Inhibitory Effect of Selective Cyclooxygenase-2 Inhibitor Lumiracoxib on Human Organic Anion Transporters Hoat1 and Hoat3

Home , Aminopterin

Drug Metab. Pharmacokinet. 25 (5): 450–455 (2010). Regular Article Inhibitory Effect of Selective Cyclooxygenase-2 Inhibitor Lumiracoxib on Human Organic Anion Transporters hOAT1 and hOAT3

Yuichi UWAI*,HiroakiHONJO and Kikuo IWAMOTO Laboratory of Clinical Pharmacodynamics, School of Pharmacy, Aichi Gakuin University, Nagoya, Japan

Full text of this paper is available at http://www.jstage.jst.go.jp/browse/dmpk

Summary: Nonsteroidal anti-inflammatory drugs (NSAIDs) delay renal excretion of antifolate methotrexate by inhibiting human organic anion transporters hOAT1 (SLC22A6) and hOAT3 (SLC22A8). In this study, we performed uptake experiments using Xenopus laevis oocytes to assess the inhibitory effect of selective cyclooxygenase-2 inhibitors on hOAT1 and hOAT3. The uptake of methotrexate into oocytes was increased by the injection of hOAT1 and hOAT3 cRNA, and transport was strongly inhibited by lumiracoxib. The apparent 50z inhibitory concentrations of lumiracoxib were estimated to be 3.3 mM and 1.9 mM for uptake of p- aminohippurate by hOAT1 and of estrone sulfate by hOAT3, respectively. Eadie-Hofstee plot analysis showed that lumiracoxib inhibited hOAT1 and hOAT3 in a competitive manner. For other cyclooxygenase-2 inhibitors celecoxib, etoricoxib, rofecoxib and valdecoxib, slight to moderate inhibition of hOAT3 only was observed. These findings show that lumiracoxib has inhibitory potential toward hOAT1 and hOAT3, comparable to that of nonselective NSAIDs.

Keywords: hOAT1; hOAT3; selective cyclooxygenase-2 inhibitor; lumiracoxib; methotrexate; transport; drug interaction

(SLC22A8) are thought to be responsible for the renal Introduction tubular uptake of methotrexate from blood, because they Organic anion transporters expressed in the renal have been shown to be expressed in the basolateral proximal tubule regulate blood concentrations of various membrane of proximal epithelial cells and to transport drugs by mediating tubular secretion of antibiotics, an- methotrexate.5–7) In addition, several laboratories have tivirals, diuretics and antitumor agents.1) Accordingly, reported that NSAIDs such as ibuprofen, indomethacin, renal organic anion transporters result in drug interac- ketoprofen, loxoprofen, phenylbutazone, piroxicam and tion when their substrate and inhibitor are co-ad- salicylate have strong inhibitory effects on methotrexate ministered to a patient. For instance, it is well known uptake by hOAT1 and hOAT3,6,7) suggesting that hOAT1 that probenecid delays renal excretion of cephalosporins, and hOAT3 are involved, at least in part, in the interac- and it is accepted that renal organic anion transporters tion between methotrexate and NSAIDs, and this eluci- are involved in this drug interaction.2) Among the drug dation at the molecular level is useful for establishing interactions involving renal organic anion transporters, optimal drug therapy. the interaction between antifolate methotrexate and non- Recently, selective cyclooxygenase-2 inhibitors have steroidal anti-inflammatory drugs (NSAIDs) is thought to been developed, and these agents are increasingly be one of the most serious, because this interaction may prescribed instead of conventional NSAIDs. As shown in cause severe toxicity or death when high-dose Figure 1, celecoxib and valdecoxib possess a sul- methotrexate is used to treat malignancies.3,4) fonamide group, and etoricoxib and rofecoxib have a In the mid-1990s, cDNA of renal organic anion methylsulfone moiety. Lumiracoxib is weakly acidic be- transporters was isolated, and since then its function and cause of a carboxylic acid group. Although the influence expression have been characterized. Human organic of selective cyclooxygenase-2 inhibitors on hOAT1 and anion transporters hOAT1 (SLC22A6) and hOAT3 hOAT3 is of interest and its evaluation is important, to

Received; May 18, 2010, Accepted; July 2, 2010, J-STAGE Advance Published Date: September 22, 2010 *To whom correspondence should be addressed: Yuichi UWAI,Ph.D.,Laboratory of Clinical Pharmacodynamics, School of Pharmacy, Aichi Gakuin University, 1-100, Kusumoto-cho, Chikusa-ku, Nagoya-shi 464-8650, Japan. Tel. ＋81-52-757-6765, Fax. ＋81-52-757-6799, E-mail: yuuwai＠dpc.agu.ac.jp

450 Inhibition of hOAT1 and hOAT3 by Lumiracoxib 451

Fig. 1. Chemical structures of selective cyclooxygenase-2 inhibitors our knowledge, no report on the subject has been pub- gentamicin at 189C. Three days after injection, the up- lished. In the present study, we assessed the inhibitory ef- take reaction was initiated by incubating the oocytes in fects of selective cyclooxygenase-2 inhibitors on hOAT1 500 mL uptake buffer (96 mM NaCl, 2 mM KCl, 1.8 mM and hOAT3 by performing uptake experiments using the CaCl2,1mMMgCl2 and 5 mM HEPES; pH 7.4) with each Xenopus laevis oocyte expression system. radiolabeled compound at room temperature in the absence or presence of a selective cyclooxygenase-2 inhibi- Materials and Methods tor for the indicated periods. Celecoxib, etoricoxib, Materials: [3H]Methotrexate (27.7 Ci/mmol) was rofecoxib and valdecoxib were dissolved in dimethyl sul- purchased from Moravek Biochemicals (Brea, CA, USA). foxide, and the final concentration in uptake buffer was [3H]p-Aminohippurate (4.53 Ci/mmol) and [3H]estrone 0.1%. The uptake reaction was terminated by adding 2 sulfate (54.3 Ci/mmol) were obtained from PerkinElmer mL ice-cold uptake buffer to each well, and the oocytes Life Science (Boston, MA, USA). Lumiracoxib was from were washed three times with 2 mL ice-cold buffer. After LKT Laboratories, Inc. (St. Paul, MN, USA). Celecoxib, washing, each oocyte was transferred to a scintillation etoricoxib, rofecoxib and valdecoxib were obtained from counting vial and solubilized in 150 mL of 10% sodium Toronto Research Chemicals Inc. (Toronto, Canada). Un- lauryl sulfate. Two milliliters of scintillation cocktail labeled p-aminohippurate and estrone sulfate were pur- Clear-sol II (Nacalai Tesque, Kyoto, Japan) were added to chased from Wako Pure Chemical Industries (Osaka, each solubilized oocyte, and radioactivity was deter- Japan) and Sigma-Aldrich (St. Louis, MO, USA), respec- mined using a liquid scintillation counter. tively. All other chemicals used were of the highest purity Kinetic analysis: The apparent 50% inhibitory con- available. centration (IC50) of lumiracoxib for hOAT1 and hOAT3 Uptake experiment using Xenopus laevis oocytes was estimated by non-linear least-squares regression anal- expressing hOAT1 or hOAT3: pBK-CMV plasmid ysis of the competition curve with a one-compartment vectors containing hOAT1 or hOAT3 cDNA were a kind model according to the following equation: A＝100× gift from Prof. Ken-ichi Inui (Kyoto University Hospital, IC50/(IC50＋[I])＋B,whereAistheuptakeamountofp- Kyoto, Japan). An uptake experiment using Xenopus laevis aminohippurate or estrone sulfate (% of control), [I] is the oocytes was performed as previously reported.8) Briefly, concentration of lumiracoxib, and B is the non-specific capped RNA encoding hOAT1 or hOAT3 was tran- organic anion uptake (% of control). scribed from Xba I-linearized pBK-CMV containing The kinetic parameters of p-aminohippurate transport hOAT1 or hOAT3, respectively, with T3 RNA poly- by hOAT1 and of estrone sulfate transport by hOAT3 merase. After 50 nL water or cRNA (25 ng) was injected were calculated using non-linear least-squares regression into defolliculated oocytes, the oocytes were maintained analysis from the following Michaelis-Menten equation: V in modified Barth's medium (88 mM NaCl, 1 mM KCl, ＝Vmax×[S]/(Km＋[S]), where V is the transport rate 0.33 mM Ca(NO3)2, 0.4 mM CaCl2, 0.8 mM MgSO4,2.4 (pmol/oocyte/hr), Vmax is the maximum velocity by the mM NaHCO3 and 5 mM HEPES) containing 50 mg/L saturable process (pmol/oocyte/hr), [S] is the concentra- 452 Yuichi UWAI, et al.

tion of p-aminohippurate or estrone sulfate (mM) and Km uptake amounts increased in a time-dependent manner. is the Michaelis-Menten constant (mM). In the presence of lumiracoxib at 100 mM, the accumula- Statistical analysis: Data were analyzed by the un- tion of methotrexate in oocytes expressing hOAT1 and paired t test or planned comparison using the Bonferroni hOAT3 was reduced to that of the control oocytes (Fig. test with GraphPad Prism, version 5.0 (GraphPad 2). This finding means that lumiracoxib inhibited the Software, San Diego, CA, USA). Differences were consi- transport of methotrexate by hOAT1 and hOAT3. dered significant at Pº0.05. Concentration dependence of inhibitory effects of lumiracoxib on hOAT1 and hOAT3: To exa- Results mine the influence of lumiracoxib on hOAT1 and Transport of methotrexate by hOAT1 and hOAT3 in detail, the dose dependence of the inhibitory hOAT3 in the presence of lumiracoxib: To inves- effect of lumiracoxib was investigated. As tracers for tigate whether the transport of methotrexate by hOAT1 hOAT1 and hOAT3, their typical substrates, p-aminohip- and hOAT3 was affected by lumiracoxib, we measured purate and estrone sulfate, respectively, were used in- the accumulation of methotrexate in oocytes exposed to stead of methotrexate. As shown in Figure 3,hOAT1 the compound. The injection of hOAT1 and hOAT3 and hOAT3 were inhibited by lumiracoxib in a dose- cRNA stimulated the uptake of methotrexate, and the dependent manner. The apparent IC50 values were calcu-

Fig. 2. Time-dependent uptake of methotrexate by hOAT1 (A) and by hOAT3 (B) in the presence of lumiracoxib (A) Water-injected oocytes (open circle) were incubated with 36.1 nM [3H]methotrexate for the indicated periods. Oocytes injected with hOAT1cRNAwereincubatedwith36.1nM[3H]methotrexate for the indicated periods in the absence (closed circle) or presence (closed triangle) of 100 mM lumiracoxib. (B) Water-injected oocytes (open circle) were incubated with 36.1 nM [3H]methotrexate for the indicated periods. Oocytes injected with hOAT3 cRNA were incubated with 36.1 nM [3H]methotrexate for the indicated periods in the absence (closed circle) or presence (closed triangle) of 100 mM lumiracoxib. The uptake amounts of [3H]methotrexate in each oocyte were determined. Each point represents the mean±S.E.M.of9to10oocytes.

Fig. 3. Concentration dependence of inhibitory effects of lumiracoxib on p-aminohippurate uptake by hOAT1 (A) and on estrone sulfate uptake by hOAT3 (B) (A) Oocytes injected with hOAT1 cRNA were incubated with 221 nM [3H]p-aminohippurate in the absence (control) or presence of lumiracoxib at various concentrations for 1 hr. (B) Oocytes injected with hOAT3 cRNA were incubated with 18.4 nM [3H]estrone sulfate in the absence (control) or presence of lumiracoxib at various concentrations for 1 hr. The uptake amounts of [3H]p-aminohippurate or [3H]estrone sulfate in each oocyte were determined and represented as % of control. Each point represents the mean±S.E.M.of9to10oocytes. Inhibition of hOAT1 and hOAT3 by Lumiracoxib 453

Fig. 4. Concentration-dependent uptake of p-aminohippurate by hOAT1 (A) and of estrone sulfate by hOAT3 (B) in the presence of lumiracoxib (A) Oocytes injected with hOAT1 cRNA were incubated with [3H]p-aminohippurate at various concentrations for 1 hr in the absence (open circle) or presence (closed circle) of 20 mM lumiracoxib. hOAT1-mediated uptake of [3H]p-aminohippurate was determined by subtracting its uptake amount in water-injected oocytes from that in oocytes injected with hOAT1 cRNA. (B) Oocytes injected with hOAT3 cRNA were incubated with [3H]estrone sulfate at various concentrations for 1 hr in the absence (open circle) or presence (closed circle) of 5 mM lumiracoxib. hOAT3-mediated uptake of [3H]estrone sulfate was determined by subtracting its uptake amount in water-injected oocytes from that in oocytes injected with hOAT3 cRNA. Inset: Eadie-Hofstee plots of the data; V, uptake rate (pmol/oocyte/hr); S, concentration of [3H]p-aminohippurate or [3H]estrone sulfate (mM). Each point represents the mean±S.E.M. of 8 to 10 oocytes.

lated to be 3.3±1.1 mM and 1.9±0.4 mM for hOAT1 valdecoxib on methotrexate transport by hOAT1 and hOAT3, respectively (mean±S.E.M. from 3 indepen- and hOAT3: Figure 5 shows the influence of dent experiments). celecoxib, etoricoxib, rofecoxib and valdecoxib on Inhibition manner of lumiracoxib toward hOAT1 and hOAT3. No statistical change in hOAT1- hOAT1 and hOAT3: To elucidate the inhibition mediated uptake of methotrexate by any cyclooxygenase- mode of lumiracoxib against hOAT1 and hOAT3, Eadie- 2 inhibitor was detected by planned comparison with the Hofstee plot analysis was performed. As shown in Figure Bonferroni test. On the other hand, celecoxib slightly but 4A, lumiracoxib affected the slope of the Eadie-Hofstee significantly inhibited hOAT3 (Pº0.05), and, moderate plots of hOAT1-mediated p-aminohippurate transport, inhibition of hOAT3 was observed in the presence of butitsy-interceptwasnotinfluenced.Furthermore,the etoricoxib, rofecoxib and valdecoxib (Pº0.001). K value of p-aminohippurate transport by hOAT1 was m Discussion significantly increased by lumiracoxib from 3.6±0.2 to 13.3±0.5 mM(mean±S.E.M. from 3 separate experi- NSAIDs are some of the most commonly prescribed 9) ments, Pº0.0001). On the other hand, the Vmax values medicines, but they often induce drug interactions. In were calculated to be 7.1±0.5 and 6.3±0.5 particular, severe toxicity has been observed by co-ad- pmol/oocyte/hr in the absence or presence of lumiracox- ministration of NSAIDs and high-dose methotrexate, and ib, respectively, and no statistical difference between one of the reasons is thought to be that NSAIDs prevent these values was recognized (P＝0.3701). renal tubular uptake of methotrexate by hOAT1 and The results for hOAT3 showed a similar tendency to hOAT3.6,7) Accordingly, it is important to find or explore hOAT1 (Fig. 4B). The slope of the Eadie-Hofstee plot of NSAIDs which do not affect drug metabolic enzymes and estrone sulfate uptake by hOAT3 was influenced by lu- drug transporters. In this study, we examined the interac- miracoxib. The Km values were estimated to be 4.7±1.2 tion of selective inhibitors of cyclooxygenase-2 with and 12.0±1.9 mM(mean±S.E.M. from 3 separate ex- hOAT1 and hOAT3, and it was shown that lumiracoxib periments) in the absence or presence of lumiracoxib, re- inhibited the uptake of methotrexate by these transport- spectively, and the difference was statistically significant ers. In addition, inhibition of hOAT3 by celecoxib, (P＝0.0315). On the other hand, no significant change of etoricoxib, rofecoxib and valdecoxib was observed. the Vmax values was observed (without lumiracoxib, 6.2± To obtain more information on the inhibition of 0.7; with lumiracoxib, 6.4±0.5 pmol/oocyte/hr; P value hOAT1 and hOAT3 by lumiracoxib, we examined the ＝0.8370). These findings imply that lumiracoxib inhibit- IC50 and inhibition mode of lumiracoxib. The IC50 was edhOAT1andhOAT3inacompetitivemanner. calculated to be 3.3 mM and 1.9 mM for hOAT1 and Effect of celecoxib, etoricoxib, rofecoxib and hOAT3, respectively. These values are comparable to 454 Yuichi UWAI, et al.

Fig. 5. Effect of various cyclooxygenase-2 inhibitors on methotrexate uptake by hOAT1 (A) and hOAT3 (B) (A) Water-injected oocytes were incubated with 36.1 nM [3H]methotrexate for 1 hr. Oocytes injected with hOAT1 cRNA were incubated with 36.1 nM [3H]methotrexate for 1 hr in the absence or presence of celecoxib, etoricoxib, rofecoxib and valdecoxib at the indicated concentrations with 0.1% dimethyl sulfoxide. (B) Water-injected oocytes were incubated with 36.1 nM [3H]methotrexate for 1 hr. Oocytes injected with hOAT3 cRNA were incubated with 36.1 nM [3H]methotrexate for 1 hr in the absence or presence of celecoxib, etoricoxib, rofecoxib and valdecoxib at the indicated concentrations with 0.1% dimethyl sulfoxide. The uptake amounts of [3H]methotrexate in each oocyte were determined. Each column represents the mean±S.E.M. of 9 to 10 oocytes. *P valueº0.05, significantly different from the values of hOAT3 cRNA-injected oocytes. ***P valueº0.001, significantly different from the values of hOAT3 cRNA-injected oocytes.

IC50 or Ki values for the transporters of traditional the results shown in Figure 3, it is suggested that NSAIDs, such as diclofenac, indomethacin, ibuprofen, meaningful inhibition of hOAT1 and hOAT3 by lu- ketoprofen, naproxen, loxoprofen and tolmetin.7,10) miracoxib may not occur with a 400-mg daily dose. Har- Eadie-Hofstee plot analysis showed that the inhibition tmann et al. reported that no effect on methotrexate manner of lumiracoxib toward hOAT1 and hOAT3 is pharmacokinetics was observed with a 400 mg daily dose competitive (Fig. 4). Takeda et al. and Maeda et al. of lumiracoxib.17) They also measured the plasma concen- reported that NSAIDs including indomethacin, trations of lumiracoxib, and reported that its mean maxi- mefenamic acid, sulindac and pranoprofen inhibited mum concentration was 7175 ng/mL (24.4 mM), and that hOAT3 in a competitive manner.6,11) It seemed that there its trough concentrations with median values ranged be- was no difference between lumiracoxib and nonselective tween 28.5 (97.0 nM) and 53.1 ng/mL (181 nM17)). The NSAIDs in the inhibitory characteristics toward hOAT1 maximum unbound concentration of lumiracoxib was and hOAT3. calculated to be 488 nM in these patients. The absence of It was also of interest whether lumiracoxib inhibits interaction between methotrexate and lumiracoxib in the hOAT1 and hOAT3 in the clinical setting. Compared clinical trial by Hartmann et al. is considered to be with ibuprofen and naproxen, advantages of lumiracoxib reasonable. Lumiracoxib is mainly metabolized by have been shown in terms of gastrointestinal and cytochrome P450 (CYP) 2C9,15) and polymorphisms exist cardiovascular safety12,13); however, the countries where in its gene, affecting the metabolic activity. Using insect lumiracoxib is prescribed are limited due to its hepato- cell microsomes, Tang et al.18) showed that allelic variants toxicity.14) Lumiracoxib shows pharmacological effects at reduced celecoxib hydroxylase activity. The possibility a daily dosage of 100 mg to 400 mg,15) and Scott et al. should not be excluded that the unbound level of lu- reported that its concentration in plasma at the steady miracoxib could reach the inhibitory range of hOAT1 state ranged from 141 mg/L (0.5 mM) to 6635 mg/L (22.6 and hOAT3 when the cyclooxygenase-2 inhibitor is ad- mM) when 400 mg lumiracoxib was orally administered ministered at the usual dosage to patients with low once daily to patients with rheumatoid arthritis.16) From CYP2C9 activity, followed by its mutation, hepatic Figure 3, it can be seen that lumiracoxib in this concen- failure and drug interaction. However, the frequencies of tration range inhibits hOAT1 and hOAT3 moderately; the allele with reduced CYP2C9 activity are reported to however, protein binding of lumiracoxib in plasma be very low, suggesting that the incidence of interaction should be considered, and is estimated to be 98%.16) Us- between methotrexate and lumiracoxib in patients with ing this percentage, the free level of lumiracoxib was cal- CYP2C9 mutation would be rare, if it occurred at all.19) culated to be 9.6 nM to 450 nM in plasma. Taken Figure 5 shows that methotrexate uptake by hOAT1 together with the unbound plasma concentrations and was not inhibited by celecoxib, etoricoxib, rofecoxib or Inhibition of hOAT1 and hOAT3 by Lumiracoxib 455 valdecoxib. The tested concentrations of the compounds organic anion transporters hOAT1 and hOAT3: comparison were comparable with the IC50 of lumiracoxib for with methotrexate. Drug Metab. Pharmacokinet., 25: 163–169 hOAT1, meaning that the inhibitory effects of the 4 cy- (2010). clooxygenase-2 inhibitors on hOAT1 are weaker than 9) Brouwers, J. and de Smet, P.: Pharmacokinetic-phar- that of lumiracoxib. On the other hand, inhibition of macodynamic drug interactions with nonsteroidal anti-inflam- hOAT3 by celecoxib, etoricoxib, rofecoxib and valdecox- matory drugs. Clin. Pharmacokinet., 27: 462–485 (1994). 10) Nozaki, Y., Kusuhara, H., Kondo, T., Iwaki, M., Shiroyanagi, Y., ib was observed. Due to the very low solubility of Nakayama, H., Horita, S., Nakazawa, H., Okano, T. and Sugiya- celecoxib, etoricoxib, rofecoxib and valdecoxib in the ma, Y.: Species difference in the inhibitory effect of nonsteroidal uptake buffer used in the present study, unfortunately, anti-inflammatory drugs on the uptake of methotrexate by hu- we could not perform a more detailed examination of man kidney slices. J. Pharmacol. Exp. Ther., 322: 1162–1170 their inhibitory effects. However, these phenomena are (2007). interesting, because no difference in the inhibitory effect 11) Maeda, A., Tsuruoka, S., Kanai, Y., Endou, H., Saito, K., on hOAT1 and hOAT3 was recognized with anionic lu- Miyamaoto, E. and Fujimura, A.: Evaluation of the interaction miracoxib. between nonsteroidal anti-inflammatory drugs and In conclusion, this study shows that lumiracoxib ex- methotrexate using human organic anion transporter 3-trans- hibited an inhibitory effect on hOAT1 and hOAT3. The fected cells. Eur. J. Pharmacol., 596: 166–172 (2008). IC values and inhibition mode of lumiracoxib are simi- 12) Schnitzer, T. J., Burmester, G. R., Mysler, E., Hochberg, M. C., 50 Doherty, M., Ehrsam, E., Gitton, X., Krammer, G., Mellein, B., lar to those of conventional NSAIDs. The clinical phar- Matchaba,P.,Gimona,A.andHawkey,C.J.:TARGETStudy macokinetics of lumiracoxib suggest that optimal ad- Group: Comparison of lumiracoxib with naproxen and ministration would not induce interactions between the ibuprofen in the therapeutic arthritis research and gastrointesti- selective cyclooxygenase-2 inhibitor and the substrates of nal event trial (TARGET), reduction in ulcer complications: ran- hOAT1 and hOAT3, including methotrexate. In addi- domised controlled trial. Lancet, 364: 665–674 (2004). tion, hOAT3 was inhibited by celecoxib, etoricoxib, 13) Farkouh, M. E., Kirshner, H., Harrington, R. A., Ruland, S., rofecoxib and valdecoxib. Verheugt, F. W., Schnitzer, T. J., Burmester, G. R., Mysler, E., Hochberg,M.C.,Doherty,M.,Ehrsam,E.,Gitton,X.,Kram- Acknowledgments: We thank Prof. Ken-ichi Inui mer, G., Mellein, B., Gimona, A., Matchaba, P., Hawkey, C. J. (Kyoto University Hospital, Kyoto, Japan) for kindly and Chesebro, J. H.: TARGET Study Group: Comparison of lu- providing pBK-CMV plasmid vectors containing hOAT1 miracoxib with naproxen and ibuprofen in the therapeutic arthritis research and gastrointestinal event trial (TARGET), or hOAT3. cardiovascular outcomes: randomised controlled trial. Lancet, References 364: 675–684 (2004). 14) Laine,L.,White,W.B.,Rostom,A.andHochberg,M.:COX-2 1) El-Sheikh, A. A., Masereeuw, R. and Russel, F. G.: Mechanisms selective inhibitors in the treatment of osteoarthritis. Semin. of renal anionic drug transport. Eur. J. Pharmacol., 585: Arthritis Rheum., 38: 165–187 (2008). 245–255 (2008). 15) Rordorf, C. M., Choi, L., Marshall, P. and Mangold, J. B.: Clini- 2) Shitara, Y., Sato, H. and Sugiyama, Y.: Evaluation of drug-drug cal pharmacology of lumiracoxib: a selective cyclo-oxygenase-2 interaction in the hepatobiliary and renal transport of drugs. inhibitor. Clin. Pharmacokinet., 44: 1247–1266 (2005). Annu. Rev. Pharmacol. Toxicol., 45: 689–723 (2005). 16) Scott, G., Rordorf, C., Reynolds, C., Kalbag, J., Looby, M., 3) Thyss, A., Milano, G., Kubar, J., Namer, M. and Schneider, M.: Milosavljev, S., Weaver, M., Huff, J. P. and Ruff, D. A.: Phar- Clinical and pharmacokinetic evidence of a life-threatening in- macokinetics of lumiracoxib in plasma and synovial fluid. Clin. teraction between methotrexate and ketoprofen. Lancet, 327: Pharmacokinet., 43: 467–478 (2004). 256–258 (1986). 17) Hartmann,S.N.,Rordorf,C.M.,Milosavljev,S.,Branson,J.M., 4) Maiche, A. G.: Acute renal failure due to concomitant action of Chales,G.H.,Juvin,R.R.,Lafforgue,P.,LeParc,J.M.,Taverni- methotrexate and indomethacin. Lancet, 327: 1390 (1986). er, C. G. and Meyer, O. C.: Lumiracoxib does not affect 5) Motohashi, H., Sakurai, Y., Saito, H., Masuda, S., Urakami, Y., methotrexate pharmacokinetics in rheumatoid arthritis patients. Goto,M.,Fukatsu,A.,Ogawa,O.andInui,K.:Geneexpression Ann. Pharmacother., 38: 1582–1587 (2004). levels and immunolocalization of organic ion transporters in the 18) Tang,C.,Shou,M.,Rushmore,T.H.,Mei,Q.,Sandhu,P., human kidney. J. Am. Soc. Nephrol., 13: 866–874 (2002). Woolf, E. J., Rose, M. J., Gelmann, A., Greenberg, H. E., De 6)Takeda,M.,Khamdang,S.,Narikawa,S.,Kimura,H., Lepeleire, I., Van Hecken, A., De Schepper, P. J., Ebel, D. L., Hosoyamada, M., Cha, S. H., Sekine, T. and Endou, H.: Charac- Schwartz, J. I. and Rodrigues, A. D.: In-vitro metabolism of terization of methotrexate transport and its drug interactions celecoxib, a cyclooxygenase-2 inhibitor, by allelic variant forms with human organic anion transporters. J. Pharmacol. Exp. Ther., of human liver microsomal cytochrome P450 2C9: correlation 302: 666–671 (2002). with CYP2C9 genotype and in-vivo pharmacokinetics. Phar- 7) Uwai,Y.,Taniguchi,R.,Motohashi,H.,Saito,H.,Okuda,M. macogenetics, 11: 223–235 (2001). and Inui, K.: Methotrexate-loxoprofen interaction: involvement 19) Kirchheiner, J. and Brockmoller, J.: Clinical consequences of of human organic anion transporters hOAT1 and hOAT3. Drug cytochrome P450 2C9 polymorphisms. Clin. Pharmacol. Ther., Metab. Pharmacokinet., 19: 369–374 (2004). 77: 1–16 (2005). 8) Uwai, Y. and Iwamoto, K.: Transport of aminopterin by human