The role of different P-glycoproteins in hepatobiliary secretion of fluorescently labeled short-chain phospholipids
1999; Elsevier BV; Volume: 40; Issue: 11 Linguagem: Inglês
10.1016/s0022-2275(20)32418-4
ISSN1539-7262
AutoresC.M.G. Frijters, Coosje J. Tuijn, Roelof Ottenhoff, Bart N. Zegers, Albert K. Groen, Ronald P.J. Oude Elferink,
Tópico(s)Lipid Membrane Structure and Behavior
ResumoClass III P-glycoproteins (Pgps) mediate biliary phosphatidylcholine (PC) secretion. Recent findings that class I P-glycoproteins are able to transport several short-chain phospholipid analogues raises questions about the role of these Pgps in physiological lipid transport. We investigated the biliary secretion of C6-7-nitro-2,1,3-benzoxadiazol-4-yl (NBD)-labeled ceramide and its metabolites in Mdr1a/b and Mdr2 knockout mice compared to control mice. Biliary secretion of these NBD-lipids was unaffected in Mdr1a/b –/– mice. Thus neither Mdr1a nor Mdr1b Pgp mediates biliary secretion of these lipids. In contrast, secretion of all three NBD-labeled short-chain phospholipids was significantly reduced in Mdr2 –/– mice. As in vitro studies revealed that Mdr2 Pgp is not able to translocate these lipid analogues, we hypothesized that Mdr2 –/– mice had a reduced PC content of the exoplasmic canalicular membrane leaflet so that extraction of the short-chain lipid probes from this membrane by canalicular bile salts was impaired. To investigate this possibility we studied the bile salt-mediated extraction of natural sphingomyelin (SM) and NBD-labeled short-chain SM from small unilamellar vesicles of different lipid composition. Natural SM could be extracted by the bile salt tauroursodeoxycholate from vesicles containing PC, cholesterol (CHOL), and SM (1:2:2) but not from vesicles containing only SM and CHOL (3:2). NBD-labeled short-chain SM could be extracted from vesicles containing PC while its extraction from pure SM:CHOL vesicles was reduced by 65%. These data confirm that the efficiency of NBD-SM extraction depends on the lipid composition and suggest that the canalicular membrane outer leaflet of Mdr2 –/– mice has a reduced PC content.—Frijters, C. M. G., C. J. Tuijn, R. Ottenhoff, B. N. Zegers, A. K. Groen, and R. P. J. Oude Elferink. Role of different P-glycoproteins in hepatobiliary secretion of fluorescently labeled short-chain phospholipids. J. Lipid Res. 1999. 40: 1950–1957. Class III P-glycoproteins (Pgps) mediate biliary phosphatidylcholine (PC) secretion. Recent findings that class I P-glycoproteins are able to transport several short-chain phospholipid analogues raises questions about the role of these Pgps in physiological lipid transport. We investigated the biliary secretion of C6-7-nitro-2,1,3-benzoxadiazol-4-yl (NBD)-labeled ceramide and its metabolites in Mdr1a/b and Mdr2 knockout mice compared to control mice. Biliary secretion of these NBD-lipids was unaffected in Mdr1a/b –/– mice. Thus neither Mdr1a nor Mdr1b Pgp mediates biliary secretion of these lipids. In contrast, secretion of all three NBD-labeled short-chain phospholipids was significantly reduced in Mdr2 –/– mice. As in vitro studies revealed that Mdr2 Pgp is not able to translocate these lipid analogues, we hypothesized that Mdr2 –/– mice had a reduced PC content of the exoplasmic canalicular membrane leaflet so that extraction of the short-chain lipid probes from this membrane by canalicular bile salts was impaired. To investigate this possibility we studied the bile salt-mediated extraction of natural sphingomyelin (SM) and NBD-labeled short-chain SM from small unilamellar vesicles of different lipid composition. Natural SM could be extracted by the bile salt tauroursodeoxycholate from vesicles containing PC, cholesterol (CHOL), and SM (1:2:2) but not from vesicles containing only SM and CHOL (3:2). NBD-labeled short-chain SM could be extracted from vesicles containing PC while its extraction from pure SM:CHOL vesicles was reduced by 65%. These data confirm that the efficiency of NBD-SM extraction depends on the lipid composition and suggest that the canalicular membrane outer leaflet of Mdr2 –/– mice has a reduced PC content.—Frijters, C. M. G., C. J. Tuijn, R. Ottenhoff, B. N. Zegers, A. K. Groen, and R. P. J. Oude Elferink. Role of different P-glycoproteins in hepatobiliary secretion of fluorescently labeled short-chain phospholipids. J. Lipid Res. 1999. 40: 1950–1957. In mice, biliary phospholipid secretion is controlled by the canalicular secretion of bile salts and by the activity of Mdr2 P-glycoprotein (Pgp). This ABC-transporter translocates phosphatidylcholine (PC) from the inner leaflet to the outer leaflet of the hepatocyte canalicular membrane, thereby supplying PC molecules for extraction by bile salts in the canalicular space (1Smit J.J.M. Schinkel A.H. Oude Elferink R.P.J. Groen A.K. Wagenaar E. van Deemter L. Mol C.A.A.M. Ottenhoff R. Van der Lugt N.M.T. van Roon M.A. Van der Valk M.A. Offerhaus G.J.A. Berns A.J.M. Borst P. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease.Cell. 1993; 75: 451-462Google Scholar, 2Oude Elferink R.P. Groen A.K. The role of mdr2 P-glycoprotein in biliary lipid secretion. Cross-talk between cancer research and biliary physiology.J. Hepatol. 1995; 23: 617-625Google Scholar, 3Frijters C.M.G. Groen A.K. Oude Elferink R.P.J. Mdr2 p-glycoprotein-mediated lipid secretion and its relevance to biliary drug transport.Adv. Drug Delivery Rev. 1997; 25: 201-215Google Scholar, 4Oude Elferink R.P.J. Ottenhoff R. Van Wijland M.J.A. Smit J.J.M. Schinkel A.H. Groen A.K. Regulation of biliary lipid secretion by mdr2-P-glycoprotein in the mouse.J. Clin. Invest. 1995; 95: 31-38Google Scholar, 5Ruetz S. Gros P. Phosphatidylcholine translocase: A physiological role for the mdr2 gene.Cell. 1994; 77: 1071-1081Google Scholar). The gene products of murine Mdr1a and Mdr1b and human MDR1, have been identified as amphipathic drug export pumps. Disruption of the Mdr1a and Mdr1b genes in mice had no effect on the phospholipid composition of bile (6Schinkel A.H. Smit J.J.M. Van Tellingen O. Beijnen J.H. Wagenaar E. van Deemter L. Mol C.A.A.M. Van der Valk M.A. Robanus-Maandag E.C. Teriele H.P.J. Berns A.J.M. Borst P. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs.Cell. 1994; 77: 491-502Google Scholar, 7Schinkel A.H. Mayer U. Wagenaar E. Mol C.A. van Deemter L. Smit J.J. Van der Valk M.A. Voordouw A.C. Spits H. Van Tellingen O. Zijlmans J.M. Fibbe W.E. Borst P. Normal viability and altered pharmacokinetics in mice lacking mdr1-type (drug-transporting) p-glycoproteins.Proc. Natl. Acad. Sci. USA. 1997; 94: 4028-4033Google Scholar), suggesting that these Pgps are not important for biliary phospholipid secretion. Instead, these Pgps are known to be involved in the biliary secretion of amphipathic drugs (6Schinkel A.H. Smit J.J.M. Van Tellingen O. Beijnen J.H. Wagenaar E. van Deemter L. Mol C.A.A.M. Van der Valk M.A. Robanus-Maandag E.C. Teriele H.P.J. Berns A.J.M. Borst P. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs.Cell. 1994; 77: 491-502Google Scholar, 8Smit J.W. Schinkel A.H. Muller M. Weert B. Meijer D.K. Contribution of the murine mdr1a p-glycoprotein to hepatobiliary and intestinal elimination of cationic drugs as measured in mice with an mdr1a gene disruption.Hepatology. 1998; 27: 1056-1063Google Scholar). In several studies Pgp-mediated translocation of fluorescent phospholipids across membrane bilayers was studied in vitro. The fatty acid on the sn-2 position of these lipids is replaced by a short acyl-chain (6 C-atoms) with 7-nitro-2,1,3-benzoxadiazol-4-yl group (NBD) as a fluorophore. The important advantage of these lipid analogues is that they can be readily inserted into biological membranes by spontaneous lipid transfer from exogenous carriers like bovine serum albumin (BSA) or liposomes. Ruetz and Gros (5Ruetz S. Gros P. Phosphatidylcholine translocase: A physiological role for the mdr2 gene.Cell. 1994; 77: 1071-1081Google Scholar) showed an ATP-dependent translocation of NBD-PC in the Mdr2 transfected yeast but not in Mdr1a transfected yeast. Ruetz and Gros (9Ruetz S. Gros P. Enhancement of mdr2-mediated phosphatidylcholine translocation by the bile salt taurocholate: implications for hepatic bile formation.J. Biol. Chem. 1995; 270: 25388-25395Google Scholar) and Nies, Gatmaitan, and Arias (10Nies A.T. Gatmaitan Z. Arias I.M. Atp-dependent phosphatidylcholine translocation in rat liver canalicular plasma membrane vesicles.J. Lipid Res. 1996; 37: 1125-1136Google Scholar) further demonstrated that this transport can be stimulated by bile salts. Smit et al. (1Smit J.J.M. Schinkel A.H. Oude Elferink R.P.J. Groen A.K. Wagenaar E. van Deemter L. Mol C.A.A.M. Ottenhoff R. Van der Lugt N.M.T. van Roon M.A. Van der Valk M.A. Offerhaus G.J.A. Berns A.J.M. Borst P. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease.Cell. 1993; 75: 451-462Google Scholar) provided evidence for translocation of natural PC in intact fibroblasts that express a MDR3 transgene. These results confirmed the proposed PC translocation function of the orthologues Mdr2 Pgp (mouse), Pgp3 (rat), and MDR3 Pgp (human) and they explain the absence of biliary PC in Mdr2 knockout mice (1Smit J.J.M. Schinkel A.H. Oude Elferink R.P.J. Groen A.K. Wagenaar E. van Deemter L. Mol C.A.A.M. Ottenhoff R. Van der Lugt N.M.T. van Roon M.A. Van der Valk M.A. Offerhaus G.J.A. Berns A.J.M. Borst P. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease.Cell. 1993; 75: 451-462Google Scholar). These mice have no Mdr2 Pgp but do express Mdr1a and Mdr1b Pgp in their canalicular membrane. Van Helvoort et al. (11van Helvoort A. Smith A.J. Sprong H. Fritzsche I. Schinkel A.H. Borst P. Van Meer G. Mdr1 p-glycoprotein is a lipid translocase of broad specificity, while mdr3 p-glycoprotein specifically translocates phosphatidylcholine.Cell. 1996; 87: 507-517Google Scholar) loaded MDR1 and MDR3 transfected cells with fluorescent phospholipid precursors like C6-NBD-diacylglycerol or C6-NBD-ceramide and studied the translocation of their metabolites, C6-NBD-(phosphatidylcholine (PC) and phosphatidylethanolamine (PE)) or C6-NBD-(sphingomyelin (SM) and glucosylceramide (GlucCer)) respectively. In their method fluorescent lipids, translocated to the outer membrane leaflet, were extracted using BSA as acceptor in the medium. The outcome of their studies was that human MDR1 and mouse Mdr1a were able to translocate NBD-PC, NBD-SM, and NBD-GlucCer, while the human MDR3 Pgp-mediated translocation was restricted to NBD-PC. Crawford, Vinter, and Gollan (12Crawford J.M. Vinter D.W. Gollan J.L. Taurocholate induces pericanalicular localization of C6-NBD-ceramide in isolated hepatocyte couplets.Am. J. Physiol. 1991; 260: G119-G132Google Scholar) reported on the biliary secretion of NBD-ceramide and its metabolites NBD-GlucCer and NBD-SM in a perfused rat liver system. Similar to endogenous phospholipid secretion, the output of these short-chain phospholipid analogues was bile salt-dependent. From these observation the question rises whether the translocation of labeled short-chain phospholipid analogues by various Pgps is representative for hepatobiliary lipid secretion of natural lipids. Because mouse Mdr1a Pgp appears to be a translocator of NBD-SM, we tested whether Mdr1a or 1b P-glycoproteins control the hepatobiliary secretion of NBD-SM. We studied the secretion of NBD-ceramide and its metabolites in an in situ mouse liver perfusion system using mice with different P-glycoprotein gene disruptions. Because Mdr1a and 1b have overlapping substrate specificity and are both expressed in the canalicular membrane, we used mice in which both genes are disrupted (Mdr1a/b –/–) (7Schinkel A.H. Mayer U. Wagenaar E. Mol C.A. van Deemter L. Smit J.J. Van der Valk M.A. Voordouw A.C. Spits H. Van Tellingen O. Zijlmans J.M. Fibbe W.E. Borst P. Normal viability and altered pharmacokinetics in mice lacking mdr1-type (drug-transporting) p-glycoproteins.Proc. Natl. Acad. Sci. USA. 1997; 94: 4028-4033Google Scholar). We also used Mdr2 knockout mice (Mdr2 –/–) (1Smit J.J.M. Schinkel A.H. Oude Elferink R.P.J. Groen A.K. Wagenaar E. van Deemter L. Mol C.A.A.M. Ottenhoff R. Van der Lugt N.M.T. van Roon M.A. Van der Valk M.A. Offerhaus G.J.A. Berns A.J.M. Borst P. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease.Cell. 1993; 75: 451-462Google Scholar), with no hepatobiliary PC secretion and control mice with the same genetic background (FVB) as the knockouts. We studied the bile salt-mediated extraction of SM and its NBD-labeled analogue from membranes of different composition, which may resemble the canalicular membrane surface. Egg-yolk phosphatidylcholine, cholesterol, N-2-hydroxyethylpiperazine-N′-2-ethane sulfonic acid (HEPES), and bovine serum albumin (BSA) (essentially fatty acid-free) were obtained from Sigma (St. Louis, MO). Buttermilk sphingomyelin was obtained from Matreya (Pleasant Gap, PA). Krebs-bicarbonate buffer contained 120 mm NaCl, 24 mm NaHCO3, 1.2 mm KH2PO4, 4.8 mm KCl, 1.2 mm MgSO4, and 1.3 mm CaCl2; pH 7.4. Tauroursodeoxycholic acid (TUDC, 99%) was obtained from Calbiochem-Novabiochem (La Jolla, CA). Methanol, HPLC-grade, was obtained from J. T. Baker (Deventer, The Netherlands). Triethylamine, 99%, and O-phosphoric acid, 85%, were obtained from Merck (Darmstadt, Germany). C6-NBD-ceramide (6-[(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]hexanoyl-sphingosine) and C6-NBD-sphingomyelin (6-[(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino-hexanoyl]sphingosylphosphocholine) and C6-NBD-C16-PC (2-(6-[(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]hexanoyl-1-hexadecanoyl-sn-glycero-3-phosphocholine) were obtained from Molecular Probes (Eugene, OR). C6-NBD-glucosyl-ceramide was a kind gift from Dr. G. van Meer, Dept. of Cell Biology and Histology, Academic Medical Center, Amsterdam. C6-NBD-ceramide was bound to bovine serum albumin, adapting the method previously described by Crawford et al. (12Crawford J.M. Vinter D.W. Gollan J.L. Taurocholate induces pericanalicular localization of C6-NBD-ceramide in isolated hepatocyte couplets.Am. J. Physiol. 1991; 260: G119-G132Google Scholar). C6-NBD-ceramide (109 nmol dissolved in 25 μl ethanol) was added to 1.25 ml Krebs-bicarbonate, pH 7.45, containing 10 mm HEPES and 30 mg/ml bovine serum albumin (essentially fatty acid-free). After dialyzing the solution 4 times against this medium at 4°C, aliquots were prepared finally containing 6.6 nmol C6-NBD-ceramide bound to bovine serum albumin in a 1:5 molar ratio. Mouse surgery and perfusion technique and solution preparation were performed as described previously (13Groen A.K. Van Wijland M.J. Frederiks W.M. Smit J.J. Schinkel A.H. Oude Elferink R.P. Regulation of protein secretion into bile: studies in mice with a disrupted mdr2 p-glycoprotein gene.Gastroenterology. 1995; 109: 1997-2006Google Scholar, 14Frijters C.M.G. Tuijn C.J. Hoek F.J. Groen A.K. Elferink R.P.J.O. Zegers B.N. Reversed-phase liquid chromatographic method for the determination of 7-nitrobenz-2-oxa-1,3-diazol-4-yl-labelled lipid analogues.J. Chromatogr. B. 1998; 710: 9-16Google Scholar). Briefly, under anesthesia (hypnorm (fentanyl/fluanisone), 1 ml/kg, Janssen Pharmaceuticals, Beerse, Belgium and diazepam 10 mg/kg, Roche, Mijdrecht, The Netherlands) the vena cava superior, the gallbladder, and portal vein were cannulated. Perfusion was performed in orthograde direction with Krebs-bicarbonate buffer in a perfusion cabinet thermostatted at 37°C. The perfusion medium was gassed with carbogen (5% CO2, 95% O2, Hoek Loos, Schiedam, the Netherlands) in an oxygenator. Directly after start of the perfusion with Krebs-bicarbonate buffer, tauroursodeoxycholic acid (TUDC) was infused through a three-way connector attached to the portal vein cannula, at a steady rate of 500 nmol × min-1 × 100 g body weight-1 during the whole experimental period to maintain constant biliary excretion. After a stabilization period of 20 min, a bolus of C6-NBD-ceramide (6.6 nmol) was introduced into the perfusion medium by infusion through the portal vein cannula. Bile samples were collected at 5-min intervals, diluted ten times with distilled water and immediately frozen at –20°C. C6-NBD-C16-PC was dissolved in eluent (85% 0.45 mm triethylammoniumphosphate in methanol and 15% 0.45 mm triethylammoniumphosphate at pH 3.0) and was used as internal standard. Before injection the samples were diluted four times with internal standard solution that resulted in a final amount of 1.2 pmol per injection. Fluorescent lipids were detected using a recently developed reversed-phase HPLC method. The HPLC system used consisted of a Gynkotek 480 HPLC gradient system (Germering, Germany) connected to a Rheodyne 7125 injection valve (Rheodyne, Cotati, CA) with a 20 μl injection loop and a Jasco FP920 Fluorimeter (Tokyo, Japan); excitation wavelength set to 470 nm and emission wavelength on 530 nm. The LC columns (Inertsil ODS-2, 5 μm, 100 mm × 3 mm ID and Spherisorb CN, 3 μm, 100 mm × 4.6 mm ID) used were obtained from Chrompack (Bergen op Zoom, The Netherlands). Data acquisition was performed on a computer with Gynkosoft integration software. The eluent flow-rate was 0.6 ml/min. Gradient runs were performed using (A) 0.45 mm triethylammoniumphosphate at pH 3.0 and (B) 0.45 mm triethylammoniumphosphate in methanol. The initial eluent composition was 15% A and 85% B; after 0.5 min this was linearly changed to 100% B in 9 min and kept for 6 min at 100%. Finally the system was reset to its initial composition in 5.5 min and stabilized for 8 min before the next analysis was started. In order to study the behavior of C6-NBD-SM and buttermilk SM in small unilamellar vesicles (SUV) exposed to bile salts, three batches of liposomes with different compositions were prepared. The first batch consisted of PC and CHOL (molar ratio 3:2); the second batch contained PC, CHOL, and SM (molar ratio 1:2:2) and the third batch was prepared by mixing SM and CHOL (3:2). The liposomes were prepared by mixing stock solutions of lipids in chloroform–methanol 4:1 in the indicated ratios and the total amount of lipid in each preparation was 81.25 μmol. All three batches were spiked with NBD-SM (135 pmol). After evaporation of the solvent under a stream of nitrogen, the lipid film was hydrated with 5 ml buffer (10 mm HEPES/135 mm NaCl/0.05% NaN3 (pH 7.4)), and sonicated for 30 min in a Branson B12 probe-type sonicator (80 watts) on ice under a stream of nitrogen. The suspensions were filtered through a sterile 0.45 μm and a 0.22 μm Millex-HA or Millex-GP filter (Millipore, Molsheime, France) subsequently, aliquoted and stored under argon at 4°C for not longer than 3 days. Total phospholipid (PC and SM) and cholesterol concentration was determined in the SUV aliquots using the enzymatic methods described in the bile and perfusate analysis section. Just before use, SUV suspensions were diluted to a total lipid concentration of 8.4 mm in 10 mm HEPES/135 mm NaCl/0.05% NaN3 (pH 7.4) containing 10 mm tauroursodeoxycholic acid (TUDC) and incubated for 10 min at 37°C. The vesicles (0.5 ml) were then chromatographed over a Sephacryl S-300 gel permeation column (elution buffer 10 mm HEPES/135 mm NaCl/0.05% NaN3 (pH 7.4) containing 10 mm tauroursodeoxycholic acid) using a flow rate of 0.7 ml/min. Fractions of 2 min were collected for 60 min in which total phospholipid and cholesterol was determined to identify the vesicular fraction and the micellar fraction. Online detection of the fluorescence signal (Jasco FP920 Fluorimeter, Tokyo, Japan) allowed determination of the percentage of NBD-SM fluorescence associated with the vesicular or micellar fraction. The amount of unlabeled SM and PC in the peak fractions comprising the micellar and vesicle fractions (two fractions of 2 min each) was determined by straight-phase HPLC combined with evaporative light scattering detection. Prior to injection on the HPLC column the lipids in the gel permeation column fractions were extracted according to the method described by Bligh and Dyer (15Bligh E.G. Dyer W.J. A rapid method of total lipid extraction and purification.Can. J. Biochem. Physiol. 1959; 37: 911-917Google Scholar). After evaporation of the organic solvents, the lipid film was dissolved in isopropanol–hexane–water 58:40:2. Essentially the same HPLC set-up was used as described in the bile and perfusate analysis section. Detection was performed using a PL-EMD 960 Evaporative Light Scattering Detector (Polymer Laboratories, Heerlen, The Netherlands). The lipids from the gel permeation fractions were separated using a Chromsphere 3 Si column (150 mm × 3.0 mm ID) obtained from Chrompack (Bergen op Zoom, The Netherlands). The eluent flow-rate was 0.4 ml/min. Gradient runs were performed using (A) isopropanol (B) hexane, and (C) water. The initial eluent composition was 58% A, 40% B, and 2% C; after 0.5 min this was linearly changed to 52% A, 40% B, and 8% C in 7 min and kept for 8 min in this composition. Finally, the system was reset to its initial composition in 5.5 min and stabilized for 8 min before the next analysis was started. Results are reported as mean ± SD from at least four animals in each experimental group (controls (n = 7), Mdr2 –/– (n = 4), Mdr1 a/b –/– (n=4)). Statistical significance was determined where appropriate by Student's t-test for unpaired data. The uptake of C6-NBD-ceramide bound to albumin was determined by measuring disappearance of the fluorescence in the perfusate after passage through the liver. Five fractions of 1 min each were collected directly after administration of the single dose of C6-NBD-ceramide followed by 10-min samples. Within the 5-min period directly after administration of C6-NBD-ceramide, a peak in fluorescence was detected; in the later fractions fluorescence was only slightly above background. The rapid peak fraction probably represented the labeled lipid that was not taken up by the liver. HPLC analysis confirmed that in this peak the fluorescent signal was from the C6-NBD-ceramide (data not shown). The low signal in the later fractions was from NBD-Cer with a small amount of NBD-SM. No NBD-GlucCer was detected in the perfusate during the entire experiment. The uptake of C6-NBD-ceramide by the liver turned out to be very efficient. About 10% of the injected dose of 6.6 nmol C6-NBD-ceramide could be recovered in the perfusion medium after passage through the liver. The uptake values did not differ among the different genotypes. The fractional uptake values for controls, Mdr1 a/b (–/–) and Mdr2 (–/–) were 89.7 ± 2.2% (n = 7), 91.9 ± 1.7% (n = 4), and 93.4 ± 1.0% (n = 4) of total administered dose. The first fluorescent lipid excreted into bile after injection was NBD-C6-ceramide. At 10–15 min after administration a sharp increase in biliary concentration was observed which gradually decreased to low steady state levels (Fig. 1A). The bulk of ceramide was excreted within a 50-min period after the administration. In contrast, the metabolites of NBD-C6-ceramide, NBD-C6-glucosylceramide, and NBD-C6-sphingomyelin showed a delayed biliary excretion with kinetics similar to the ceramide (Fig. 1B). The total amount of NBD-labeled lipids secreted during a period of 100 min after administration of the NBD-ceramide was highest in the Mdr1a/b (–/–) mice (5.0 ± 1.7% of the administered dose; n = 4). This was, however, not significantly different from the total biliary secretion in control mice (3.7 ± 1.9%; n = 7). Significantly less total NBD-lipid was secreted in Mdr2 (–/–) mice bile during this period (0.75 ± 0.1%; n = 4). Analysis of bile for the individual fluorescent lipid species revealed that there was no significant difference in biliary secretion between controls and Mdr1a/b (–/–) (Fig. 2). The 100-min cumulative secretion in controls was: 521.7 ± 313.1 (NBD-ceramide); 123 ± 60 (NBD-glucosylceramide), and 42 ± 21 pmol/100 g body weight (NBD-sphingomyelin). In Mdr1 a/b (–/–) mice, the secretion values of these lipid species were 748 ± 314, 176 ± 82, and 41 ± 16 pmol/100 g body weight, respectively. However, the fluorescent lipid secretion values in Mdr2 (–/–) were decreased significantly for all three lipid species. NBD-ceramide values only reached 92 ± 16 pmol/100 g body weight, NBD-glucosylceramide reached 34.0 ± 10 and NBD-sphingomyelin reached 17 ± 3 pmol/100 g body weight.Fig. 2.Biliary secretion of NBD-labeled lipids in mice with different P-glycoprotein genotypes. The values represent the cumulative fluorescence in bile during the 100-min post-administration of NBD-ceramide (error bars indicate standard deviation). Values from Mdr2 (–/–) and Mdr1 a/b (–/–) mice were tested for statistically significant differences with data from control mice by Student's t-test; * P < 0.05.View Large Image Figure ViewerDownload (PPT) The reduced NBD-lipid secretion observed in Mdr2 (–/–) mice was somewhat unexpected because these lipids are not substrates for Mdr2 Pgp and this reduction could be a secondary effect of the absence of Mdr2 Pgp. The outer membrane leaflet of canalicular membranes probably contains phosphatidylcholine, sphingomyelin, and cholesterol. It is expected that, due to the absence of supply by Mdr2 Pgp, the PC content of this membrane leaflet in Mdr2 (–/–) mice is drastically reduced. We hypothesized that such alteration of the lipid composition of the canalicular outer membrane leaflet in Mdr2 –/– mice causes reduced bile salt-mediated NBD-lipid extraction. To test this hypothesis, we compared the behavior of buttermilk SM and NBD-labeled short-chain SM in membranes with different lipid composition exposed to bile salts, in an in vitro extraction assay. Briefly, lipid vesicles were prepared containing PC-CHOL (3:2 molar ratio), PC-SM-CHOL (1:2:2), or SM-CHOL (3:2) and trace amounts of NBD-SM were added to all. The vesicles were preincubated for 10 min at 37°C with tauroursodeoxycholate (10 mm) and then applied to a gel permeation column, which was eluted with a buffer containing 10 mm TUDC. Analysis of the eluted fractions for both NBD-labeled and nonlabeled lipid species revealed the fractional distribution of lipid over the micellar and vesicular fraction. Figure 3 shows the online detection of fluorescent signal and the amount of cholesterol and phospholipid in fractions collected from the gel permeation column after application of the SUVs of different composition in TUDC. The lipids eluted in two peak fractions. The first peak contains the large vesicular structures (vesicular peak) and the second peak contains the smaller micelles (micellar peak). As expected, the amount of lipids ending up in the micellar peak varied with the composition of the vesicles. PL and cholesterol were best extracted from SUVs composed of PC-CHOL; about 15–20% of total amount of lipid that was loaded on the column was recovered in the micellar fraction (Fig. 3B). When SM was also present in the SUVs (PC:SM:CHOL (1:2:2)) the amount of lipids retrieved in the micellar peak was only 4–5% of the total amount loaded on the column (Fig. 3D). In SUV containing exclusively SM and CHOL, no phospholipid (i.e., SM) and cholesterol could be detected in the micellar peak (Fig. 3E). Because the enzymatic assay that was used could not distinguish between PC and SM, we analyzed the main peak fractions with HPLC and evaporative light scattering detection. These data showed that in case of PC-SM-CHOL vesicles, the micellar fraction predominantly contained PC. The ratio PC:SM in the vesicular peak was 3.5 ± 0.6 times lower than in the micellar peak. The data also confirmed the absence of SM in the micellar fraction when the SM-CHOL vesicles were used. The online detection of NBD-SM extracted by bile salt from the different SUVs is depicted in Fig. 3A, C, and E, respectively, and showed a behavior that was markedly different from the natural long-chain buttermilk SM. In contrast to natural SM, the extraction of NBD-SM was equally high in PC-CHOL- and PC-SM-CHOL-containing vesicles. The most striking difference was the identification of NBD-SM-associated fluorescence in the micellar fraction in CHOL-SM vesicles. Although extraction from these vesicles was significantly reduced compared to the SUVs containing PC, still 35% of the NBD-SM could be retrieved in the micelles (Fig. 4). Although the mechanism of biliary phospholipid secretion is not fully understood, two major factors driving this process can be distinguished. First, activity of Mdr2 P-glycoprotein in the canalicular membrane is an absolute requirement for phospholipid secretion. Second, bile salt secretion into the canalicular lumen is necessary to evoke lipid secretion and the efficiency of this extraction increases with hydrophobicity of the bile salt species. These observations have been combined in a hypothetical model describing the mechanism of lipid secretion (reviewed in refs. 3Frijters C.M.G. Groen A.K. Oude Elferink R.P.J. Mdr2 p-glycoprotein-mediated lipid secretion and its relevance to biliary drug transport.Adv. Drug Delivery Rev. 1997; 25: 201-215Google Scholar, 4Oude Elferink R.P.J. Ottenhoff R. Van Wijland M.J.A. Smit J.J.M. Schinkel A.H. Groen A.K. Regulation of biliary lipid secretion by mdr2-P-glycoprotein in the mouse.J. Clin. Invest. 1995; 95: 31-38Google Scholar, 16Oude Elferink R.P.J. Tytgat G.N.J. Groen A.K. The role of mdr2 P-glycoprotein in hepatobiliary lipid transport.FASEB J. 1997; 11: 19-28Google Scholar). In short, PC molecules are delivered to the canalicular membrane leaflet facing the cytoplasm. Mdr2 Pgp (MDR3 in humans) translocates these molecules in an ATP-driven process to the leaflet facing the canalicular lumen where they vesiculate in biliary vesicles. The latter process requires bile salts to be present
Referência(s)