Leukemia (1999) 13, 578–584  1999 Stockton Press All rights reserved 0887-6924/99 $12.00 http://www.stockton-press.co.uk/leu Role of MRP1 in multidrug resistance in acute myeloid leukemia O Legrand, R Zittoun and J-P Marie

EA1529, Universite´ Paris 6, Formation de Recherche Claude Bernard, and Service d’He´matologie, Hoˆpital Hoˆtel-Dieu, AP-HP, Paris, France

The best characterized resistance mechanism in adult acute drugs among which are doxorubicin, daunorubicin, idarub- myeloid leukemia (AML) is the one mediated by the MDR1 gene icin, mitoxantrone and etoposide.13,14 At least in transfectants, which has been shown to be associated with poor outcome. However, alternative proteins such as the more recently recog- MRP1 is able to act as an energy-dependent drug pump, nized multidrug-associated protein (MRP1), may also contrib- mainly localized in the plasma membrane, that extrudes drugs ute to the resistance to anthracyclines and etoposide in AML. from the cell against a concentration gradient.14–17 It has been Recently, the role of this protein was discussed and was shown that MRP1 is identical to a GSH-S conjugate carrier unclear in AML. However, recent data concerning the func- described in many cells, because the endogenous leukotriene tionality and the modulation of the activity of MRP1 may eluci- C4 and other GSH-S conjugates are transported by MRP1.18–22 date its role in comparison with other mechanisms of resist- ance. In this paper, we will review these recent data concerning Recently, it was shown that ATP-dependent uptake of the the role of MRP1 in adult AML. unmodified drug vincristine by membrane vesicles derived Keywords: MDR; MRP1; acute myeloid leukemia; functional test; from MRP1 transfected HeLa cells is dependent on the pres- prognostic factor ence of GSH. It is possible that GSH interacts directly with MRP1 and that this interaction is necessary for transport. In addition, GSH depletion by buthionine sulphoximine (BSO) Introduction increases the accumulation of DNR and Rhodamine 123 (Rh123) in several MRP1-positive cell lines.21 Therefore, The use of the first cytostatic agents produced a dramatic MRP1 is a multiple organic anion transporter, it transports cer- improvement in the management of acute myeloid leukemia tain glutathione conjugates and may be partly dependent on (AML), but drug resistance eventually arose during chemo- intracellular glutathione levels for the transport of anthracy- therapy. Among different mechanisms of resistance, the clines. Marbeuf-Gueye et al23 studied the transport kinetics of appearance of the multidrug resistance (MDR) phenotype is a series of anthracyclines in tumor cell lines that overexpress the most frequently observed in many cancer cell lines MRP1 and Pgp. They showed that the transport kinetics of (including leukemia). The best characterized resistance mech- anthracyclines by MRP1 and Pgp are very similar except for anism is the one mediated by the MDR1 gene which has been OH-DOX and OH-DNR which had a lower VMax (maximal shown to be associated with poor outcome.1 However, alter- efflux rate) in the case of MRP1-mediated transport suggesting native proteins such as the multidrug-associated protein a role for the amino group in the interaction with glutathione. (MRP1), may also contribute to the resistance to anthracy- On the other hand, Lorico et al24 compared the sensitivity clines and etoposide in AML.2,3 The role of MRP1 is still being to chemotherapeutic drugs and toxins of wild-type W9.5 discussed and is unclear in AML.4–11 In this paper, we will embryonic stem cells and of single and double MRP1 gene review recent data concerning the role of MRP1 in adult AML. knockout cells derived there from. MRP1 expression was tot- ally abrogated in the double knockout cell line and partially abrogated in the single knockout cell line. RT-PCR analyses Structure and function of MRP1 demonstrated that MDR1 gene was not expressed. The cyto- toxic activities of etoposide, teniposide, vincristine, doxorub- The MRP1 gene is located on chromosome 16p13.1. MRP1, icin, daunorubicin and sodium arsenite were significantly like Pgp belongs to the ATP binding cassette (ABC) super- greater in double knockout cells than in parenteral wild-type family of membrane transport protein. It is a glycosylated ES cells. These findings indicate that baseline MRP1 transmembrane protein of 190 kDa.2 Bakos et al’s12 experi- expression has the capacity to exert a protective role against ment and the alignment of the MRP1 sequence with the the toxicity of multiple chemotherapeutic agents and natural human cystic fibrosis transmembrane conductance (CFTR) toxins. In the same way, deletion of the MRP1 gene was found suggest that human MRP1 contains a tandem repeat of six previously in five of 13 AML patients with inv(16) and was transmembrane helices, each followed by a nucleotide bind- associated with longer overall survival and disease-free sur- ing domain and that the C-terminal membrane-bound region vival.25 This aberration was suggested as being the reason for is glycosylated. The N-terminal region of MRP1 contains an the better prognosis of patients with inv(16) in this small series additional membrane-bound, glycosylated area with four or of patients. However, lack of MDR1 gene expression might five transmembrane helices, which seems to be a character- be an alternative explanation for the good prognosis of these istic feature of MRP-like ATP-binding cassette transporters. patients with inv(16).25 Transfection experiments with different eukaryotic Zaman et al26 have shown that in the MRP-overexpressing expression vectors containing full-length complementary cell line GLC4/ADR the MRP1 gene copy number is increased DNAs of the MRP1 gene have shown that MRP1 can confer, about 25-fold, as compared to the parental cell line. However, just like Pgp, resistance to a broad range of natural product no AML patients analyzed by Burger et al,27,28 including those who had high MRP mRNA expression showed evidence for MRP1 gene amplification. Therefore, they concluded that the increase in MRP1 expression in AML patients was not associa- Correspondence: O Legrand, Hoˆpital Hoˆtel Dieu, 1 place du parvis Notre Dame, Service d’he´matologie, 75 181 Paris Cedex 04, France; ted with amplification of the cognate gene. 29 Fax: 33 1 42 34 82 54 Ma et al demonstrated that MRP1 is highly phosphoryl- Received 29 September 1998; accepted 9 December 1998 ated and that phosphate groups are metabolically active and MRP1 in AML O Legrand et al 579 undergo cycles of phosphorylation and dephosphorylation in bone marrow cells, while we have shown that MRP1 activity the cell. Serine is the single amino acid phosphorylated in was higher in CD34− than in CD34+ leukemia cells.11 MRP1. The effect of H-7 and staurosporine (two protein kinase Several studies on MRP1 expression in AML have been inhibitors) can reduce the phosphorylation of this protein. It reported. The incidence of MRP1 gene overexpression in AML has also been found that, in the presence of these agents, there patients was obtained by RNA or protein techniques. Messen- is a major increase in drug accumulation and concomitant ger RNA was quantified using either RT-PCR or by RNase pro- inhibition in drug efflux of resistant cells. Therefore, certain tection assay. Protein was recently detected by immunocyto- phosphate groups of protein MRP1 play an important role in chemistry, Western blot and flow cytometry with monoclonal modulating drug accumulation in resistant cells. antibodies (MRPm6, MRPr1, QRL-1 and QRL-3).4,6–8,10,11,28,39–43 The DNA containing promoter activity has been sequenced All these antibodies recognize an internal epitope of in its entirety and found to contain multiple putative regulat- MRP1.16,17 Few comparative analyses between these tech- ory sites. Studies using a number of different deletion mutants niques have been published. Filipits et al44 showed that there suggest that the major promoter activity of MRP1 resides in a was a fair correlation between results obtained with QCRL- highly GC-rich region of 914 nucleotides (nucleosides −91 to 1/QCRL-3 and those obtained with MRPr1 by immunocytoch- +103). Promoter activity contained within this sequence, how- emistry for AML patients. In other studies we have demon- ever, is modulated by both positive and negative regulatory strated that three techniques (RT-PCR, in situ hybridation and elements.30 The functional analysis of the nucleotide binding protein detection by flow cytometry) gave similar results in domains of MRP1 identify several mutations that are critical normal hematopoietic cells and in fresh leukemic cells.11,39 for MRP-mediated multidrug resistance.31 Recently, besides MRP1 at least four other homologues of Analyses of the subcellular distribution of MRP1 in cell lines these proteins (these new genes were called MRP2 to MRP5) revealed that approximatively 80% of MRP1 in the SW-1573 have been identified.45 However, MRP3 and MRP5 were only and HeLa transfectants were associated with plasma mem- overexpressed in a few cell lines, and the RNA levels did not brane, while the comparable localization in the drug-selected seem to correlate with resistance to either doxorubicin or cis- H69AR cells contained only approximatively 50% of the pro- platin. MRP4 was not overexpressed in any of the cell lines tein. The remaining MRP1 was codistributed in endocytotic which were analyzed. MRP2 (or cMOAT) was substantially vesicles. The relatively high proportion of MRP1 associated overexpressed in several cell lines, and MRP2 RNA levels cor- with these fractions in H69AR cells may contribute to the lack relate with cisplatin but not doxorubicin resistance in a subset of an observable accumulation defect in these cells when of resistant cell lines. Sequence analysis revealed that the pro- compared with the transfectants.2 In AML cells MRP1 staining tein segment used to generate the MRPm6 antibody was (by immunocytochemistery) was primarily cytoplasmic.32 derived from the most homologous portion located in the C- Several publications have shown that the resistance of cell terminal part of MRP1. In contrast, a more MRP1-specific seg- lines selected by drugs is related to the sequential emergence ment in the N-terminal part of the protein was used for the of MRP1 and MDR1 gene overexpression. At clinically rel- MRPr1 antibody. According to den Boer et al46 this may evant concentrations of doxorubicin or homoharringtonine, in explain the absence of a significant correlation, for 168 chil- human myeloid leukemia cell lines, the resistance was related dren with leukemia, between the fluorescence index using to MRP1 overexpression, but not to MDR1 expression.33,34 MRPm6 and MRPr1. Only when cell lines were exposed to weak concentrations The percentage of MRP1 overexpression reported ranged for prolonged time periods or when selected for relatively from 7 to 30%. This is difficult to compare because of hetero- high-level drugs resistance, did Pgp/MDR1 overexpression genous patient populations and techniques (RT-PCR, immuno- become apparent. In the same way, in other cell lines (murine cytochemistry, flow cytometry, Western blot etc), and differ- leukemia, small cell lung cancer cells) Pgp expression ences with regard to the definition of overexpression. In occurred only in relatively high concentrations of drugs.35–37 addition, only one technique was used and the functionality In the light of these results, MRP1 gene overexpression was of MRP1 was not assessed in these studies.4,6–8,10,11,28,42,43 For probably an early event in the development of drug resistance, MDR1/Pgp different consensus recommendations showed and MRP1 and Pgp could be co-expressed in AML cells. These improved data reproducibility and therefore comparison of observations suggest that: (1) MRP1 overexpression is prob- results between different studies.47–49 For example, these rec- ably an early event in the development of drug resistance; ommendations emphasized the importance of the presen- (2) drug resistance is usually multifactoral. Therefore, clinical tation of data as a continuous variable and the need for mul- trials which modulate only Pgp might have limited success. tiple techniques to be employed to ensure accurate All these data suggest that the role of MRP1 in resistance to measurement of Pgp expression. Therefore, as for MDR1/Pgp, chemotherapy, and particularly in AML, need to be con- we hope that laboratories involved in this area of clinical sidered. Therefore, several studies assessed the expression, research will act quickly to develop guidelines for the tech- function and prognostic value of MRP1 in adult AML. Here niques used to determine the role of MRP1 in the AML resist- we will review all these data. ance to chemotherapy. On the other hand, in several studies, MRP1 expression was higher in AML at relapse compared to at the time of diag- MRP1 expression in AML nosis.4,6,7 An analysis of sequential MRP1 expression in 13 patients indicated an increase of expression at relapse in six For MDR1 several publications have shown that there was a patients as compared to the time of diagnosis.8 Schneider et correlation between Pgp activity and CD34 expression.38 al6 also showed that, in 62 patients, samples of AML obtained However, with regard to MRP1 the few results are conflicting at the time of relapse contained approximatively a two-fold and include very few samples. In normal haematopoietic bone higher level of MRP1. Analysis of 13 paired samples, the first marrow CD34 cells, the expression was similar to the basal obtained at diagnosis and the second at relapse, showed that level in parental HL60 cell line.39 In contrast, van der Kolk et MRP1 expression was increased at the time of relapse in more al40 found activity of MRP1 in normal CD34+ hematopoietic than 80% of patients,4 in contrast to unchanged level of MRP1 in AML O Legrand et al 580 Table 1 Different probes for evaluating the role of MRP1

Probes Specificity Characteristics

Rhodamine123 MRP1 The transport of Rh123 by MRP1 is a debated question Pgp Transported by MRP1 in a GSH-dependent manner Carboxyfluorescein diacetate MRP1 By binding directly to MRP1 Transported by MRP1 in a GSH-independent manner Calcein-AM/calcein MRP1/Pgp Calcein-AM is a substrate both of MRP1 and Pgp MRP1 Calcein is only a substrate of MRP1 By binding directly to MRP1 or Pgp Transported by MRP1 in a GSH-independent manner Daunorubicin MRP1 Transported by MRP1 in a GSH-dependent manner Pgp

MDR1 expression. For Zhou et al,7 the incidence of MRP1 be transported by MRP1 in a GSH-dependent manner.51–53 gene overexpression tended to be higher in relapsed patients However, according to other authors, Rh123 is not very than in newly diagnosed patients (38 vs 18%). efficiently effluxed by MRP1.11,54–56 But this discrepancy is perhaps due to technical problems. Indeed, Minderman et al57 showed that under a 1 h reincubation in drug-free medium, MRP1 activity in AML no retention difference of the Rh123 was detected between parenteral and MRP1-expressing resistant cells. But upon In several publications, the expression and function of Pgp did extending the reincubation time to 4 h, an efflux of Rh123 not correlate.1,50 For this reason, the functional role appears in the resistant lines became apparent, and was even more more informative than the quantification of these proteins in pronounced after 24 h.57 For Twentyman et al,53 COR-L23/P AML samples for this protein. Therefore, as for MDR1, func- and COR-L23/R cell lines accumulated Rh123 at similar rates tional MRP1 assays must be used to validate mRNA or protein for the first 30 min, then the curves diverged. For van der Kolk measurements and to quantify the effect of MRP1 or the mag- et al,40 Rh123 was specific in nonhematopoietic cells, but not nitude of the effect of a blocker of the MRP1-mediated drug in hematopoietic cells for studying Pgp-mediated export and efflux on the intracellular drug concentration. could be inhibited by MK571 in hematopoietic cells. How- The majority of data concerning functional tests has been ever, all these studies involved only cell lines, while in fresh obtained in cell lines and few data are available in clinical leukemic cells the results are debated. Van der Kolk40 found AML samples. However, in cancer cell lines, MRP1 activity that Rh123 was a substrate of MRP1, but we did not.11 These has been analyzed. Different compounds (Table 1) and modu- two studies did not use the same modulators (MK571 and lators (Table 2) are available to study MRP1 activity. probenecid, respectively). In van der Kolk’s study,40 there was Rh123 acts as a fluorescent substrate for Pgp but may also no correlation between MRP1 expression (measured by West-

Table 2 Modulators of MRP1

Modulators Specificity Analysed cells Action

Probenecid MRP1 Cell lines. By binding directly to MRP1 Fresh leukemic cells MK-571 MRP1 Cell lines. By binding directly to MRP1 Fresh leukemic cells BSO MRP1 Cell lines By reducing the intracellular glutathione (GSH) BCNU MRP1 Cell lines By inhibition of the GSH redox cycle and therefore reducing the intracellular GSH PAK-104P Pgp Cell lines By binding directly to MRP1 MRP1 VX-170 Pgp Cell lines By binding directly to MRP1 MRP1 Indomethacin MRP1 Cell lines Inhibition glutathione-S-transferase ± by binding directly to MRP1 Ethacrynic acid MRP1 Cell lines Inhibition glutathione-S-transferase ± by binding directly to MRP1 Imidazothiazole derivates MRP1 Cell lines (N276-12,-14,-17) NBD Non-specific Cell lines A vacuolar H+-ATPase inhibitor Genistein Non-specific Cell lines Specific inhibitor of tyrosine kinase MRP1 in AML O Legrand et al 581 ern blot) and Rh123 efflux-blocking factor of MK-571 even doxorubicin transport and restored intracellular concen- though a correlation was found between MRP1 expression trations in vitro in HT1080/DR4 cell line.67 and carboxyfluorescein diacetate (CF) F efflux blocking factor Another substrate, the diuretic drug ethacrynic acid, which of MK-571. is an efficient inhibitor of glutathione-S-transferases, is a high- CF and calcein-acetoxymethylester (calcein-AM) are non- affinity substrate and inhibitor of the glutathione-S-conjugate fluorescent compounds, which permeate the plasma mem- pump associated with MRP1. This implies that ethacrynic acid brane, and upon cleavage of the ester bonds by intracellular may modulate drug resistance of tumor cells not only by esterases are transformed into the fluorescent compound inhibiting glutathione S-transferase activity, but also by which might be a substrate for MRP1 but not for Pgp.58,59 inhibiting the export of drug conjugate from the cell by Thus, while glutathione may be involved in MRP1-mediated MRP1.68 Indomethacin is also reported as a specific inhibitor resistance to some chemotherapeutic agents, it is not neces- of MRP1 possibly functioning also by inhibition of gluta- sary for the efflux of substrates such as these two com- thione-S-transferase or, alternatively, by direct competition pounds.59,60 Another substrate, DiOC2(3), which is a substrate with the drug at the transport site.69 Pyridine analog, [2[4- of Pgp, is not a substrate of MRP1.57 (diphenylmethyl)-1-piperazinyl]ethyl 5(trans-4,6-dimethyl- Among some of these compounds are substrates of both 1,3,2-dioxyaphosphorinan-2-yl)-2,6-dimethyl-4-(3 nitrophenyl)- MRP1 and Pgp, such as calcein-AM, Rh123, DNR. Although 3-pyridinecarboxylate P-oxide] (PAK-104P) that reverses Pgp- MRP1 has been identified as an organic anion transporter and mediated MDR, directly interacted with MRP1 and completely Pgp as a transporter of certain positively charged compounds, reversed the vincristine resistance in MRP1-mediated MDR C- there was considerable overlap in resistance spectrum, sug- A120 cells.65,70 VX-710, a pipecolinate derivate (a novel non- gesting that both proteins transport important anticancer macrocyclic ligand of the FK-506-binding protein FKBP12) is agents in AML treatment such as the anthracyclines and eto- a potent modulator of Pgp-mediated multidrug resistance, but poside. Essodaigui et al61 performed a kinetic analysis of the also affects multidrug resistance in MRP1-expressing cells and efflux of calcein-AM, a common neutral substrate for both may exert its action, at least in part, by binding directly to 71 proteins, and compared it with the kinetics of efflux of calcein, MRP1. Three newly synthetized imidazothiazole derivatives which is only effluxed by MRP1. They showed that calcein is (N276-12, N276-14 and N276-17) can be useful chemosensit- 72 much less efficiently transported by MRP1 than by calcein- izing agents against MRP1 and Pgp. AM. Compared to previously published data for anthra- Other modulators such as genistein, a specific inhibitor of cyclines, the kinetic data for MRP-mediated calcein-AM tyrosine kinase, or 7-chloro-4-notrobenz-2-oxa-1,3-diazole + pumping are most similar to those for the neutral hydroxy- (NBD), a vacuolar H -ATPase, have been reported to modu- 73–75 daunorubicin. late MRP1 of a non-specific manner. However, genistein In the same way different compounds are available to showed major differences in effects on Rh123 vs DNR trans- modulate MRP1 activity (Table 2). A specific inhibitor of port in the MRP1-mediated MDR cell lines: the accumulation of DNR was increased, whereas the accumulation of Rh123 MRP1-mediated transport is MK-571, a leukotriene D4 recep- 73 tor antagonist.19 This compound has been shown to increase was decreased by genistein. Recently, van der Kolk et al40 and ourselves11 have demon- cytotoxicity of anticancer drugs in a MRP1-positive myeloid strated that MRP1 was functional in fresh acute myeloid leu- leukemic cell line. Probenecid, an inhibitor of organic anion kemic blast cells using CF and calcein-AM, respectively. We transport, reversed resistance to DNR in a concentration- have also seen that some substrates of MRP1 were also sub- dependent manner in HL60/AR and H69AR tumor cell lines strates of Pgp (calcein-AM, CF, ±Rh123 etc).76 When the com- which overexpressed MRP1.62 This effect of probenecid was bination of the two inhibitors MRP1 and Pgp was used in the associated with an increased accumulation of DNR and with calcein-AM uptake and in the Rh123 efflux assays, a higher the correction of the altered subcellular distribution of DNR. blocking factor was observed by van der Kolk et al40 and our- The concentration of probenecid that reversed MDR is clini- selves,11 than when both inhibitors were used separately. cally achievable in vivo, without major toxicity. These results Using calcein-AM as substrate and probenecid and CsA as suggest that probenecid is a good modulator for MRP1 activity modulator(s) of MRP1 and Pgp, respectively, we have shown and a potential candidate for clinical use to reverse MDR- in 50 AML patients, that one third of de novo untreated AML 62 associated MRP1. In contrast, probenecid did not reverse exhibited activity of both MRP1 and Pgp, 27% with activity MDR in tumor cell lines which overexpressed Pgp. Probene- of MRP1 only, 17% with activity of Pgp only and 23% without cid decreased the release of GSH from resistant cells into activity of MRP1 and Pgp (manuscript in preparation). the medium. In tumor cell lines, CO-L23/R, GLC4/ADR and MOR/R, which overexpressed MRP1, the resistance to DNR, VCR and MRP1 expression and activity as a prognostic factor Rh123 can be partially reversed by exposing these cells to The correlation between MRP1 gene expression or activity buthionine sulphoximine (BSO), an inhibitor of glutathione and treatment outcome is poorly documented in AML. How- (GSH) synthesis. This effect of BSO on drug resistance was ever, in the majority of studies, only one technique was associated with an increased intracellular accumulation of employed and different techniques were used (including RT- DNR and Rh123, owing to the inhibition of the enhanced PCR, immunocytochemistery, flow cytometry) to measure drug efflux. In contrast, the accumulation of DNR was not MRP1 expression.4–11 In addition, few studies analyzed the increased in a Pgp-mediated cell line exposed to BSO. In functionality of this protein. Therefore, as for MDR1/Pgp, con- addition, increase of cellular GSH levels in BSO-treated cell census recommendations and analysis are needed to improv- lines restored the accumulation deficit of DNR and Rh123, ing data reproductibility. The results of these studies are sum- but did not affect calcein or CF accumulation.59,60,63–66 In the marized in Table 3. Unfortunately, several of these studies are same way, N,N-bis(2-chloroethyl)-N-nitrosourea (BCNU), difficult to compare because of the heterogenous patient which decreased intracellular GSH pool by inhibition of the populations and the differences regarding the definition of GSH redox cycle, significantly inhibited MRP1-mediated overexpression. MRP1 in AML O Legrand et al 582 Table 3 MRP1 as a prognostic factor in AML

First author/(Ref.) Techniques used Patients Achievement of Duration of Duration of (No. of patients) CR CR overall survival

Hart4 RT-PCR De novo AML P = 0.07 ND ND (45 patients) te Boekhorst5 RNase protection 35 newly primary AML NS ND NS assay Schneider6 RT-PCR 29 de novo AML NS NS ND 14 at relapse Zhou7 RT-PCR Newly and relapsed AML P = 0.005 ND ND (52 patients) Filipits8 ICC De novo AML NS NS NS (80 patients) Lohri9 RT-PCR De novo untreated AML NS NS NS (57 patients) Leith10 Flow cytometry De novo untreated AML NS NS NS (352 patients) Legrand11 RT-PCR De novo untreated AML NS ND ND Flow cytometry NS ND ND functional test P = 0.03 ND ND (flow cytometry) (53 patients) Legranda functional test De novo untreated AML P = 0.009b P = 0.03b P = 0.05b (50 patients)

aUnpublished data. bWhen combined activity of MRP1 and Pgp was analyzed, while the activity of Pgp and MRP1 separately influenced the achievement of CR, but not the duration of DFS or OS.

Filipits et al8 did not observe correlations between the level References of MRP1 (by immunocytochemistry) and overall survival or DFS of patients. However, the fact that most patients surviving 1 Marie JP, Zhou DC, Gurbuxani S, Legrand O, Zittoun R. MDR1/P- more than 2 years did not express high MRP1 levels warrants glycoprotein in haematological neoplasms. Eur J Cancer 1996; additional studies on the association of MRP1 expression of 32A: 1034–1038. leukemic cells and long-term survival of patients. Zhou et al7 2 Cole SPC, Bhardwaj G, Gerlach JH, Mackie JE, Grant CE, Almquist 4 KC, Stewart AJ, Kurz EU, Ducan AMV, Deelay RG. Over- and Hart et al, using RT-PCR assay, showed that over- expression of a transporter gene in a multidrug-resistant human expression of MRP1 was a prognostic factor for achievement lung cancer cell line. Science 1992; 258: 1650–1654. of CR. In contrast, other studies mentioned that the achieve- 3 Cole SPC, Deelay RG. Multidrug resistance-associated protein: ment of CR, duration of remission and overall survival were sequence correction. Science 1993; 260: 879. independent of the level of MRP1 mRNA expression.5,6,9 In 4 Hart SM, Ganeshaguru K, Hoffbrand AV, Prentice HG, Mehta AB. 50 patients, we have shown that the combined functionality Expression of the multidrug resistance-associated protein (MRP) in acute leukaemia. Leukemia 1994; 8: 2163–2168. of MRP1 and Pgp was a prognostic factor for duration of CR 5 te Boekhorst PA, Lowenberg B, van Kapel J, Nooter K, Sonneveld and survival, but not MRP1 or Pgp separately (manuscript in P. Multidrug resistant cells with high proliferative capacity deter- preparation). This report suggests that MRP1 and Pgp together mine response to therapy in acute myeloid leukemia. Leukemia need to be considered and that clinical trials that select modu- 1995; 9: 1025–1031. lated Pgp are likely to achieve limited success. 6 Schneider E, Cowan KH, Bader H, Toomey S, Schwartz GN, Karp JE, Burke PJ, Kaufmann SH. Increased expression of the multidrug resistance-associated protein gene in relapsed acute leukemia. Blood 1995; 85: 186–193. Conclusions 7 Zhou DC, Zittoun R, Marie J-P. Expression of multidrug resistance- associated protein (MRP) and multidrug resistance (MDR1) genes Several experiments have shown that MRP1, as Pgp, also con- in acute myeloid leukemia. Leukemia 1995; 9: 1661–1666. tribute to the resistance to anthracyclines and etoposide in cell 8 Filipits M, Suchomel RW, Zo¨chbauer S, Brunner R, Lechner K, lines. However, the role of MRP1 in AML is a question for Pirker R. Multidrug resistance-associated protein in acute myeloid leukemia: no impact on treatment outcome. Clin Cancer Res debate. As for MDR1/Pgp, the major obstacle to definitive 1997; 3: 1419–1425. conclusions is the difficulty in standardizing techniques. How- 9 Lohri A, van Hille B, Bacchi M, Fopp M, Joncourt F, Reuter J, ever now, we can assess the functionality of MRP1 and the Cerny T, Fey MF, Herrman R. Five putative drug resistance para- respective roles of MRP1 and Pgp in fresh leukemic samples. meters (MDR1/P-glycoprotein, MDR-associated protein, gluta- This allows us to elucidate the role of MRP1 as a prognostic thione-S-transferase, bcl-2 and topoisomerase IIalpha) in 57 newly factor and as a factor of resistance to chemotherapy in AML diagnosed acute myeloid leukaemias. Swiss group for Clinical Cancer Research (SAKK). Eur J Haematol 1997; 59: 206–215. in comparison with Pgp. However, as for MDR1/Pgp, it will 10 Leith CP, Kopecky KJ, Chen IM, Slovak M, Head DR, Weick J, be important to expand the consensus recommendations to Appelbaum FR, Willman CL. Frequency and clinical significance include MRP1. of expression of the multidrug resistance proteins, MDR1, MRP MRP1 in AML O Legrand et al 583 and LRP in acute myeloid leukemia patients less than 65 years 29 Ma L, Krishnamachary N, Center MS. Phosphorylation of multi- old. A Southwest Oncology Group study. Blood 1997; 90: 389a drug resistance associated protein gene encoded protein P190. (Abstr.). Biochemistry 1995; 34: 3338–3343. 11 Legrand O, Simonin G, Perrot JY, Zittoun R, Marie J-P. Pgp and 30 Zhu Q, Center MS. Cloning and sequence analysis of promoter MRP activities using calcein-AM are prognostic factors in adult region of the MRP gene of HL60 cells isolated for resistance to acute myeloid leukemia patients. Blood 1998; 91: 4480–4488. adriamycin. Cancer Res 1994; 54: 4488–4492. 12 Bakos E, Hegedus T, Hollo Z, Welker E, Tusnady GE, Zaman GJ, 31 Zhu Q, Sun H, Center MS. Functional analysis of the nucleotide Flens MJ, Varadi A, Sarkadi B. Membrane topology and glycosyl- binding domains of the multidrug resistance protein (MRP). Oncol ation of the human multidrug resistance-associated protein. J Biol Res 1997; 9: 229–236. Chem 1996; 271: 12322–12326. 32 Nooter K, Burger H, Stoter G. Multidrug resistance-associated pro- 13 Zaman GR, Flens MJ, van Leusden MR. The human multidrug tein (MRP) in haematological malignancies. Leuk Lymphoma resistance-associated protein MRP is a plasma membrane drug- 1996; 20: 381–387. efflux pump. Proc Natl Acad Sci USA 1994; 91: 8822–8826. 33 Slapak CA, Mizunuma N, Kufe DW. Expression of the multidrug 14 Grant CE, Valdimarsson G, Hipfner DR, Almquist KC, Cole SPC, resistance associated protein and P-glycoprotein in doxorubicin- Deeley RG. Overexpression of multidrug resistance-associated selected human myeloid leukemia cells. Blood 1994; 84: 3113– protein (MRP) increase resistance to natural product drugs. Cancer 3121. Res 1994; 54: 357–361. 34 Zhou DC, Ramond S, Viguie´ F, Faussat AM, Zittoun R, Marie J- 15 Cole SPC, Sparks KE, Fraser K, Loe DW, Grant CE, Wilson GM. P. Progressive resistance to homoharringtonine in human myelo- Pharmacological characterization of multidrug resistant MRP- leukemia K562 cells: relationship to sequential emergence of MRP transfected human tumor cells. Cancer Res 1994; 54: 5902–5910. and MDR1 gene overexpression and MDR1 gene translation. Int 16 Flens MJ, Izquierdo MA, Scheffer GL, Fritz JM, Meijer CJLM, J Cancer 1996; 65: 365–371. Scheper RJ, Zaman GJR. Immunochemical detection of the multi- 35 Brock I, Hipfner DR, Nielsen BS, Jensen PB, Deeley RG, Cole SPC, drug resistance-associated protein MRP in human multidrug-resist- Sehested M. Sequential coexpression of the multidrug resistance ant tumor cells by monoclonal antibodies. Cancer Res 1994; 54: gene MRP and mdr1 and their products in VP-16 (etoposide)- 4557–4563. selected H69 small cell lung cancer cells. Cancer Res 1995; 55: 17 Hipfner DR, Gauldie SD, Deeley RG, Cole SPC. Detection of the 459–462.

Mr 190 000 multidrug protein, MRP, with monoclonal antibodies. 36 Slapak CA, Daniel JC, Levy SB. Sequential emergence of distinct Cancer Res 1994; 54: 5788–5792. resistance phenotypes in murine erythroleukemia cells under adri- 18 Leier I, Jedlitschky G, Buchholz U, Cole SPC, Deeley RG, Keppler amycin selection: decreased anthracycline uptake precedes D. The MRP gene encodes an ATP-dependent export pump for increased P-glycoprotein expression. Cancer Res 1990; 50:

leukotriene C4 and structurally related conjugates. J Biol Chem 7895–7901. 1994; 25: 27807–27810. 37 Richon VM, Weich N, Leng L, Kiyokawa H, Ngo L, Rifkind RA, 19 Gekeler V, Ise W, Sanders KH, Ulrich WR, Beck J. The leukotriene Marks PA. Characteristics of erythroleukemia cells selected for LTD4 receptor antagonist MK571 specifically modulates MRP vincristine resistance that have accelerated inducer-mediated dif- associated multidrug resistance. Biochem Biophys Res Commun ferentiation. Proc Natl Acad Sci USA 1991; 88: 1666–1670. 1995; 208: 345–352. 38 te Boekhorst P, de Leeuw K, Schoester M. Predominance of func- 20 Mu¨ller M, Meijer C, Zaman GJR, Borst P, Scheper RJ, Mulder NH, tional multidrug resistance (MDR1) phenotype in CD34+ acute de Vries EG, Jansen PL. Overexpression of the gene encoding the myeloid leukemia cells. Blood 1993; 82: 3157–3162. multidrug resistance-associated protein results in increased ATP- 39 Legrand O, Perrot J-Y, Tang RP, Simonin G, Gurbuxani S, Zittoun dependent glutathione S-conjugate transport. Proc Natl Acad Sci R, Marie J-P. Expression of the multidrug resistance-associated USA 1994; 91: 13033–13037. protein (MRP) mRNA and protein in normal peripheral blood and 21 Versantvoort CH, Broxterman HJ, Bagrij T, Scheper RJ, Twenty- bone marrow haematopoietic cells. Br J Haematol 1996; 94: man PR. Regulation by glutathione of drug transport in multidrug- 23–33. resistant human lung tumour cell lines overexpressing multidrug 40 van der Kolk DM, de Vries EG, Koning JA, van den Berg E, Muller resistance-associated protein. Br J Cancer 1995; 72: 82–89. M, Vellenga E. Activity and expression of the multidrug resistance 22 Zaman GJ, Lankelma J, van Tellingen O, Beijnen J, Dekker H, proteins MRP1 and MRP2 in acute myeloid leukemia cells, tumor Paulusma C, Oude Elferink RP, Baas F, Borst P. Role of glutathione cell lines, and normal hematopoietic CD34+ peripheral cells. Clin in the export of compounds from cells by the multidrug-resistance- Cancer Res 1998; 4: 1727–1736. associated protein. Proc Natl Acad Sci USA 1995; 92: 7690–7694. 41 Legrand O, Simonin G, Zittoun R, Marie J-P. Both P-gp and MRP 23 Marbeuf-Gueye C, Broxterman HJ, Dubru F, Priebe W, Garnier- contribute to drug resistance in AML. Leukemia 1998; 12: Suillerot A. Kinetics of anthracycline efflux from multidrug resist- 1327–1328. ance protein-expressing cancer cell compared with P-glyco- 42 Ross DD, Doyle LA, Schiffer CA, Lee EJ, Grant CE, Cole SPC, protein-expressing cancer cells. Mol Pharmacol 1998; 53: 141– Deeley RG, Yang W, Tong Y. Expression of multidrug resistance- 147. associated protein (MRP) mRNA in blast cells from acute myeloid 24 Lorico A, Rappa G, Flavell RA. Double knockout of the MRP gene leukemia patients. Leukemia 1996; 10: 48–58. leads to increased drug sensitivity in vitro. Cancer Res 1996; 56: 43 Schuurhuis GJ, Broxterman HJ, Ossenkeppele GJ, Baak JPA, Eek- 5351–5355. man CA, Kuiper CM, Feller N, van Heijningen THM, Klumper E, 25 Kuss BJ, Deeley RG, Cole SPC, Willman CL, Kopecky KJ, Wolman Pierters R, Lankelma J, Pinedo HM. Functional multidrug resist- SR, Eyre HJ, Lane SA, Nancarrow JK, Whitmore SA, Callen DF. ance phenotype associated with combined overexpression of Deletion of gene for multidrug resistance in acute myeloid leu- Pgp/MDR1 and MRP together with beta-D-arabinosylcytosine sen- kaemia with inversion in chromosome 16: prognostic impli- sitivity may predict clinical response in acute myeloid leukemia. cations. Lancet 1994; 343: 1531–1534. Clin Cancer Res 1995; 1: 81–88. 26 Zaman GJR, Versantvoort CHM, Smit JJM, Eijdems EWHM, de 44 Filipits M, Suchomel RW, Lechner K, Pirker R. Immunocytochem- Haas M, Smith AJ, Broxterman HJ, Mulder NH, de Vries EGE, Baas ical detection of the multidrug resistance-associated protein and F, Borst P. Analysis of the expression of MRP, the gene for a new P-glycoprotein in acute myeloid leukemia: impact of antibodies, putative transmembrane drug transporter in human multidrug sample source and disease status. Leukemia 1997; 11: 1073– resistant lung cancer cell lines. Cancer Res 1993; 53: 1747–1750. 1077. 27 Burger H, Nooter K, Sonneveld P, van Wingerden KE, Zaman GJR, 45 Kool M, de Haas M, Scheffer GL, Scheper RJ, van Eijk MJT, Juijn Stoter G. High expression of the multidrug resistance-associated JA, Baas F, Borst P. Analysis of expression of cMOAT (MRP2), protein (MRP) in chronic and prolymphocytic leukemia. Br J MRP3, MRP4, and MRP5, homologues of the multidrug resistance- Haematol 1994; 88: 348–356. associated protein gene (MRP1), in human cancer cell lines. Can- 28 Burger H, Nooter K, Zaman GJR, Sonneveld P, van Wingerden cer Res 1997; 57: 3557–3547. KE, Oostrum RG, Stoter G. Expression of the multidrug resistance- 46 den Boer ML, Pieters R, Kazemier KM, Rottier MM, Zwaan CM, associated protein (MRP) in acute and chronic leukemias. Leuke- Kaspers GJ, Janka-Schaub G, Henze G, Creutzig U, Scheper RJ, mia 1994; 8: 990–997. Veerman AJ. Relationship between major vault protein/lung resist- MRP1 in AML O Legrand et al 584 ance protein multidrug resistance-associated protein, P-glyco- rescein in multidrug-resistance-protein-overexpressing murine and protein expression, and drug resistance in childhood leukemia. human leukemia cells. Eur J Biochem 1997; 243: 219–224. Blood 1998; 91: 2092–2098. 61 Essodaigui M, Broxterman HJ, Garnier-Suillerot A. Kinetic analysis 47 Beck WT, Grogan TM, Willman CL, Cordon-Cardo C, Parham of calcein and calcein-acetoxymethylester efflux mediated by the DM, Kuttesch JF, Andreeff M, Bates SE, Berard CW, Boyett JM, multidrug resistance protein and P-glycoprotein. Biochemistry Brophy NA, Broxterman HJ, Chan HSL, Dalton WS, Dielt M, Fojo 1998; 37: 2243–2250. AT, Gascoyne RD, Head D, Houghton PJ, Kumar Svrivastava D, 62 Gollapudi S, Kim CH, Tran BN, Sangha S, Gupta S. Probenecid Lehnert M, Leith CP, Paietta E, Pavelic ZP, Rimsza L, Roninson reverses multidrug resistance in resistance-associated protein- IB, Sikic BI, Twentyman PR, Warnke R, Weinstein R. Methods to overexpressing HL60/AR and H69/AR cells but not in P-glyco- detect P-glycoprotein-associated multidrug resistance in patient protein-overexpressing HL60/Tax and P388/ADR cells. Cancer tumors: consensus recommendations. Cancer Res 1996; 56: Chemother Pharmacol 1997; 40: 150–155. 3010–3020. 63 Versantvoort CH, Bagrij T, Wright KA, Twentyman PR. On the 48 Marie J-P, Huet S, Faussat A-M, Perrot J-Y, Chevillard S, Barbu V, relationship between the probenecid-sensitive transport of dauno- Bayle C, Boutonnat J, Calvo F, Campos-Guyotat L, Colosetti P, rubicin or calcein and the gluthatione status of cells overexpress- Cazin J-L, De Cremoux P, Delvincourt C, Demur C, Drenou B, ing the multidrug resistance-associated protein (MRP). Int J Cancer Fenneteau O, Feuillard J, Garnier-Suillerot A, Genne P, Gorisse 1995; 63: 855–862. M-C, Gosselin P, Jouault H, Lacave R, Le Calvez G, Le´glise M-C, 64 Vanhoefer U, Yin MB, Harstrick A, Seeber S, Rustum YM. Carba- Le´once S, Manfait M, Maynadie´ M, Merle-Be´ral H, Merlin J-L, moylation of glutathione reductase by N,N-bis(2-chloroethyl)-N- Mousseau M, Morjani H, Picard F, Pinguet F, Poncelet P, Racadot nitrosourea associated with inhibition of multidrug resistance pro- E, Raphael M, Richard B, Rossi J-F, Schlegel N, Vielh P, Zhou DC, tein (MRP) function. Biochem Pharmacol 1997; 53: 801–809. Robert J. French Network of the Drug Resistance Intergroup, and 65 Chen ZS, Mutoh M, Sumizawa T, Furukawa T, Haraguchi M, Tani Drug Resistance Network of ‘Assistance Publique-Hoˆpitaux de A, Akiyama S. Reversal of heavy metal resistance in multidrug- Paris’. Multicentric evaluation of the MDR phenotype in leukemia. resistant human KB carcinoma cells. Biochem Biophys Res Com- Leukemia 1997; 11: 1086–1094. mun 1997; 236: 586–590. 49 Marie J-P, Legrand O, Perrot J-Y, Chevillard S, Huet S, Robert J. 66 Feller N, Broxterman HJ, Wahrer DC, Pinedo HM. ATP-dependent Measuring multidrug resistance expression in human malig- efflux of calcein by the multidrug resistance protein (MRP): no nancies: elaboration of concensus recommendations. Semin Hem- inhibition by intracellular glutathione depletion. FEBS Lett 1995; atol 1997; 34: 63–71. 368: 385–389. 50 Bailly JD, Muller C, Jaffre´zou JP, Demur C, Gassar G, Bordier C, 67 Vanhoefer U, Yin MB, Harstrick A, Seeber S, Rustum YM. Carba- Laurent G. Lack of correlation between expression and function of moylation of glutathione reductase by N,N-bis(2-chloroethyl)-N- Pgp in acute myeloid leukemia cells. Leukemia 1995; 9: 799–807. nitrosourea associated with inhibition of multidrug resistance pro- 51 Davey MW, Hargrave RM, Davey RA. Comparison of drug tein (MRP) function. Biochem Pharmacol 1997; 53: 801–809. accumulation in P-glycoprotein-expressing and MRP-expressing 68 Zaman GJ, Cnubben NH, van Bladeren PJ, Evers R, Borst P. Trans- human leukaemia cells. Leukemia Res 1996; 20: 657–664. port of the glutathione conjugate of ethacrynic acid by the human 52 Webb M, Raphael CL, Asbahr H, Erber WN, Meyer BF. The detec- multidrug resistance protein MRP. FEBS Lett 1996; 391: 126–130. tion of rhodamine 123 efflux at low levels of drug resistance. Br 69 Draper MP, Martell RL, Levy SB. Indomethacin-mediated reversal J Haematol 1996; 93: 650–655. of multidrug resistance and drug efflux in human and murine cell 53 Twentyman PR, Rhodes T, Rayner S. A comparison of rhodamine lines overexpressing MRP, but not P-glycoprotein. Br J Cancer 123 accumulation and efflux in cells with P-glycoprotein- 1997; 75: 810–815. mediated and MRP-associated multidrug resistance phenotype. 70 Chuman Y, Chen ZS, Seto K, Sumizawa T, Furukawa T, Tani A, Eur J Cancer 1994; 30A: 1360–1369. Haraguchi M, Niwa K, Yamada K, Aikou T, Akiyama S. Reversal 54 Broxterman HJ, Lankelma J, Pinedo HM, Eekman CA, Waher DC, of MRP-mediated vincristine resistance in KB cells by buthionine Ossenkoppele GJ, Schuurhuis GJ. Theorical and practical con- sulfoximine in combination with PAK-104P. Cancer Lett 1998; siderations for the measurement of P-glycoprotein function in 129: 69–76. acute myeloid leukemia. Leukemia 1997; 11: 1110–1118. 71 Germann UA, Ford PJ, Shlyakhter D, Mason VS, Harding MW. 55 Feller N, Kuiper CM, Lankelma J, Ruhdal JK, Scheper RJ, Pinedo Chemosensitization and drug accumulation effects of V-710, vera- HM, Broxterman HJ. Functional detection of MDR1/p170 and pamil, cyclosporin A, MS-209 and GF120918 in multidrug resist- MRP/p190-mediated multidrug resistance in tumour cells by flow ant HL60/ADR cells expressing the multidrug resistance-associa- cytometry. Br J Cancer 1995; 72: 543–550. ted protein MRP. Anticancer Drugs 1997; 8: 141–155. 56 Pall G, Spitaler M, Hofmann J, Thaler J, Ludescher C. Multidrug 72 Naito S, Koike K, Ono M, Machida T, Tasaka S, Kiue A, Koga H, resistance in acute leukemia: a comparison of different diagnostic Kumazawa J. Development of novel reversal agents, imidazothia- method. Leukemia 1997; 11: 1067–1072. zole derivatives, targeting MDR1 and MRP-mediated multidrug 57 Minderman H, Vanhoefer U, Toth K, Yin MB, Minderman MD, resistance. Oncol Res 1998; 10: 123–132. Wrzosek C, Slovak ML, Rustum YM. DiOC2(3) is not a substrate 73 Versantvoort CH, Rhodes T, Twentyman PR. Acceleration of MRP- for multidrug resistance protein (MRP)-mediated drug efflux. Cyto- associated efflux of rhodamine 123 by genistein and related com- metry 1996; 25: 14–20. pounds. Br J Cancer 1996; 74: 1949–1954. 58 Breeuwer P, Drocourt JL, Bunschoten N, Zwietering MH, Rom- 74 Benderra Z, Morjani H, Trussardi A, Manfait M. Role of the vacuo- bouts FM, Abee T. Characterization of uptake and hydrolysis of lar H+-ATPase in daunorubicin distribution in etoposide-resist- fluorocein diacetate and carboxyfluorocein diacetate by intra- ance MCF7 cells overexpressing the multidrug-resistance associa- cellular esterases in saccharomyces cerevisiae, which result in ted protein. Int J Oncol 1998; 12: 711–715. accumulation of fluorescent product. Appl Environ Microbiol 75 Benderra Z, Morjani H, Trussardi A, Manfait M. Characterization 1995; 61: 1614–1619. of H+-ATPase-dependent activity of multidrug resistance-associa- 59 Hollo Z, Homolya L, Hegedus T, Sarkadi B. Transport properties ted protein in homoharringtonine-resistant human leukemic K562 of the multidrug resistance-associated protein (MRP) in human cells. Leukemia 1998; 12: 1539–1544. tumour cells. FEBS Lett 1996; 383: 99–105. 76 Homolya L, Hollo Z, Germann UA, Pastan I, Gottesman MM, Sar- 60 Draper MP, Martell RL, Levy SB. Active efflux of the free acid form kadi B. Fluorescent cellular indicators are extruded by the multid- of the fluorescent dye 2Ј,7Ј-bis(2-carboxyethyl)-5(6)-carboxyfluo- rug resistance protein. J Biol Chem 1993; 268: 21493–21501.