ATP-dependent Transport of Bile Salts by Rat Multidrug Resistance-associated Protein 3 (Mrp3)*

We have previously shown that cloned rat multidrug resistance-associated protein 3 (Mrp3) has the ability to transport organic anions such as 17β-estradiol 17-β-d-glucuronide (E217βG) and has a different substrate specificity from MRP1 and MRP2 in that glutathione conjugates are poor substrates for Mrp3 (Hirohashi, T., Suzuki, H., and Sugiyama, Y. (1999) J. Biol. Chem. 274, 15181–15185). In the present study, the involvement of Mrp3 in the transport of endogenous bile salts was investigated using membrane vesicles from LLC-PK1 cells transfected with rat Mrp3 cDNA. The ATP-dependent uptake of [3H]taurocholate (TC), [14C]glycocholate (GC), [3H]taurochenodeoxycholate-3-sulfate (TCDC-S), and [3H]taurolithocholate-3-sulfate (TLC-S) was markedly stimulated by Mrp3 transfection in LLC-PK1 cells. The extent of Mrp3-mediated transport of bile salts was in the order, TLC-S > TCDC-S > TC > GC. The K m andV max values for the uptake of TC and TLC-S wereK m = 15.9 ± 4.9 μm andV max = 50.1 ± 9.3 pmol/min/mg of protein and K m = 3.06 ± 0.57 μm andV max = 161.9 ± 21.7 pmol/min/mg of protein, respectively. At 55 nm[3H]E217βG and 1.2 μm[3H]TC, the apparent K m values for ATP were 1.36 and 0.66 mm, respectively. TC, GC, and TCDC-S inhibited the transport of [3H]E217βG and [3H]TC to the same extent with an apparent IC50 of ∼10 μm. TLC-S inhibited the uptake of [3H]E217βG and [3H]TC most potently (IC50 of ∼1 μm) among the bile salts examined, whereas cholate weakly inhibited the uptake (IC50 ∼75 μm). Although TC and GC are transported by bile salt export pump/sister of P-glycoprotein, but not by MRP2, and TCDC-S and TLC-S are transported by MRP2, but not by bile salt export pump/sister of P-glycoprotein, it was found that Mrp3 accepts all these bile salts as substrates. This information, together with the finding that MRP3 is extensively expressed on the basolateral membrane of human cholangiocytes, suggests that MRP3/Mrp3 plays a significant role in the cholehepatic circulation of bile salts.

Rat Mrp3, initially referred to as MRP-like protein 2, has been cloned from the liver of cMOAT/MRP2-deficient rats (EHBR) as an inducible protein under cholestatic conditions (15). Recently, König et al. (16) showed that MRP3 is located on the basolateral membrane of human hepatocytes with exceptionally high expression in the livers of patients whose cMOAT/ MRP2 function is deficient (Dubin-Johnson syndrome) and from a patient with primary biliary cirrhosis. In addition, Kool et al. (17) reported high expression of MRP3 on the lateral membrane of intrahepatic bile duct epithelial cells (cholangiocytes) and the basolateral membrane of the hepatocytes surrounding the portal tracts. It may be plausible that MRP3/ Mrp3 compensates for the impaired function of cMOAT/MRP2 in the liver. In addition, because Northern blot analysis indicated the extensive expression of MRP3/Mrp3 in intestinal tissues of humans and rats, it is possible that MRP3/Mrp3 is also responsible for the intestinal transport of organic anions (2,15,16,18). Our recent findings are consistent with this hypothesis. Recently, we demonstrated that rat Mrp3 mediates the transport of some organic anions, including glucuronides (e.g. E 2 17␤G) and methotrexate (19). Moreover, the substrate specificity of Mrp3 was shown to differ markedly from that of MRP1 and cMOAT/MRP2 in that glutathione conjugates (e.g. 2,4-dinitrophenyl-S-glutathione and leukotriene C 4 ) are poor substrates for Mrp3 (19).
In the present study, we examined whether bile salts, another class of organic anions, are transported via Mrp3 to identify further physiological substrates for Mrp3. At present, at least two different pathways for the biliary excretion of bile salts have been reported (6); non-sulfated bile salts such as taurocholate (TC), cholate (CA), and glycocholate (GC) are transported via bile salt export pump/sister of P-glycoprotein (BSEP/spgp) (20), whereas in vivo studies suggest that the biliary excretion of sulfated bile salts such as taurochenodeoxycholate-3-sulfate (TCDC-S) and taurolithocholate-3-sulfate (TLC-S) is mediated by cMOAT/Mrp2 (21). The transport studies using membrane vesicles prepared from rat Mrp3-transfected LLC-PK1 cells have enabled us to show that both nonsulfated and sulfated bile salts are physiological substrates for Mrp3.  (22). Chenodeoxycholate-3-sulfate was synthesized as described previously (23). TC and CA were purchased from Calbiochem and Wako Ltd. (Osaka, Japan), respectively. GC, TLC-S, lithocholate-3-sulfate, and E 2 17␤G were purchased from Sigma.
Membrane vesicles were prepared from 1 ϫ 10 8 of the Mrp3-and vector-transfected LLC-PK1 cell populations as described previously (25). The membrane vesicles were frozen in liquid nitrogen and then transferred to a freezer (Ϫ100°C) until required. The transport studies were performed using the rapid filtration technique (26). Briefly, 16 l of transport medium (10 mM Tris-HCl, 250 mM sucrose, 10 mM MgCl 2 , pH 7.4) containing radiolabeled compounds, with or without unlabeled substrate, were preincubated at 37°C for 3 min and then rapidly mixed with a 4-l membrane vesicle suspension (10 g of protein). The reaction mixture contained 5 mM ATP or AMP and ATP-regenerating system (10 mM creatine phosphate, 100 g/ml creatine phosphokinase). The transport reaction was stopped by the addition of a 1-ml ice-cold stop solution containing 250 mM sucrose, 0.1 M NaCl, and 10 mM Tris-HCl (pH 7.4). The stopped reaction mixture was filtered through a 0.45-m mixed cellulose ester filter (Millipore Corp., Bedford, MA) and then washed twice with 5 ml of stop solution. Radioactivity retained on the filter was determined using a liquid scintillation counter (LSC-3500, Aloka Co., Tokyo, Japan). The ATP-dependent uptake of ligands was calculated by subtracting the ligand uptake in the absence of ATP from that in the presence of ATP.
Preparation of Antibody-The polyclonal antibody was obtained by immunizing rabbits with a maltose-binding protein fusion protein containing the 136 amino acids corresponding to 838 -973 of the deduced rat Mrp3 amino acid sequence. The pMAL-c2 expression vector (New England Biolabs, Inc., Beverly, MA) was used for the expression of the fusion protein. After the purification by amylose resin, rabbits were immunized with 250 g of fusion protein mixed with Freund's complete adjuvant (Sigma).
Western Blotting-Crude membrane fractions were prepared from the liver of Sprague-Dawley (SD) rats and EHBR as described previously (27). Membrane fractions from transfected LLC-PK1 cells (60 g of protein) or crude liver membrane (30 g of protein) were loaded on a 7.5% polyacrylamide slab gel containing 0.1% SDS and then transferred onto a nitrocellulose filter by electroblotting. The filter was blocked with Tris-buffered saline containing 0.05% Tween-20 and 5% bovine serum albumin for 2 h at room temperature and probed overnight at 4°C with polyclonal anti-Mrp3 antibody (dilution 1:1000). Antibody was visualized with [ 125 I]anti-rabbit antibody (Amersham Pharmacia Biotech), exposed to Fuji imaging plates (Fuji Photo Film Co., Ltd., Kanagawa, Japan) for 3 h at room temperature, and analyzed with an imaging analyzer (BAS 2000, Fuji Photo Film Co., Ltd.).

Uptake of Bile Salts into Membrane Vesicles-
The expression of rat Mrp3 in the transfected cells was confirmed by Western blot analysis using the antiserum (Fig. 1). The antibody recognized an approximately 190-kDa band in liver membrane vesicles from EHBR (Fig. 1, lane 1). The expression level in this mutant rat liver was much higher than that in SD rats (Fig. 1, lane 2). The immunoreactive 190-kDa band was also observed in the vesicles from Mrp3-transfected, but not from vectortransfected, LLC-PK1 cells ( Fig. 1, lanes 3 and 4). Fig. 2. The ATP-dependent uptake of these bile salts was stimulated by the transfection of Mrp3, indicating that both sulfated and non-sulfated bile salts are substrates for rat Mrp3.   (Fig. 4). Kinetic parameters (K m and V max ) were found to be K m ϭ 15.9 Ϯ 4.9 M and V max ϭ 50.1 Ϯ 9.3 pmol/min/mg of protein for TC (Fig. 4a) and K m ϭ 3.06 Ϯ 0.57 M and V max ϭ 161.9 Ϯ 21.7 pmol/min/mg of protein for TLC-S (Fig. 4b) (19).

The uptake of [ 3 H]TC, [ 14 C]GC, [ 3 H]TCDC-S, and [ 3 H]TLC-S by the membrane vesicles from Mrp3-or vector-transfected LLC-PK1 cells is shown in
Osmotic Sensitivity of [ 3 H]TC Uptake into Membrane Vesicles-To confirm that the vesicle-associated uptake of [ 3 H]TC reflects transport into a vesicular space, rather than binding to the vesicle surface, the uptake of [ 3 H]TC was measured in the presence of several concentrations of sucrose in the transport medium (Fig. 6). The uptake of [ 3 H]TC was reduced on increasing the sucrose concentration in the uptake medium (Fig. 6). Moreover, the y intercept for the relationship between the amount of TC associated with the vesicles versus the reciprocal of the sucrose concentration in the medium, which indicates the binding of [ 3 H]TC to the vesicle surface, was similar in the presence (9.6 l/mg of protein) and absence (11.4 l/mg of protein) of ATP. This suggests that the major part of the ATPdependent uptake of [ 3 H]TC is because of the uptake and not because of adsorption to the vesicle surface. DISCUSSION We have previously shown that rat Mrp3 is a primary active transporter for several organic anions (19). Although Mrp3 recognizes glucuronide conjugates as good substrates, the substrate specificity of Mrp3 is in marked contrast to that of MRP1 and cMOAT/MRP2 in that glutathione conjugates are poor substrates for Mrp3 (19). In the present study, we examined the transport of bile salts via rat Mrp3 to investigate the physiological substrates for this transporter by using membrane vesicles prepared from LLC-PK1 transfected with cloned rat Mrp3 cDNA.
Cumulative evidence indicated that MRP1 and cMOAT/ MRP2 mediate the transport of glucuronide and sulfate conju-gates of bile salts but not monovalent bile salts (5,6,9,28), e.g. transport studies with membrane vesicles isolated from cDNA-transfected cells revealed that TLC-S is a substrate for human MRP1 (29). Moreover, several in vivo results suggest that glucuronide and sulfate conjugates of bile salts are extruded into the bile with the aid of Mrp2; the biliary excretion of intravenously administered 3␣-sulfates of taurochenodeoxycholate, taurolithocholate and glycolithocholate (21), and 3-Oglucuronides of cholate and lithocholate (30) was impaired in GY/transport-deficient rats compared with control rats. A similar phenomenon has been also reported in EHBR (31). Moreover, we have provided direct evidence that both TLC-S and TCDC-S are substrates for cMOAT/Mrp2; the ATP-dependent transport of these sulfated bile salts was observed in isolated bile canalicular membrane vesicles (CMVs) from SD rats, but not from EHBR (40). In contrast, it has been demonstrated that the transport of monovalent bile salts across the bile canalicular membrane is mediated by BSEP/spgp (28,32). This conclusion is based on the finding that the transport activity of monovalent bile salts is comparable between normal and Mrp2deficient rats (21,30,31). Recently, Gerloff et al. (20) demonstrated that spgp, which shows higher homology with MDR1 rather than the MRP family, is endowed with the BSEP function. As shown in Figs. 2 and 3, Mrp3 exhibits a unique ability to transport bile salts; the ATP-dependent uptake of both monovalent (TC and GC) and sulfated (TLC-S and TCDC-S) bile salts was markedly stimulated by Mrp3 transfection in LLC-PK1 cells. Kinetic analysis revealed that the K m value of TLC-S for Mrp3 was 3.1 M in Mrp3-transfected LLC-PK1 cells (Fig. 4), indicating that Mrp3 has also similar high affinity for TLC-S as observed for cMOAT/MRP2 in CMVs (0.9 M) (40). The ATP-dependent transport of TC was also saturable (K m ϭ 15.9 M) (Fig. 4) with K m values severalfold higher than those determined in CMVs (2.1 M) (32) and membrane vesicles from Sf9 cells expressing BSEP/spgp (5.5 M) (20). The K m value for ATP determined by measuring the Mrp3-mediated transport of [ 3 H]E 2 17␤G (1.36 mM) (Fig. 5) was in a similar range with that of human MRP1 (0.39 mM) (33), rat cMOAT/Mrp2 (0.26 mM) (34), and rabbit cMOAT/Mrp2 (0.62 mM) (14).
We also determined the mutual inhibitory effect between bile salts and E 2 17␤G in membrane vesicles expressing Mrp3 (Table I). The inhibitory effect of E 2 17␤G on the ATP-dependent uptake of [ 3 H]TC was minimal even at 75 M (Table I) (20). The efficiency of the transport of CA is much less than that of TC in CMVs due to the low affinity (K m ϭ 84 M), 2 and in addition, no significant effect of spgp transfection was observed in membrane vesicles from Sf9 cells (20). TLC-S potently inhibited the transport of [ 3 H]E 2 17␤G and [ 3 H]TC at lower concentrations (Table I), consistent with its high affinity to Mrp3, as discussed in the previous paragraph. The less inhibition potency of TCDC-S compared with that of TLC-S also agreed with our recent findings for cMOAT/Mrp2; in CMVs, the K m value for TCDC-S (8.8 M) was significantly higher than that for TLC-S (0.9 M) (40). We previously found that Mrp3 is expressed in the liver of EHBR, whose cMOAT/Mrp2 function is hereditarily defective, but not in the liver of normal SD rats (15). The expression of Mrp3 is induced in SD rat liver by phenobarbital treatment and by treatment, which increases the plasma bilirubin concentration and/or its glucuronide (e.g. cholestasis induced by common bile duct ligation and by ␣-naphthylisothiocyanate treatment) (41). These results are in accordance with immunohistochemical studies showing high expression of MRP3 in human liver specimens from patients whose cMOAT/MRP2 function is deficient (Dubin-Johnson syndrome) and from a patient with primary biliary cirrhosis compared with normal liver samples (16). Considering that MRP3 is located on the basolateral membrane of human hepatocytes (16,17), it is possible that MRP3/ Mrp3 may play an important role in protecting hepatocytes from endogenous bile salts by extruding them from hepatocytes into the blood. In bile-duct ligated rat liver, down-regulation of Na ϩ /taurocholate cotransporting polypeptide, a transporter responsible for the Na ϩ -dependent uptake of bile salts into hepatocytes, has been reported (35). The down-regulation of Na ϩ / taurocholate cotransporting polypeptide and up-regulation of Mrp3 may be regarded as a regulatory mechanism to reduce the intrahepatic concentration of bile salts. In addition, because the expression of MRP3 on the (baso)lateral membrane of cholangiocytes has been postulated 3 (17), it is possible that MRP3, together with apical sodium-dependent bile acid transporter expressed in these intrahepatic biliary epithelial cells (36), may also participate in cholehepatic circulation of bile salts (37).
MRP3/Mrp3 is also extensively expressed in human and rat intestinal tissues under physiological conditions (2,15,16,18). By using antiserum against human MRP3, we could detect MRP3 on the basolateral membrane of small intestinal epithelial cells in humans. 3 Although the localization of Mrp3 in rat small intestine still remains to be clarified, it is possible that Mrp3 is also expressed on the basolateral membrane of these epithelial cells in rats. Although numerous studies have suggested that several kinds of transporters on the brush border membrane mediate the intestinal uptake of bile salts (reviewed in Ref. 38), little is known about the basolateral secretion of bile salts. There is only a report by Weinberg et al. (39), which indicated the presence of sodium-independent organic anion exchange protein using basolateral membrane vesicles. It may be that Mrp3 plays a role in the enterohepatic circulation of bile salts by transporting them from enterocytes into circulating blood. Further studies on the role of Mrp3 in intestinal tissues will provide us with new insights into the mechanism of enterohepatic circulation of bile salts.
In conclusion, we demonstrated that not only sulfated bile salts (TCDC-S and TLC-S) but also monovalent bile salts (TC and GC) are good substrates for rat Mrp3. The substrate specificity of Mrp3 is in marked contrast to that of MRP1 and cMOAT/MRP2, in that monovalent bile salts are not significantly transported by MRP1 and cMOAT/MRP2. It is possible that Mrp3 acts as an inducible transporter to compensate for the reduced function of cMOAT/MRP2 and/or BSEP/spgp and extrudes bile salts into the blood. In addition, Mrp3 may be involved in the cholehepatic and enterohepatic circulation by extruding bile salts from cholangiocytes and enterocytes, respectively, into the blood.