Hepatobiliary transporter expression in percutaneous liver biopsies of patients with cholestatic liver diseases

Hepatobiliary Transporter Expression in Percutaneous Liver Biopsiesof Patients With Cholestatic Liver Diseases

GERNOT ZOLLNER,1 PETER FICKERT,1 RAINER ZENZ,2 ANDREA FUCHSBICHLER,2 CONNY STUMPTNER,2 LUKAS KENNER,2 PETER FERENCI,3

RUDOLF E. STAUBER,1 GUENTER J. KREJS,1 HELMUT DENK,2 KURT ZATLOUKAL,2 AND MICHAEL TRAUNER1

SEE EDITORIAL ON PAGE 758

Reduced hepatobiliary transporter expression could explainimpaired hepatic uptake and excretion of bile salts and otherbiliary constituents resulting in cholestasis and jaundice. Be-cause little is known about alterations of hepatobiliary transportsystems inhumancholestatic liverdiseases, itwas theaimof thisstudy to investigate such potential changes. Hepatic mRNA lev-els in hepatobiliary transport systems for bile salts (NTCP,BSEP), organic anions (OATP2, MRP2, MRP3), organic cations(MDR1), phospholipids (MDR3), and aminophospholipids(FIC1) were determined in 37 human liver biopsies and controllivers by competitive reverse-transcription polymerase chain re-action (RT-PCR). Transporter tissue distribution was investi-gated by immunofluorescence microscopy. In patients with in-flammation-induced icteric cholestasis (mainly cholestaticalcoholic hepatitis), mRNA levels of NTCP, OATP2, and BSEPwere reduced by 41% (P < .001), 49% (P < .005), and 34% (P <.05) compared with controls, respectively. In addition, NTCPand BSEP immunostaining was reduced. MRP2 mRNA levelsremained unchanged, but canalicular immunolabeling for MRP2was also decreased. mRNA expression of MRP3, MDR1, MDR3,and FIC1 remained unchanged. In contrast to the alterations oftransporterexpression in inflammation-induced icteric cholesta-

sis, transporter expression did not change in anicteric cholesta-sis caused by primary biliary cirrhosis (PBC) stages I and II. Inconclusion, reduced expression of hepatobiliary transport sys-tems for bile salts and other organic anions may contribute toinflammation-induced cholestasis in humans. Reduction oftransporter gene expression can occur at the mRNA level asobserved for NTCP, OATP2, and BSEP. However, reducedMRP2 immunostaining in the presence of conserved MRP2mRNA levels suggests an additional role for posttranscriptional/posttranslational mechanisms. (HEPATOLOGY 2001;33:633-646.)

Hepatobiliary transport systems mediate hepatic uptakeand biliary excretion of bile salts, non–bile salt organic anions(e.g., bilirubin and its conjugates), organic cations (e.g.,drugs), and lipids (e.g., cholesterol, phospholipids).1 Recentcloning of these transport systems in humans and rodents hasgreatly enhanced our understanding of the molecular basis ofbile formation (Fig. 1).2-4 Experimental animal studies of cho-lestasis have shown that reduced expression of hepatobiliarytransport systems may explain impaired bile excretory func-tion resulting in cholestasis.3,4 In addition, mutations of trans-porter genes can cause hereditary cholestatic syndromes inhumans, further indicating that alterations of transporter ex-pression and function may play an important role in thepathogenesis of cholestasis.3,4 Despite significant advances inour understanding of the regulation of hepatobiliary trans-porter expression in rodents,3,4 very little is known aboutalterations of these transport systems in acquired cholestaticliver diseases in humans.5-9

Therefore, the aim of the present study was to test the hy-pothesis that reduced expression of hepatobiliary transportsystems may contribute to the pathogenesis of cholestasis andjaundice in adults. Using competitive reverse-transcriptionpolymerase chain reaction (RT-PCR) with homologous inter-nal RNA standards,10 we studied alterations of a panel of 8major hepatocellular transport systems at the mRNA level inroutine percutaneous liver biopsies from cholestatic patients.In addition, transporter tissue distribution was investigatedby immunofluorescence microscopy. This study demon-strates that reduced expression of transport systems for he-patic uptake and excretion of bile salts and other organicanions (e.g., bilirubin) may play an important role in thepathogenesis of acquired cholestatic liver diseases in humans.

MATERIALS AND METHODS

Materials

The following reagents were used: AMV reverse transcriptase, Kle-now polymerase, Sp6, and T7 RNA polymerase and restriction en-zymes (Boehringer Mannheim, Mannheim, Germany); AmpliTaq

Abbreviations: RT-PCR, reverse-transcription polymerase chain reaction; PBC, pri-mary biliary cirrhosis; CHC, chronic hepatitis C; HCV, hepatitis C virus; AST, aspartatetransaminase; ALT, alanine transaminase; GGT, g-glutamyl transpeptidase; ALP, alka-line phosphatase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; NTCP, Na1/taurocholate cotransporter; OATP, organic anion transporting protein; BSEP, bile saltexport pump; MDR1, multidrug export pump; MDR3, phospholipid flippase; MRP2,canalicular conjugate export pump; MRP3, basolateral MRP homologue; PFIC, progres-sive familial intrahepatic cholestasis; FIC1, putative aminophospholipid transporter;LST-1, liver-specific organic anion transporter.

From the 1Division of Gastroenterology and Hepatology, Department of Internal Med-icine and 2Department of Pathology, Karl-Franzens University Graz; and 3Department ofInternal Medicine IV, University of Vienna, Austria.

Received August 16, 2000; accepted December 11, 2000.Presented in part at the 50th annual meeting of the American Association for the Study

of Liver Diseases, November 1999, Dallas, TX, and the 7th European GastroenterologyWeek, November 1999, Rome, Italy; and published in abstract form (HEPATOLOGY 1999;30:467A and Gut 1999;45[Suppl V]:A1-A2).

Supported by grant 7171 from the Jubilee Funds of the Austrian National Bank (toM.T.), the Joseph Skoda Prize from the Austrian Society of Internal Medicine (to M.T.),and grant S7401-MOB from the Austrian Science Foundation (to K.Z.).

Address reprint requests to: Michael Trauner, M.D., Division of Gastroenterology andHepatology, Department of Internal Medicine, Karl-Franzens University, Auenbrugger-platz 15, A-8036 Graz, Austria. E-mail: [email protected]; fax: (43) 316-385-3062.

Copyright © 2001 by the American Association for the Study of Liver Diseases.0270-9139/01/3303-0020$35.00/0doi:10.1053/jhep.2001.22646

633

DNA Polymerase (Perkin-Elmer, Branchburg, NJ); Prime RNase In-hibitor (5 Prime 3 Prime Inc., Boulder, CO); deoxynucleosidetriphosphates (Pharmacia, Uppsala, Sweden); pCR II vector and OneShot competent cells (Invitrogen, Groningen, the Netherlands); andQuantum Prep Plasmid Miniprep and Maxiprep kits (Bio-Rad Labo-ratories, Hercules, CA). All other materials were of the highest puritycommercially available and were purchased from Merck (Darmstadt,Germany).

Tissue Specimens

Liver tissue specimens were obtained during routine diagnosticliver biopsies with a 1.4-mm Menghini or a 17-gauge Tru-cut biopsyneedle. An aliquot (approximately one third) of the biopsy specimenwas immediately snap-frozen and stored in liquid nitrogen until ex-traction of total RNA and immunofluorescence microscopy. The re-maining material was used for routine histology. In addition, normalliver tissue was obtained from patients undergoing surgical resectionof liver metastases. All patients had given their informed consent forthe study, and the experimental protocol was approved by the EthicsCommittee of the Karl-Franzens University in Graz in accordancewith the ethical guidelines of the 1975 Declaration of Helsinki.

Patient Characteristics

Thirty-seven liver specimens derived from 4 patient groups (con-trols without histological evidence of liver disease, inflammation-

induced cholestasis, primary biliary cirrhosis [PBC], chronic hepa-titis C [CHC]) were studied (Table 1). In the first group, control liversamples (total of 13) were from unaffected areas from surgical liverresections of hepatic metastases (n 5 6) and from patients (n 5 7)who underwent liver biopsy for staging of hematologic malignancy,suspected liver fibrosis, or exclusion of Wilson’s disease withouthistologic evidence of liver disease. Importantly, these controls hadnot received any relevant drug therapy (known to influence trans-porter expression) before liver biopsy.

The second group comprised patients with inflammation-inducedacute icteric cholestasis (total of 11 patients), mostly caused by cho-lestatic alcoholic hepatitis (n 5 7). The diagnosis of alcoholic hepa-titis was based on a positive history of alcohol consumption above 60g/d, characteristic laboratory features, and liver histology. In addi-tion, 2 patients had cholestatic autoimmune hepatitis as defined byhigh-titer antinuclear and/or anti–smooth muscle antibodies andcharacteristic liver histology, and 2 patients had drug-induced cho-lestasis (caused by phenylbutazone and the synthetic antiandrogenflutamide).

The third group (8 patients) consisted of patients with chronicanicteric cholestasis caused by PBC (4 stage I, 4 stage II, according toLudwig11). The diagnosis of PBC was based on established criteriaincluding high-titer antimitochondrial (M2) antibodies and compat-ible liver histology. In addition, 3 patients with histologically moreadvanced (but still anicteric) stage IV PBC were studied. However,these patients were statistically analyzed separately from stage I/IIPBC because of potential confounding variables of severe liver fibro-sis and cirrhosis with an increased amount of fibrous tissue at theexpense of hepatocytes.

To determine the specificity of potential changes in transporterexpression in cholestasis, a fourth, noncholestatic group (5 patients)with CHC was studied. These patients were hepatitis C virus (HCV)antibody–positive and HCV-RNA–positive, and had moderately ele-vated transaminase levels and biopsy-proven chronic hepatitis.

The biochemical characteristics of the 4 groups studied are sum-marized in Table 1. Serum samples were analyzed for bilirubin,aspartate transaminase (AST), alanine transaminase (ALT), g-glu-tamyltranspeptidase (GGT), and alkaline phosphatase (ALP) activi-ties by routine clinical chemistry testing on a Hitachi 717 analyzer(Boehringer Mannheim, Mannheim, Germany). Serum bile salt lev-els were determined by a commercially available radioimmunoassayfor conjugated bile salts (ICN Pharmaceuticals, Orangeburg, NY).Both cholestatic groups had significantly elevated GGT and ALPserum levels with an ALT/ALP ratio , 2, while only patients withinflammatory cholestasis had marked (predominantly conjugated)hyperbilirubinemia (mean, 11 mg/dL) and significantly elevated se-rum bile salt levels (mean, 42.5 mmol/L) (Table 1). In line with thesebiochemical changes, liver histology showed typical features of in-trahepatic cholestasis (e.g., bilirubin stasis in bile canaliculi, hepato-cytes, and Kupffer cells in inflammation-induced cholestasis; feath-ery degeneration of hepatocytes in PBC). Extrahepatic biliaryobstruction was excluded by normal appearance of bile ducts onultrasound and by absence of histologic features suggestive of extra-hepatic obstruction.

FIG. 1. Hepatobiliary transport systems investigated in this study. Thisschematic representation shows 2 adjacent hepatocytes comprising a bilecanaliculus. A Na1/taurocholate cotransporter (NTCP) at the basolateralmembrane of hepatocytes (facing the sinusoidal blood) mediates hepatic bilesalt (BS2) uptake, while a family of Na1-independent organic anion trans-porting polypeptides (OATPs) also transports (non–bile salt) organic anions(OA2) and cations (OC1). OATP2 (also known as the liver-specific trans-porter, LST-1 or OATP-C) is currently the most prominent member of thisfamily. Bile salts are excreted into bile via a bile salt export pump (BSEP),while reduced glutathione (GSH) and (non–bile salt) organic anions (e.g.,conjugated bilirubin) are excreted via a canalicular conjugate export pump(MRP2). Another isoform of the multidrug resistance-related protein is lo-cated at the basolateral membrane (MRP3), where it may serve as a compen-satory overflow system under cholestatic conditions. In addition, the canalic-ular membrane contains a multidrug export pump (MDR1) and aphospholipid flippase (MDR3). A P-type adenosine triphosphatase (FIC1) isa putative aminophospholipid (PE, phosphatidylethanolamine; PS, phospha-tidylserine) transporter (from the outer to the inner leaflet of the cell mem-brane), but may also play a direct or indirect role in canalicular bile saltexcretion, because mutations of this transporter result in subtypes of hered-itary cholestasis.

TABLE 1. Biochemical Characteristics of Patient Groups

GroupTotal Number

(males/females)Age(yr) AST (U/L) ALT (U/L) GGT (U/L) ALP (U/L)

Bilirubin(mg/dL)

Bile Salts(mmol/L)

Controls 13 (7/6) 41 6 5 20 6 5 35 6 11 40 6 7 100 6 10 0.7 6 0.1 0.7 6 0.1Inflammatory cholestasis 11 (5/6) 44 6 3 79 6 9† 53 6 10 454 6 74† 314 6 52† 11 6 2.2† 42.5 6 6.4†PBC 8 (1/7) 57 6 6 27 6 4 37 6 6 261 6 78† 485 6 110† 0.7 6 0.2 4.8 6 2.7CHC 5 (4/1) 52 6 5 42 6 14 50 6 7 81 6 17* 112 6 20 0.7 6 0.1 1.6 6 0.6

NOTE. Values are means 6 SEM.* P , .05 vs. controls.† P , .005 vs. controls.

634 ZOLLNER ET AL. HEPATOLOGY March 2001

RNA Extraction

Isolation of total RNA was performed according to a proceduredescribed by Krieg et al.12 RNA was quantified spectrophotometri-cally at 260 nm, and the quality of total RNA was controlled bydenaturating formaldehyde agarose gel electrophoresis.

Determination of mRNA Copy Numbers

Steady-state mRNA levels in hepatobiliary transport systems (Fig.2) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) wereinvestigated by competitive RT-PCR using internal RNA standards

with mutated restriction sites as described by Eferl et al.13 This RT-PCR technique has been successfully applied to simultaneouslyquantify mRNA expression of several genes in minute amounts ofliver tissue (e.g., fetal mouse liver).13

Construction of Internal RNA Standards. To control for potential vari-ations in reverse transcription and PCR amplification efficiencies,internal homologous RNA standards with mutated restriction sites10

were generated for the following hepatobiliary transport systems:Na1/taurocholate cotransporter (NTCP),14 organic anion transport-ing polypeptide 2 (OATP2; also known as OATP-C or LST-1; symbolSLC21A6),15 bile salt export pump (BSEP),16 multidrug exportpump (MDR1),17 phospholipid flippase (MDR3),18 conjugate exportpump (MRP2),19 basolateral MRP homologue (MRP3),20 the puta-tive aminphospholipid transporter, FIC1,21 and GAPDH.22 Primersfor cDNA synthesis and PCR are listed in Table 2. The sequence ofthe homologous internal RNA standards differed only by 2 to 8 basepairs from the wild-type sequence, resulting in a mutated restrictionsite. Standards were constructed either by removing an existing re-striction site by filling in with Klenow polymerase (MDR3, MRP2,MRP3, GAPDH),23 or by introducing a new restriction site by over-lap-extension24 or megaprimer PCR25 using mutagenic primers (Ta-ble 2). Mutated PCR products were cloned into pCR II vectors, andtheir successful mutagenesis was checked by restriction digest andsequencing with an Abi Prism automatic sequencer (Perkin Elmer,Branchburg, NJ). To obtain RNA standards, the cloned DNA stan-dards were linearized and transcribed into sense RNA in vitro usingSp6 or T7 RNA polymerase following a standard protocol.26 Quan-tification of in vitro transcripts was performed using video densitom-etry with an Intas Digit-Store system (Goettingen, Germany) andRFLP-Scan software (Scanalytics, Billerica, MA). Copy numbers (i.e.,numbers of RNA molecules) were calculated by comparing bandintensities of each transcript to MS2 RNA (Invitrogen, Groningen,the Netherlands) band intensity.

Reverse Transcription and Competitive PCR. Reverse transcription ofwild-type and standard RNA was performed simultaneously withinone tube for all transporters and GAPDH following the protocol ofthe cDNA Cycle Kit for RT-PCR (Invitrogen) using 300 ng total RNA,specific downstream primers, and defined copy numbers of standardRNAs for 8 hepatobiliary transporters and GAPDH. PCR was per-formed using one tenth of the cDNA products in a 50-ml reactioncontaining 13 PCR buffer (60 mmol/L Tris HCl, 15 mmol/L[NH4]2SO4, 1.5 mmol/L MgCl2 [pH 8.5]), 0.25 mmol/L of eachdNTP, 1 mmol/L of each primer, and 5 mL dimethylsulfoxide. Thevials were heated to 94°C, then AmpliTaq DNA polymerase wasadded, and subsequently cycled in an OmniGene system (Hybaid,Middle-Essex, UK) for 45 cycles at 94°C for 1 minute, at 60°C for 1minute, and at 72°C for 1 minute. The program ended with a finalextension step at 72°C for 4 minutes. Control experiments wereconducted to exclude contamination by genomic DNA or PCR prod-ucts. PCR products were separated on ethidium bromide–stained 1%agarose gels. The specificity of the developed PCR technique wasconfirmed by the product of predicted size and by sequence analysisand comparison with published sequences. To distinguish standardand wild-type amplicons, equal amounts of PCR products were di-gested with the corresponding restriction enzyme (Table 2) for atleast 4 hours, and restriction fragments were run on ethidium bro-mide–stained 2% agarose gels.

Determination of Steady-State mRNA Levels. After video densitome-try, the mRNA copy number was calculated by comparison of stan-dard and wild-type band intensities. In addition, corrections weremade for relative band sizes and heterodimer formation as de-scribed.13 Because reverse transcription and amplification by PCRwere performed simultaneously in identical tubes (Fig. 2A), anypotential variations in cDNA synthesis and PCR reaction equallyaffected standard and wild-type amplicons.27 Cross-validation of thisRT-PCR technique with quantitative Northern blotting has demon-strated that the accuracy of both methods is similar.13 As shown inFig. 2B, the relationship between the amount of input test RNA andmeasured copy numbers was linear over a range of 0.05 to 1 mg,

FIG. 2. Competitive RT-PCR with mutated RNA standards. (A) Principleof competitive RT-PCR assay. (1) 300 ng of total RNA (containing the wild-type mRNAs), defined copy numbers of RNA standards for each of the 8transporters and GAPDH, and specific downstream primers were used forcDNA synthesis. Mutated RNA standards were constructed by introducing anew restriction site (X), which permits the later distinction of standard andwild-type derived PCR-products. Alternatively, for some transporter stan-dards, an already-existing restriction site was removed (not shown). Standardand wild-type RNAs were simultaneously reverse-cotranscribed into cDNAsin the same reaction tube, thus allowing corrections for differences of cDNAsynthesis efficiency. (2) Standard and wild-type cDNAs were coamplifiedtogether in the same vial by PCR to control for any variations in the amplifi-cation process. (3) Restriction digests were performed to distinguish standard(digestible)- and wild-type (nondigestible)-derived PCR products. (4) Re-sulting fragments were separated by agarose gel electrophoresis. Gels weredigitalized and band intensities were measured using RFLP-Scan software.After correction for relative band size and heterodimer formation, copy num-bers of wild-type RNA were calculated from the ratio of wild-type to standardband intensities. (B) Linearity of RT-PCR method. To determine the relation-ship between input test RNA and the PCR product, titration curves wereperformed for each transporter using increasing amounts of wild-type RNA(0.05 to 1 mg) and constant amounts of standard RNA. NTCP titration curveshown here as an example. Note the linear relationship between the amountof input test RNA and measured NTCP copy numbers (means of 3 indepen-dent measurements 6 SD).

HEPATOLOGY Vol. 33, No. 3, 2001 ZOLLNER ET AL. 635

confirming the high sensitivity of this RT-PCR method.27 Because ofsample-to-sample variations in RNA quality, transporter copy num-bers were expressed as a ratio relative to GAPDH mRNA levels,7

which allowed corrections for differences in tissue handling duringliver biopsies, storage time, and variations in the efficiency of RNAextraction and RNA quality as described previously.5,7,9

Immunofluorescence Staining and ConfocalLaser Scanning Microscopy

Double-labeling immunofluorescence microscopy was performedon cryosections of liver biopsies (4 mm thick) fixed in aceton andchloroform for 10 minutes each at room temperature using the poly-clonal antibodies against MRP2 (in a dilution of 1:100 kindly sup-plied by Dr. D. Keppler, Heidelberg, Germany), BSEP (1:50; kindlysupplied by Drs. M. Muller and P. Jansen, Groningen, the Nether-lands), NTCP (1:25; kindly supplied by Drs. B. Stieger and P. Meier,Zurich, Switzerland), and the monoclonal antibody, C219 (1:30;Signet, Dedham, MA). The monoclonal antibodies, Ks 8.7, Ks 18.04(1:30; Progen, Heidelberg, Germany), and the polyclonal antibody,50K160 (1:50), were used to visualize the cytokeratin 8/18 net-work.28 Fluoresceine isothiocyanate–conjugated goat anti-mouseIgG (Zymed, San Francisco, CA) and tetramethylrhodamine isothio-cyanate–conjugated swine anti-rabbit Ig (Dako, Glostrup, Denmark)were used as secondary antibodies. Negative controls were per-formed by omitting the primary antibodies. Fluorescent staining wasvisualized using a MRC 600 (BioRad, Richmond, CA) laser-scanningconfocal device attached to a Zeiss Axiophot. The fluorescent imageswere collected using the confocal photomultiplier tube using stan-dardized conditions. For dual labeling, separate excitation wave-lengths (488 nm for fluoresceine isothiocyanate; 568 nm for tetra-methylrhodamine isothiocyanate) from a krypton/argon ion laserwere used. Separate filter tubes allowed acquisition and storage of theimages of the identical optical focal plane within the cell. Overlap-ping of green (fluoresceine isothiocyanate) and red (tetramethylrho-damine isothiocyanate) dyes resulted in a yellow signal. Only sam-ples that were prepared in parallel in all steps were compared.

Statistical Analysis

cDNA synthesis and PCR were run in triplicate. Individual trans-porter mRNA expression levels are given relative to GAPDH mRNA.All mean values are expressed 6 SD unless stated otherwise. Differ-

ences among patient groups were analyzed by the Mann-Whitneyrank sum test (2-tailed), or multivariate ANOVA with posttesting,when appropriate, using the SigmaStat statistic program (Jandel Sci-entific, San Rafael, CA). Regression analysis was performed using theInstat statistic program (Graph Pad Software, Palo Alto, CA). P , .05was considered significant.

RESULTS

Hepatobiliary Transporter mRNA Expression in Control Livers.To analyze mRNA expression of a panel of hepatobiliarytransport systems (Fig. 1) in routine liver biopsies, competi-tive RT-PCR with homologous internal RNA standards wasestablished. This technique permitted the simultaneous deter-mination of mRNA copy numbers of 8 major hepatobiliarytransport systems using only 300 ng of total RNA. The level oftransporter mRNA expression (copy number per 300 ng totalRNA given as mean 6 SD) in normal control livers decreasedin the following order: OATP2 (1.7 3 106 6 2.5 3 105) .MRP2 (4.9 3 105 6 7.4 3 104) . BSEP (1.7 3 105 6 2.4 3104) . NTCP (1.6 3 105 6 1.8 3 104) . MDR3 (1.2 3 105 61.3 3 104) . MDR1 (1.1 3 105 6 1.3 3 104) . FIC1 (6.3 3104 6 9.9 3 103) . MRP3 (3.4 3 104 6 5.6 3 103). OATP1(mostly expressed in the brain29) was only detected at verylow expression levels in control livers (not shown) and wasexcluded from further analysis because quantification wasonly possible using 1 mg of total RNA.8 Interestingly, trans-port systems involved in hepatic uptake and biliary excretionof organic anions (OATP2 and MRP2) showed the highestlevel of mRNA expression, followed by “pure” bile salt trans-porters (BSEP and NTCP). No sex differences in expressionlevels (data not shown) were observed for any of the transportsystems investigated.

Hepatobiliary Transporter mRNA Expression in Cholestatic LiverDiseases. To study potential alterations of transporter mRNAexpression in adult cholestatic liver diseases, routine liverbiopsies from patients with inflammation-induced ictericcholestasis (as a result of cholestatic alcoholic hepatitis, cho-lestatic autoimmune hepatitis, drug-induced cholestasis) and

TABLE 2. Primers for Competetive RT-PCR and Generation of Internal RNA Standards With Mutated Restriction Site

Gene Name(gene symbol)

UpperPrimer

LowerPrimer Restriction Site Mutagenic Primer Reference

NTCP(SLC10A1)

95-115 688-667 Sac I (459) ins 445-468; 59-GATGACCACGAGCTuCCACCTTCTG-39 14

OATP2(SLC21A6)

90-109 683-664 Hind III (333) ins 329-338; 59-TGAAuAGCTTTGCCACATTTC-39 15

BSEP(ABCB11)

3267-3296 4021-4002 BamH I (3576) ins 3568-3587; 59-AATGTCCGuGATCCTCCGCTC-39 16

MDR1(ABCB1)

834-860 1376-1350 Xho I (1071) ins 1063-1086; 59-GCTAACCCuTCGAGATTTTGGCCAT-391086-1063; 59-ATGGCCAAAATCuTCGAGGGTTAGC-39

17

MDR3(ABCB4)

81-97 570-549 Sal I (460) re * 18

MRP2(ABCC2)

2793-2814 3478-3455 Acc I (3115) re * 19

MRP3(ABCC3)

3260-3304 3940-3917 Sma I (3705) re * 20

FIC1(ATP8B1)

2499-2522 3010-2989 BamH I (2921) ins 2931-2911; 59-CTGCGCAGAGuGATCCATTGAA-39 21

GAPDH 401-424 857-834 Ksp I (456) re * 22

NOTE. Base-pair numbering corresponds to published cDNA or mRNA sequences. Restriction enzymes and positions of restriction sites are given; “ins” fornewly inserted restriction site and “re” for removed restriction site. Mutagenic primers with exchanged nucleotides are written in bold. Bar indicates restrictionsite. Note that only 1 mutagenic primer is needed for megaprimer PCR.

* Standards for MDR3, MRP2, MRP3, and GAPDH were mutated by Klenow polymerase without the use of mutagenic primers.


anicteric cholestasis caused by stage I and II PBC were inves-tigated. In addition, transporter expression was investigatedin liver biopsies from patients with CHC to control for thespecificity of potential changes in transporter expression incholestatic patients. Compared with GAPDH mRNA expres-sion in controls (2.02 3 106 6 1.16 3 106 copies per 300 ngtotal RNA), only minor, statistically insignificant increaseswere observed in all disease groups to a similar extent, and nodifferences in GAPDH expression were observed among thegroups with inflammatory cholestasis (2.7 3 106 6 1.27 3106 copies), PBC (2.81 3 106 6 1.32 3 106 copies), and CHC(2.61 3 106 6 1.81 3 106 copies). NTCP mRNA levels weresignificantly reduced by 41% 6 21% in patients with inflam-matory cholestasis compared with control subjects (P , .001)(Fig. 3A). In addition, NTCP mRNA levels were reduced to asimilar degree in 2 patients (on standard immunosuppressionwith cyclosporine, prednisolone, and azathioprine) with cho-lestasis caused by biliary stricture after liver transplantation(data not shown). No changes were seen in patients with PBC(stage I/II) and CHC (Fig. 3A). NTCP mRNA levels also re-mained unchanged in 3 patients with stage IV PBC (data notshown) who were statistically analyzed separately from stageI/II PBC because of potential confounding variables of cirrho-sis. Similar to previous studies in human liver,5,7,9 a widescatter of mRNA expression levels was observed in controland cholestatic livers. However, individual mRNA expressionlevels were highly reproducible when measured in triplicate,indicating that this wide distribution reflects interindividualvariations in transporter expression. Although an overlap ofNTCP mRNA levels among patient groups was observed, asignificant inverse correlation was found between hepaticNTCP mRNA and serum bile salt levels of the patients (r 52.77, r2 5 .54, P , .001), suggesting a causal relationshipbetween retention of bile salts and down-regulation of NTCPexpression (Fig. 3B). However, no further correlation couldbe observed between NTCP mRNA levels and either serumlevels of bilirubin, GGT, ALP, AST, and ALT, or any otherbiochemical parameter.

mRNA levels of basolateral OATP2 (SCL21A6) were alsosignificantly reduced by 49% 6 26% in patients with inflam-matory cholestasis compared with controls (P , .005) (Fig.4A); again OATP2 mRNA levels were reduced to a similardegree in 2 patients with cholestasis caused by biliary strictureafter liver transplantation (data not shown). No changes wereseen in patients with PBC (stage I/II) and CHC (Fig. 4A).OATP2 mRNA levels also remained unchanged in 3 patientswith stage IV PBC (data not shown). As with NTCP, a signif-icant inverse correlation was found between hepatic OATP2mRNA levels and serum bile salt levels (r 5 2.76, r2 5 .58,P , .0001) (Fig. 4B). In contrast to NTCP, OATP2 mRNAexpression also inversely correlated with serum levels of totalbilirubin (r 5 2.58, r2 5 .34, P , .0001) (data not shown),while no correlation could be observed with GGT, ALP, ortransaminases.

Compared with control subjects, canalicular BSEP mRNAlevels were reduced by 34% 6 21% in patients with inflam-matory cholestasis (P , .05), while no significant changeswere observed in patients with stage I/II PBC and CHC (Fig.5A). BSEP mRNA levels also did not change significantly in 3patients with stage IV PBC (data not shown). A considerableoverlap of BSEP mRNA expression was found between con-trols and patients with icteric cholestasis. In contrast to NTCP

and OATP2, no significant correlation between BSEP mRNAand serum bile salt levels (r 5 2.19, r2 5 .04, P 5 .314) or anyother biochemical parameter was observed. However, patientswith a more profound reduction of NTCP and OATP2 mRNAlevels were also at the lower end of the BSEP mRNA expres-sion range, suggesting a coordinated reduction of these trans-port systems.

Interestingly, no changes of MRP2 (canalicular bilirubinconjugate export pump) mRNA levels were observed in pa-

FIG. 3. NTCP mRNA expression in liver diseases. After isolation of totalRNA from percutaneous routine liver biopsies, cDNA synthesis and compet-itive RT-PCR were performed as described in Materials and Methods. NTCPmRNA copy numbers were normalized to those of GAPDH, to correct fordifferences in efficiency of RNA extraction and RNA quality. Each data pointrepresents the mean of 3 independent competitive RT-PCRs per patient. (A)Compared with controls (n 5 13), NTCP mRNA levels were significantlyreduced by 41% in patients with inflammatory cholestasis (IC) (n 5 11) (P ,.001), while no changes were seen in patients with PBC (n 5 8) and CHC(n 5 5). Columns represent mean values 1 SD. (B) Inverse correlation be-tween NTCP mRNA expression and serum bile salt levels. Individual NTCPmRNA expression (from patients in whom serum bile salt levels were avail-able) was plotted against serum bile salts levels (logarithmic scale). A signif-icant inverse correlation was found between hepatic NTCP mRNA levels andserum bile salt levels of the patients (r 5 2.77, r2 5 .54, P , .001), suggestinga causal relationship between retention of bile salts and down-regulation ofNTCP expression.


tients with icteric inflammatory cholestasis (Fig. 5B), despitethe presence of pronounced hyperbilirubinemia and jaundicein these patients (mean serum level of predominantly conju-gated bilirubin, 11 mg/dL). As expected, no significantchanges were seen in anicteric stage I/II PBC and CHC (Fig.5B). MRP2 mRNA levels also did not change significantly in 3(anicteric) patients with stage IV PBC (data not shown).mRNA levels of MDR1, MDR3, MRP3, and FIC1 also re-mained preserved in all groups investigated (Table 3). How-

ever, FIC1 mRNA levels were reduced by 50% in a patientwith benign recurrent intrahepatic cholestasis (data notshown) caused by mutations of this transport system.21

Immunofluorescence Microscopy for Hepatobiliary Transport Sys-tems in Control and Cholestatic Livers. Given the limited amountof routine liver biopsy material, it was not possible to studytransporter protein expression by Western analysis. There-fore, tissue immunofluorescence labeling was performed(whenever frozen material was available after RNA extrac-tion) on liver biopsies from 19 patients (5 controls, 7 chole-static alcoholic hepatitis, 2 cholestatic autoimmune hepatitis,1 drug [flutamide]-induced cholestasis, 3 stage I/II PBC, and 1cholestasis caused by biliary stricture after liver transplanta-tion) to assess tissue distribution of hepatobiliary transport

FIG. 5. BSEP and MRP2 mRNA expression in liver diseases. After isola-tion of total RNA from percutaneous routine liver biopsies, cDNA synthesisand competitive RT-PCR were performed as described in Materials and Meth-ods. Transporter mRNA copy numbers were normalized to those of GAPDH,to correct for differences in efficiency of RNA extraction and RNA quality.Columns represent mean values 1 SD. Each data point represents the mean of3 independent competitive RT-PCRs per patient. (A) BSEP mRNA levels werereduced by 34% in patients with inflammatory cholestasis (P , .05), while nochanges were seen in PBC (n 5 8) and CHC (n 5 5). (B) No significantdifferences were observed for MRP2 among patient groups studied, even inpatients with inflammatory icteric cholestasis with markedly elevated serumlevels of bilirubin (mean, 11 mg/dL). Controls (n 5 13), inflammatory cho-lestasis (IC) (n 5 11), PBC (n 5 8), CHC (n 5 5).

FIG. 4. OATP2 mRNA expression in liver diseases. After isolation of totalRNA from percutaneous routine liver biopsies, cDNA synthesis and compet-itive RT-PCR were performed as described in Materials and Methods. OATP2mRNA copy numbers were normalized to those of GAPDH, to correct fordifferences in efficiency of RNA extraction and RNA quality. Each data pointrepresents the mean of 3 independent competitive RT-PCRs per patient. (A)Compared with controls (n 5 13), OATP2 mRNA levels were significantlyreduced by 49% in patients with inflammatory cholestasis (IC) (n 5 11) (P ,.005), while no changes were seen in patients with PBC (n 5 8) and CHC(n 5 5). Columns represent mean values 1 SD. (B) Inverse correlation be-tween OATP2 mRNA expression and serum bile salt levels. Individual OATP2mRNA expression (from patients in whom serum bile salt levels were avail-able) was corrected for GAPDH and was plotted against serum bile salt levels(logarithmic scale). A significant inverse correlation was found between he-patic OATP2 mRNA levels and serum bile salt levels of the patients (r 5 2.76,r2 5 .58, P , .0001), suggesting a causal relationship between retention ofbile salts and down-regulation of OATP2 expression.


systems. In control livers, a distinct basolateral staining pat-tern for NTCP (Fig. 6A) and a canalicular staining pattern forBSEP (Fig. 7A and 7B) and MRP2 (Fig. 8A and 8B) wereobserved. Optimal fixation using acetone and chloroform for10 minutes each at room temperature was crucial to obtain asharp canalicular signal for BSEP and MRP2. A distinct cana-licular staining pattern was also obtained with the C219 anti-body (Fig. 9A and 9B) recognizing both MDR1 and MDR3 inaddition to BSEP.30 No lobular gradients were observed innormal liver for any of the transport systems investigated.

NTCP immunostaining was markedly reduced in liversfrom patients with inflammatory cholestasis caused by chole-static alcoholic hepatitis (Fig. 6B), cholestatic autoimmunehepatitis (Fig. 6C), or drug (flutamide)-induced cholestasis(Fig. 6D), and with obstructive cholestasis (not shown), sug-gesting that the reduction in NTCP mRNA levels (see above)also affects NTCP protein expression. In contrast, no reduc-tion in NTCP staining was seen in tissue from patients withstage I/II PBC (Fig. 6E) and CHC (not shown), again consis-tent with the lack of NTCP mRNA changes in these livers (seeabove).

A marked disruption of the normal canalicular BSEP stain-ing pattern with a reduction of the number of BSEP-positivecanaliculi was observed in patients with inflammatory cho-lestasis (Fig. 7C-7H) and obstructive cholestasis (not shown),but not in patients with stage I/II PBC (Fig. 7I and 7J) andCHC (not shown). Although an overall reduction of BSEP-positive canaliculi was observed in patients with inflamma-tory cholestasis, BSEP staining appeared to be preserved oreven enhanced in canaliculi that remained BSEP-positive (Fig.7F and 7H). Canalicular MRP2 immunolabeling was also re-duced in patients with inflammatory cholestasis (Fig. 8C-8H)and obstructive cholestasis (not shown), whereas no changeswere seen in stage I/II PBC (Fig. 8I and 8J) and CHC (notshown). In contrast to the reduction of BSEP and MRP2, can-alicular staining with the C219 antibody (recognizing con-served epitopes of MDR1, MDR3, and BSEP30) was preservedor even enhanced in patients with inflammatory cholestasis(Fig. 9C-9H) and biliary obstruction (not shown), indicatingdifferential changes in canalicular transport systems in cho-lestasis. Again, no changes were seen in stage I/II PBC (Fig. 9Iand 9J) and CHC (not shown). Importantly, tissue distribu-tion of all transport systems investigated in the present studyremained unchanged in stage I/II PBC livers, even in peripor-tal areas with cholate-stasis.

DISCUSSION

Experimental studies in rats have shown reduced expres-sion of basolateral uptake systems (Ntcp, Oatp1) and canalic-ular export pumps (Bsep, Mrp2), which may contribute toimpaired hepatic uptake and excretion of bile salts and otherorganic anions (e.g., bilirubin) in cholestasis.3,4 On the otherhand, maintained or even increased expression of other he-patocellular transport systems (e.g., Mdr1, Mrp3) may repre-sent a protective mechanism in an attempt to limit hepatocel-lular accumulation of bile salts and other potentially toxicbiliary constituents during experimental cholestasis.3,4 Muta-tions of transporter genes in humans can result in defectivetransporter expression and function as exemplified by re-duced or even absent expression of MRP2 in Dubin-Johnsonsyndrome,31 FIC1 in progressive familial intrahepatic cho-lestasis (PFIC)-1,21 BSEP in PFIC-2,32 and MDR3 in PFIC-3.33,34 These findings suggest that altered expression of hepa-tobiliary transport systems could also play an important rolein the pathogenesis of acquired cholestatic liver diseases.However, systematic studies of transporter expression havenot been performed as yet in adult cholestatic liver diseases,because such studies have been limited mainly by the smallamounts of liver tissue available from percutaneous liver bi-opsies, restricting the use of Northern and Western blot anal-ysis.5-9 Using a competitive RT-PCR assay and immunofluo-rescence microscopy for an entire transporter panel, thepresent study demonstrates that reduced expression of baso-lateral (NTCP, OATP2) and canalicular transport systems(BSEP, MRP2) may be relevant for the pathogenesis of ac-quired cholestasis in adults (Fig. 10). While reduction of thesetransporters may contribute to cholestasis, others (MDR1,MDR3, MRP3, and FIC1) remained preserved (Fig. 10), sug-gesting a differential regulation of hepatobiliary transport sys-tems in human cholestasis.

All cholestatic patients included in this study had clear-cutcholestatic liver injury as defined by an increase of serum ALPover 2-fold the upper limit of normal and an ALT/ALP ratio of,2.36 Patients with cholestasis all had intrahepatic causessuch as PBC or inflammation-induced cholestasis (cholestaticvariants of alcoholic, autoimmune, and drug-induced hepati-tis), reflecting the clinical indications for percutaneous liverbiopsies at our institutions. Inflammation-induced cholestasisrepresents clinically relevant and typical examples of cho-lestasis and may be mediated by proinflammatory cytokines,thus pathophysiologically resembling the well-established an-imal model of endotoxin-induced cholestasis.37,38 Therefore,the results obtained in patients with inflammation-inducedcholestasis allow conclusions to be drawn concerning thepathogenesis of acquired cholestasis in humans.

Reduced NTCP mRNA levels and immunostaining in pa-tients with inflammation-induced icteric cholestasis could re-sult from inhibitory effects of proinflammatory cytokines onNTCP promoter activity, similar to inhibition of the rat Ntcppromoter by tumor necrosis factor a and interleukin-1b.42,43

mRNA levels of the recently cloned OATP2 (SCL21A6) werealso significantly reduced, consistent with recent findings incommon bile duct–ligated rats.44 Hepatic NTCP and OATP2mRNA levels correlated inversely with serum bile salt levels,suggesting a causal relationship between retention of bile saltsand down-regulation of these basolateral bile salt–uptake sys-tems.45 Interestingly, bile salts have recently been identified asnatural ligands for the farsenoid X receptor regulating tran-

TABLE 3. mRNA Expression of MDR1, MDR3, MRP3, and FIC1in Liver Diseases

Group MDR1 MDR3 MRP3 FIC1

Controls 100 6 7% 100 6 8% 100 6 13% 100 6 12%Inflammatory

cholestasis 114 6 10% 89 6 13% 93 6 18% 94 6 14%PBC 98 6 20% 141 6 26% 95 6 13% 92 6 31%CHC 144 6 19% 93 6 16% 134 6 25% 63 6 20%

NOTE. Transporter mRNA levels of MDR1, MDR3, MRP3, and FIC1 werenormalized to those of GAPDH, to correct for differences in efficiency of RNAextraction and RNA quality. Values are expressed as percentage of controls(means 6 SEM). No significant changes were observed for these transportersamong patient groups studied.


scription of genes involved in synthesis and transport of bilesalts.46,47 Further studies will have to demonstrate whetherdown-regulation of NTCP and OATP2 gene expression couldalso be mediated by bile salts via interaction with the farsenoidX receptor.47,48 Teleologically, reduced expression of basolat-eral bile salt–uptake systems would be predicted to protectthe hepatocyte by limiting the hepatocellular accumulation oftoxic bile salts during cholestasis.39,45 Down-regulation ofOATP2 could also interfere with hepatic organic anion anddrug clearance in cholestatic liver diseases.

In addition to reduced expression of basolateral NTCP andOATP2, immunolabeling of canalicular BSEP and MRP2, aswell, were reduced in patients with inflammation-inducedcholestasis, suggesting a coordinated reduction of these trans-port systems. The moderate reduction of BSEP mRNA expres-sion (by 34%) in patients with inflammation-induced cho-lestasis is consistent with reduced BSEP mRNA levels (by 32%to 49%) in prolonged experimental cholestasis in rats.49 How-ever, in contrast to NTCP and OATP2, no correlation betweenBSEP mRNA and serum bile salt levels was observed, and amarked overlap of BSEP mRNA expression was found be-tween control subjects and cholestatic patients. These find-ings suggest that posttranscriptional/posttranslational mech-anisms could play an additional role in the reduction of BSEPexpression in cholestasis. Interestingly, mRNA levels of thecanalicular (bilirubin) conjugate export pump MRP2 re-mained unchanged, despite pronounced (predominantly con-jugated) hyperbilirubinemia (mean serum bilirubin level, 11mg/dL). However, at the protein level, canalicular MRP2staining was markedly reduced, suggesting that posttranscrip-tional/posttranslational mechanisms may be involved in thereduction of MRP2. Unchanged MRP2 mRNA expression ininflammation-induced cholestasis was unexpected, becausethis condition is usually associated with systemic as well as intra-hepatic activation of proinflammatory cytokines,37,38 which causepronounced down-regulation of Mrp2 in rats.30,50,51 Our finding isfurther supported by a recent study demonstrating no changesin MRP2 mRNA levels in human liver slices exposed to endo-toxin,52 indicating clear species differences between humansand rats. In addition, a preliminary study in humans has alsodemonstrated reduced MRP2 immunostaining without majorchanges in mRNA expression in livers from patients with ob-structive cholestasis,53 further suggesting that MRP2 may beregulated mostly at the posttranscriptional level.

While expression of these transport systems was reduced ininflammatory cholestasis, mRNA levels of MDR1 (multidrugexport pump) and MDR3 (phospholipid flippase) were main-tained and canalicular staining with the C219 antibody recog-nizing all P-glycoproteins including MDR1 and 330 was pre-served or even enhanced. Maintainance or up-regulation of

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™FIG. 6. NTCP immunofluorescence staining in human liver biopsies. In-

direct immunofluorescence staining was performed with a specific primaryantibody against NTCP on aceton and chloroform-fixed frozen sections frompercutaneous routine liver biopsies prepared in parallel as described in Ma-terials and Methods. The fluorescent images were collected under standard-ized conditions by confocal laser scanning microscopy. Normal basolateralstaining pattern of NTCP in control liver (A). Reduced immunolabeling forNTCP in inflammatory cholestasis caused by cholestatic alcoholic hepatitis(B), cholestatic autoimmune hepatitis (C), and drug (flutamide)-inducedcholestasis (D). Normal basolateral staining pattern of NTCP in stage II PBC(E). CON, control liver; ALC, cholestatic alcoholic hepatitis; AIH, cholestaticautoimmune hepatitis; DRUG, drug-induced cholestasis. Bar 5 20 mm.


FIG. 7. BSEP immunofluores-cence staining in human liver biop-sies. Immunofluorescence stainingwas performed with specific primaryantibodies against BSEP (in red) andcytokeratin 8/18 (in green) on acetonand chloroform-fixed frozen sec-tions from routine percutaneousliver biopsies prepared in parallel asdescribed in Materials and Methods.The fluorescent images were col-lected under standardized condi-tions by confocal laser scanning mi-croscopy. Double (A, C, E, G, I) andcorresponding single labeling (B, D,F, H, J) are shown. Overlapping ofpericanalicular cytokeratin sheet (ingreen) with canalicular BSEP immu-nolabeling (in red) results in a yellowsignal. Normal canalicular stainingpattern of BSEP in control liver (A,B). Reduced canalicular immunola-beling for BSEP in inflammatorycholestasis caused by cholestatic al-coholic hepatitis (C, D), cholestaticautoimmune hepatitis (E, F), anddrug (flutamide)-induced cholesta-sis (G, H). The red dot in (G and H)(left margin) represents nonspecificstaining of a leukocyte in the sinu-soidal space with the secondaryantibody. Note empty cells withoutcytokeratin filament network con-taining Mallory bodies in cholestaticalcoholic hepatitis (C; right margin).Although an overall reduction ofBSEP-positive canaliculi was ob-served, BSEP staining was preservedor even appeared enhanced in somecanaliculi that remained BSEP-posi-tive (F, H). Normal canalicular stain-ing pattern of BSEP in stage II PBC(I, J). Bar 5 20 mm. CON, controlliver; ALC, cholestatic alcoholic hep-atitis; AIH, cholestatic autoimmunehepatitis; DRUG, drug-inducedcholestasis.


FIG. 8. MRP2 immunofluores-cence staining in human liver biop-sies. Immunofluorescence stainingwas performed with specific primaryantibodies against MRP2 (in red) andcytokeratin 8/18 (in green) on acetonand chloroform-fixed frozen sec-tions from percutaneous routineliver biopsies prepared in parallel asdescribed in Materials and Methods.The fluorescent images were col-lected under standardized condi-tions by confocal laser scanning mi-croscopy. Double (A, C, E, G, I) andcorresponding single labeling (B, D,F, H, J) are shown. Overlapping ofpericanalicular cytokeratin sheet (ingreen) with canalicular MRP2 immu-nolabeling (in red) results in a yellowsignal. Normal canalicular stainingpattern of MRP2 in control liver(A, B). Reduced canalicular immu-nolabeling for MRP2 in inflamma-tory cholestasis caused by cholestaticalcoholic hepatitis (C, D), cholestaticautoimmune hepatitis (E, F), anddrug (flutamide)-induced cholesta-sis (G, H). Normal canalicular stain-ing pattern of MRP2 in stage II PBC(I, J). Bar 5 20 mm. CON, controlliver; ALC, cholestatic alcoholic hep-atitis; AIH, cholestatic autoimmunehepatitis; DRUG, drug-inducedcholestasis.


FIG. 9. P-glycoprotein immuno-fluorescence staining in human liverbiopsies. Double-labeling immuno-fluorescence staining was performedwith a specific C219 antibody (rec-ognizing all P-glycoproteins includ-ing MDR1 and MDR3 in addition toBSEP) (in red) and cytokeratin 8/18(in green) on aceton and chloroform-fixed frozen sections from routinepercutaneous liver biopsies preparedin parallel as described in Materialsand Methods. The fluorescent im-ages were collected under standard-ized conditions by confocal laserscanning microscopy. Double (A, C,E, G, I) and corresponding single la-beling (B, D, F, H, J) are shown.Overlapping of pericanalicular cyto-keratin sheet (in green) with canalic-ular transporter immunolabeling (inred) results in a yellow signal. Nor-mal canalicular staining pattern incontrol liver (A, B). Maintained oreven enhanced canalicular stainingpattern in inflammatory cholestasiscaused by cholestatic alcoholic hep-atitis (C, D), cholestatic autoim-mune hepatitis (E, F), and drug (flu-tamide)-induced cholestasis (G, H).Normal canalicular staining patternin stage II PBC (I, J). Bar 5 20 mm.CON, control liver; ALC, cholestaticalcoholic hepatitis; AIH, cholestaticautoimmune hepatitis; DRUG, drug-induced cholestasis.


MDR P-glycoproteins could represent a protective mecha-nism against cholestasis-induced oxidative stress.30 In con-trast to experimental models of cholestasis, mRNA levels ofMRP3 (possibly serving as a compensatory overflow systemvia the basolateral membrane under cholestatic conditions54)were not increased in patients with inflammatory cholestasis,although this does not exclude changes at the protein level.Because MRP3 is expressed in both hepatocytes and cholan-giocytes,53-55 alterations of MRP3 expression may be morepronounced in long-standing obstructive cholestasis withmarked bile duct and ductular proliferation.56 In line with thishypothesis, increased MRP3 expression has been observedmainly at the basolateral domain of cholangiocytes of prolif-erated bile ducts in patients with obstructive cholestasis,suggesting a role for MRP3 in cholehepatic cycling of con-jugated bile salts.53 mRNA levels of FIC1 - a canaliculartransporter mutated in hereditary cholestasis21 - also re-mained unchanged, indicating that alterations of FIC1 mRNAexpression may not play a role in the acquired cholestaticdisorders investigated in this study. Importantly, conservedexpression of these transport systems in the present studyindicates that reduction of NTCP, OATP2, and BSEP mRNA isnot a nonspecific finding secondary to hepatocellular injury,liver fibrosis, or influx of inflammatory cells with a reductionof hepatocyte-specific RNA.

In contrast to the alterations in inflammatory cholestasis,no changes in transporter mRNA expression and immunoflu-orescence staining were observed in precirrhotic stages ofPBC (stage I-II). These findings indicate that alterations ofhepatobiliary transporter expression (e.g., down-regulationby proinflammatory cytokines produced in the inflamed por-tal tracts) may not play a central role in the pathogenesis ofearly PBC in which interlobular bile ducts are the primarytargets of an immune attack.11 However, patients with PBChad anicteric cholestasis, while all patients with inflamma-tion-induced cholestasis had more severe icteric cholestasiswith higher serum bilirubin and bile salt levels. Therefore,down-regulation of transporters could be a rather late event in

PBC, secondary to loss of interlobular bile ducts with progres-sive cholestasis and resulting liver damage. TransportermRNA levels also remained unchanged in 3 patients withanicteric stage IV PBC (data not shown). Further studies willhave to determine whether hepatobiliary transporter expres-sion changes during icteric stages of PBC. However, jaundiceoccurs late in the clinical course of PBC, and potential changesof transporter expression in icteric PBC may reflect conse-quences of late-stage disease rather than giving mechanisticinsights into the causes of cholestasis in these patients. Basedon the data from Mdr2 (2/2) knockout mice developing achronic nonsuppurative cholangitis, it has been hypothesizedthat absent MDR3 (human orthologue of mouse Mdr2) couldbe an important factor in the pathogenesis of vanishing bileduct syndromes such as PBC.57 Consistent with a previousreport,9 no changes of MDR3 mRNA levels were seen in PBC,indicating that alterations of MDR3 mRNA expression do notplay a role in the pathogenesis of this disease.

In summary, our findings demonstrate a coordinated re-duction of basolateral (NTCP, OATP2) and canalicular trans-port systems (BSEP, MRP2) in acquired cholestatic liver dis-eases, which may contribute to increased serum bile salt levelsand impaired clearance of drugs and organic anions duringcholestasis. Reduced expression of these transport systems inhumans can occur at least in part at the mRNA level as ob-served for NTCP, OATP2, and, to a lesser extent, for BSEP.Normal MRP2 mRNA levels despite reduced canalicularMRP2 immunostaining in severely cholestatic and jaundicedpatients suggest that posttranscriptional/posttranslationalmechanisms (e.g., changes in translational efficiency, im-paired intracellular sorting/targeting to the canalicular mem-brane, retrieval from the canalicular membrane with sub-sequent lysosomal breakdown)51,58,59 may also play animportant role in the regulation of hepatobiliary transportsystems in humans. Altered hepatobiliary transporter expres-sion in cholestasis may be an important target for specificpharmacotherapeutic interventions aimed at stimulation andrestoration of defective transporter expression and func-tion.5,6,35,60

Acknowledgment: Specific primary antibodies were gener-ously provided by Drs. Bruno Stieger and Peter Meier, Univer-sity Hospital, Zurich, Switzerland (NTCP); Dr. DietrichKeppler, Deutsches Krebsforschungszentrum, Heidelberg,Germany (MRP2); and Drs. Michael Muller and Peter Jansen,University Hospital Groningen, the Netherlands (BSEP). Theauthors also acknowledge their surgical colleagues who per-formed the liver resections; Dr. Birgit Murer for her assistancein coordinating the procurement of the liver biopsy samples;and Dr. Franz Quehenberger for his help with the statisticalanalysis.

REFERENCES

1. Oude Elferink RPJ, Meijer DKF, Kuipers F, Jansen PLM, Groen AK,Groothuis GMM. Hepatobiliary secretion of organic compounds; molec-ular mechanisms of membrane transport. Biochim Biophys Acta 1995;1241:215-268.

2. Muller M, Jansen PLM. Molecular aspects of hepatobiliary transport.Am J Physiol 1997;272:G1285-G1303.

3. Trauner M, Meier PJ, Boyer JL. Molecular pathogenesis of cholestasis.N Engl J Med 1998;339:1217-1227.

4. Trauner M, Meier PJ, Boyer JL. Molecular regulation of hepatocellulartransport systems in cholestasis. J Hepatol 1999;31:165-178.

FIG. 10. Alterations of hepatobiliary transport systems in inflammation-induced cholestasis. Reduced expression of basolateral (NTCP, OATP2) andcanalicular transport systems (BSEP, MRP2) may contribute to cholestasisand jaundice by impairing hepatic uptake and excretion of bile salts and otherorganic anions (e.g., conjugated bilirubin). Down-regulation of these trans-port systems can occur at least in part at the mRNA level as observed forNTCP, OATP2, and, to a lesser extent, for BSEP. Normal MRP2 mRNA levelsdespite reduced canalicular MRP2 immunostaining suggest a role for post-transcriptional/posttranslational mechanisms. Unchanged expression ofother transport systems (MDR1, MDR3, MRP3, and FIC1) indicates thatchanges of these transport systems may not play a role in the pathogenesis ofinflammation-induced cholestasis. In addition, maintained expression ofMDR1 and MRP3 may protect hepatocytes from further accumulation ofpotentially toxic biliary constituents and cholestasis-induced oxidative stress.


5. Prieto J, Qian C, Garcia N, Diez J, Medina JF. Abnormal expression ofanion exchanger genes in primary biliary cirrhosis. Gastroenterology1993;105:572-578.

6. Medina JF, Martinez-Anso E, Vazquez JJ, Prieto J. Decreased anion ex-changer 2 immunoreactivity in the liver of patients with primary biliarycirrhosis. HEPATOLOGY 1997;25:12-17.

7. Shneider BL, Fox VL, Schwarz KB, Watson CL, Ananthanarayanan M,Thevananther S, Christie DM, et al. Hepatic basolateral sodium-depen-dent-bile acid transporter expression in two unusual cases of hyper-cholanemia and in exrahepatic biliary atresia. HEPATOLOGY 1997;25:1176-1183.

8. Kullak-Ublick GA, Beuers U, Fahney C, Hagenbuch B, Meier PJ, Paum-gartner G. Identification and functional characterization of the promoterregion of the human organic anion transporting polypeptide gene. HEPA-TOLOGY 1997;26:991-997.

9. Dumoulin FL, Reichel C, Sauerbruch T, Spengler U. Semiquantitation ofintrahepatic MDR3 mRNA levels by reverse transcription/competitivepolymerase chain reaction. J Hepatol 1997;26:852-856.

10. Becker-AndrE M, Hahlbrock K. Absolute mRNA quantification using thepolymerase chain reaction (PCR). A novel approach by a PCR aidedtranscript titration assay (PATTY). Nucleic Acids Res 1989;17:9437-9446.

11. Ludwig J, Dickson ER, MacDonald GSA. Staging of chronic non-suppu-rative destructive cholangitis (syndrome of primary biliary cirrhosis).Virchows Arch A 1978;379:103.

12. Krieg P, Amtmann E, Sauer G. The simultaneous extraction of high-molecular-weight DNA and RNA from solid tumors. Anal Biochem 1983;134:288-294.

13. Eferl R, Sibilia M, Hilberg F, Fuchsbichler A, Kufferath I, Guertel B, ZenzR, et al. Functions of c-jun in liver and heart developement. J Cell Biol1999;145:1049-1061.

14. Hagenbuch B, Meier PJ. Molecular cloning, chromosomal localization,and functional characterization of a human Na1/bile acid cotransporter.J Clin Invest 1994;93:1326-1331.

15. Abe T, Kakyo M, Tokui T, Nakagomi R, Nishio T, Nakai D, Nomura H, etal. Identification of a novel gene family encoding human liver-specificorganic anion transporter LST-1. J Biol Chem 1999;274:17159-17163.

16. Strautnieks SS, Bull LN, Knisely AS, Kocoshis SA, Dahl N, Arnell H, SokalE, et al. A gene encoding a liver-specific ABC transporter is mutated inprogressive familial intrahepatic cholestasis. Nat Genet 1998;20:233-238.

17. Chen CJ, Chin JE, Ueda K, Clark DP, Pastan I, Gottesman MM, RoninsonIB. Internal duplication and homology with bacterial transport proteinsin the mdr1 (P-glycoprotein) gene from multidrug-resistant human cells.Cell 1986;47:381-389.

18. van der Bliek AM, Kooiman PM, Schneider C, Borst PA. Sequence ofmdr3 cDNA encoding a human P-glycoprotein. Gene 1988;71:401-411.

19. Taniguchi M, Wada M, Kohno K, Nakamura T, Kawabe T, Kawakami M,Kagotani K, et al. A human canalicular multispecific organic anion trans-porter (cMOAT) gene is overexpressed in cisplatin-resistant human can-cer cell lines with decreased drug accumulation. Cancer Res 1996;56:4124-4129.

20. Kiuchi Y, Suzuki H, Hirohashi T, Tyson CA, Sugiyama Y. cDNA cloningand inducible expression of human multidrug resistance associated pro-tein 3 (MRP3). FEBS Lett 1998;433:149-152.

21. Bull LN, van Eijk MJT, Pawlikowska L, DeYoung JA, Juijn JA, Liao M,Klomp LWJ, et al. A gene encoding a P-type ATPase mutated in two formsof hereditary cholestasis. Nat Genet 1998;18:219-224.

22. Arcari P, Martinelli R, Salvatore F. The complete sequence of a full lengthcDNA for human liver glyceraldehyde-3-phosphate dehydrogenase: evi-dence for multiple mRNA species. Nucleic Acids Res 1984;12:9179-9189.

23. Ausubel FM, Brent R, Kingston RE Moore DD Seidman JG, Simth JA,Struhl K. Current protocols in molecular medicine. New York: JohnWiley & Sons, Inc., 1998.

24. Ho SN, Hunt HD, Horton RM, Pullen JK, Pease LR. Site-directed mu-tagenesis by overlap extension using the polymerase chain reaction.Gene 1989;77:51-59.

25. Sarkar G, Sommer SS. The “megaprimer” method of site-directed mu-tagenesis. Biotechniques 1990;8:404-407.

26. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratorymanual. New York: Cold Spring Harbor Laboratory Press, 1989.

27. Zimmermann K, Mannhalter JW. Technical aspects of quantitative com-petetive PCR. Biotechniques 1996;21:268-279.

28. Hutter H, Zatloukal K, Winter G, Stumptner C, Denk H. Disturbance ofcytokeratin homeostasis in griseofulvin-intoxicated mouse liver. Lab In-vest 1993;69:576-582.

29. Kullak-Ublick GA, Hagenbuch B, Stieger B, Schteingart CD, HofmannAF, Wolkoff AW, Meier PJ. Molecular and functional characterization ofan organic anion transporting polypeptide cloned from human liver.Gastroenterology 1995;109:1274-1282.

30. Vos TA, Hooiveld GJEJ, Koning H, Childs S, Meijer DKF, Moshage H,Jansen PLM, et al. Up-regulation of the multidrug resistance genes, mrp1and mdr1b, and down-regulation of the organic anion transporter, mrp2,and the bile salt transporter, spgp, in endotoxemic rat liver. HEPATOLOGY

1998;28:1637-1644.31. Kartenbeck J, Leuschner U, Mayer R, Keppler D. Absence of the canalic-

ular isoform of the MRP gene encoded conjugate export pump from thehepatocytes in Dubin-Johnson syndrome. HEPATOLOGY 1996;23:1061-1066.

32. Jansen PLM, Strautnieks SS, Jacquemin E, Hadchouel M, Sokal EM,Hooivield GJEJ, Koning JH, et al. Hepatocanalicular bile salt exportpump deficiency in patients with progressive familial intrahepatic cho-lestasis. Gastroenterology 1999;117:1370-1379.

33. Deleuze JF, Jacuemin E, Dubuisson C, Crestell D, Dumont M, Erlinger S,Bernard O, et al. Defect of multidrug resistance 3 gene expression in asubtype of progressive familial intrahepatic cholestasis. HEPATOLOGY

1996;23:904-908.34. de Vree JML, Jacquemin E, Sturm E, Cresteil D, Bosma PJ, Aten J, Deleuze

JF, et al. Mutations in the MDR3 gene cause progressive familial intrahe-patic cholestasis. Proc Natl Acad Sci U S A 1998;95:282-287.

35. Prieto J, Garcia N, Marti-Climent JM, Penuelas I, Richter JA, Medina JF.Assessment of biliary bicarbonate secretion in humans by positron emis-sion tomography. Gastroenterology 1999;117:167-172.

36. Brenichou C. Criteria of drug-induced liver disorders. Report of an in-ternational consensus meeting. J Hepatol 1990;11:272-276.

37. Trauner M, Fickert P, Stauber RE. Inflammation-induced cholestasis. JGastroenterol Hepatol 1999;14:946-959.

38. Crawford JM, Boyer JL. Clinicopathology conferences: inflammation-induced cholestasis. HEPATOLOGY 1998;28:253-260.

39. Gartung C, Ananthanarayanan M, Rahman MA, Schuele S, Nundy S,Soroka CS, Stolz A, et al. Down-regulation of expression and function ofthe rat liver Na1/bile acid cotransporter in extrahepatic cholestasis. Gas-troenterology 1996;110:199-209.

40. Moseley RH, Wang W, Takeda H, Lown K, Shick L, AnanthanarayananM, Suchy FJ. Effect of endotoxin on bile acid transport in rat liver: apotential model for sepsis-associated cholestasis. Am J Physiol 1996;271:G137-G146.

41. Simon FR, Fortune J, Iwahashi M, Gartung C, Wolkoff A, Sutherland E.Ethinyl estradiol cholestasis involves alterations in expression of liversinusoidal transporters. Am J Physiol 1997;271:G1043-G1052.

42. Trauner M, Arrese M, Lee H, Boyer JL, Karpen SJ. Endotoxin downregu-lates rat hepatic ntcp gene expression via decreased activity of criticaltranscription factors. J Clin Invest 1998;101:2092-2100.

43. Denson LA, Auld KL, Schiek DS, McClure MH, Mangelsdorf DJ, KarpenSJ. Interleukin-1 supresses retionoid transactivation of two hepatic trans-porter genes involved in bile formation. J Biol Chem 2000;275:8835-8843.

44. Kakyo M, Unno M, Tokui T, Nakagomi R, Nishio T, Iwasashi H, Nakai D,et al. Molecular characterization and functional regulation of a novel ratliver–specific organic anion transporter rlst-1. Gastroenterology 1999;117:770-775.

45. Gartung C, Schuele S, Schlosser SF, Boyer JL. Expression of the rat liverNa1/taurocholate cotransporter is regulated in vivo by retention of bili-ary constituents but not their depletion. HEPATOLOGY 1997;25:284-290.

46. Makishima M, Okamoto AY, Repa JJ, Tu H, Learned RM, Luk A, Hull MV,et al. Identification of a nuclear receptor for bile acids. Science 1999;284:1362-1365.

47. Sinal CJ, Tohkin M, Miyata M, Ward JM, Lambert G, Gonzales FJ. Tar-geted disruption of the nuclear receptor FXR/RXR impairs bile acid andlipid homeostasis. Cell 2000;102:731-744.

48. Karpen SJ, Sun AQ, Kudish B, Hagenbuch B, Meier PJ, AnanthanarayananM, Suchy FJ. Multiple factors regulate the rat liver basolateral sodium–dependent bile acid cotransporter gene promotor. J Biol Chem 1996;271:15211-15221.

49. Lee JM, Trauner M, Soroka CJ, Stieger B, Meier PJ, Boyer JL. The molec-ular expression of the bile salt export pump, sister of P-glycoprotein, isselectively preserved in cholestatic liver injury. Gastroentrology 2000;118:163-172.


50. Trauner M, Arrese M, Soroka CJ, Ananthanarayanan M, Koeppel TA,Schlosser SF, Suchy FJ, et al. The rat canalicular conjugate export pump(Mrp2) is down-regulated in intrahepatic and obstructive cholestasis.Gastroenterology 1997;113:255-264.

51. Paulusma CC, Kothe MJ, Bakker CT, Bosma PJ, van Bokkhoven I, vanMarle J, Bolder U, et al. Zonal down-regulation and redistribution of themultidrug resistance protein 2 during bile duct ligation in rat liver. HEPA-TOLOGY 2000;31:684-693.

52. Elferink MG, Olinga P, Merema MT, Mueller M, Slooff M, Meijer DK.Perturbation of NTCP, BSEP and MRP2 in human liver slices exposed toLPS [Abstract]. HEPATOLOGY 2000;32(Suppl 2):A1075.

53. Shoda J, Oda K, Kamiya J, Nimura Y, Suzuki H, Sugiyama Y, Miyazaki H,et al. The molecular and immunohistochemical expression of multidrugresistance-associated protein (MRP3) in the liver and intestine of patientswith cholestasis and its biological significance [Abstract]. Gastroenterol-ogy 2000;118:A933.

54. Koenig J, Rost D, Cui Y, Keppler D. Characterization of the humanmultidrug resistance protein isoform MRP3 localized to the basolateralhepatocyte membrane. HEPATOLOGY 1999;29:1156-1163.

55. Kool M, van der Linden M, de Hass M, Scheffer GL, de Vree JML, SmithAJ, Jansen G, et al. MRP3, an organic anion transporter able to transport-anti cancer drugs. Proc Natl Acad Sci U S A 1999;96:6914-6919.

56. Ogawa K, Suzuki H, Hirohashi T, Ishikawa T, Meier PJ, Hirose K,Akizawa T, et al. Characterization of inducible nature of MRP3 in ratliver. Am J Physiol 2000;278:G438-G446.

57. Smit JJ, Schinkel AH, Oude Elferink RP, Groen AK, Wagenaar E, vanDeemeter L, Mol CA, et al. Homozygous disruption of the murine mdr2P-glycoprotein gene leads to a complete absence of phospholipid frombile and to liver disease. Cell 1993;75:451-462.

58. Boyer JL, Soroka CJ. Vesicle targeting to the apical domain regulates bileexcretory function in isolated rat hepatocyte couplets. Gastroenterology1996;109:1600-1611.

59. Kubitz R, Wettstein M, Warskulat U, Haeussinger D. Regulation of themultidrug resistance protein 2 in the rat liver by lipopolysaccharide anddexamethasone. Gastroenterology 1999;116:401-410.

60. Fickert P, Zenz R, Zollner G, Pojer C, Fuchsbichler A, Zatloukal K, DenkH, et al. Regulation of hepatocellular transporter expression depends onbile acid hydrophobicity in mice [Abstract]. Gastroenterology 2000;118(Suppl 2):A934.


Hepatobiliary transporter expression in percutaneous liver biopsies of patients with cholestatic liver diseases

Documents