Introduction

Top Abstract Introduction Results Discussion References

A number of ATP-binding cassette transporters are expressed constitutively in the canalicular membrane of the hepatocyte and mediate the active unidirectional transport of solutes into bile. These transporters include the multidrug resistance (MDR¹) genes and the multidrug resistance-associated protein (MRP) genes. MRP2/Mrp2² mediates the biliary excretion of a wide variety of amphipathic anionic substrates, primarily glucuronide, sulfate, and glutathione conjugates (Keppler and Konig, 1997). The MDR gene family consists of two members in human (MDR1 and MDR3) and three members in rodents (Mdr1a, Mdr1b, and Mdr2). MDR1 transports many hydrophobic therapeutic drugs, and its overexpression confers multiple drug resistance to tumor cells (Ling, 1993; Gottesman and Pastan, 1993). In contrast, Mdr2 expressed in yeast secretory vesicles mediates translocation of a fluorescent phosphatidylcholine analog from the inner to the outer hemileaflet of the membrane (Ruetz and Gros, 1994), while MDR3 promotes a similar translocation of phosphatidylcholine in fibroblasts from transgenic mice (Smith et al., 1994). Homozygous disruption of murine Mdr2 leads to a complete absence of phospholipids in bile without affecting the composition or biliary excretion of bile salts (Smit et al., 1993). Thus, Mdr2/MDR3 translocates phosphatidylcholine from the inner to the outer hemileaflet of the canalicular membrane, followed by its release from the outer hemileaflet into bile.

Biliary lipids are postulated to contribute to the hepatobiliary excretion of certain hydrophobic organic anions by forming phospholipid/cholesterol vesicles and phospholipid/bile salt mixed micelles that associate with these organic anions. This "micellar sink" thus decreases the concentration of the monomeric form of the organic anions in bile, which would otherwise limit their transport (Scharschmidt and Schmid, 1978). Indocyanine green (ICG) is a water-soluble tricarbocyanine dye containing two polar sulfonate groups and a quaternary ammonium group. It is a relatively hydrophobic organic anion that has a high hepatic extraction ratio, and it is excreted extensively unchanged in bile (Cherrick et al., 1960). Characterization of the distribution of ICG in model bile or rat bile has shown that it is extensively (90-100%) associated with phospholipid vesicles and mixed lipid/bile salt micelles (Scharschmidt and Schmid, 1978; Tazuma and Holzbach, 1987). In view of the critical dependence of biliary secretion of phospholipid on Mdr2 expression, and the marked association of ICG with phospholipid vesicles and mixed micelles in bile, we postulated that Mdr2 is important for biliary excretion of ICG. We therefore examined the biliary excretion of ICG, estradiol-17-(-D-glucuronide) (E₂17G), a hydrophobic anion and a high-affinity substrate for Mrp2, and endogenous glutathione, a hydrophilic anion, in wild-type and Mdr2^/ mice.

Experimental Procedures

Materials. [³H]E₂17G (47.3 Ci/mmol, 97% purity) was obtained from PerkinElmer Life Science Products (Boston, MA). E₂17G, taurocholate, estradiol-3-sulfate-17(-D-glucuronide) (E₂3SO₄17G), reduced glutathione (GSH), and ICG were obtained from Sigma Chemical Co. (St. Louis, MO), and GSH reductase (yeast) and 5,5'-dithio-bis-(2-nitrobenzoic acid) from Calbiochem (La Jolla, CA). E₂17G was dissolved in buffer/propylene glycol/ethanol (10:4:1, v/v/v).

Animals. FVB/NJ wild-type and Mdr2^/ female mice (5-7 weeks old) (Jackson Lab, Bar Harbor, ME) had free access to food and water and were maintained on a 12-h, automatically timed light/dark cycle in a high efficiency particulate air-filtered laminar flow unit for at least 7 days before use. All procedures involving animals were conducted in accordance with guidelines for the Care and Use of Laboratory Animals.

	Results

Top Abstract Introduction Results Discussion References

Biliary Excretion of ICG in Wild-Type and Mdr2^/ Mice. Relatively young mice (6-8 weeks old) were used to minimize liver disease in Mdr2^/ mice (Smit et al., 1993). The size of the liver (mean ± S.D.) of Mdr2^/ mice (1.85 ± 0.16 g) was about twice that of wild-type mice (0.95 ± 0.16 g), although their body weights (mean ± S.D.) were similar (23.5 ± 2.2 and 25.1 ± 1.3 g in wild-type and Mdr2^/, respectively). Mdr2^/ mice produced more bile than wild-type mice, both in vivo (data not shown) and in the perfused liver (Fig. 1). The biliary excretion of ICG was profoundly decreased in Mdr2^/ mice, even though the effects of ICG (0.4 µmol) on bile flow were minimal. The differences in peak biliary concentration and cumulative excretion over 50 min of ICG between wild-type and Mdr2^/ mice were about 26.3- and 5.8-fold, respectively (Fig. 1, left panels). Subtraction of the ICG in the perfusate outflow from the dose administered indicated that hepatic uptake of ICG was similar in livers from wild-type and Mdr2^/ mice (data not shown).

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Bile flow, biliary concentration, and cumulative biliary excretion of ICG (left panels) and E217G (right panels) in the single-pass perfused liver from wild-type (closed symbols) and Mdr2/ (open symbols) mice. ICG (0.4 µmol) or 0.1 µmol of E217G was administered via the portal vein. Values represent the mean ± S.E. (n = 3).

Biliary Excretion of E₂17G in Mdr2^/ Mice. Although the concentration of E₂17G equivalents in bile was significantly (p < 0.05) lower in Mdr2^/ mice than in wild-type mice, the increased bile flow in Mdr2^/ mice minimized the difference in its biliary excretion (Fig. 1, right panels). Following a noncholestatic dose of E₂17G (0.1 µmol), the cumulative biliary excretion of E₂17G equivalents at 60 min was 55 and 64% of the dose in wild-type and Mdr2^/ mice, respectively (p > 0.05). A significantly (p < 0.05) higher proportion of [³H]E₂17G equivalents in bile was identified as E₂3SO₄17G in wild-type mice than in Mdr2^/ mice (15.5 ± 2.6 versus 2.5 ± 0.5%, respectively).

Biliary Excretion of GSH in Mdr2^/ Mice. The biliary excretion of endogenous total GSH was significantly decreased by 65% in Mdr2^/ mice relative to wild-type mice. The biliary secretory rate of total GSH in wild-type controls and Mdr2^/ mice was 67.5 ± 11.3 and 23.7 ± 1.9 nmol/min/liver, respectively, while the concentration in bile was 2.6 ± 0.2 and 0.3 ± 0.1 mM, respectively.

	Discussion

Top Abstract Introduction Results Discussion References

While the biliary excretion of ICG in Mdr2^/ mice was reduced to about 10% of that in wild-type mice, a similar proportion of the dose of E₂17G was secreted in bile of Mdr2^/ mice (64% of the dose) and in wild-type mice (55% of the dose). The basis for the decreased sulfation of E₂17G in Mdr2^/ mice is not known and has not been described for other substrates. It may reflect a longer residence time of E₂17G in liver of wild-type mice, permitting further sulfation. The biliary excretion of total GSH in Mdr2^/ mice was decreased to about one third of that in wild-type controls, consistent with initial reports (Smit et al., 1993). Thus, the biliary excretion of the most hydrophobic organic anion, E₂17G, was slightly increased, while biliary excretion of ICG, an anion of intermediate hydrophobicity, was markedly decreased, and that of the most hydrophilic anion, GSH, was inhibited to an intermediate extent. Several mechanisms for these disparate effects, particularly the decreased biliary excretion of ICG in Mdr2^/ mice, seem plausible.

Decreased Expression of Mrp2 in Mdr2^/ Mice. According to a preliminary report (Lammert et al., 1999), expression of Mrp2 is decreased 35% in Mdr2^/ mice 8 to 22 weeks of age. Mrp2 contributes significantly to the GSH in bile (Paulusma et al., 1999), which is almost completely absent in bile of Mrp2-deficient rats (Oude Elferink et al., 1989). The 65% decrease in biliary excretion of GSH in Mdr2^/ mice could therefore be attributed to decreased expression of Mrp2. Since the biliary excretion of the high-affinity MRP2 substrates E₂17G and bilirubin glucuronides (Cui et al., 1999; Kamisako et al., 1999) was not decreased in Mdr2^/ mice (Fig. 1; Smit et al., 1993), it is unlikely that decreased expression of Mrp2 is the primary mechanism for the markedly decreased biliary excretion of ICG in Mdr2^/ mice. Finally, as discussed below, several authors have suggested that ICG is not an Mrp2 substrate. Decreased biliary excretion of GSH in Mdr2^/ most likely reflects the cholangitis and hepatocyte damage that results when, in the absence of phospholipid, bile salts present in high concentrations in bile are unable to form mixed micelles and elute phospholipids and cholesterol from adjacent membranes (Smit et al., 1993).

Mdr2 Directly Translocates ICG across the Canalicular Membrane. Recently, Smith et al. (2000) reported an increased directional transport of several MDR1 substrates (digoxin, paclitaxel, and vinblastine) across polarized monolayers of MDR3-transfected cells, indicating that MDR3 transports a range of drugs in addition to phosphatidylcholine. Several reports have postulated two different pathways for canalicular secretion of organic anions and have suggested that ICG and rose bengal do not share the same transport system as that for bromosulfophthalein and dibromosulfophthalein, i.e., Mrp2 (Mahu et al., 1977; Sathirakul et al., 1993). The biliary excretion of ICG in Mrp2-deficient rats is 25 to 75% of that in normal rats (Verkade et al., 1990; Jansen et al., 1993) and is relatively well preserved compared with other Mrp2 substrates. Sathirakul et al. (1993) demonstrated that the decreased excretion of ICG in Mrp2-deficient rats is probably due to alterations in its intracellular binding and transport, and not to decreased transport across the canalicular membrane, and they suggested that Mrp2 is not the primary transporter for canalicular secretion of ICG. Thus, an alternate pathway for ICG biliary excretion could reflect Mdr2-mediated transport. However, ICG and rose bengal do not affect biliary lipid excretion (Verkade et al., 1990; Yamashita et al., 1992; Takikawa et al., 1993), which argues against these anions interacting directly with Mdr2. Alternatively, the decreased intracellular transport of ICG noted by Sathirakul et al. (1993) was attributed to decreased intracellular binding to ligandin(s) resulting from accumulation of bilirubin glucuronides in Mrp2-deficient rats, and increased partitioning of ICG into a "deep" compartment (organelles). While biliary excretion of bilirubin glucuronides is not altered in Mdr2^/ mice, plasma bilirubin is elevated (Smit et al., 1993) and could influence the intracellular binding of ICG, thus disrupting its biliary excretion.

Biliary Excretion of ICG Requires the Presence of Biliary Phospholipid Vesicles and Mixed Micelles. Several lines of evidence indicate that incorporation into phospholipid vesicles and mixed micelles is important for the biliary excretion of ICG. Gel permeation chromatography, ultrafiltration, and ultracentrifugation have been used to examine the association of a series of organic anions of varying hydrophobicity with biliary lipid particle species, including phospholipid/cholesterol vesicles, mixed bile salt-lipid micelles, and simple bile salt micelles. ICG is consistently found in association with phospholipid vesicles and mixed micelles, as are other relatively hydrophobic organic anions (Scharschmidt and Schmid, 1978; Tazuma and Holzbach, 1987; Verkade et al., 1990). Furthermore, the biliary excretion of ICG is markedly stimulated (138%) by infusion of taurocholate, a micelle-forming bile salt, whereas infusion of dehydrocholate, a nonmicelle-forming bile salt, stimulated ICG biliary excretion by only 55%. Bile flow was stimulated 195 and 297% by infusion of taurocholate and dehydrocholate, respectively (Vonk et al., 1974). Thus, the increased biliary excretion of ICG seen upon infusion of taurocholate is not due simply to the increase in bile flow but rather reflects the increased formation of mixed micelles, consistent with data demonstrating that loading of Mdr2-expressing yeast secretory vesicles with taurocholate, but not with taurodehydrocholate, enhances Mdr2-mediated phosphatidylcholine translocation activity (Ruetz and Gros, 1995). Furthermore, the biliary secretion of protoporphyrin is reduced by 90% in Mdr2^/ mice, whereas that of the more hydrophilic coproporphyrins I and III is affected to a much lesser extent (Beukeveld et al., 1996). Protoporphyrin in human and rat bile is associated with phospholipid vesicles, while coproporphyrins I and III are associated with bile salt micelles. Ultrarapid cryofixation of livers in situ has shown that in Mdr2^/ mice, the number of phospholipid vesicles within the canalicular lumen is 12% of that in wild-type mice (Crawford et al., 1997). The similar reduction between phospholipid vesicle numbers and biliary excretion of ICG and protoporphyrin in Mdr2^/ mice strongly supports the hypothesis that phospholipid vesicles and mixed micelles contribute to the biliary excretion of these organic anions and serve as a micellar sink. The chemical/biochemical mechanism for the high association of ICG with these components of bile is not known, nor is the basis for the necessity of this association for its biliary elimination.

/