Search

Hyperbilirubinemia is a frequent side effect induced by long-term therapy with the antibiotic fusidate. The aim of this study was to elucidate the molecular mechanisms of fusidate-induced hyperbilirubinemia by investigating its influence on hepatic transport systems in the canalicular membrane. Using canalicular membrane vesicles from rat liver, we determined the effect of fusidate on the adenosine 5'-triphosphate (ATP)-dependent transport of substrates of the apical conjugate export pump, multi-drug resistance protein 2 (Mrp2, symbol Abcc2) and the bile salt export pump (Bsep, symbol Abcb11). Fusidate inhibited the ATP-dependent transport of the Mrp2 substrates 17beta-glucuronosyl estradiol and leukotriene C4, and the transport of cholyltaurine by Bsep with Ki values of 2.2+/-0.3, 7.6+/-1.3, and 5.5+/-0.8 microM, respectively. To elucidate the in vivo implication of these findings, the effect of fusidate treatment on the elimination of intravenously administered tracer doses of 17beta-glucuronosyl estradiol and cholyltaurine into bile was studied in rats. Treatment with fusidate (100 micromol/kg body weight) reduced the biliary excretion rate of 17beta-glucuronosyl [3H]estradiol and [3H]cholyltaurine by 75 and 80%, respectively. Extended treatment of rats with fusidate (100 micromol/kg body weight, three times daily i.p. for 3 days) reduced hepatic Mrp2 protein levels by 61% (P

The membrane proteins mediating the ATP-dependent transport of lipophilic substances conjugated to glutathione, glucuronate, or sulfate have been identified as members of the multidrug resistance protein (MRP) family. Several isoforms of these conjugate export pumps with different kinetic properties and domain-specific localization in polarized human cells have been cloned and characterized. Orthologs of the human MRP isoforms have been detected in many different organisms. Studies in mutant rats lacking the apical isoform MRP2 (symbol ABCC2) indicate that anionic conjugates of endogenous and exogenous substances cannot exit from cells at a sufficient rate unless an export pump of the MRP family is present in the plasma membrane. Several mutations in the human MRP2 gene have been identified which lead to the absence of the MRP2 protein from the hepatocyte canalicular membrane and to the conjugated hyperbilirubinemia of Dubin–Johnson syndrome. Overexpression of recombinant MRP2 confers resistance to multiple chemotherapeutic agents. Because of its function in the terminal excretion of cytotoxic and carcinogenic substances, MRP2 as well as other members of the MRP family, play an important role in detoxification and chemoprevention.

The secretion of bilirubin conjugates from hepatocytes into bile represents a decisive step in the prevention of hyperbilirubinemia. The bilirubin conjugates, monoglucuronosyl bilirubin (MGB) and bisglucuronosyl bilirubin (BGB), were previously suggested to be endogenous substrates for the apical multidrug resistance protein (MRP2), a member of the adenosine triphosphate (ATP)-binding cassette family of transporters (symbol ABCC2), also termed canalicular multispecific organic anion transporter. We have characterized this ATP-dependent transport using membrane vesicles from human embryonic kidney (HEK) cells expressing recombinant rat as well as human MRP2. MGB and BGB, 3H-labeled in the glucuronosyl moiety, were synthesized enzymatically with recombinant UDP-glucuronosyltransferase 1A1, and stabilized with ascorbate. Rates for ATP-dependent transport of MGB and BGB (0.5 μmol/L each) by human MRP2 were 183 and 104 pmol × mg protein−1 × min−1, respectively. Km values were 0.7 and 0.9 μmol/L for human MRP2, and 0.8 and 0.5 μmol/L for rat MRP2, with MGB and BGB as substrates, respectively. Leukotriene C4 and 17β-glucuronosyl estradiol, which are both known high-affinity substrates for human MRP2, inhibited [3H]MGB transport with IC50 values of 2.3 and 30 μmol/L, respectively. Cyclosporin A competitively inhibited human and rat MRP2-mediated transport of [3H]MGB, with Ki values of 21 and 10 μmol/L, respectively. Our results provide direct evidence that recombinant MRP2, cloned from rat as well as human liver, mediates the primary-active ATP-dependent transport of the bilirubin conjugates MGB and BGB.

The polarized rat hepatoma/human fibroblast hybrid cell line, WIF-B, forms apical vacuoles into which cholephilic substances are secreted. We studied expression, localization, and function of the apical conjugate export pump, Mrp2, in WIF-B cells. Mrp2, the apical isoform of the multidrug resistance protein, alternatively termed canalicular Mrp (cMrp) or canalicular multispecific organic anion transporter (cMoat), is a 190-kd membrane glycoprotein mediating adenosine triphosphate (ATP)-dependent transport of glucuronides, glutathione S-conjugates, and other amphiphilic anions across the hepatocyte canalicular membrane into bile. Expression of the rat mrp2 gene in WIF-B cells was shown by reverse-transcription polymerase chain reaction (PCR), followed by sequencing of the amplified 789-bp fragment. Immunoblotting, using antibodies reacting with the amino-terminal or with the carboxyl-terminal sequence of rat Mrp2, detected the 190-kd glycoprotein in WIF-B cell homogenates. Immunofluorescence microscopy localized Mrp2 to the apical membrane domain. Preloading of WIF-B cells with a membrane-permeable ester of the calcium-dependent fluorescent indicator, Fluo-3, was followed by Mrp2-mediated secretion of the amphiphilic anion, Fluo-3, into the apical vacuoles. This transport was potently inhibited by cyclosporin A added to the culture medium. Direct measurements of ATP-dependent transport into Mrp2-containing plasma membrane vesicles in comparison with Mrp2-deficient vesicles established that Fluo-3 is transported by Mrp2 with a Km value of 3.7 micromol/L. Our results indicate that the polarized WIF-B cells express the rat ortholog of the apical conjugate-transporting ATPase, Mrp2. The function of Mrp2 as well as the action of inhibitors can thus be analyzed by use of the fluorescent amphiphilic anion, Fluo-3.

The membrane proteins mediating the ATP-dependent transport of glutathione S-conjugates and related amphiphilic anions have been identified as the multidrug resistance proteins MRP1 and MRP2. These 190-kDa membrane glycoproteins were cloned in recent years and shown to be unidirectional, ATP-driven, export pumps with an amino acid identity of 49% in humans. MRP1 is detected in the plasma membrane of many cell types, including erythrocytes; whereas MRP2, also termed canalicular MRP (cMRP) or canalicular multispecific organic anion transporter (cMOAT), has been localized to the apical domain of polarized epithelia, such as the hepatocyte canalicular membrane and kidney proximal tubule luminal membrane. Physiologically important substrates of both transporters include glutathione S-conjugates, such as leukotriene C4, as well as bilirubin glucuronides. 17 beta-glucuronosyl estradiol and glutathione disulfide. Both transporters have been associated with multiple drug resistance of malignant tumors because of their capacity to pump drug conjugates and drug complexes across the plasma membrane into the extracellular space. The substrate specificity of MRP1 and MRP2 studied in inside-out oriented membrane vesicles is very different from MDR1 P-glycoprotein. MRP1 and MRP2 may be termed conjugate transporting ATPases, functioning in detoxification and, because of their role in glutathione disulfide export, in the defense against oxidative stress.

The human multidrug resistance protein (MRP1) confers resistance of cells to a number of different cytostatic drugs and functions as an export pump for glutathione S-conjugates, glucuronides and other amphiphilic anions. The present study details for the first time MRP1-mediated ATP-dependent transport of various glutathione S-conjugates of the bifunctional alkylating agents chlorambucil and melphalan. In membrane vesicles prepared from cells expressing recombinant MRP1, the conjugates were transported at rates in the following order: monoglutathionyl chlorambucil > bisglutathionyl chlorambucil > monohydroxy monoglutathionyl chlorambucil and monoglutathionyl melphalan > monohydroxy monoglutathionyl melphalan. In addition, we show that membranes from chlorambucil-resistant GST-alpha-overexpressing CHO cells as well as from their parental cells express the hamster homologue of MRP1. With both CHO cell membrane preparations, we observed ATP-dependent transport of monoglutathionyl chlorambucil and of leukotriene C4, a glutathione S-conjugate and high-affinity substrate of MRP1. The transport rates measured in the resistant cells were only two- to three-fold higher than those measured in the control cells. These results together with cytotoxicity assays comparing MRP1-overexpressing cell pairs with the CHO cell pair indicate that, although MRP1-mediated transport is active, it may not be the rate-limiting step in chlorambucil resistance in these cell lines. Images Figure 3

A membrane glycoprotein of 190 kDa has been identified previously by photoaffinity labeling as a candidate for the ATP-dependent export pump for leukotriene C4 in mastocytoma cells [Leier, I., Jedlitschky, G., Buchholz, U. & Keppler, D. (1994) Eur. J. Biochem. 220, 599–606]. The present study indicates that this protein represents the murine homolog of the human multidrug resistance protein (MRP). In immunoblot analyses several polyclonal anti-MRP antibodies and one monoclonal antibody recognized the protein of 190-kDa in plasma membranes of mastocytoma cells. Immunoprecipitation after photoaffinity labeling with [3H]leukotriene C4 precipitated the labeled 190-kDa glycoprotein. Deglycosylation by glycopeptide N-glycosidase F of mastocytoma membrane proteins was performed in comparison with membranes from MRP-overexpressing cells and resulted in a reduction of the molecular mass of 190 kDa by about 20 kDa in all membrane preparations. The expression of the murine mrp gene in the mastocytoma cells was analyzed by amplification and sequencing of two mrp cDNA fragments in the first nucleotide binding domain (182 bp) and in a domain proximal to the 3′-end (291 bp). The deduced amino acid sequences of these fragments were identical with murine Mrp and 86.7% and 89.7% identical with the corresponding sequences of human MRP. These results indicate that the ATP-dependent release of leukotriene C4 by murine mastocytoma cells is mediated by murine Mrp.

The identification of the multidrug resistance protein (MRP) as a conjugate export pump in several cell types suggested its involvement in the long-known glutathione-S-conjugate transport across erythrocyte membranes. We investigated the ATP-dependent transport of glutathione S-conjugates in human erythrocyte and erythroleukemia cell membrane vesicles using the endogenous conjugate leukotriene C4 (LTC4), known to be a high-affinity substrate for MRP, in addition to S-(2,4-dinitrophenyl)glutathione. The kinetic parameters, including the Km value for LTC4 of 118 +/- 5 nM and the inhibition constants for transport of both substrates for the quinoline-based inhibitor MK 571, were similar to those obtained for transport mediated by recombinant MRP. Direct photoaffinity labeling of human erythrocyte membranes with [3H]LTC4 revealed a major binding protein of about 190 kDa which was immunoprecipitated by an anti-MRP serum. The radiolabeling of this protein was specifically suppressed by the transport inhibitor MK 571. Several additional anti-MRP sera detected the protein of about 190 kDa in human erythrocyte and erythroleukemia cell membranes. These data identify for the first time the glutathione-S-conjugate transporting protein in erythrocyte membranes.

We have previously shown that the multidrug resistance protein (MRP) mediates the ATP-dependent membrane transport of the endogenous glutathione conjugate leukotriene C4 (LTC4) and of structurally related anionic conjugates of lipophilic compounds [Jedlitschky, Leier, Buchholz, Center and Keppler (1994) Cancer Res. 54, 4833-4836; Leier, Jedlitschky, Buchholz, Cole, Deeley and Keppler (1994) J. Biol. Chem. 269, 27807-27810]. We demonstrate in the present study that MRP also mediates the ATP-dependent transport of GSSG, as shown in membrane vesicles from human leukaemia cells overexpressing MRP (HL60/ADR cells) or HeLa cells transfected with an MRP expression vector (HeLa T5 cells) in comparison with the respective parental or control cells. The Km value for ATP-dependent transport of GSSG was 93 +/- 26 microM (mean value +/- S.D., n=5) in membrane vesicles from HeLa T5 cells. GSH, at a concentration of 100 microM, was not a substrate for any significant ATP-dependent MRP-mediated transport. The transport of GSSG was competitively inhibited by LTC4, by the leukotriene D4 receptor antagonist 3-([{3-(2-[7-chloro-2-quinolinyl]ethenyl)phenyl}-{(3-dimethylamino-3- oxopropyl)-thio}-methyl]thio)propanoic acid (MK 571) and by S-decylglutathione, with K1 values of 0.3, 0.6 and 0.7 microM respectively. These studies identify MRP as the membrane glycoprotein which mediates the ATP-dependent export of GSSG from these cells.

We have previously shown that the multi-drug resistance protein (MRP) mediates the ATP-dependent membrane transport of glutathione S-conjugates and additional amphiphilic organic anions. In the present study we demonstrate the expression of MRP in hepatocytes where it functions in hepatobiliary excretion. Analysis by reverse transcription-PCR of human and normal rat liver mRNA resulted in two expected cDNA fragments of MRP. Four different antibodies against MRP reacted on immunoblots with the glycoprotein of about 190 kD from human canalicular as well as basolateral hepatocyte membrane preparations. A polyclonal antibody directed against the carboxy-terminal sequence of MRP detected the rat homolog of MRP in liver. Double immunofluorescence microscopy and confocal laser scanning microscopy showed the presence of human MRP and rat Mrp in the canalicular as well as in the lateral membrane domains of hepatocytes. The transport function of the mrp gene-encoded conjugate export pump was assayed in plasma membrane vesicles with leukotriene C4 as a high-affinity glutathione S-conjugate substrate. The deficient ATP-dependent conjugate transport in canalicular membranes from TR- mutant rat hepatocytes was associated with a lack of amplification of one of the mrp cDNA fragments and with a selective loss of Mrp on immunoblots of canalicular membranes. Double immunofluorescence microscopy of livers from transport-deficient TR- mutant rats localized Mrp only to the lateral but not to the canalicular membrane. Our results indicate that the absence of Mrp or an isoform of Mrp from the canalicular membrane is the basis for the hereditary defect of the hepatobiliary excretion of anionic conjugates by the transport-deficient hepatocyte.

Amphiphilic carboxylates acting as peroxisome proliferators and hypolipidemic drugs induce enzymes of peroxisomal lipid beta-oxidation, certain drug-metabolizing enzymes in the liver, and a number of additional proteins. The peroxisome proliferators represent a well-established class of non-genotoxic hepatocarcinogens. In this study we characterized the hepatic elimination of the peroxisome proliferator nafenopin. In the rat in vivo, 1 hr after intravenous administration of [3H]nafenopin, approx. 40% of injected radioactivity was recovered in bile. HPLC analysis of bile samples revealed that only about 10% of the radioactivity recovered in bile was associated with non-metabolized nafenopin and approx. 90% with more polar metabolites. One of the main metabolites formed in the liver and excreted into bile was identified as nafenopin glucuronide by beta-glucuronidase-catalysed reconversion to nafenopin. In mutant rats deficient in the canalicular transport of leukotriene C4 and related amphiphilic anion conjugates, recovery of [3H]nafenopin-derived radioactivity in bile was reduced to 4% of the injected dose. Although nafenopin glucuronide could not be detected in bile, it was a major metabolite in the liver from these mutant rats. Using membrane vesicles enriched in bile canalicular membranes from normal rats, transport of nafenopin glucuronide was shown to be a primary-active ATP-dependent process which was inhibited by leukotriene C4 and S-dinitrophenyl glutathione with IC50 values of 0.2 and 12 microM, respectively. ATP-dependent transport was not detectable for non-conjugated nafenopin. In canalicular membrane vesicles prepared from the mutant rats, the rate of ATP-dependent transport of nafenopin glucuronide was less than 10% of the transport observed in vesicles from normal rats. These data indicate that conjugation and subsequent transport by the ATP-dependent export carrier for leukotriene C4 and related conjugates is a major pathway for the elimination of nafenopin and structurally-related peroxisome proliferators.

The biosynthesis of leukotriene C4 (LTC4) must be followed by an export of this mediator into the extracellular space where it interacts with receptors. Using mastocytoma cells we have demonstrated the existence of a primary-active, ATP-dependent transport mediating this export of LTC4 [Schaub, T., Ishikawa, T. & Keppler, D. (1991) FEBS Lett. 279, 83-86]. The following inhibitors served to characterize further this transport system in plasma membrane vesicles from mastocytoma cells: Probenecid, an inhibitor of organic anion transport, induced half-maximal inhibition of the ATP-dependent LTC4 transport at 71 microM. Cyclosporin A and its non-immunosuppressive analog PSC 833 inhibited the ATP-dependent transport with Ki values of 4.5 microM and 30 microM, respectively. The LTD4 receptor antagonist 3-([(3-(2-[7-chloro-2-quinolinyl]ethenyl)phenyl)-[(3-dimethylamino-3- oxopropyl)-thio]-methyl]thio)propanoic acid (MK 571) was the most potent competitive inhibitor of the export carrier with a Ki value of 0.8 microM. The transport inhibitor MK 571 served as competitor in the photoaffinity labeling of LTC4-binding membrane proteins using [3H]LTC4 as the photolabile ligand. Proteins with molecular masses of about 190 kDa and 35 kDa were predominantly labeled. In addition, a minor [3H]LTC4 labeling was observed in the molecular mass range of 100 kDa. The [3H]LTC4 labeling of the 190-kDa protein was competed for by MK 571. The labeled proteins resisted extraction from the membrane with 2% sodium taurocholate suggesting that they are integral membrane proteins. Treatment of the membrane proteins with peptide N-glycosidase F resulted in the appearance of an additional labeled polypeptide of about 140 kDa suggesting that the 190-kDa protein is a glycoprotein. Photoaffinity labeling with 8-azido[alpha-32P]ATP predominantly labeled the LTC4-binding 35-kDa protein. The [3H]LTC4-labeled 190-kDa protein showed a mean isoelectric point at pH 6.3 with a range of pH 5.8-6.7, while the 35-kDa protein had an isoelectric point at pH 6.8. Specific labeling of a 190-kDa membrane glycoprotein by the glutathione conjugate LTC4, which is competed for by a potent inhibitor of the ATP-dependent LTC4 export carrier, pinpoints its involvement in the ATP-dependent transport of LTC4 and related conjugates.

The uptake of tritiated cysteinyl leukotrienes (LTC4, LTD4, LTE4) and LTB4 was investigated in freshly isolated rat hepatocytes and different hepatoma cell lines under initial-rate conditions. Leukotriene uptake by hepatocytes was independent of an Na+ gradient and a K+ diffusion potential across the hepatocyte membranes as established in experiments with isolated hepatocytes and plasma membrane vesicles. Kinetic experiments with isolated hepatocytes indicated a low-Km system and a non-saturable system for the uptake of cysteinyl leukotrienes as well as LTB4 under the conditions used. AS-30D hepatoma cells and human Hep G2 hepatoma cells were deficient in the uptake of cysteinyl leukotrienes, but showed significant accumulation of LTB4. Moreover, only LTB4 was metabolized in Hep G2 hepatoma cells. Competition studies on the uptake of LTE4 and LTB4 (10 nM each) indicated inhibition by the organic anions bromosulfophthalein, S-decyl glutathione, 4,4'-diisothiocyanato-stilbene-2,2'-disulfonate, probenecid, docosanedioate, and hexadecanedioate (100 microM each), but not by taurocholate, the amphiphilic cations verapamil and N-propyl ajmaline, and the neutral glycoside ouabain. Cholate and the glycoside digitoxin were inhibitors of LTB4 uptake only. Bromosulfophthalein, the strongest inhibitor of leukotriene uptake by hepatocytes, did not inhibit LTB4 uptake by Hep G2 hepatoma cells under the same experimental conditions. Leukotriene-binding proteins were analyzed by comparative photoaffinity labeling of human hepatocytes and Hep G2 hepatoma cells using [3H]LTE4 and [3H]LTB4 as the photolabile ligands. Predominant leukotriene-binding proteins with apparent molecular masses in the ranges of 48-58 kDa and 38-40 kDa were labeled by both leukotrienes in the particulate and in the cytosolic fraction of hepatocytes, respectively. In contrast, no labeling was obtained with [3H]LTE4 in Hep G2 cells. With [3H]LTB4 a protein with a molecular mass of about 48 kDa was predominantly labeled in the particulate fraction of the hepatoma cells, whereas in the cytosolic fraction a labeled protein in the range of 40 kDa was detected. Our results provide evidence for the existence of distinct uptake systems for cysteinyl leukotrienes and LTB4 at the sinusoidal membrane of hepatocytes; however, some of the inhibitors tested interfere with both transport systems. Only LTB4, but not cysteinyl leukotrienes, is taken up and metabolized by the transformed hepatoma cells.

Transport processes control not only synthesis and release of LTC4 but also the elimination and excretion of LTC4 and its metabolites. 1. (i) A primary-active ATP-dependent export carrier mediates the release of LTC4 from a leukotriene-generating cell, as exemplified by mastocytoma cells, and as measured in mastocytoma plasma membrane vesicles (2). 2. (ii) Release of cysteinyl leukotrienes into the blood circulation is followed by a rapid elimination with an initial half-life of 38 sec in rats and 4.0 min in man, as measured with the labeled, representative LTC4 catabolite, N-acetyl-LTE4. 3. (iii) 11C-labeled N-acetyl-LTE4 can serve for non-invasive studies on cysteinyl leukotriene elimination and excretion by the liver and kidney in the intact organism using positron emission tomography. An impairment of leukotriene transport from the liver across the canalicular membrane into bile, studied in mutant rats and in extrahepatic cholestasis, leads to compensatory diversion of cysteinyl leukotriene elimination to the kidney. N-Acetyl-LTE4 labeled with a short-lived positron-emitting isotope provides quantitative insight into the pathways of cysteinyl leukotriene elimination in vivo. 4. (iv) Cysteinyl leukotriene export from the liver into bile is mediated by an ATP-dependent primary-active export carrier. This decisive step in cysteinyl leukotriene elimination has been characterized in hepatocyte canalicular membrane vesicles (3). The leukotriene exporter is deficient in transport mutant rats. The leukotriene carrier is distinct from other ATP-dependent export carriers identified in this membrane domain, such as the ATP-dependent bile salt export carrier (25) and the multidrug export carrier (27).

Bilirubin conjugates are secreted from hepatocytes into bile. Monoglucuronosyl bilirubin is an endogenous substrate for the multidrug resistance protein 2, which is located in rat hepatocyte canalicular membrane. We have characterized this ATP-dependent transport using rat canalicular membrane vesicles. Monoglucuronosyl bilirubin, 3 H-labeled in the glucuronosyl moiety, was synthesized enzymatically using recombinant UDP-glycosyltransferase 1A1, and was stabilized with ascorbate. The rate for ATP-dependent transport of monoglucuronosyl bilirubin (at 0.5 μM) was 7.3±1.1 pmol mg protein −1 min −1 and the K m value was 1.3±0.4 μM. This is the first time to demonstrate this kinetic constant of monoglucuronosyl bilirubin for the rat hepatocyte canalicular membrane. The K m value is similar to one for recombinant rat multidrug resistance protein 2.

Several members of the multidrug resistance protein (MRP) family mediate the ATP-dependent transport of amphiphilic anions across membranes. The substrate specificity of recombinant human MRP1 has been most extensively defined by use of inside-out membrane vesicles. Substrates include the glutathione S-conjugate leukotriene C4, 17 beta-glucuronosyl estradiol, glucuronosyl bilirubin, glutathione disulfide, in addition to the fluorescent lipophilic anion Fluo-3. These substances are also substrates for the apical isoform MRP2, also termed canalicular multispecific organic anion transporter, cMOAT, which shares only 49% amino acid sequence identity with MRP1. The K(m) of leukotriene C4 for MRP2 is 10-fold higher than for MRP1, and the K(m) of 17 beta-glucuronosyl estradiol is 4.8-fold higher for MRP2 than for recombinant human MRP1. Human as well as rat MRP2 confer multidrug resistance to polarized MDCKII cells permanently expressing the recombinant glycoprotein in their apical plasma membrane. Resistance of cells transfected with human and rat MRP2 to etoposide was enhanced 5-fold and 3.8-fold, and resistance to vincristine was enhanced 2.3-fold and 6.0-fold, respectively. Conjugate-transporting members of the MRP family with a related sequence and a similar function have been detected recently. In addition to several MRP isoforms (MRP1-6) and orthologs in mammals (human, rat, rabbit, mouse), MRP family members have been identified in the nematode Caenorhabditis elegans, in the yeast Saccharomyces cerevisiae, and in the plant Arabidopsis thaliana. These conjugate export pumps of the MRP family play a widespread role in detoxification, drug resistance, and, because of the role in the export of glutathione disulfide by MRP1 and MRP2, in the defense against oxidative stress.

Publisher Summary This chapter describes the multidrug resistance protein (MRP1) that has been cloned and sequenced from multidrug-resistant human lung cancer cells and identified as an integral membrane glycoprotein belonging to the superfamily of adenosine triphosphate-binding cassette (ABC) transporters. Primary active unidirectional adenosine triphosphate (ATP)-dependent transport of amphiphilic anions—in particular, of conjugates of lipophilic substances with glutathione, glucuronate, or sulfate, are the recognized functions of MRP1. An isoform of MRP1 with a related sequence and a similar function has been cloned from human and rat and localized predominantly to the hepatocyte canalicular membrane. This isoform is termed “MRP2” or “canalicular MRP” (cMRP) or “canalicular multispecific organic anion transporter” (cMOAT). Canalicular (apical) MRP is deficient in human Dubin–Johnson syndrome and in two mutant rat strains lacking the ATP-dependent transport of anionic conjugates across the hepatocyte canalicular membrane.

Bilirubin is secreted from the liver into bile mainly as monoglucuronosyl and bisglucuronosyl conjugates. We demonstrate for the first time that ATP-dependent transport of both bilirubin glucuronides is mediated by the multidrug resistance protein (MRP1) as well as by the distinct canalicular (apical) isoform MRP2, also termed cMRP or cMOAT (canalicular multispecific organic anion transporter). In membrane vesicles from MRP1-transfected HeLa cells mono[3H]glucuronosylbilirubin and bis[3H]glucuronosylbilirubin (each at 0.5 μM) were transported with rates of 5.3 and 3.1 pmol/min per mg of protein respectively. Rat hepatocyte canalicular membrane vesicles, which contain Mrp2 (the rat equivalent of MRP2), transported mono[3H]glucuronosylbilirubin and bis[3H]glucuronosylbilirubin at rates of 8.9 and 8.5 pmol/min per mg of protein, whereas membrane vesicles from mutant liver lacking Mrp2 showed no transport of the conjugates. In membrane vesicles from human hepatoma Hep G2 cells, which predominantly expressed MRP2, transport rates were 8.3 and 4.4 pmol/min per mg of protein for monoglucuronosylbilirubin and bisglucuronosylbilirubin respectively. ATP-dependent transport of the glutathione S-conjugate [3H]leukotriene C4, an established high-affinity substrate for MRP1 and MRP2, was inhibited by both bilirubin glucuronides with IC50 values between 0.10 and 0.75 μM. The ratios of leukotriene C4 transport and bilirubin glucuronide transport, determined in the same membrane vesicle preparation, indicated substrate specificity differences between MRP1 and MRP2 with a preference of MRP2 for the glucuronides.

The MRP gene encodes a 190-kDa integral membrane glycoprotein which functions as a primary-active ATP-dependent export pump for amphiphilic anions. The MRP gene-encoded conjugate export pump and its canalicular isoform represent the transport activity which has been described earlier as multispecific organic anion transporter, non-bile acid organic anion transporter, glutathione S-conjugate export pump, or leukotriene export pump. Analyses of the substrate specificity of the human MRP pump were performed in plasma membrane vesicles from MRP-overexpressing drug-selected cells (7) and cells transfected with an MRP expression vector (8). Substrates for MRP include thioether-linked conjugates of lipophilic compounds with glutathione, cysteinyl glycine, cysteine, and N-acetyl cysteine, but also glutathione disulfide, and glucuronate conjugates such as etoposide glucuronide. This broad-specificity ATP-dependent export pump is not only overexpressed in several multidrug resistant tumor cells and tissues, but is also present in most normal cells and tissues. The expression of cMRP and MRP in human liver and of cMrp and its homolog Mrp in rat liver was demonstrated by reverse transcription PCR, cDNA sequencing, and immunoblotting (13). The important function of the cMRP gene-encoded broad-specificity conjugate export pump in hepatobiliary excretion is illustrated by the selective absence of this canalicular isoform from the hepatocyte canalicular membrane in transport-deficient mutant rats. This altered lack of cMrp is the basis for the hereditary detect of the hepatobiliary excretion of anionic conjugates in the mutant animals (13). The absence of this canalicular Mrp in the mutants is analogous to the defect in the human Dubin-Johnson syndrome which is characterized by an impaired excretion of conjugated anions across the canalicular membrane.

Cyclosporins are potent tools to inhibit several primary-active, ATP-dependent export carriers. This has been demonstrated in membrane vesicle transport assays for CsA and for its non-immunosuppressive analog PSC 833. Inhibition in the low micromolar and in the nanomolar concentration range is shown for the three distinct ATP-dependent export carriers in the liver canalicular membrane mediating the secretion into bile of leukotrienes (LTC4, other cysteinyl leukotrienes, and related conjugates), bile salts (taurocholate), and amphiphilic, mostly cationic substances (daunorubicin and other P-glycoprotein substrates). Competitive inhibition by cyclosporins is most potent for ATP-dependent taurocholate transport with Ki values of 0.2 and 0.6 μ m for CsA and PSC 833, respectively. This inhibition is in agreement with in vivo studies in the rat demonstrating a block at the canalicular membrane in the hepatobiliary elimination of labeled taurocholate. The data suggest that cholestasis, as a side effect during CsA therapy, is largely due to inhibition of the ATP-dependent bile salt export carrier in the canalicular membrane. Inhibition by cyclosporins is less effective with respect to ATP-dependent leukotriene transport, both during biosynthetic release from mastocytoma cells and during hepatobiliary excretion. The Ki values for the former were 4.5 and 30 μ m , and the Km/Ki ratios only 0.015 and 0.002 for CsA and PSC 833, respectively. Distinct transporters are inhibited by the cyclosporins with different potency and structurally modified cyclosporins may serve to induce preferential inhibition of a selected transporter. This is illustrated by the inhibition of the multidrug export carrier with daunorubicin as substrate using PSC 833 as inhibitor with a Ki value of 0.3 μ m in an in vitro membrane transport system.

Chain shortening via beta-oxidation from the omega-end has been recognized as the major pathway for the degradation of cysteinyl leukotrienes as well as leukotriene B4 (LTB4). The metabolic compartmentation of this pathway was studied using peroxisomes purified from normal and clofibrate-treated rat liver. beta-Oxidation products of omega-carboxy-LTB4, including omega-carboxy-dinor-LTB4 identified by gas chromatography-mass spectrometry, were formed by the isolated peroxisomes. The reaction was dependent on CoA, ATP, and NAD and was stimulated by FAD. NADPH was necessary for the further metabolism of omega-carboxy-dinor-LTB4. Together with microsomes a degradation of omega-carboxy-LTB4 also proceeded in isolated mitochondria in the presence of CoA, ATP, and carnitine. beta-Oxidation of the cysteinyl leukotriene omega-carboxy-N-acetyl-leukotriene E4 was observed only with isolated peroxisomes in combination with lipid-depleted microsomes. Direct photoaffinity labeling using omega-carboxy-[3H] LTB4 and omega-carboxy-N-[3H]acetyl-LTE4 served to identify peroxisomal leukotriene-binding proteins. The bifunctional protein (EC 4.2.1.17 and 1.1.1.35) and 3-ketoacyl-CoA thiolase (EC 2.3.1.16) of the peroxisomal beta-oxidation system were the predominantly labeled polypeptides as revealed by precipitation with monospecific antibodies. In vivo studies with N-acetyl-[3H2]LTE4, N-acetyl-[3H8]LTE4, and N-[14C]acetyl-LTE4 after treatment with the peroxisome proliferator clofibrate indicated formation and biliary excretion of large amounts of metabolites more polar than omega-carboxy-tetranor-N-acetyl-LTE3 including omega-carboxy-tetranor-delta 13-N-acetyl-LTE4 and omega-carboxy-hexanor-N-acetyl-LTE3. Increased formation of beta-oxidized catabolites of N-acetyl-LTE4 and LTB4 was also observed in hepatocytes isolated after clofibrate treatment. Our results indicate that peroxisomes play a major role in the beta-oxidation of leukotrienes from the omega-end. Whereas omega-carboxy-LTB4 was beta-oxidized both in isolated peroxisomes and mitochondria, the cysteinyl leukotriene omega-carboxy-N-acetyl-LTE4 was exclusively degraded in peroxisomes.

The proinflammatory cysteinyl leukotrienes are inactivated in primates by (a) intravascular degradation, (b) hepatic and renal uptake from the blood circulation, (c) intracellular metabolism of leukotriene E4 (LTE4), and (d) biliary and renal excretion of LTC4 degradation products. We have analyzed cysteinyl leukotriene metabolites excreted into bile and urine of the monkey Macaca fascicularis and of man. In both species, hepatobiliary leukotriene elimination predominated over renal excretion. In a representative healthy human subject at least 25% of the administered radioactivity were recovered from bile and 20% from urine within 24 h. In monkey and man intravenous administration of 14,15-3H2-labeled LTC4 resulted in the biliary and urinary excretion of labeled LTE4, omega-hydroxy-LTE4, omega-carboxy-LTE4, omega-carboxy-dinor-LTE4, and omega-carboxy-tetranor-dihydro-LTE4. Small amounts of N-acetyl-LTE4 were detected in human urine only. Oxidative metabolism of LTE4 proceeded more rapidly in the monkey resulting in the formation of higher relative amounts of omega-oxidized leukotrienes in this species as compared to man. [3H]H2O amounted to less than 2% of the administered dose in monkey and human bile and urine samples. Incubation of isolated human hepatocytes with [3H2]LTC4, [3H2]LTD4, and [3H2]LTE4 showed that only [3H2]LTE4 underwent intracellular oxidative metabolism resulting in the formation of omega- and beta-oxidation products. N-Acetylated LTE4 derivatives were not detected as products formed by human hepatocytes. By a combination of reversed-phase high-performance liquid chromatography and radioimmunoassay, endogenous LTE4 and N-acetyl-LTE4 were detected in human urine in concentrations of 220 +/- 40 and 24 +/- 3 pM, corresponding to 12 +/- 1 and 1.5 +/- 0.2 nmol/mol creatinine, respectively (mean +/- SEM; n = 10). Endogenous LTD4 and LTE4 were detected in human bile (n = 3) in concentrations between 0.2-0.9 nM. Our results demonstrate that LTD4 and LTE4 are major LTC4 metabolites in human bile and/or urine and may serve as index metabolites for the measurement of endogenously generated cysteinyl leukotrienes. Moreover, omega-oxidation and subsequent beta-oxidation from the omega-end contribute to the metabolic degradation of LTE4 not only in monkey but also in man.

ω-Oxidation with subsequent β-oxidation from the ω-end is the major pathway for inactivation and degradation of leukotrienes. Oxidative degradation of leukotriene E 4 (LTE 4 ), N -acetyl-LTE 4 , and LTB 4 was inhibited by the ω-trifluoro analogs of LTE 4 , ω-trifluoro-LTE 4 (ω-F 3 -LTE 4 ), and (1, S ,2 R )-5-{3-[1-hydroxy-15, 15,15 - trifluoro - 2 - (2 - 1H - tetrazol - 5 - ylethylthio) pentadeca-3( E ),5( Z )-dienyl]phenyl}-1H-tetrazole (LY 245769). The latter substance inhibited the oxidative degradation of LTE 4 and N -acetyl-LTE 4 in the rat in vivo by 50% at a dose of 7 μmol/kg body weight. In rat hepatocyte cultures both ω-trifluoro analogs interfered with the ω-oxidation of N -acetyl-LTE 4 and LTB 4 with IC 50 values of about 4 μ m . Both analogs inhibited the ω-hydroxylation in isolated rat liver microsomes with IC 50 values between 16 and 37 μ m . This inhibition is apparently competitive. In addition, in liver cytosol, the conversion of the ω-hydroxylated leukotrienes to ω-carboxy-LTE 4 and ω-carboxy-LTB 4 was inhibited by both compounds. ω-Trifluoro analogs of leukotrienes provide a new tool for interfering with the inactivation of leukotrienes in the ω-oxidation pathway.

ATP-dependent para -aminohippurate transport by the apical multidrug resistance protein MRP2. Background Para-aminohippurate (PAH), a widely used model substrate for organic anion transport in proximal tubule epithelia, was investigated as a substrate for the apical multidrug resistance protein MRP2 (symbol ABCC2). This ATP-dependent export pump for anionic conjugates and additional amphiphilic anions was cloned recently and localized to the apical membrane of proximal tubules in human and rat kidney. Methods Membrane vesicles from HEK-MRP2 cells containing recombinant human MRP2 and from control vector-transfected HEK-Co cells were incubated with various concentrations of [ 3 H]PAH, and the net ATP-dependent transport into inside-out vesicles was determined. Comparative studies were performed with membrane vesicles containing recombinant human MRP1. Results Transport rates at 10 μmol/L PAH were 21.9 ± 1.9 and 1.6 ± 0.4 pmol × mg protein -1 × min -1 (means ± SEM, N = 10) with membrane vesicles from HEK-MRP2 and HEK-Co cells, respectively. The K m value for PAH was 880 μmol/L. The high-affinity substrate leukotriene C 4 and the inhibitor of MRP-mediated transport, MK571, inhibited MRP2-mediated transport of PAH (100 nmol/L) with IC 50 values of 3.3 and 4.0 μmol/L, respectively. The nephrotoxic mycotoxin ochratoxin A inhibited MRP2-mediated PAH transport with an IC 50 value of 58 μmol/L. Ochratoxin A was itself a substrate for MRP2. Conclusions PAH is a good substrate for the ATP-dependent export pump MRP2. The localization and function of MRP2 indicate that this unidirectional transport protein contributes to the secretion of PAH and other amphiphilic anions into the lumen of kidney proximal tubules.

Background/Aims: Inhibition of bile salt transport across the hepatocyte during cholestasis induced by cyclosporin A has been shown. However, the contribution of the different bile salt transport systems in liver to cholestasis has remained controversial. Methods: The sensitivity of different bile salt transport systems in liver to cyclosporin-induced inhibition was determined by transport assays in plasma membrane vesicles and by in vivo studies in the rat. Results: Cyclosporin A-induced inhibition of sodium-dependent uptake of bile salts across the sinusoidal membrane, of potential-dependent, and of adenosine triphosphate (ATP)-dependent bile salt transport across the canalicular membrane exhibited inhibition constants ( K l ) of 5, 70, and 0.2 μmol/L, respectively. The nonimmunosuppressive cyclosporin analogue PSC 833 also preferentially inhibited the ATP-dependent bile salt transport with an inhibition constant of 0.6 μmol/L. Cyclosporin A and its analogue PSC 833 [(3′-oxo-4-butenyl-4-methyl-Thr 1 )-(Val 2 )-cyclosporin] (25 mg/kg each) served as tools to interfere with [ 14 C]taurocholate secretion into bile in vivo, causing an accumulation of [ 14 C]-taurocholate taurocholate in liver and reducing bile flow to 50%. In mutant rats deficient in the transport of leukotriene C 4 and related conjugates across the canalicular membrane, bile flow was reduced to 14%. Conclusions: The cyclosporins preferentially inhibit the ATP-dependent bile salt export carrier in the canalicular membrane. This inhibition reduces bile salt-dependent bile flow and causes intrahepatic cholestasis.

The multidrug resistance-associated protein (MRP) is the product of an ATP-binding cassette transporter gene overexpressed in some tumor cells resistant to antineoplastic agents. We studied the transport function of MRP in membrane vesicles prepared from HeLa cells transfected with an MRP expression vector and overexpressing this 190-kDa membrane glycoprotein. ATP-dependent primary-active transport into the vesicles was demonstrated for leukotriene C4 (LTC4), LTD4, LTE4, and S-(2,4-dinitrophenyl)glutathione with relative rates, at a substrate concentration of 50 nM, of 1.0, 0.27, 0.14, and 0.16, respectively. The endogenous glutathione conjugate LTC4 had the highest affinity for this transporter with a Km of 97 nM. The Km for ATP was 19 microM. Direct photoaffinity labeling with [3H]LTC4 labeled a 190-kDa membrane protein predominantly in the MRP-transfected HeLa cells. ATP-dependent LTC4 transport was effectively inhibited by the LTD4 receptor antagonist MK 571, whereas cyclosporin A and, particularly, its analog PSC 833 were much less potent. The respective Ki values were 0.6, 5, and 27 microM, respectively. In addition, MK 571 preferentially inhibited photoaffinity labeling of the 190-kDa protein in the MRP transfectants. Our results provide direct evidence that the MRP gene encodes a primary-active ATP-dependent export pump for conjugates of lipophilic compounds with glutathione and several other anionic residues. We conclude that the biosynthetic release of LTC4 from cells is mediated by the 190-kDa product of the MRP gene.