that the trapping of MTX in the liver due to reduced biliary (via risk of increased toxicity due to genetic polymorphisms or Abcc2 and Abcg2) and sinusoidal (via Abcc3) elimination coincidental inhibition affecting activity of a single gene or of MTX. This in turn led to the increased hepatic formation proteinismuchreduced.Thisisclearlyadvantageousinthe of the toxic MTX metabolite 7OH-MTX (Fig. 2B and D), natural protection from xenobiotic toxins. It also means that showing that Abcc2, Abcc3, and Abcg2 are very important it is comparatively safe to use a drug such as MTX in the for limiting the hepatic formation of this toxic metabolite by clinic, as the risk of unpredictable toxicity due to dysfunc- keeping liver MTX levels low. Moreover, 7OH-MTX is also tioning of one detoxifying protein is relatively limited. efficiently transported out of the liver for fecal or urinary Clearly, drugs of which the toxicity is critically dependent excretion by Abcc2, Abcc3, and Abcg2. Together, these on the function of just one detoxifying protein bear consid- processes result in a reduced exposure of the body to the erably higher risks, and they should probably be avoided in highly toxic 7OH-MTX. It is therefore possible that de- clinical practice. Nevertheless, as many drugs (and food creased activity and/or expression of ABCC2, ABCC3, components) may be substrates or inhibitors of these three and/or ABCG2 may lead to accumulation of MTX and ABC transporters, coadministration of MTX with drugs that

Mol Cancer Ther 2009;8(12). December 2009

3358 Methotrexate in Abcc2;Abcc3;Abcg2−/− Mice

are also ABCC2, ABCC3, and ABCG2 substrates should be Acknowledgments done with caution. Furthermore, many mutations and poly- We thank Dilek Iusuf, Evita van de Steeg, Jurjen Lagas, and Robert van Waterschoot for critical reading of the manuscript; Rob Lodewijks, Enver morphisms of ABCC2 and ABCG2 that can influence drug Delic, and Hans Tensen for excellent technical assistance; Martin van der pharmacokinetics exist (17, 36). Especially in patients with Valk, Ji-Ying Song, and Nadine Meertens for histologic analysis; and impaired function of ABCC2 and/or ABCG2, the effect of George Scheffer and Jean-Marc Fritschy for kindly providing antibodies. ABCC3 on MTX and 7OH-MTX pharmacokinetics may be more important than previously anticipated. References − − Interestingly, in Abcc3;Abcg2 / mice, the fecal excretion of 1. Borst P, Oude Elferink RP. Mammalian ABC transporters in health and disease. Annu Rev Biochem 2002;71:537–92. MTX and 7OH-MTX was significantly higher than in wild- 2. Borst P, Zelcer N, van de Wetering K. MRP2 and 3 in health and disease. type mice. The mildly increased liver accumulation of MTX Cancer Lett 2006;234:51–61. and 7OH-MTX early after administration in these mice due 3. Schinkel AH, Jonker JW. Mammalian drug efflux transporters of the to combined absence of Abcc3 and Abcg2 in the liver (Sup- ATP binding cassette (ABC) family: an overview. Adv Drug Deliv Rev – plementary Table S3) may have caused increased excretion 2003;55:3 29. 4. Kruh GD, Belinsky MG, Gallo JM, Lee K. Physiological and pharmaco- of these compounds into the intestine over 24 hours, most logical functions of Mrp2, Mrp3 and Mrp4 as determined from recent stud- likely via Abcc2 (13). In addition, absence of Abcc3 in the ies on gene-disrupted mice. Cancer Metastasis Rev 2007;26:5–14. intestinal wall of these mice may lead to reduced reabsorp- 5. Deeley RG, Westlake C, Cole SP. Transmembrane transport of endo- tion of MTX and perhaps 7OH-MTX from the intestine (13). and xenobiotics by mammalian ATP-binding cassette multidrug resistance − − proteins. Physiol Rev 2006;86:849–99. As a consequence, enterocytes of Abcc3;Abcg2 / mice may 6. van Herwaarden AE, Schinkel AH. The function of breast cancer resis- becomeexposedtohigherlevelsofMTXandthesemice tance protein in epithelial barriers, stem cells and milk secretion of drugs could therefore be more prone to gastrointestinal toxicity. and xenotoxins. Trends Pharmacol Sci 2006;27:10–6. As this could mean that patients with decreased ABCC3 ex- 7. Assaraf YG. The role of multidrug resistance efflux transporters in pression or activity in the intestine may be more susceptible antifolate resistance and folate homeostasis. Drug Resist Updat 2006; 9:227–46. to gastrointestinal toxicity (a common side effect of MTX 8. Zamek-Gliszczynski MJ, Nezasa K, Tian X, et al. Evaluation of the role treatment (37), it would be interesting to test this hypothesis of multidrug resistance-associated protein (Mrp) 3 and Mrp4 in hepatic ba- in more detail in the future. solateral excretion of sulfate and glucuronide metabolites of acetamino- phen, 4-methylumbelliferone, and harmol in Abcc3−/− and Abcc4−/− Kidney failure is another common toxicity in patients mice. J Pharmacol Exp Ther 2006;319:1485–91. treated with high doses of MTX (38). This is thought to be 9. Zamek-Gliszczynski MJ, Hoffmaster KA, Nezasa K, Tallman MN, primarily due to the formation of crystalline deposits of Brouwer KL. Integration of hepatic drug transporters and phase II metab- MTX and especially of the poorly water-soluble metabolite olizing enzymes: mechanisms of hepatic excretion of sulfate, glucuronide, and glutathione metabolites. Eur J Pharm Sci 2006;27:447–86. 7OH-MTX in renal tubules (37). Because in the Abcc2;Abcc3; −/− 10. Wielinga P, Hooijberg JH, Gunnarsdottir S, et al. The human multidrug Abcg2 mice the formation of 7OH-MTX and, as a result, resistance protein MRP5 transports folates and can mediate cellular resis- the urinary excretion of 7OH-MTX was dramatically in- tance against antifolates. Cancer Res 2005;65:4425–30. creased compared with wild-type mice [and the other (com- 11. Chen ZS, Lee K, Walther S, et al. Analysis of methotrexate and folate pound) knockout strains], this suggests that ABCC2, transport by multidrug resistance protein 4 (ABCC4): MRP4 is a compo- nent of the methotrexate efflux system. Cancer Res 2002;62:3144–50. ABCC3, and ABCG2 together are very important in the pro- 12. Vlaming ML, Pala Z, van Esch A, et al. Functionally overlapping roles tection of the kidneys from 7OH-MTX exposure and that of Abcg2 (Bcrp1) and Abcc2 (Mrp2) in the elimination of methotrexate and they may prevent MTX- and 7OH-MTX-related toxicities 7-hydroxymethotrexate in vivo. Clin Cancer Res 2009;15:3084–93. in patients. 13. Vlaming ML, Pala Z, van Esch A, et al. Impact of Abcc2 (Mrp2) and Abcc3 (Mrp3) on the in vivo elimination of methotrexate and its main toxic For practical reasons, this study was done in female mice. metabolite 7-hydroxymethotrexate. Clin Cancer Res 2008;14:8152–60. In females, the liver expression of Abcg2/ABCG2 protein is 14. Kitamura Y, Hirouchi M, Kusuhara H, Schuetz JD, Sugiyama Y. In- lower than in males (39), and the effect of Abcg2 on MTX creasing systemic exposure of methotrexate by active efflux mediated and 7OH-MTX pharmacokinetics may therefore be lower by multidrug resistance-associated protein 3 (mrp3/abcc3). J Pharmacol Exp Ther 2008;327:465–73. as well. Whether this is indeed the case should be investi- 15. Masuda M, I'izuka Y, Yamazaki M, et al. Methotrexate is excreted into gated in future studies. the bile by canalicular multispecific organic anion transporter in rats. Can- The data presented here illustrate the dramatic changes in cer Res 1997;57:3506–10. the pharmacokinetics of both MTX and 7OH-MTX when 16. Huang Y. Pharmacogenetics/genomics of membrane transporters in two or more ABC transporters are absent. In the future, cancer chemotherapy. Cancer Metastasis Rev 2007;26:183–201. these results may be helpful in predicting responses to 17. Maeda K, Sugiyama Y. Impact of genetic polymorphisms of transporters on the pharmacokinetic, pharmacodynamic and toxicological properties of MTX treatment in patients based on individual expression anionic drugs. Drug Metab Pharmacokinet 2008;23:223–35. or activity data of ABCC2, ABCC3, and/or ABCG2. Fur- 18. Rau T, Erney B, Gores R, Eschenhagen T, Beck J, Langer T. High-dose − − thermore, the Abcc2;Abcc3;Abcg2 / mice we generated will methotrexate in pediatric acute lymphoblastic leukemia: impact of ABCC2 polymorphisms on plasma concentrations. Clin Pharmacol Ther 2006;80: be excellent tools to determine the relative and overlapping 468–76. effects of these ABC transporters on the pharmacokinetics of 19. Ranganathan P, Culverhouse R, Marsh S, et al. Methotrexate (MTX) a wide range of existing and newly developed drugs. pathway gene polymorphisms and their effects on MTX toxicity in Cauca- sian and African American patients with rheumatoid arthritis. J Rheumatol 2008;35:572–9. Disclosure of Potential Conflicts of Interest 20. Warren RB, Smith RL, Campalani E, et al. Genetic variation in efflux transporters influences outcome to methotrexate therapy in patients with No potential conflicts of interest were disclosed. psoriasis. J Invest Dermatol 2008;128:1925–9.

Mol Cancer Ther 2009;8(12). December 2009

Molecular Cancer Therapeutics 3359

21. Vlaming ML, Mohrmann K, Wagenaar E, et al. Carcinogen and anti- 31. Tian X, Swift B, Zamek-Gliszczynski MJ, Belinsky MG, Kruh GD, cancer drug transport by Mrp2 in vivo: studies using Mrp2 (Abcc2) knock- Brouwer KL. Impact of basolateral multidrug resistance-associated protein out mice. J Pharmacol Exp Ther 2006;318:319–27. (Mrp) 3 and Mrp4 on the hepatobiliary disposition of fexofenadine in per- – 22. Zelcer N, van de Wetering K, de Waart R, et al. Mice lacking Mrp3 fused mouse livers. Drug Metab Dispos 2008;36:911 5. (Abcc3) have normal bile salt transport, but altered hepatic transport of 32. Gorlick R, Bertino JR. Clinical Pharmacology and Resistance to Dihy- endogenous glucuronides. J Hepatol 2006;44:768–75. drofolate Reductase Inhibitors. Antifolate Drugs in Cancer Therapy. Toto- – 23. Jonker JW, Buitelaar M, Wagenaar E, et al. The breast cancer resis- wa (NJ): Humana Press; 1999, p. 37 57. tance protein protects against a major chlorophyll-derived dietary photo- 33. Chu X, Strauss J, Mariano M, et al. Characterization of mice lacking toxin and protoporphyria. Proc Natl Acad Sci U S A 2002;99:15649–54. the multidrug resistance protein Mrp2 (Abcc2). J Pharmacol Exp Ther – 24. van de Wetering K, Zelcer N, Kuil A, et al. Multidrug resistance pro- 2006;317:579 89. teins 2 and 3 provide alternative routes for hepatic excretion of morphine- 34. Zeng H, Chen ZS, Belinsky MG, Rea PA, Kruh GD. Transport of meth- glucuronides. Mol Pharmacol 2007;72:387–94. otrexate (MTX) and folates by multidrug resistance protein (MRP) 3 and 25. Soontornmalai A, Vlaming ML, Fritschy JM. Differential, strain-specific MRP1: effect of polyglutamylation on MTX transport. Cancer Res 2001; – cellular and subcellular distribution of multidrug transporters in murine 61:7225 32. choroid plexus and blood-brain barrier. Neuroscience 2006;138:159–69. 35. De Graaf D, Sharma RC, Mechetner EB, Schimke RT, Roninson IB. 26. Ogihara H, Saito H, Shin BC, et al. Immuno-localization of H+/peptide P-glycoprotein confers methotrexate resistance in 3T6 cells with deficient cotransporter in rat digestive tract. Biochem Biophys Res Commun 1996; carrier-mediated methotrexate uptake. Proc Natl Acad Sci U S A 1996;93: – 220:848–52. 1238 42. 27. Jonker JW, Wagenaar E, Van Eijl S, Schinkel AH. Deficiency in the 36. Suzuki H, Sugiyama Y. Single nucleotide polymorphisms in multidrug organic cation transporters 1 and 2 (Oct1/Oct2 [Slc22a1/Slc22a2]) in resistance associated protein 2 (MRP2/ABCC2): its impact on drug dispo- – mice abolishes renal secretion of organic cations. Mol Cell Biol 2003;23: sition. Adv Drug Deliv Rev 2002;54:1311 31. 7902–8. 37. Pratt WB, Ruddon RW, Ensminger WD, Maybaum J. Antimetabolites. – 28. van Waterschoot RA, van Herwaarden AE, Lagas JS, et al. Midazolam The Anticancer Drugs. pp. 74 101New York-Oxford: Oxford University metabolism in cytochrome P450 3A knockout mice can be attributed to Press; 1994. up-regulated CYP2C enzymes. Mol Pharmacol 2008;73:1029–36. 38. Green MR, Chamberlain MC. Renal dysfunction during and after high- – 29. van Tellingen O, van der Woude HR, Beijnen JH, van Beers CJ, dose methotrexate. Cancer Chemother Pharmacol 2009;63:599 604. Nooyen WJ. Stable and sensitive method for the simultaneous deter- 39. Merino G, van Herwaarden AE, Wagenaar E, Jonker JW, Schinkel AH. mination of N5-methyltetrahydrofolate, leucovorin, methotrexate and Sex-dependent expression and activity of the ATP-binding cassette trans- 7-hydroxymethotrexate in biological fluids. J Chromatogr 1989;488: porter breast cancer resistance protein (BCRP/ABCG2) in liver. Mol Phar- 379–88. macol 2005;67:1765–71. 30. Bardelmeijer HA, Beijnen JH, Brouwer KR, et al. Increased oral bio- 40. Zelcer N, Reid G, Wielinga P, et al. Steroid and bile acid conjugates are availability of paclitaxel by GF120918 in mice through selective modula- substrates of human multidrug-resistance protein (MRP) 4 (ATP-binding tion of P-glycoprotein. Clin Cancer Res 2000;6:4416–21. cassette C4). Biochem J 2003;371:361–7.

Mol Cancer Ther 2009;8(12). December 2009

Abcc2 (Mrp2), Abcc3 (Mrp3), and Abcg2 (Bcrp1) are the main determinants for rapid elimination of methotrexate and its toxic metabolite 7-hydroxymethotrexate in vivo

Maria L. H. Vlaming, Anita van Esch, Zeliha Pala, et al.

Mol Cancer Ther 2009;8:3350-3359. Published OnlineFirst December 8, 2009.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-09-0668

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2009/12/15/1535-7163.MCT-09-0668.DC1

Cited articles This article cites 38 articles, 19 of which you can access for free at: http://mct.aacrjournals.org/content/8/12/3350.full#ref-list-1

Citing articles This article has been cited by 11 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/8/12/3350.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/8/12/3350. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.