54
q ELSEVIER Biochimica et Biophysica Acta 1241 (1995) 215-268 Btt Biochi f ic~a et Biophysica A~ta Hepatobiliary secretion of organic compounds; molecular mechanisms of membrane transport Ronald P.J. Oude Elferink a,*, Dirk K.F. Meijer b, Folkert Kuipers c, Peter L.M. Jansen d, Albert K. Groen a, Geny M.M. Groothuis b Department of Gastrointestinal and Lit'er Diseases, Center for Lit'er and Intestinal Research, Academic Medical Center F0-116, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands b . . . Department of Pharmacokmetws and Drug Dehvery, Universi~ Centre for Pharmacy, Groningen Institute for Drug Studies, Groningen , The Netherlands c Department of Pediatrics, UniL'ersio' ofGroningen, Groningen, The Netherlands d Department of Gastroenterology and Hepatology, Uniuersity ofGroningen, Groningen, The Netherlands Received 21 December 1994; revised 15 Febuari 1995; accepted 23 Febuari 1995 Contents 1. Introduction .................................................... 216 2. Methods for studying hepatic transport processes ................................ 216 2.1. In vivo studies ................................................ 216 2.2. In vitro studies ................................................ 217 2.3. Methods used for isolation of transport proteins .............................. 219 2.4. The use of animal models .......................................... 221 3. Hepatobiliary transport of organic anions .................................... 221 3.1. Sinusoidal uptake of organic anions .................................... 222 3.2. Transcellular transport of organic anions .................................. 225 3.3. Canalicular transport ............................................ 226 4. Hepatobiliary6 transport of organic cations ................................... 235 4.1. Sinusoidal uptake of organic cations .................................... 235 4.2. Intracellular sequestration of organic cations ................................ 239 4.3. Canalicular transport of organic cations .................................. 241 4.4. The role of P-glycoproteins ......................................... 242 5. Hepatobiliary transport of lipids ......................................... 247 5. I. Composition and physical form of biliary lipids .............................. 247 5.2. The origin and precursor pools of biliary lipids .............................. 248 5.3. Intracellular trafficking of biliary lipids ................................... 249 5.4. Regulation of biliary lipid secretion: the role of bile salts ......................... 250 5.5. Regulation of biliary lipid secretion: the role of mdr2 P-glycoprotein ................... 251 5.6. Functional aspects of biliary lipid secretion ................................ 253 * Corresponding author. Fax: +31 20 6917033. 0304-4157/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0304-4157(95)00006-2

Hepatobiliary secretion of organic compounds; molecular mechanisms of membrane transport

  • Upload
    uva

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

q

ELSEVIER Biochimica et Biophysica Acta 1241 (1995) 215-268

Btt Biochi f ic~a et Biophysica A~ta

Hepatobiliary secretion of organic compounds; molecular mechanisms of

membrane transport

Ronald P.J. Oude Elferink a,*, Dirk K.F. Meijer b, Folkert Kuipers c, Peter L.M. Jansen d,

Albert K. Groen a, Geny M.M. Groothuis b

Department of Gastrointestinal and Lit'er Diseases, Center for Lit'er and Intestinal Research, Academic Medical Center F0-116, Meibergdreef 9, 1105

AZ Amsterdam, The Netherlands b . . .

Department of Pharmacokmetws and Drug Dehvery, Universi~ Centre for Pharmacy, Groningen Institute for Drug Studies, Groningen , The

Netherlands

c Department of Pediatrics, UniL'ersio' ofGroningen, Groningen, The Netherlands

d Department of Gastroenterology and Hepatology, Uniuersity ofGroningen, Groningen, The Netherlands

Received 21 December 1994; revised 15 Febuari 1995; accepted 23 Febuari 1995

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

2. Methods for studying hepatic transport processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

2.1. In vivo studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

2.2. In vitro studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217

2.3. Methods used for isolation of transport proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

2.4. The use of animal models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

3. Hepatobiliary transport of organic anions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

3.1. Sinusoidal uptake of organic anions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222

3.2. Transcellular transport of organic anions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

3.3. Canalicular transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226

4. Hepatobil iary6 transport of organic cations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235

4.1. Sinusoidal uptake of organic cations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235

4.2. Intracellular sequestration of organic cations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

4.3. Canalicular transport of organic cations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241

4.4. The role of P-glycoproteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242

5. Hepatobiliary transport of lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247

5. I. Composi t ion and physical form of biliary lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247

5.2. The origin and precursor pools of biliary lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248

5.3. Intracellular trafficking of biliary lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249

5.4. Regulation of biliary lipid secretion: the role of bile salts . . . . . . . . . . . . . . . . . . . . . . . . . 250

5.5. Regulation of biliary l ipid secretion: the role of mdr2 P-glycoprotein . . . . . . . . . . . . . . . . . . . 251

5.6. Functional aspects of biliary lipid secretion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

* Corresponding author. Fax: +31 20 6917033.

0 3 0 4 - 4 1 5 7 / 9 5 / $ 0 9 . 5 0 © 1995 Elsevier Science B.V. All rights reserved

SSDI 0 3 0 4 - 4 1 5 7 ( 9 5 ) 0 0 0 0 6 - 2

216 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215 268

6. Regulation of canalicular transport and its consequence for development of cholestasis . . . . . . . . . . . . 254

6.1. Regulation of transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254

6.2. Cholestasis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255

7. Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258

1. Introduction

One of the main functions of the liver, the formation of

bile, involves the vectorial transport of compounds such as

bile salts, phospholipids, cholesterol and other organic

compounds. Bile flow is critically dependent on the trans-

port of bile salts and organic anions because it is an

osmotically driven process; the secretion of bile salts

provides an osmotic gradient between canalicular lumen

and plasma, which causes a water flow through the tight

junctions sealing the adjacent hepatocytes. Thus any im-

pairment of the transport of major biliary components will

lead to compromised bile flow, i.e., cholestasis. It is

therefore important to understand the complex series of

events responsible for the uptake of compounds into the

hepatocyte, transport through the cell and secretion into the

canalicular lumen. Once these processes are delineated one

can try to understand the etiology of diverse forms of

intrahepatic cholestasis. Due to the critical dependence of

bile formation on transport processes it is conceivable that,

in most cases, intrahepatic cholestasis will be the conse-

quence of a disturbance in one or more steps of hepatobil-

iary transport.

Besides the maintenance of bile flow, transport pro-

cesses are of importance for the disposition of a wide array

of endo-and xenobiotics. The disposition of drugs from the

body as reflected by plasma clearance may largely depend

on their hepatic metabolism and biliary clearance. The

elucidation of the mechanisms of hepatic uptake and bil-

iary disposition will therefore contribute to a better defini-

tion of the plasma disappearance profiles of existing and

newly developed pharmaceuticals. In addition, proper un-

derstanding of hepatic transport may lead to the develop-

ment of compounds that can be specifically targeted to

various cell types in the liver.

In this review we will give an overview of the present

knowledge of the processes involved in biliary secretion of

organic solutes. In order to limit the scope of this review

we will only deal with transport of 'cholephilic' com-

pounds, i.e., those compounds that undergo significant

hepatic uptake and are secreted into bile in high concentra-

tions. In this respect, three classes of compounds can be

discerned: organic anions (including bile salts), organic

cations and lipids. Furthermore, our aim is primarily to

discuss the present knowledge of the mechanisms involved

in these transport processes, rather than to extensively

describe the behavior of individual compounds. Evidently,

emphasis will be put on recent developments in this field;

many excellent reviews have been devoted to this subject

in the past [ 1 - 11 ].

2. Methods for studying hepatic transport processes

Hepatobiliary transport can be studied using various in

vivo and in vitro techniques. Because each of these tech-

niques has its own requirements, possibilities and limita-

tions, a proper insight in the process studied often requires

their combined use.

2.1. In vivo studies

In the intact organism the hepatobiliary transport of

drugs or endogenous cholephils should ideally be evalu-

ated by taking an adequate number plasma, urine and bile

samples after intravenous injection of the compound of

interest. Subsequent compartmental analyses of the con-

centrations will provide data on the rate constants for

uptake and excretion, thereby assessing quantitatively the

involvement of the liver in the clearance of the compound

under study under optimal conditions of liver function. For

obvious ethical and toxicological reasons, the majority of

the in vivo studies on drug transport are performed in

laboratory animals, mostly the rat. Under anesthesia a

limited number of blood samples can be obtained together

with bile samples via a cannula in the common bile duct

and urine samples using a cannula in the urinary bladder.

Although these experiments allow the study of drug trans-

port under optimal liver function they are limited to a time

period of about 4 - 6 hours. In addition, effects of anesthe-

sia on drug transport can seriously influence the results

[12]. Therefore, techniques were developed to study hepa-

tobiliary transport of drugs in the unrestrained, non-

anesthetized rat using permanent cannulas in the jugular

vein and the common bile duct and in the urine bladder,

although the latter is technically more difficult [13-16].

When a permanent cannula of the bile duct is combined

with cannulation of the duodenum, bile secretion into the

duodenum is maintained. In this way bile samples can be

collected without long-term interruption of the enterohep-

atic circulation.

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 217

In man, pharmacokinetic studies are usually limited to

analysis of plasma disappearance curves and urinary excre-

tion data.This implies that the involvement of the liver

cannot be rigorously assessed, due to the unknown extra-

hepatic distribution and elimination. Sampling bile directly

from the bile duct can only be performed in patients with

temporarily implanted bile catheters after major liver or

gallbladder surgery. This limits these kinetic studies to

patients, who may have abnormal liver function due to

their disease, to surgery and/or concomitant medication

[17,18]. In addition, only incomplete recovery of bile can

be achieved. Balloon-occludable multilumen duodenal

tubes [19] and triple-lumen-tubes [20] were used to mea-

sure biliary excretion in normal subjects. Recently novel

techniques were developed to quantify biliary excretion in

healthy, unanesthetised humans: Vonk et al. [21] used an

encapsulated nylon thread, swallowed by a fasting subject.

After dissolution of the capsule the thread extends into the

duodenum and duodenal bile can be sampled upon recov-

ery of the thread. With this technique relative concentra-

tions of bile components can be measured [22,23] but

absolute excretion rates can not be measured. Using the

'duodenal-marker-perfusion technique' developed by An-

gelin et al. [24] more quantitative data on biliary excretion

can be obtained, and the kinetic interactions of digitoxine

with verapamil [25] and with quinine and quinidine [26]

were established.

In vivo hepatic transport studies in man are complex,

due to the influence of extrabepatic elimination and the

impossibility to study the influence of blood and bile flow.

Therefore, in vitro studies are necessary to obtain more

detailed kinetic data on the mechanisms involved in uptake

and excretion processes by the liver.

2.2. In uitro studies

The isolated perfused liuer

Studying drug transport in the isolated perfused liver

[27-31] combines the advantages of excluding extrahep-

atic influences, maintaining an intact liver structure along

with the opportunity to manipulate medium flow, composi-

tion and temperature as well as perfusion direction. This

allows, for example, studies on the influence of protein

binding [28,32,33], on the acinar heterogeneity of hepato-

cytes with regard to transport functions [34-36], on flow-

dependency of drug clearance [37] and on the influence of

temperature on hepatic clearance [38]. Multiple samples

can be taken from both the inflow and the outflow medium,

allowing pharmacokinetic analysis by compartmental mod-

els [39] and multiple indicator dilution studies [31,40].

Adequate oxygen supply to the liver can be achieved with

Krebs solutions supplied at high perfusion rates [27,41].

Addition of washed (human or rat) red blood cells to the

medium allows studies under more physiological flow

rates [42,43]. This is, however, technically more compli-

cated and results may be influenced by drug binding to the

blood cells. As alternative oxygen carriers, fluorocarbons

have been added to the medium, but toxic effects such as

ballooning of the sinusoids [44] and depression of cy-

tochrome P-450 activity were reported [45]. Livers can be

perfused in a single pass fashion, allowing optimal control

of the inflowing medium composition. Alternatively, the

medium can be recirculated, which mimicks the in vivo

situation and allows simple kinetic compartmental analy-

sis. The latter method is less expensive when costly sub-

strates or medium components are to be used [46].

Usually livers are perfused through the portal vein only.

As a consequence the biliary plexus, which in vivo is

mainly supplied with blood via the arteria hepatica, is not

adequately perfused. This does not, however, seem to

influence biliary transport function to a large extent, be-

cause in the combined arterial-portal venous perfusion

[47,48] only a modestly increased bile flow is reported

[48]. Whether the denervation of the perfused liver influ-

ences transport processes has not been estabfished yet.

The limited number of experiments thal~ can be per-

formed with a single perfused liver preparation and the

constraints of an acceptable liver viability of about 3 -4 h

initiated the development of other techniques, such as

isolated liver cells, hepatocyte couplets, plasma membrane

vesicles, and, more recently, the precision-cut liver slices.

Liter slices

The use of liver slices to study liver function was

virtually abandoned when the succesfull isolation of viable

hepatocytes was reported by Berry and Friend in 1968

[49]. Technical limitations in the preparation of slices thin

enough to allow sufficient oxygen and substrate supply to

the inner cell layers have led to alternative approaches.

The recent development of the Krumdieck tissue slicer,

allows a reproducible production of 200-300 /zm thin

slices [50]. Cell damage as result of the slicing procedure

seems rather limited. This technique initiated a revival in

the use of liver slices in research on hepatic metabolism

and transport. The advantages of this method are evident:

the liver structure with the various parenchymal and sinu-

soidal liver cells remains intact and the hepatocytes retain

their polarity,as in the isolated perfused liver, but now

many experiments can be performed in one liver. In addi-

tion, the localisation of periportal and perivenous cells in

the acinus is maintained. However, in contrast to the intact

liver, where acinar gradients of substrates are inevitable

due to the unidirectional blood flow, all cells in the slices

are exposed to the same substrate concentrations [36].

Potentially deleterious enzymes, present in collagenase

preparations, are not needed for the preparation of the

slices, in contrast to the usual hepatocyte preparation tech-

niques. Most importantly, the technique is easily applica-

ble to the livers of various species including that of man,

which renders this technique very useful to study drug

transport in human liver in vitro and to assess interspecies

differences in drug metabolism and transport [51,52]. The

218 R.P..I. Oude Elferink et aL / Biochirnica et Biophysica Acta 1241 (1995) 215-268

cells of the inner cell-layers seem to be adequately sup-

plied with oxygen and substrates [53] and the slices can be

cultured for at least 24 h. The slices can still be conve-

niently used after 24 h of cold storage in UW ('University

of Wisconsin'; [54,55]) organ preservation solution (Olinga,

to be published). As with isolated hepatocytes, cryopreser-

vation results in appreciable loss in metabolic function. Up

to now liver slices are mainly employed for drug

metabolism and toxicity studies [56,57], and studies on

drug transport until now are limited to the excretion of bile

salts [58].

The applicability of this technique to the study on

mechanisms of drug transport in man makes it an attractive

experimental tool. Human liver slices can be succesfully

prepared from pieces of livers that are discarded after

tranplantation of a part of a donor liver and from liver

tissue obtained after partial hepatectomy. The liver tissue

may be stored for up to 24 h in an adequate storage

solution before slicing without loss of viability. Using

human liver slices drug metabolism rates have been ob-

tained which were comparable to those in isolated cells

[57,59], but detailed transport studies with human liver

slices have not been published yet.

Isolated hepatocytes

Isolated hepatocytes are widely used to investigate the

mechanisms of hepatic transport and often isolated cells

show transport characteristics similar to those in isolated

perfused livers [39,60-62]. Nevertheless, damage to plasma

membrane proteins during perfusion with collagenase (that

often contains variable amounts of trypsin) cannot be

excluded. The ease of performing experiments with multi-

ple identical cell samples, excluding the influence of other

liver cell types, makes this technique very versatile in

determining the K m and Vma ~ as well as the driving forces

for uptake of a compound under study. Competition by

other compounds and the influence of protein binding can

be easily defined [61,63-65].

The loss of cell polarity and the redistribution of

canalicular membrane proteins over the entire cell surface

[66,67] creates a problem for the study of canalicular

transport processes and makes this model less suitable for

such experiments. It has, nevertheless, been demonstrated

that under the proper conditions canalicular transport func-

tions can be measured in freshly isolated hepatocytes

[68,69].

Uptake studies are usually performed with freshly iso-

lated cells. Although the cell polarity is partly recovered

during culture [67], the uptake rate in cultured hepatocytes

may rapidly decrease during culturing of the cells [70,71].

In the case of bile salt uptake this has been shown to be, at

least partly, due to a decrease in the mRNA levels of the

Na-taurocholate transporter during culture [72]. Possibly

also the loss of available membrane area as a result of

spreading and attachment of the cells to the culture dish

may play a role here [73]. Coculturing the cells with a liver

epithelial cell-line [74], or addition of tocopherol and

dexamethasone [75], retards this loss in transport rate.

Such experimental conditions may enable studies on

chronic effects of drugs and on mechanisms and regulation

of transport [65].

Isolated hepatocytes can be stored for at least 24 h in

organ preservation solution, i.e., the University of Wiscon-

sin (UW) solution or modifications thereof [54,55], without

loss of transport functions. Cryopreservation of liver cells

generally leads to a considerable loss of metabolically

viable cells after thawing [76].

Separation of isolated rat hepatocytes into fractions

enriched in periportal and perivenous cells can be achieved

by a brief, zone-selective perfusion with digitonin [77,78]

or by fluorescence activated cell sorting after selective

labelling of one of the zones [79]. The obtained cell

fractions can be used for studies into the zonal distribution

of metabolic functions including transport [36,80].

Hepatocytes can be isolated from human donor liver

tissue and from liver tissue obtained after partial hepatec-

tomy. The reproducibility of the isolation procedure how-

ever, with respect to yield and viability is generally less

than with rat liver. This seems to be mainly due to the

variability in the human liver tissue specimens [81]. How-

ever, after removal of non-viable cells by Percoll density

separation, human hepatocytes have been used for drug

metabolism and toxicity studies [81,82]. Published drug

transport studies with human hepatocytes are still scarce,

but the available data show that these cells are an appropri-

ate model to study interspecies differences and the mecha-

nisms of hepatic transport in man [81,83].

In the past, hepatocyte derived cell-lines have been

tested as a model to study drug transport, but with little

success due to decreased or even absent transport functions

[84,85]. More recently, cell lines have been obtained which

display a higher differentiation state. An important exam-

ple is the WIF-B cell, a hepatoma-derived hybrid cell line.

The majority of these cells (> 70%) form large canalicular

spaces in which organic anions like fluorescein and FITC-

labeled glycocholate are concentrated, suggesting that tran-

scellular transport of organic anions is functional in these

cells [86]. Similar results were obtained by Petzinger et al.

[87,88] who produced hybrid cells by fusion of primary rat

hepatocytes with Reuber hepatoma cells. These cells too

form canaliculi into which NBD-labeled cholate is concen-

trated. Analysis of the bile salt uptake revealed, however,

that these cells lack the sodium-dependent taurocholate

uptake system that is expressed in normal hepatocytes.

Thus, although for some purposes such cells can be highly

valuable, the expression of the various transport systems

cannot be predicted.

Canalicular secretion is successfully being studied in

couplets of hepatocytes: pairs of cells that are not sepa-

rated from each other during collagenase treatment, and

that have conserved a bile canalicular vacuole. Cell prepa-

rations can be enriched in couplets by limiting the collage-

R.P.J. Oude EIferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215 268 219

membranes seems to be inevitable with the presently avail-

able techniques [103]. Recently, conventionally purified

canalicular membranes [104] were further purified by free

flow electrophoresis [105]. Despite its limitations the use

of hepatocyte plasma membrane vesicles has been invalu-

able to the characterization of hepatic transport mecha-

nisms. Interference of intracellular events such as toxic

influence on liver cell function, metabolism and binding to

intracellular binding proteins and organelles is excluded. In

additon, both the internal and external medium can be

manipulated and this enables the study on the electrogenic

features, the energy dependency and the ion requirements

of transport [ 106].

Plasma membrane vesicles can also be prepared from

human livers [107-109] and sodium-dependent transport

of taurocholate was demonstrated in human basolateral

vesicles in accordance with isolated human cells [81] and

rat basolateral vesicles [110] and hepatocytes [111].

Fig. 1. The use of hepatocyte couplets for functional characterization of organic anion transport. Hepatocyte couplets from TR- rats (see Section 3.3) were analysed by phase contrast (panel A). Subsequently the cells were incubated with carboxydichlorofluorescein diacetate and analysed by confocal scanning laser microscopy. Note that no secretion of fluores- cence is observed in the canalicular vacuole due to the defect in organic anion secretion (panel B). As a control the couplets were also incubated with the fluorescent bile salt, FITC-glycocholate. Line scanning reveals concentrative secretion of the bile salt in the canalicular vacuole (panel C). From Ref. [124]

nase digestion and subsequent centrifugal elutriation of the

obtained cell suspension [89,90]. These couplets are espe-

cially suitable to study biliary excretion processes using

fluorescence microscopy [91,92], confocal microscopy (see

Fig. 1) [93,94] and time lapse video recording [95]. Using

such techniques the contractile nature of the canaliculi,

comparable to the canaliculi in vivo, has been demon-

strated [96,97]. Even the separation of periportal and

perivenous couplets was reported recently using elutriation

centrifugation [98]. Another important advantage of cou-

plets is the possibility to study routing and localization of

canalicular (transport) proteins. A beginning has been made

with studies into the influence of various conditions, like

hormones and second messengers, on these processes

[99,100].

Plasma membrane vesicles

The development of techniques to prepare sealed mem-

brane vesicles from the sinusoidai (basolateral) and

canalicular domains of the plasma membrane, allowed the

study of transport mechanisms at both poles of the hepato-

cyte separately [ 101,102]. For a proper interpretation of the

data care must be taken to define the origin and the

orientation (right side out or inside out) of the vesicles.

Cross contamination with the mutual domain specific

plasma membrane elements as well as other (intracellular)

2.3. Methods" used fi)r isolation ~?f" transport proteins

AffiniO' chromatography

In the past many groups have approached the isolation

of hepatic transport proteins by affinity chromatography

[112-114]. Suitable ligands were immobilized to a matrix

and solubilized fractions of hepatic membranes were ap-

plied to such columns and eluted by addition of the free

ligand. In the case of transport proteins this approach may

be hindered by a number of intrinsic problems.

Firstly, one has to be sure that the solubilized protein is

still functionally active, c.q. able to bind its ligand properly

with a low dissociation constant. It is a well known fact

that many integral membrane proteins lose their activity

upon complete solubilization. It is crucial for this approach

that the putative transporter protein is completely solubi-

lized. If solubilization is not complete, membrane frag-

ments may be formed that contain more than one protein.

Then, if the desired protein binds to the column due to its

affinity for the ligand, other (abundant) proteins may be

also retained by 'piggy-back' isolation due to their ubiqui-

tous presence in membrane vesicles. In the past this effect

has probably led to erroneous assignment of a transport

function to major binding proteins present in hepatic mem-

brane preparations.

Secondly, amphipathic ligands, like bil~rubin and bile

salts, have often been used for affinity chromatography.

The inherent problem of such compounds is that they also

aspecifically bind to hydrophobic sites on proteins. Since

the purification involves membrane proteins, many of such

sites will be present in the preparations. Therefore the

chance of aspecific binding of proteins to these columns is

considerable.

Thirdly, a problem arises that is general to any form of

purification of transport proteins; the purification process

can only be evaluated by reconstitution of the protein into

liposomes followed by the measurement ,of its transport

220 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

capacity. In many cases this represents an insensitive assay

for which too little protein is obtained to get correct data.

With these problems in mind it is not surprising that in

the past 'carrier' proteins have been isolated that later

turned out to have either another cellular localization than

the plasma membrane or exhibited another enzymatic func-

tion rather than the transport of cholephilic compounds.

Although well known examples exist of proteins which

have multiple localizations within the cell or have multiple

enzymic functions, the finding of bifunctional proteins

must be approached with great care and suspicion in view

of the pitfalls mentioned above.

Photoaffinity labeling of putative transport proteins Photo-reactive, radioactive, derivatives of ligands that

are transported across the hepatic membrane have been

synthesized in the past to identify proteins involved in the

transport process [115]. When incubated with membranes

in a sufficiently low concentration, proteins that specifi-

cally bind these compounds (among which must be the

carrier proteins) will bind the radioactive ligand to a higher

extent than other proteins. Upon illumination, the specific

binding proteins will therefore be labeled ro a relatively

higher extent than other proteins and this may help in the

purification and identification of these proteins. Although

elegant, this technique also suffers from significant prob-

lems. Transport proteins may indeed be labeled to higher

specific radioactivity, but their abundancy may be too low

to recognize them among other binding proteins. Con-

versely, aspecific binding proteins may reach only low

specific radioactivity but, due to their abundancy, they still

may come out as the most prominent labeled species. In

addition, similar to the situation with affinity chromatog-

raphy, often ligands are used with amphipathic properties.

Such substrates tend to bind aspecifically to proteins or

even may insert into membranes. In this context it is

important to note that in the past photo-affinity labeling

with radioactive 7,7 azo-derivative of taurocholate has led

to the assignment of a canalicular 100 kDa protein as the

protein involved in taurocholate transport across the

canalicular membrane [116]. However, when total proteins

from the canalicular membrane are analyzed by SDS-PAGE

it is clear that multiple 100 kDa proteins are abundantly

present in this membrane. Indeed, many proteins in this

molecular mass region have been identified in the canalic-

ular membrane at present among which are the ecto-ATPase

and dipeptidylpeptidase IV. Therefore labeling of this 100

kDa protein could in fact represent aspecific binding to

one or more highly abundant proteins. Although the pro-

tein preparation, upon purification and reconstitution gave

rise to potential-dependent taurocholate transport, it has

recently become clear that in highly purified canalicular

membranes no potential-dependent taurocholate transport

can be observed in spite of the fact that the particular 100

kDa protein was exclusively present in this fraction [105]

(see also Section 3). In view of the problems mentioned

above the assignment of a transport function to these

proteins on basis of photo-affinity labeling must be ap-

proached with caution.

Expression cloning with Xenopus laevis oocytes

The recognized capacity of Xenopus laevis oocytes to

efficiently translate foreign genetic information combined

with their ability to assemble oligomeric receptor/channel

complexes and insert them into the plasma membrane, has

greatly stimulated the cloning of plasma membrane recep-

tors and transport proteins [117]. Recently, this technique

has also been implemented in the field of hepatic trans-

porters and within a relative short period several cDNAs

have been isolated, the RNA of which upon injection in

oocytes give rise to transport functions.

In this approach the oocytes are used as an expression

system, while the actual cloning of a cDNA is achieved by

standard procedures. As a preliminary step in cloning

studies, the microinjection of total poly(A) + mRNA from

liver homogenate into the oocytes serves to provide evi-

dence for the presence of translatable mRNA encoding the

desired transport function while establishing the ability of

the oocyte to express the biologically active protein. At

this important stage several possibilities exist: injection of

total mRNA gives a transport signal above the background

and this means that after size fractionation, establishment

of a cDNA library and testing of pools of clones from the

library can only increase the signal due to the higher

abundancy of the signal. Thus, if a transport signal is

obtained with total poly(A) + mRNA further strategies for

cloning are straightforward, though laborious. However,

for several reasons a signal may be absent. The assay used

to assess transport activity is of crucial importance: it must

be sensitive enough to monitor activity in a few oocytes

and it must be rapid and simple enough to handle large

numbers when clones are to be tested. It is clear that a

transport assay involving uptake into the cells is consider-

ably easier to perform than an assay that exclusively

reflects a secretory process. Another important problem is

that the endogenous transport activity of the oocytes must

be low; otherwise the signal derived from the injected

mRNA will disappear in the background [118,119]. The

latter problem has clearly impeded the cloning of some

transport functions, like bile salt secretion, of which con-

siderable background is already present in the oocyte

[119]. Bile salt secretion as well as the secretion of non-bile

salt organic anions seems to be a function that is ubiqui-

tously expressed in many cell types including plants and

bacteria [ 120-122].

Once a cDNA encoding a putative transporter is ob-

tained, the sequence can be used to produce antibodies so

that the plasma membrane localization can be verified. In

addition the cDNA can be expressed in more suitable

mammalian cell types to verify its function [123]. Even at

this stage one has to be careful with interpretation of the

results. Using the suitable vectors and cell types one may

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 221

obtain extremely high expression levels of a single protein

and this may derange the target cell. Therefore the demon-

stration of a transport function by expression of a cDNA

per se is not sufficient and has to be accompanied by a

certain assessment of the specific activity of the protein

and this specific activity has to be in the same order of

magnitude as that in the normal hepatocyte.

2.4. The use o f animal models

Animal models have proven to be invaluable for studies

into hepatic transport processes. The advantage of animal

models is that transport can be studied in an integrated

model that is defined by the absence, malfunction or

overexpression of a single gene. These animal models are

especially valuable for the study of canalicular transport

mechanisms. Two types of animal models can be discerned

in this respect; animal strains with a spontanous mutation

that gives rise to disturbances in transport and transgenic

animal models. The TR rat (Section 3.3) and the LEC rat

(Section 3.3.6) are examples of the first group. These

animals exist within the population of control rats (like the

Wistar and Sprague Dawley colonies) and have been dis-

covered by serendipitous observation of phenotypic abnor-

malities. Even in models where the relevant gene and its

mutation are not known, as is the case in the TR- and the

LEC rat, a host of information can be obtained by compar-

ison of transport processes in control and mutated animals.

Using hepatocytes and plasma membrane vesicles from

TR- rats it was discovered that canalicular organic anion

transport is a primary active, ATP-dependent mechanism

[68,124]; the protein(s) involved in this process have,

however, still not been characterized. The mdr2 knockout

mouse (Section 5.3.2) is an example of a transgenic animal

model. In this model the situation is quite reverse in that

the mdr2 gene had been isolated and characterized already

more than ten years ago. Its function remained elusive

until the phenotype of the mdr2 knockout mouse demon-

strated the involvement of mdr2 Pgp in canalicular secre-

tion of phosphatidylcholine [125]. This gene was inacti-

vated in embryonic stem cells by homologous recombina-

tion of the normal mdr2 gene with a construct that lacked

the promotor region and the first two exons of the gene

(see Ref. [126] for this technique). After selection of

clones with an inactivated gene the embryonic stem cells

were injected into blastocysts which gave rise to chimeric

mice. Chimers in which the inactivated gene had under-

gone gerrnline transmission were used to breed hetero-

zygotes and these were in turn used to breed homozygotes.

Although these procedures are more and more becoming

standard techniques, it must be born in mind that the

production of such mouse models is laborious and is only

successful if a number of criteria is met with: the protein

involved should not have a vital function during embryo-

genesis and fetal development because otherwise no viable

animals will be obtained. Preferably, the gene to be dis-

rupted must have a single tissue-specific function; if the

protein has a ubiquitous expression and if its function is

vital in several tissues, the animal will suffer from a severe

pathology or even not be viable. In such cases primary and

secondary consequences of gene disruption will not be

easily discriminated. In addition, the function to be charac-

terized must preferably be mediated by a single gene

product. Thus if the function is mediated by a number of

homologous proteins with overlapping substrate specifici-

ties, the disruption of the gene of one of these proteins is

not likely to give rise to a characteristic phenotype. As an

example, knockout mice have also been produced for the

m d r l a gene in the mouse [127]. Because the m d r l a and

the m d r l b gene have a strongly overlapping substrate

specificity for amphipatic neutral and cationic compounds,

double knockouts for both genes will have to be produced

before information can be obtained about the involvement

of these P-glycoproteins in hepatobiliary transport of these

compounds. The mentioned criteria certainly hold for a

number of hepatocyte-specific (transport) functions so that

in the future more knockout mouse models may be ex-

pected in this field of research.

3. Hepatobiliary transport of organic anions

As far as transhepatic transport is concerned the cate-

gory of organic anions has to be divided into bile salts and

non-bile salt organic anions. Bile salts, quantitatively, are

by far the most important organic anions in bile and, in

addition, are transported by mechanisms different from

that of most other organic anions. Although the term

'organic anions' may suggest a homogenous group of

compounds, this notion is incorrect: compounds that are

traditionally included in this group vary widely in their

physicochemical characteristics and it is very difficult to

describe common features apart from their anionic groups,

Bile salts, in most mammalian species conjugated to glycine

or to taurine, display a single negative charge at physio-

logical pH. The presence of one, two or three hydroxyl

groups at different positions of the bile salt molecule lead

to a considerable difference in hydrophilicity of the vari-

ous bile salt species identified. Taurocholate, a trihydroxy

bile salt abundant in human and rat bile, has been used

mainly in transport studies. Most, but not all, cholephilic

non-bile salt organic anions have more than one negative

charge at physiological pH and are, to a certain extent,

amphiphilic. The latter feature seems to be important for

hepatic uptake while the multianionic ,character of a

molecule increases the efficiency of canalicular secretion.

Examples of endogenous non-bile salt organic anions are

bilirubin glucuronides, (oxidized and reduced) glutathione,

glucuronides of estrogens, leukotrienes, and sulfate and

glucuronide conjugates of thyroid hormone. Many exoge-

nous organic anions or anionic conjugates are secreted into

222 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

bile too; examples include some cefalosporins, glu- ou t

curonidated, sulfated and glutathione-conjugated farmaca S" . . . . . ...a

like the conjugates of aminacetophen.

3.1. Sinusoidal uptake of organic anions

Bile salts

Functionally, two different mechanisms have been dis-

cerned for uptake of organic anions: a Na+-dependent and

a Na+-independent system. The Na+-dependent system is

driven by the sodium gradient that is built up by the

activity of Na÷K+-ATPase which is exclusively present in

the basolateral membrane of the hepatocyte [128].

There is ample evidence that taurocholate is taken up by

hepatocytes in a Na+-dependent fashion [129-131]. How-

ever, non-bile salt compounds like bumetanide, phalloidin

are also substrates for sodium-dependent uptake [131,132].

Hepatic uptake of bile salts is affected by the length and

charge of the side chain as well as the number of hydroxyl

groups [131,133-135]. Conjugation of bile salts with tau-

rine increases the affinity for the uptake system. Reduction

of the number of hydroxyl groups on the steroid moiety

increases the inhibitory potency on taurocholate transport,

although the actual sodium-dependent uptake of unconju-

gated mono- and dihydroxy bile salts might be much less

efficient than that of taurocholate. Several compounds like

cyclosporin A and BSP were reported to inhibit tauro-

cholate transport non-competitively with low K i [131,136].

Conflicting data have been obtained as to whether the

sodium-dependent uptake involves an electroneutral or an

electrogenic mechanism. Neutral cotransport was sug-

gested by comparisons of sodium and taurocholate fluxes

in isolated hepatocytes [137], by the absence of a stimula-

tory effect of valinomycin and an outwardly directed K +-

gradient on the Na+-dependent uptake of taurocholate into

sinusoidal membrane vesicles [138] and by the lack of an

effect of taurocholate on membrane voltages and resis-

tances in hepatocytes [139]. Electrogenic transport is sup-

ported by the finding of Hill coefficients for sodium of

about 2 in isolated hepatocytes [129,140,141] and the

observation in sinusoidal membrane vesicles that permeant

anions stimulate the uptake of taurocholate [110]. More

recently, Lidofsky et al. added evidence to this by demon-

strating both in the perfused liver and in isolated and

cultured hepatocytes that uptake of taurocholate leads to

depolarization of the membrane potential [142]. Using

microelectrode impalement, Weinman and Weeks provided

additional evidence for the electrogenic character of sinu-

soidal taurocholate uptake. They observed taurocholate

induced depolarization under conditions of minimized cor-

recting K ÷ and C1- conductances in isolated rat hepato-

cytes [143]. Wehner [144], however, suggested from his

results that the depolarization observed in cultured hepato-

cytes is caused by an induced Na ÷ conductance. Using

microelectrode impalement in 24 h cultured hepatocytes he

observed that taurocholate-induced depolarization was not

in

cytosol

photoaffinity labeling

48 - 54 kDa

epoxide hydrolase

48 kDa 6 transmembrane

domains

1 Ntcp

39 kDa

7 transmsmbrane domains

Fig. 2. Postulated sinusoidal sodium-dependent uptake systems for tauro-

cholate (TC). The upper transporter represents protein(s) that were identi-

fied by photoaffinity labeling with azido-derivatives of taurocholate and

which have a molecular mass between 48 and 54 kDa [145-147]. A

transport function has never been directly demonstrated for these proteins

(stippled arrow for transport). The middle transporter represents the

protein isolated by the group of Levy [ 114,148-150], which has sequence

identity with micrsornal epoxide hydrolase. This putative transporter was

reconstituted but did not display strict Na + dependency (as indicated by

the stippled arrow for Na+-transport. The third transporter represents

Ntcp, which was recently cloned by the group of Meier [152,153,669].

Upon expression in oocytes this transporter displays strict Na+-depend -

ency. In vitro translation of the mRNA in the presence of microsomes

yields a protein of 39 kDa. However, it was shown very recently [669]

that this gene product has an apparent molecular mass of 51 kDa in rat

liver, probably due to more extensive glycosylation.

reversible upon withdrawal of the bile salt from the culture

medium.

At the protein level several basolateral candidate pro-

teins have been isolated and postulated to catalyze the

sodium-dependent uptake of taurocholate Fig. 2. Firstly,

photoaffinity labeling with reactive taurocholate deriva-

tives have lighted up protein(s) in the molecular mass

range of 48 to 54 kDa which were postulated (subunits of)

taurocholate transporting proteins [145-147]. These pro-

teins have, however, never been isolated and reconstituted

to demonstrate transporting activity in proteoliposomes.

The group of Levy et al. [114] has isolated a sinusoidal

membrane protein of 48 kDa by means of affinity chro-

matography on a glycocholate-column. Upon reconstitu-

tion this protein preparation gave rise to both Na+-depen -

dent and Na+-independent transport suggesting that trans-

port through this protein is not strictly dependent on

sodium cotransport. Subsequently, a monoclonal antibody

was generated against this preparation [148]. Immunopu-

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 223

rification of the 49 kDa protein from sinusoidal rat liver

membranes followed by reconstitution also demonstrated

taurocholate transport [149]. More recently, this group

discovered that the protein which reacts with this antibody

is similar if not identical to the microsomal epoxide hydro-

lase [150]. Although the extent of similarity has to be

further delineated these findings pose problems on the

interpretation of previous results. Evidently, it may be

postulated that epoxide hydrolase is localized both on the

plasma membrane and in the endoplasmic reticulum, which

has been demonstrated for other proteins. However, at the

same time two entirely different functions have to be

assigned to the protein, i.e., stereospecific hydration of

alkene and arene oxides to trans-dihydrols vs. Na+-depen -

dent transmembrane transport of bile salts. Epoxide hydro-

lase is a member of a group of very abundant + 50 kDa

proteins in the endoplasmic reticulum. Thus plasma mem-.

brane preparations may well be contaminated with this

protein and so may a glycocholate affinity purified prepa-

ration from such membranes. This may in turn indicate

that the antibody reacts with epoxide hydrolase instead of

the bile salt transporter. It remains to be explained why the

immunoaffinity purified protein displays bile salt transport-

ing properties. As has been demonstrated for a similar

situation in canalicular preparations, immunopurified

preparations may not be absolutely pure and may, probably

depending on the conditions of solubilization, contain con-

taminating proteins of similar molecular mass. Since the

microsomal epoxide hydrolase has been cloned [151],

transfection of of cells with this cDNA will reveal the

possible bile salt transporting capacities of this protein in

near future.

Recently, a Na+-dependent taurocholate uptake system

from rat liver has been expressed in Xenopus laeuis

oocytes [152] and a full-length cDNA was subsequently

cloned using this expression system [153]. This cDNA

encodes for a protein (called Ntcp for Na+/taurocholate

cotransport polypeptide) that has an (unglycosylated)

molecular mass of 33-35 kDa. Hydrophobicity analysis

reveals seven putative transmembrane domains and the

cDNA shows no homology with epoxide hydrolase. High

stringency blotting of several rat tissues only revealed

expression in the liver. Recently, the ileal Na+-dependent

bile salt transporter from the hamster was cloned and this

cDNA displays considerable amino acid sequence identity

and predicted structural similarity with Ntcp. When trans-

lated in the presence of microsomes, Ntcp cDNA gives rise

to a glycosylated polypeptide of about 39 kDa. This is

significantly lower than the molecular mass of the putative

transporters that were identified by photoaffinity labeling,

which are in the region of 48 kDa. The difference may be

explained by a more extensive glycosylation of Ntcp in

hepatocytes giving rise to a higher molecular mass. Indeed,

using recently developed antibodies against this protein,

Stieger er al. [669] demonstrated that the molecular mass

of the gene product in rat liver is 51 kDa. They also

demonstrated that the protein is specifically localized in

the sinusoidal membrane [669]. Taurocholate transport into

oocytes that had been injected with Ntcp cDNA was found

to be strictly Na+-dependent and was inhibited by tau-

rochenodeoxycholate and BSP, but not by taurodehydro-

cholate. Inhibition by the loop-diuretic bumetanide which

was also thought to be a substrate for the Na+-dependent

transporter [154] was relatively weak [15211. This adds

evidence to the results from Honscha et al. [155] who

showed that different fractions from size-fractionated rat

liver poly(A*)-RNA give rise to taurocholate and

bumetanide transport, respectively, when injected into

Xenopus oocytes. These results suggest that bumetanide

and taurocholate are taken up via separate Na+-dependent

transporters.

Using a probe complementary to the rat Ntcp, Hagen-

buch et al. [156] also cloned the homologous human cDNA

(NTCP) and mRNA from this clone injected in Xenopus

oocytes also gave rise to Na÷-dependent taurocholate up-

take in these cells. There is 88% similarity between the

human and rat amino acid sequences. The gene for NTCP

was localized on chromosome 14. The human transporter

has a distinctly higher affinity for taurocholate than the rat

protein (Kms 6.3 /xM and 25 #M respectively). The

inhibition pattern of a number of bile salts confirmed the

earlier observation that taurine-and glycine-conjugated bile

salts are better substrates than unconjugated bile salts

[156].

In an ontogenic study, Boyer et al. [157] showed that

the mRNA for Ntcp is absent through most of gestation

and is first detected toward the end of gestation on fetal

days 18-21. The expression of Ntcp further increased after

birth up to a 5-told higher level at adulthood compared to

the expression just before birth. This perinatal develop-

ment coincides with earlier transport studies by Suchy et

al. [158] who only found Na+-dependent bile salt transport

in basolateral plasma membranes prepared from fetal rat

livers just before birth. It was furthermore shown that Ntcp

mRNA is absent from HepG2 cells [157] and from dedif-

ferentiated primary hepatocytes (i.e., rat hepatocytes alter

three days culture [159]). These cells also lack functional

Na+-dependent taurocholate uptake. Altogether these data

suggest that Ntcp/NTCP is a major determinant of bile

salt uptake into hepatocytes.

Non-bile salt organic anions

Many non-bile salt organic anions such as bilirubin,

bromosulphthalein (BSP), indocyanine green (ICG),

pravastatin and dinitrophenylglutathione, are substrates for

one or more Na+-independent uptake mechanisms

[11,160-162]. For an extensive review on the uptake of

individual cholephilic organic anions see [I,10]. In many

studies BSP was used as a model compound for sinusoidal

uptake. The transport of this compound into rat liver

plasma membrane vesicles was shown to be Na~-indepen -

dent and electrogenic [163,164]. At present it appears that

224 R.P.J. Oude El/Prink et al. / Biochimica et Biophysica Acta 1241 (1995) 215 268

out in

• ' c-'oso'yt, s in t

BBBP O A ss kDa

OABP OA 8-subunit mitochondrial

F1 -ATPase

55 kDa

OA BTL 37 kDa

o a t p

O A 71 kDa

10 transmambrana domains

CI"

Fig. 3. Postulated uptake systems for sinusoidal sodium-independem

uptake of organic anions. The first two proteins (BBBP [ 113,154,165,166]

and OABP [112,168,169]) have never been reconstituted to demonstrate a

transport function. Bilitranslocase (BTL; [ 1 I, 174-176]) has been reconsti-

tuted and mediates electrogenic transport of BSP, but its sequence is

unknown. The last protein (oatp) has been cloned from both rat [177,178]

and human [670,671] liver and upon expression this 71 kDa protein

mediates high affinity, Na+-independent transport of BSP and bile salts.

Transport via this protein can be stimulated by CI-.

the assumption of a single Na+-independent system for

uptake of organic anions is an oversimplification. The

number of polypeptides involved in the uptake of non-bile

salt organic anions may very well be more than one and

overlapping substrate specificities may further confuse the

situation.

Several groups have postulated the identification of

polypeptides involved in basolateral transport of organic

anions (see Fig. 3). Using affinity chromatography with

immobilized bilirubin and BSP as ligands, a 55 kDa

protein was isolated from rat liver plasma membranes

[113,154]. Antibodies raised against this BSP/bilirubin

binding protein (BBBP) inhibited uptake of BSP and

bilirubin into isolated rat hepatocytes [165] as well as

HepG2 cells [166]. Thus far the protein has, however, not

been reconstituted to test its capacity for organic anion

transport.

Wolkoff et al. [112] also isolated a 55 kDa protein by

affinity chromatography on BSP-glutathione-agarose and

raised an antiserum against this organic anion binding

protein (OABP). Immunological studies with this anti-

serum revealed expression of the protein in many tissues

like heart, brain and intestine. Comparison of BBBP and

OABP by immunochemical techniques suggested that these

preparations represent different proteins [167]. Subse-

quently, anti-OABP was used for cloning of the cDNA for

this protein [168]. The sequence of the isolated cDNA

clone turned out to be identical to the /3-subunit of mito-

chondrial FrATPase. Though unexpected, this finding was

explained by demonstrating in an immunofluorescence

study that an antibody against the /3-subunit of mitochon-

drial FrATPase also reacted with the hepatocyte plasma

membrane. In HepG2 cells, however, only intracellular

localization of immunoreactive material was observed;

these cells were shown to lack chloride-sensitive, high

affinity uptake of BSP [169]. The OABP protein prepara-

tion has not been reconstituted to verify its transporting

capacity.

More or less independent from the characterization of

this polypeptide the same group characterized the mecha-

nism of BSP transport into cultured primary rat hepato-

cytes. They observed that uptake is chloride-dependent

[170], but not driven by a chloride-gradient [171,172]. The

presence of chloride ions strongly increased the affinity of

BSP binding to the cells [172]. Micromolar concentrations

of ATP in the medium rapidly and reversibly inhibited

BSP uptake by more than 60%, suggesting that this trans-

port is subject to regulation via purinergic receptors [173].

A third protein, termed bilitranslocase (BTL), has been

purified from rat liver plasma membranes. This prepara-

tion, which contains a single polypeptide of 37 kDa upon

SDS-PAGE, was used to raise monoclonal and polyclonal

antibodies. Furthermore, the protein was reconstituted in

liposomes. The proteoliposomes were shown to transport

BSP by an electrogenic mechanism [11,174]. BTL was

furthermore shown to be present in HepG2 cells by

Marchegiano et al. [175] who also observed saturable BSP

uptake in these cells.

The relative role of these proteins (BBBP, OABP and

BTL) in the uptake of organic anions by hepatocytes has

still to be demonstrated. In an attempt to address this

question, Torres et al, [176] recently measured the effect of

anti-BBBP and anti-BTL antibodies on BSP uptake into rat

liver plasma membrane vesicles. Transport was measured

both in the presence and absence of an inside-positive

membrane potential and it was observed that anti-BTL

antibodies inhibited the electrogenic uptake while anti-

BBBP antibodies inhibited the electroneutral uptake of

BSP. They extrapolated their results and calculated that

both BTL and BBBP are important for transport, the

relative importance being determined by the prevailing

BSP concentration (Km, BT L 5 /zM and Km,BBBP 20 /zM). This calculation does, however, not take into account the

fact that the hepatocyte membrane potential is negative

inside ( + 36 mY) and the possible presence of at least one

other transport system (see below).

Finally, a cDNA was recently isolated from a rat liver

library by Jacquemin et al. [177,178] encoding an organic

anion transporting polypeptide (OATP). Upon injection in

Xenopus laevis oocytes, mRNA derived from this clone

gave rise to Na+-independent BSP transport into these

R.P.J. Oude EIferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215 268 225

cells. Similar to what was noted with isolated hepatocytes

[172], BSP uptake in these transfected oocytes was Cl--de-

pendent: however, this dependency was only noted when

albumin-bound BSP was used as substrate and the effect of

chloride was only observed at low BSP concentrations

[178]. Maximal uptake velocity was identical in the pres-

ence and absence of albumin but the K m for BSP de-

creased more than ten-fold by the omission of albumin.

The effect of chloride may actually represent a stimulation

of the extraction of BSP from albumin. In addition to BSP,

OATP was capable of mediating Na+-independent

(tauro)cholate uptake. Surprisingly, this cDNA encoded a

polypeptide with a substantially higher molecular mass

than the putative transporters mentioned above. Cell-free

translation yielded a protein of 59 kDa and in the presence

of canine pancreatic microsomes a glycosylated product of

71 kDa was observed. Northern blot analysis of mRNA

from different tissues revealed expression in liver and

kidney. Low stringency blotting conditions also revealed

that multiple forms of OATP may be present in this tissue,

indicating that a family of homologous polypeptides could

be involved in the uptake process.

3.2. Transcellular transport of organic anions

lntracellular binding proteins and interaction with mem-

branes

After uptake across the basolateral membrane, organic

anions have to be directed to the canaliculus in order to

achieve vectorial secretion. Pharmacokinetic studies have

shown that the hepatocyte cytosol constitutes a storage

compartment for many organic anions [179]. This is due to

the presence of proteins with high affinity for organic

compounds. Three classes of proteins should be mentioned

in this context: glutathione S-transferases (including lig-

andins), 3-c~-hydroxysteroid dehydrogenase ( Y ' - b i n d i n g

protein) and fatty acid binding proteins [180]. Glutathione

S-transferase is present in the cytosol in concentrations up

to 0.2 mM and is the most important intracellular binding

protein for non-bile salt organic anions; Y'-binding protein

or 3-c~-hydroxysteroid dehydrogenase is an important

binder of bile salts. Binding of organic anions to these

cytosolic binding proteins may reduce the efflux rate from

the cells at both the sinusoidal and canalicular level and

also may reduce accumulation in intracellular membranes

and organelles.

Besides binding to proteins, bile salts and more hy-

drophobic organic anions can interact with intracellular

membranes. This may constitute an additional storage

compartment. Unconjugated bilirubin, which is very hy-

drophobic due to internal hydrogen bonding of the car-

boxyl groups, has been demonstrated to rapidly associate

with unilamellar phospholipid vesicles [181]. It was sug-

gested that lhe transfer of bilirubin and similar hydropho-

bic organic anions between intracellular membranes may

actually be an important mechanism of transport through

the cell [181].

Intracellular L'esicular transport

Vesicles play a role in a number of different hepatic

transport processes. The best characterized vesicular trans-

port through the hepatocyte is that of polymeric IgA.

Vesicular transport has, however, also been implicated for

biliary lipids, bile salts and a variety of other organic

amphipaths. Several groups have shown that secretion of

phospholipid and cholesterol in bile can be inhibited by

high doses of the microtubule assembly inhibitors

colchicine and vinblastine [182-188]. This issue will be

extensively discussed in Section 5.

A number of groups has demonstrated sensitivity of bile

salt and organic anion secretion to microtubule poisons

and this has been used as indirect evidence for the involve-

ment of intracellular vesicles [188-190]. Mori et al. [191]

have demonstrated that the canalicular secretion of indo-

cyanine green is inhibited by colchicine treatment, whereas

sulfobromophtalein secretion was unaffected. Dumont et

al. [192] reported that the biliary secretion of the glu-

tathione conjugate of diethylmaleate after peritoneal ad-

ministration of high doses of diethylmaleate is inhibited by

pretreatment of the rats with colchicine. A general conclu-

sion from these data must be, however, that the sensitivity

of bile salt and organic anion transport to colchicine

appears to occur only at high fluxes of these compounds.

Using an elegant approach with antibodies against bile salt,

the group of Erlinger [193,194] demonstrated the associa-

tion of bile salt with vesicles inside hepatocytes. Ayoma et

al. [ 195,196] reported the presence of several vesicle-medi-

ated pathways for organic anions. Notably phenol red in

their studies was transported by a colchicine-insensitive

but monensin-sensitive pathway [196]. At sufficiently low

concentrations, monensin rather selectively disrupts the

Golgi network [197]. Since this machinery is certainly

involved in the assembly of transport proteins destined for

the canalicular membrane, the results of Ayoma et al. [196]

could very well be explained by a primary effect on the

amount of transport protein. In view of the fact that phenol

red is highly soluble and the inhibitory effect was greater

at a high dose of the organic anion, which obviously

requires a higher capacity of transport protein, this expla-

nation seems plausible. By incubating hepatocytes with

monochlorobimane, which, after intracellular conversion to

glutathione-bimane, is an excellent substrate for the

canalicular multispecific organic anion transporler (see

Section 3.3), Oude Elferink et al. [198] demonstrated the

presence of intracellular vesicles that accumulate this fluo-

rescent organic anion. Since this phenomenon was virtu-

ally absent in the hepatocytes from TR- rats it was

concluded that this vesicular accumulation involves the

canalicular organic anion transporter. They hypothesized

that these vesicles are involved in regulation of the amount

226 R.P.J. Oude Elf erink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

of canalicular organic anion transporters on the membrane

rather than in vectorial transport of organic anions to bile.

It cannot be excluded that data which suggest a vesicu-

lar pathway actually represent translocation of bile salt or

organic anions into vesicles by de novo synthesized or

endocytosed carrier proteins which can be transported to

the plasma membrane. For the housekeeping of proteins

and lipids in the different cellular compartments cells have

extremely active vesicle trafficking. According to the group

of Hubbard [199-203] proteins which eventually end up in

the canalicular membrane first travel by a vesicular path-

way to the sinusoidal membrane where they are sorted,

endocytosed and transported to the canalicular membrane.

It is not yet clear whether all proteins take this route and

recent data suggests that even the proteins that do follow

this pathway do not all travel at the same rate and thus

could be present in different kinds of vesicles [199]. When

proteins such as the canalicular bile salt carrier also first

travel to the sinusoidal membrane then, depending on the

rates of vesicular transport, the proteins must be present

inside the cell for an appreciable period. It is therefore not

surprising that the presence of bile salts can be demon-

strated in vesicular structures. In addition it may be that a

fraction of the canalicular transport proteins is stored in

intracellular compartments for recruitment during in-

creased need for secretory capacity (see Section 6.1.1)

The conclusion from the data available on putative

intracellular transport of organic anions and bile salts via

intracellular vesicles must be that under physiological con-

ditions this represents at best a minor pathway for the

secretion of these compounds.

Sinusoidal effiux o f organic anions

Subsequent to their uptake into the hepatocyte, com-

pounds can be secreted back into the sinusoidal space.

Especially in the absence of substantial canalicular secre-

tion this is an evident process. Sinusoidal effiux has been

demonstrated for compounds like bilirubin [170], DBSP

[38] and harmol-sulfate [204]. It may be mediated by the

same carrier that catalyzes uptake of these compounds but

indirect evidence exists that it may involve separate mech-

anisms [38]. Plasma albumin strongly binds a large amount

of organic anions and thereby stimulates sinusoidal efflux

[205]. This phenomenon may at least partly explain the

predominant sinusoidal secretion of sulfate conjugates:

compared to glucuronide conjugates, sulfate conjugates

bind more avidly to albumin than the glucuronides [206].

Sinusoidal effiux of GSH is a physiologically highly

relevant process. Over 90% of GSH in the circulation

originates from the liver [207] and this plays an important

role in detoxification and protection against oxidative stress

[208]. GSH effiux can be inhibited by a number of organic

anions including BSP-glutathione and bilirubin [209]. Si-

nusoidal GSH transport is bidirectional and driven by the

plasma membrane potential [210-212]. Recently, Fernan-

dez-Checa et al. [213] isolated a mRNA size fraction from

rat liver which, upon injection in Xenopus laevis oocytes,

gives rises to GSH transport with sinusoidal character-

istics.

3.3. Canalicular transport

The secretion across the canalicular membrane is rate-

limiting in the overall secretion of many cholephilic com-

pounds and represents the most important concentrative

step. Indeed, for all classes of organic compounds (cations,

anions and bile salts) this step creates a larger concentra-

tion gradient than that observed over the sinusoidal/baso-

lateral membrane. For example, for organic anions like

dibromosulphthalein (DBSP) a bile/l iver concentration ra-

tio of 100-1000 can be reached [179]. It was assumed for

many years that canalicular transport of both bile salts and

organic anions was primarily driven by the plasma mem-

brane potential [214-216] The concentration gradients are,

however, too large to be accounted for only by the mem-

brane potential difference over the canalicular membrane,

even when a micellar sink in the biliary compartment is

postulated [217]. For amphiphilic compounds this sink

could be constituted by the biliary lipids. In recent years it

has become clear that for most classes of compounds

primary active (i.e., ATP-dependent) transport systems are

present in the canalicular membrane [2,182,218,219]. The

characterization of these transport systems has been im-

peded for a long time, primarily by the inaccessibility of

this small membrane domain. The purification of canalicu-

lar membrane vesicles [104,220] made direct transport

experiments possible. Finally, the discovery of mutant

animals with defective canalicular organic anion secretion

has stimulated the characterization of separate transporters

for organic anions and bile salts [221]

Distinct canalicular transport mechanisms fo r bile salts

and organic anions

Early studies have suggested that separate canalicular

secretion routes may exist for bile salts and organic anions.

The evidence was mostly based on studies employing

techniques such as competitive inhibition in either whole

animals or isolated perfused liver preparations. However,

interpretation of the data was often difficult because inter-

actions may occur at many levels when transport is mea-

sured in the whole organ or animal. Interaction occurs at

the level of sinusoidal/basolateral uptake and at the level

of intracellular binding, intracellular metabolism and

canalicular secretion (for extensive review of these studies

see Ref. [l ]).

The first direct evidence for distinct canalicular trans-

port systems for bile salts and non-bile salt organic anions

came with the recognition of the human Dubin-Johnson

syndrome; this is a rare congenital chronic conjugated

hyperbilirubinemia [222,223]. The hepatic clearance of

bilirubin and other cholephilic organic anions, like BSP

and indocyanine green, is impaired in these patients [224],

R.P.J. Oude Elferink et al. /Biochimica et Biophysica Acta 1241 (1995) 215-268 227

whereas bile salt clearance is normal [225]. Mutant Cor-

riedale sheep have an inherited defect that closely resem-

bles the Dubin Johnson syndrome: these animals also show

a decreased secretion of conjugated bilirubin, BSP,

iopanoic acid and indocyanine green, while taurocholate

transport is normal [226,227]. The discovery of a more

suitable experimental animal model for the Dubin-Johnson

syndrome, the GY or TR- rat, made it possible to charac-

terize this system at the biochemical level. These animals

have a phenotype which strongly resembles that of the

Corriedale sheep and the Dubin-Johnson syndrome.

G Y / T R - rats have impaired biliary secretion of bilirubin

glucuronides, BSP, ICG and many other organic anions

(see below), while taurocholate secretion is almost normal

[221,228-230].

Canalicular bile salt transport (Fig. 4)

Until recently canalicular bile salt transport was charac-

terized as a membrane potential-driven process [138]. A

protein of 100 kDa was partially purified from canalicular

rat membranes that, upon reconstitution, gave rise to po-

tential dependent taurocholate transport into proteo-

liposomes [116,231 ]. It was, however, clear that the magni-

tude of the membrane potential over the canalicular mem-

brane ( _ 35 mV) is not high enough to explain the high

bile salt concentration gradient that is actually observed in

vivo. Adachi et al. [106] were the first to demonstrate that

uptake of taurocholate into rat canalicular membrane vesi-

cles could strongly be stimulated by the addition of ATP.

This observation was soon confirmed by a number of other

groups in rat [232-234] and human liver membranes [108].

These studies point to a system that is specific for bile

salts with a K m for taurocholate in the low micromolar

range. No other nucleotides than ATP can drive transport

and vanadate inhibits transport at low concentrations (90%

inhibition at 50 p~M) suggesting z-phosphate transfer from

ATP to the transporter-polypeptide. ATP-dependent tauro-

cholate transport could be inhibited by glycocholate and by

unconjugated as well as taurine-conjugated di-and trihy-

droxy bile salts but not by (tauro)dehydrocholate [233,234].

The more recent detection of considerable ATP-depen-

dent taurocholate transport in canalicular membrane vesi-

cles questions the physiological role of the previously

demonstrated electrogenic transport. It could be that the

latter experiments have demonstrated the residual electro-

genic transport of the ATP-dependent transporter in the

absence of ATP. On the other hand it is possible that two

distinct bile salt transporters are present in the canalicular

membrane. A recent study by the group of Meier [105] has

brought considerable clarity in this confusing situation.

Using free flow electrophoresis they have further purified

the conventional canalicular membrane preparation which

still contains considerable amounts of endoplasmatic retic-

ulum. In this way two subfractions were obtained one of

which was 2-3-fold depleted of markers of the endoplas-

mic reticulum and 2.5-fold further enriched in the canalicu-

lar marker alkaline phosphatase. The other subfraction was

further enriched for ER-markers and virtually devoid of

canalicular enzyme activity. Removal of contaminating ER

resulted in a complete loss of electrogenic taurocholate

transport activity from the enriched canalicular membrane

fraction. In contrast, ATP-dependent transport activity re-

(~) ER

~ kDa ?

o TC

cytosol

T C --------_____ ?

ATP

T C - -

ADP

canaliculus

, ATP

i

ecto-

ATPase

ADP

cBAT

polypept ide

u n k n o w n

Fig. 4. Postulated transporters involved in canalicular secretion and intracellular sequestration of taurocholate. In the canalicular membrane an

ATP-dependent taurocholate transport activity is present [106,232-234], which has not been characterized at the protein level. It is distinct from the

previously characterized 100 kDa protein, which transports taurocholate in an ATP-independent, electrogenic mechanism [116,231]. The latter protein is

now thought to be localized in intracellular membranes, notably endoplasmic reticulum [105]. In addition Sippel et al. [240-242] have purified a putative

ATP-dependent bile salt transporter which turned out to be identical with ecto-ATPase. This protein is thought to mediate ATP-dependent bile salt

transport, but this function is not dependent on the ecto-ATPase function.

228 R.P.J. Oude Elferink et a l . / Biochimica et Biophysica Acta 1241 (1995) 215-268

mained associated with both the ER and the canalicular

membrane fraction. These data suggest that the electro-

genic transport activity is a distinct entity which is absent

from the canalicular membrane and resides in the endo-

plasmic reticulum, whereas the ATP-dependent transport

system appears to be the primary mediator of canalicular

taurocholate transport. Strikingly, substantial ATP-depen-

dent transport was also present in the fraction that was

de-enriched for canalicular membranes. This strongly sug-

gests that ATP-dependent transport is also present in intra-

cellular compartments. This observation may shed light on

the discussion of regulation of transport by recruitment of

transporters from intracellular stores (see Section 6). The

experiments by Kast et al. [105] must lead to reinterpreta-

tion of previous experiments from the same group concern-

ing a 100 kDa protein that would be responsible for

canalicular bile salt secretion. This protein was purified

from conventional canalicular membrane preparations and

gave rise to potential-dependent taurocholate transport upon

reconstitution in liposomes [116,231]. The conclusion from

the most recent results must be that this protein is not a

canalicular bile salt transporter.

Several groups have attempted to identify the canalicu-

lar protein that is responsible for ATP-dependent tauro-

cholate transport. Muller et al. [232] have attributed this

primary active transport to a 110 kDa protein which is

recognized by an antiserum against gp110. Immunoprecip-

itation of solubilized canalicular membranes with this anti-

serum gave rise to a fraction that, upon reconstitution,

catalyzed ATP-dependent taurocholate transport. Using this

antiserum for purification of this heavily glycosylated pro-

tein from rat liver, Becker et al. [235] demonstrated by

sequencing of ten internal peptides that the protein was

identical or at least highly similar to the cloned rat liver

ecto-ATPase [236]. Surprisingly, however, immunoprecip-

itation of this protein was not associated with removal of

any ATPase activity. Moreover the purified gp110 did not

display ATPase activity although it had been characterized

as an ATP-dependent bile salt transporter [232]. The pro-

tein displayed considerable microheterogeneity, probably

due to differences in glysosylation. Complete deglycosyla-

tion yielded a polypeptide of 48 kDa and these character-

istics strongly resemble those of canalicular proteins that

had been identified with antibodies raised in a similar

fashion: HA4 [237], C-CAMI05 [238], and another gp110

[239]).

More recently, Sippel et al. purified a 110 kDa protein

from rat canalicular membrane vesicles by bile salt affinity

chromatography. Sequencing of this protein revealed iden-

tity with the cloned ecto-ATPase. This was surprising

since the sequence of this protein predicts a single mem-

brane-spanning domain [236], which is unlikely for a

transporter protein. In order to determine its function as a

bile salt transporter, the cDNA for ecto-ATPase was trans-

fected into COS cells and these cells were used to demon-

strate taurocholate efflux [240]. Using a truncated cDNA

the authors furthermore observed that the cytoplasmic tail

of the protein is essential for its transport function. In a

subsequent report the same group [241] showed, using

modified cDNA constructs, that the exoplasmic ATPase

function of the protein was not essential for transport.

Nevertheless extracellular ATP stimulated taurocholate ef-

flux. In addition, depletion of intracellular ATP abrogated

efflux. On the other hand protein kinase C-dependent

phosphorylation of the cytoplasmic tail appeared to be

essential for activity and tyrosine kinase dependent phos-

phorylation would regulate the activity [242]. On the basis

of these results Sippel et al. [241] suggested that the

mechanism of taurocholate transport involves secretion of

ATP into the canaliculus (e.g., by the putative ATP-chan-

nel function of mdrl P-glycoprotein [243]) followed by

subsequent energization of transport by the protein that

does not involve the ecto-ATPase activity. It is difficult to

envisage how this function, which needs the presence of

ATP in the canalicular lumen, should be combined with

the massive ATPase activity that is present on the canalic-

ular membrane. As mentioned above, all these studies rely

on an assay of taurocholate transport in transfected COS

cells. In this assay radioactive taurocholate is allowed to

associate with the cells and subsequently release of label

from the cell is measured. It is difficult to envision how a

substantial amount of the added taurocholate can enter

cells by simple diffusion. Specific transport was defined by

the difference in taurocholate release in the presence and

absence of DIDS, but the obtained data in absence and

presence of DIDS were never shown. The assay that was

used did not unequivocally demonstrate transmembrane

transport of taurocholate using criteria like saturability,

time-dependency and temperature-sensitive efflux in the

presence and absence of DIDS. A further concern is that

the data conflict with the observation that ATP-dependent

taurocholate transport is only observed in inside-out vesi-

cles in the presence of ATP. In addition, the study of Kast

et al. [105], described above, demonstrated that subfrac-

tionation of canalicular plasma membranes dissociated the

immunodetectable presence of ecto-ATPase from ATP-de-

pendent taurocholate transport indicating that ecto-ATPase

at best represents a minor component of canalicular ATP-

dependent bile salt transport. Altogether it is, as yet, quite

unclear which polypeptide(s) are responsible for the ATP-

dependent translocation of bile salts into the canaliculus.

The canalicular multispecO~c organic anion transporter:

cMOAT

Using the TR rat as a standard model for the defi-

ciency of canalicular organic anion transport it was possi-

ble to define a group of substrates that is transported into

bile via a putative transport system that is absent or

inactive in this animal. Although this transport system has

not yet been characterized at the protein level nor at the

cDNA level, it has been termed canalicular multispecific

organic anion transporter (cMOAT). The characteristics

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 229

and substrate specificity are thus far entirely defined by its

absence in the TR rat. Upon description of the TR- rat,

two other rat strains, the GY [244] and the EHBR rat [245],

have independently been discovered and characterized by

two laboratories. These groups have initially investigated

different aspects of the secretory defect. Fortunately, cross

breeding experiments with the different strains have shown

that the offspring is always jaundiced and therefore the

conclusion can be drawn that the mutation in these animals

is either identical or at least allelic [246,247]. Thus, results

concerning phenotypical characteristics of these different

strains may be highly interchangable, except that the dif-

ferent genetic backgrounds may cause different observa-

tions between the EHBR rat (from the Sprague-Dawley

strain) and the TR and GY strains (from the same

original Wistar colony). Using model organic anions like

dinitrophenyl-glutathione (GS-DNP) and BSP, which are

poorly secreted by the G Y / T R rat, the mechanism of

transport has been studied in different experimental sys-

tems ranging from the intact animal [221,228,244,248,249],

isolated perfused liver [229], isolated hepatocytes

[246,250.251], hepatocyte couplets [252] and isolated

canalicular membrane vesicles [247,252-254].

Kobayashi et al. [255] were the first to show that

transport of GS-DNP in mixed rat liver plasma membrane

vesicles could be stimulated by ATP. The kinetics of

transport (linear over a period of two hours) strongly

deviated from that observed in later studies using more

purified canalicular vesicles, in which it was shown that

ATP-dependent transport can only be measured for a short

period ( < 10 rain) due to high rate of ATP hydrolysis in

these membranes, even in the presence of a regenerating

system. ATP-dependent GS-DNP transport was subse-

quently observed in freshly isolated hepatocytes [251] and

purified canalicular membranes [256] Similarly, the uptake

of cysteinyl-leukotriene [253], bilirubin-glucuronide [254],

BSP [247], and p-nitrophenyl-glucuronide [257] could

strongly be stimulated by the addition of ATP. Nishida et

al. [247] directly demonstrated that enrichment of inside-out

oriented vesicles by antibody induced density perturbation

increased the specific ATP-dependent transport activity of

the preparation with BSP as substrate. This strongly sug-

gests that the observed ATP-dependent transport acts as an

extrusion system in the canalicular membrane.

The work of several groups [247,252-254,256-258]

has demonstrated that the putative transport system

cMOAT, as studied in membrane vesicles, is temperature

dependent, osmotically sensitive and saturable with Km'S

for the organic anions in the low micromolar range (BSP:

31 /zM; GS-DNP: 71 /zM; cysteinyl-leukotriene C4:0.25

/.tM; p-nitrophenyl-glucuronide: 20 /zM; and bilirubin

glucuronide: 71 /zM). When transport is measured in

canalicular plasma membrane vesicles, the K m for GS-

DNP (71 #M) was found to be about 15-fold lower than

the apparent K m that was observed in transport experi-

ments with isolated hepatocytes (0.9 mM [250]). This

suggests that in the intact cell almost all GS-DNP is

present in a (protein-)bound form.

The transport is ATP-dependent and no other nu-

cleotides are able to drive the transport nor could this be

achieved with non-hydrolyzable analogues of ATP. Trans-

port could be inhibited by DIDS, vanadate and probenecid.

Furthermore the addition of protonophore (CCCP) or the

ionophore valinomycin did not inhibit transport [247,254]

indicating that transport was not secondarily driven by

ATP-induced H+-or K+-gradients. In fact, valinomycin

further induced ATP-stimulated uptake of bilirubin glu-

curonide suggesting that transport is electrogenic.

Nishida et al. [254] demonstrated that normal canalicu-

lar plasma membrane vesicles transport bilirubin glu-

curonide by both an ATP-and a membrane potential-de-

pendent mechanism. In TR- membrane vesicles the ATP-

dependent mechanism was absent, but the membrane po-

tential-dependent mechanism was retained. The observa-

tions of Nishida et al. are confirmed by more recent

experiments in TR rats and perfused livers [228], which

indicate that for some organic anions, like bilirubin ditau-

rate, a low affinity secretion pathway is preserved in TR

rats. It is, as yet, not clear whether these two secretory

systems represent separate systems of which apparently

only one is defective in the TR rat. Another plausible

explanation is that one system (the putative cMOAT) can

transport compounds both in an ATP-and a membrane

potential-dependent fashion and that in the TR rat the

protein is mutated in such a way that only the ATP-hy-

drolysing modality is defective. In this respect it is also of

importance to note that~ in the in vivo situation, the biliary

output of bilirubin in TR / G Y rats is only mildly de-

creased [221,259]. Apparently these animals are capable of

significant bilirubin excretion, albeit at strongly increased

serum bilirubin concentration. This also suggests that a

low affinity secretion pathway for bilirubin is still present

in the canalicular membrane of these animals. The studies

mentioned above concerning ATP-dependent bilirubin

transport contrast with Adachi et al. [260] who reported

that uptake of bilirubin-glucuronide in canalicular vesicles

is not ATP-but bicarbonate-dependent. This group also

recently reported that transport of bilirubin glucuronides in

canalicular membrane vesicles l¥om the EHBR rat is nor-

mal in contrast to the data from other groups [261], At

present it is difficult to reconcile these findings.

Studies with transport deficient rats clearly demon-

strated that the ATP-dependent transport of the compounds

described above is mediated by the same transporter or by

members of the same family of transporters. The particular

transport processes could be related to the findings of

Nicotera et al, [262,263] who had demonstrated earlier that

both GSSG and GS-DNP are able to stimulate an ATP-hy-

drolyzing activity in isolated rat liver plasma membranes.

Zimniak et al. [264] showed that this activity could be

purified as a 37 kDa protein from canalicular membranes,

using antibodies directed against affinity-purified GS-

230 R.P.J. Oude Elferink et a l . / Biochimica et Biophysica Acta 1241 (1995) 215-268

DNP-stimulated ATPase from erythrocyte membranes

[265]. However, this polypeptide was found to be normally

present in TR-rat liver plasma membranes and this makes

a relation, if any, between this ATPase and cMOAT

hypothetical. A possible explanation could be that the

isolated polypeptide may represent a subunit with ATPase

activity that is part of a larger complex necessary for

transport. It was suggested by Zimniak and Awasthi [218]

that the 37 kDa polypeptide may be a proteolytic fragment

of a larger 85-90 kDa protein. Very recently, the group of

Zimniak [266,267] reported that the above-mentioned puri-

fied 90 kDa protein displays organic anion transport capac-

ity upon reconstitution in liposomes. They furthermore

found that transport could be stimulated 2-fold by protein

kinase C-mediated phosphorylation. This would be in line

with the observation of Roelofsen et al. [268] that canalicu-

iar organic anion transport in hepatocytes can be either

directly or indirectly stimulated by protein kinase C.

In transport studies with isolated canalicular plasma

membrane vesicles little or no inhibition of bilirubin glu-

curonide or BSP transport could be observed by the addi-

tion of 10-100 /zM taurocholate [247,254]. Furthermore,

bile salt transport is not defective in canalicular membrane

preparations from TR- rats [247] (see Fig. 5). In contrast,

o e o E

0 u) (/) 40

(0

"6 E " 2 0

8O

131 E " " 60 O. z a

d) 40

E o. 20

0

20

15 ¢D

E

"3 E Q.

G

Wistar . A

, 0

0 S 10 IS 20

Wi . ~ t

--,-o

0 $ 10 15 20 4o

• m

6o

B

70

Wistar C

0 5 10 1S 20 50 60

T i m e ( ra in)

5o

o o

TR-

40

30

2O

10

o

o s lo

6o T R

4O

0

TF Jo

Io

!o

A

_ _ - - - - !

| , * -

1 S 20 60 70

r . j

B

0 S 10 IS 20 40 50

e j

C

o 5 1o 1§ 20 ,so 60

T i m e (ra in)

Fig. 5. ATP-dependent transport of organic anions (GSSG and GS-DNP) and bile salt (taurocholate) in canalicular membrane vesicles from Wistar (left)

and T R - (right) rats. Vesicles were incubated with the radioactive ligand (upper panel: 3H-GSSG; middle panel: 3H-GS-DNP; lower panel:

3H-taurocholate) and an ATP-regenerating system. Note that organic anion uptake is completely absent in TR- membranes while taurocholate transport is

fully functional. The apparent overshoot in the uptake of various compunds is caused by the depletion of ATP in the incubation medium. Data from Ref.

[667]

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 231

Table 1 Organic anions (or anionic complexes) of which biliary transport is reduced in GY/TR- rats

Number of Reference negative charges

Endogenous compounds conjugated bilirubin 2 [229] glutathione GSH 2 [250] glutathione GSSG 4 [251] cysteinyl-leukotrienes 2/3 [248] triiodothyronine-glucuronide 2 [652] coproporphyrin 1 2 [221 ] Bile salt conjugates cholate 3-O-glucuronide 2 [270] lithocholate 3-O-glucuronide 2 [270] nordeoxycholate 3-O-glucuronide 2 [249] tauro/glycolithocholate 3-sulfate 2 [244] taurochenodeoxycholate 3-sulfate 2 [244] nordeoxycholate-3-sulfate 2 [249] Exogenous compounds ceftriaxone 2 [653] ampicillin 2 [275] carboxydichlorofluorescein 2 [ 124] dibromosulfophthalein 2 [229] bromosulfophthalein-glutathione 4 [229l dinitrophenyl-glutathione 2 [250] glutathionyl-bromoisovalerylurea 2 [654] naphtol- 1-glucuronide 1 [332] indocyanine green 2 [228] gadolinium-ethoxybenzyl-DTPA 2 [655] Metals zinc [295] copper (acute i.v. administation) [295] manganese [296]

basis of the reduced secretion these compounds are consid-

ered as substrates for the putative cMOAT. Compounds

from this list which deserve special attention will be

discussed below.

Bilirubin. The most prominent substrate for cMOAT is

conjugated bilirubin. In the absence of active cMOAT,

bilirubin accumulates in the blood and its removal is partly

by renal secretion [269]. It should be noted that in the

absence of active ATP-dependent cMOAT the accumula-

tion of conjugated bilirubin in serum is relatively moderate

(33 p~M). This is caused by renal secretion and by a

considerable residual biliary secretion of bilirubin-

glucuronide. Recently Jansen et al. [228] have studied the

biliary secretion of a number of organic anions at different

infusion rates in the TR . It was observed that the residual

secretion in TR rats of some putative cMOAT substrates,

like bilirubin-ditaurate is quite high when high infusion

doses are applied and that reduced secretion is only ob-

served at low doses. It is possible that this residual biliary

secretion is caused by an alternative organic anion trans-

porter. This could be related to the observed ATP-indepen-

dent transport in TR-ra ts that may be driven by the

membrane potential [254]. Alternatively, the mutation in

the transport system in TR- rats may be such that only the

affinity for substrates like bilirubin ditaurate is strongly

decreased. This residual activity apparently cannot be used

to transport glutathione-conjugates because the secretion of

these compounds is low at all doses in the TR- rat

[228,250].

many studies demonstrated mutual transport inhibition of

several non-bile salt organic anions and in some cases this

inhibition was competitive. This suggests that cMOAT has

a rather broad substrate specificity. Since in T R - rats the

canalicular taurocholate transporter operates normally and

the secretion of compounds like bilirubin-glucuronide and

BSP is severely impaired, one can draw the conclusion that

the bile salt transporter does not accommodate these non-

bile salts as substrates. Whether the converse is also true,

unconjugated bile salts as substrates for cMOAT, is more

difficult to prove since an animal model with absent

canalicular bile salt transport is not available. It should be

stressed that the separation between bile salt and organic

anion is probably not complete. Inhibition of bilirubin-dig-

lucuronide [254] and GS-DNP [256] transport by high

concentrations of taurocholate (500-600 /zM) in canalicu-

lar membrane vesicles have been observed.

Substrates for the canalicular multispecific organic anion

transporter

Table 1 shows a list of compounds of which biliary

secretion is significantly reduced in G Y / T R - rats. On

Divalent anionic bile salt conjugates. Bile salts can be

sulfated or glucuronidated at the 3-OH position. In particu-

lar for the secondary, more hydrophobic bile salts like

lithocholate this may represent a detoxification route. This

conjugation adds an additional negative charge to the

molecule as to make it considerably more hydrophilic.

Strong evidence exists that in contrast to the monovalent

bile salts, the secretion of these divalent anionic bile salts

occurs via cMOAT. The secretion of sulfated taurolitho-

cholic acid, sulfated glycolithocholic acid and sulfated

taurochenodeoxycholic acid is considerably impaired in

the G Y / T R rat, whereas the secretion of their unsulfated

counterparts is normal [244]. The secretion of sulfated

taurocholate, however was almost normal It was also

shown that the 3-OH glucuronide of nordeoxycholic acid,

which is a divalent anion, is severely impaired in the

G Y / T R - rat, in contrast to the carboxyl-linked glu-

curonide which is a monovalent anion [249]. Similarly, the

3-OH glucuronide of lithocholate undergoes little biliary

secretion in the G Y / T R rat whereas the respective car-

boxyl-linked glucuronides are efficiently removed via bile

[270]. These data have led to the hypothesis that a sub-

strate for cMOAT must have at least two negative charges

to be transported by this system [230,246,250,270]. Fur-

thermore, in view of the fact that the more hydrophilic

232 R.P.J. Oude Elferink et al. / Biochimica et Biophysica A cta 1241 (1995) 215-268

divalent anionic bile salts are almost normally transported

in the TR- rat, the substrate apparently should also have a

core with a certain hydrophobicity. Indeed almost all

molecules of which transport is impaired in the TR- rat

have at least two negative charges (see Table 1).

Anionic antibiotics. The third generation cephalosporin,

ceftriaxone, is rapidly secreted into bile of both man and

rat [271,272]. The biliary secretion of ceftriaxone in the rat

is inhibited by ursodeoxycholate and this has led Xia et ai.

[273] to the suggestion that ceftriaxone may be secreted

via the canalicular bile salt transporter. However, secretion

of ceftriaxone is virtually absent in TR rats. In the

normal rat about 75% of the infused dose of ceftriaxone

was recovered in bile and no competition could be ob-

served by simultanous infusion of taurocholate. From this

it may be concluded that for its hepatobiliary secretion

ceftriaxone is not transported by the bile salt transporter

but via cMOAT [274]. An explanation for the inhibitory

effect of ursodeoxycholate may be that at high dose of

ursodeoxycholate divalent anionic metabolites like the 3-

O-sulfate or the 3-O-glucuronide are formed that may

inhibit ceftriaxone secretion. Verkade et al. [275] have

shown that biliary secretion of ampicillin is strongly im-

paired in the G Y / T R - rat indicating that this antibiotic is

most probably also secreted via cMOAT. Many other

antibiotics like rifamycin and amoxycillin are concentrated

into bile [276] but the responsible transport system has not

been characterized yet. It is interesting, however, to com-

pare the pharmacokinetic properties of a large group of

related compounds like the cephalosporins. Bergan [277]

described that from a group of 30 cephalosporins only 5

(cefotiam, cefoperazone, cefbuperazone, ceftriaxone and

latamoxef) were secreted into bile for more than 10%.

There was no clear relation between the extent of binding

to serum proteins and biliary secretion; high serum binding

only tended to increase the serum half-life of the

cephalosporin. This may indicate that the specificity of

hepatic carriers involved in uptake and secretion may play

a crucial role in the choice between renal and biliary

secretion. Interestingly, most of the cephalosporins that

undergo biliary secretion have two negative charges on the

molecule suggesting that this may be an important deter-

mining parameter. However, several of the compounds that

are secreted via the kidney exclusively also bear a double

negative charge, indicating that the prerequisites for biliary

secretion are clearly more complex.

The physiological role o f biliary GSSG secretion

When the liver is subjected to oxidative stress intra-

cellular GSH is rapidly oxidized to GSSG. This can occur

under conditions of reperfusion after ischemia or when

certain oxidants like paraquat or t-butylhydroperoxide reach

the liver in sufficiently high concentrations [278]. The

formation of GSSG is accompanied by its concentrative

secretion into bile and this transport competes with that of

glutathione-conjugates like GS-DNP [279]. Furthermore

this secretion is ATP-dependent and absent in TR rats

suggesting that this transport occurs via cMOAT [251].

The paradoxical consequence of high affinity canalicular

GSSG transport is that the hepatocyte extrudes GSH-

equivalents (in the form of GSSG) under conditions of

increased need. Indeed, inhibition of GSSG secretion ap-

pears to protect hepatocytes from injury during oxidative

stress. It was demonstrated that incubation of hepatocytes

in 10 mM fructose inhibits GSSG extrusion by a drastic

reduction of the cellular ATP level [251]. Silva et al. [280]

showed that this treatment prevented nitrofurantoin-in-

duced cytotoxicity in isolated hepatocytes, suggesting that

the extrusion of thiol equivalents is deleterious for the

cells. Therefore, questions arise concerning the function of

biliary GSSG secretion. It is well known that GSSG may

form mixed disulfides with thiol groups in proteins thereby

influencing their activity [281]. In this respect the extrusion

may function to eliminate potentially toxic GSSG to pre-

vent inactivation of enzyme activities. However, the activ-

ity of the GSSG-reductase cycle in hepatocytes, which

reduces GSSG at the expense of NADPH, is much higher

than the activity of GSSG secretion [282,283]. As a conse-

quence, the G S H / G S S G ratio in the hepatocyte under

normal conditions and during oxidative stress will be

determined by the balance between GSSG formation and

reduction by GSSG-reductase while canalicular efflux will

have little or no influence. Recently, Chung et al. [284]

reported that GSH under normal conditions inhibits the

activity of the GSSG reductase. They hypothesized that the

efflux of GSSG during oxidative stress principally serves

to decrease the total glutathione content of the cell, thereby

relieving GSH-mediated inhibition of the GSSG-reductase.

In this way the actual activity of the reductase would be

increased by net extrusion of thiol equivalents. At the same

time this mechanism may protect the cell from too high

NADPH consumption through the GSSG-reductase cycle

which would be needed to restore the G S H / G S S G ratio at

high total glutathione contents. Further evidence lbr this

hypothesis has as yet to be produced.

The secretion of GSH and GSH-associated compounds

Reduced glutathione (GSH) is present in millimolar

concentrations in bile of rats and a number of other species

[285]. Since the GSH concentration in plasma is in the

micromolar range, there is a large concentration gradient

from plasma to bile and this may drive water flux through

the tight junctions of the hepatocytes. A number of elegant

studies by Ballatori et al. [286,287] have indeed demon-

strated that GSH is a major determinant of the bile salt-in-

dependent bile flow in rats. In contrast to normal rats,

reduced glutathione is virtually absent from TR- bile. This

may explain the significant reduction of the bile salt

independent bile flow in this mutant animal [221]. The

R.P.J. Oude ElJerink et al. / Biochimica et Biophysica Acre 1241 (1995) 215-268 233

absence of GSH in TR bile may lead to the suggestion

that GSH secretion into bile occurs via cMOAT. There are,

however, arguments contradicting this suggestion.

Firstly, in contrast to all other organic anions that are

secreted into bile via cMOAT, there is no uphill gradient

for GSH from hepatocyte to bile. The intracellular concen-

tration of GSH in the rat varies between 5 and l0 mM and

that in bile between 1 and 5 mM [285]. However, this

probably represents an underestimation of the amount of

GSH that is secreted by the hepatocyte. Gamma-gluta-

myltranspeptidase, which is an apical ectoenzyme, is pre-

sent on the canalicular membrane and on the apical mem-

brane of bile duct epithelial cells. Thus, GSH present in

bile is actively degraded during its passage through the

biliary tree. Indeed, inhibition of gamma-GT by treatment

of rats with acivicin leads to a substantial increase in the

biliary GSH concentration, but never exceeds that of the

hepatocyte cytosol [288].

Secondly, several groups [289,290] have demonstrated

that the transport of GSH into canalicular membrane vesi-

cles is not stimulated by ATP but driven by the membrane

potential. Furthermore, Fernandez-Checa et al. [289]

showed that GSH transport is not deficient in canalicular

membrane vesicles from the EHBR ( ~ G Y / T R - ) rat.

Thirdly. several reports on organic anion transport in

canalicular membrane vesicles demonstrate that millimolar

concentrations of GSH do not, or hardly, inhibit the

(cMOAT-mediated) transport of typical organic anions like

BSP [247], GS-DNP [256] and bilirubin-glucuronide [254].

Although these arguments seem straightforward indica-

tions for the existence of a GSH transport system that is

distinct from ATP-dependent cMOAT it remains to be

demonstrated why GSH is completely absent from TR rat

bile. Fernandez-Checa showed that GS-conjugates in rather

low concentrations cis-inhibit GSH uptake in canalicular

vesicles. Therefore, they hypothesized that GSH secretion

in T R rats may be inhibited by certain endogenous

cMOAT substrates that have accumulated in TR- rat

hepatocytes [289].

The mechanism of canalicular GSH secretion may be

elucidated in the near future by the recent isolation of a

cDNA for the putative canalicular GSH transporter. With

Xenopus oocytes as expression system Yi et al. [291] were

able to isolate a 4 kb cDNA which gives rise to a 96 kDa

polypeptide that catalyzes GS-BSP insensitive effiux of

GSH from the oocytes. In previous studies this group had

already shown that sinusoidal GSH effiux can be inhibited

by BSP-glutathione whereas canalicular GSH transport is

insensitive to this compound. Furthermore, canalicular GSH

efflux could be induced by phenobarbital treatment

[210,289,292]. Northern blot analysis revealed that the

mRNA for this protein is present in liver, kidney, brain,

intestine and lung but not in heart. In addition, the mRNA

level could be induced by phenobarbital treatment. On

basis of these characteristics the authors conclude that

isolated cDNA probably represents the canalicular GSH

transporter. In an earlier study Femandez-Checa et al.

[213] had shown that, in addition to 4 kb mRNA for the

putative canalicular transporter, a 2.5kb mRNA is present

in liver that encodes for a BSP-glutathione inhibitable

GSH transporter. This mRNA may represent the sinusoidal

GSH transporter.

Healey metal ions as complexes with GSH. A number of

(heavy) metals has been shown to undergo substantial

biliary secretion. Since some of these appear to be depen-

dent on biliary secretion of glutathione they will be dis-

cussed in this section; for extensive review see [293].

It has been postulated that copper is secreted into bile as

a GSH-complex. Alexander and Aaseth [294] proposed this

mechanism on basis of the observation that depletion of

hepatic GSH leads to a marked decrease in biliary copper

secretion. However, in G Y / T R rats, which do not se-

crete GSH in their bile, secretion of endogenous copper

into bile is not impaired [295,296], implying that the

secretion of endogenous copper is independent of GSH

secretion. On the other hand, such a GSH-dependent path-

way may be of importance for the removal of copper

during i.v. copper administration [295,296]. In this model

of acute copper loading, a rapid and a slow phase of

elimination can be discerned in normal rats; the rapid

phase is absent in the mutant animals. These findings

indicate that at least two pathways must exist for biliary

copper secretion, one GSH-independent and another one

GSH-dependent and responsible for elimination of acutely

administered copper. The latter pathway does not accom-

modate copper that accumulates in the liver during dietary,

chronic copper overload. Data from several studies

[297,298] have suggested that biliary copper secretion, in

particular during hepatic copper overload, involves lysoso-

mal sequestration and subsequent exocytosis into bile. It

was very recently shown by Dijkstra et al. (J. Clin. Invest.,

in press) that rat liver membranes contain an ATP-depen-

dent, saturable, high affinity transport system 10r copper.

Its cellular localization remains, however, to be estab-

lished. In patients with Wilson's disease hepatobiliary

copper elimination is impaired and copper accumulates in

the liver. Recently, a candidate gene for Wilson's disease

was discovered and this gene encodes a P-type ATPase.

The cellular localization of this protein is not yet known

but this will shed light on the sequence of events in biliary

copper secretion. Also the characterization of a rat model

(the LEC rat, see Ref. [299]) will be of great value in the

near future in this field. This inbred mutant strain of rats

was originally isolated from a closed colony ,of Long-Evans

rats. This strain, associated with spontaneous hepatitis,

exhibits excessive hepatic copper accumulation and low

copper incorporation into ceruloplasmin [2!)9]. It was re-

cently shown that biliary copper secretion is also impaired

in these rats [300]. Thus, these rats have a phenotype

highly similar to that of patients with Wilson's disease.

Very recently, it was also shown that the defective gene in

234 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

this strain cosegregates with the rat homologue of the

human ' Wilson gene' [301 ].

There are indications that the canalicular secretion of

zinc also involves a GSH-dependent mechanism. Alexan-

der et al. [294,302] demonstrated that hepatic GSH-deple-

tion leads to a reduction in biliary zinc elimination. Fur-

thermore Houwen et al. [295] showed that biliary zinc

secretion is severely decreased in the G Y / T R - rat. The

biliary secretion route, however, does not play an impor-

tant role in the overall homeostasis of this metal [293].

There is ample evidence that the biliary secretion of

mercury and methylmercury is a GSH-dependent process

[303-305]. Recently, Dutczak et al. [306] substantiated this

contention by looking at the transport of a glutathione-

methylmercury complex in canalicular plasma membrane

vesicles. They demonstrated temperature-sensitive trans-

port into an osmotically active space. Transport was elec-

trogenic but could not be stimulated by ATP. The transport

was cis-inhibited and trans-stimulated by GSH. In con-

trast, inhibition by GSSG and a number of GS-conjugates

was very weak or absent. These data suggest that glu-

tathione-methylmercury is secreted into bile by a GSH-

specific carrier and not by cMOAT. It is interesting to note

in this respect that this complex is monovalent. Other

metal-GSH complexes like those of inorganic mercury,

lead [307] and cadmium [308] are thought to form multiva-

lent GSH-complexes and these complexes may reach bile

via cMOAT.

Oxoanions. When arsenite is injected into rats about 24-

37% of the drug is recovered in bile within 2 h [309]. This

is accompanied by a very strong (20-30-fold) increase in

the secretion of GSH in bile [310,311]. Upon administra-

tion of radioactive arsenite to the isolated perfused rat liver

Anundi et al. [312] analysed the secreted arsenite by thin

layer chromatography and found that the major part of the

radioactive arsenite was associated with a ninhydrin posi-

tive spot eluting together with glutathione. This is in

accordance with a recent NMR study by Delnomdedieu et

al. [313] in which the spontaneous in vitro association

between 3 molecules GSH with 1 molecule arsenite was

demonstrated. Similarly, the metabolite dimethylarsinite

formed a 1:1 complex with GSH. It was also demonstrated

that organic anions like BSP and indocyanine green (ICG)

are able to inhibit biliary arsenite secretion [310-312].

Recently Gyurasics et al. [314] reported similar results for

bismuth and antimony. Taken together, these results sug-

gest that arsenite (and bismuth and antimony) may un-

dergo canalicular secretion as a GSH complex, possibly

via cMOAT.

Ubiquitous expression o f organic anion secretion

Although the efficient concentrative secretion of amphi-

pathic anions seems to be a dedicated function of the

hepatocanalicular membrane, there is ample evidence that

highly similar transport systems are present in a wide

variety of cell types. It is already known for a long time

that GSSG [315] and GS-DNP [316] are pumped out of

erythrocytes and erythrocyte ghosts and it was demon-

strated that this process is ATP-dependent [317]. Similarly,

GS-DNP tranport can be observed in erythrocyte mem-

brane vesicles and this transport is competitively inhibited

by GSSG [318]. Thus, there are striking similarities in

transport characteristics between hepatocanalicular and

erythrocyte organic anion transport systems. The GSSG-

dependent ATPase activity in the rat liver plasma mem-

brane demonstrated by Nicotera et al. [262,263] had simi-

lar characteristics as that in the erythrocyte membrane

[319]. Furthermore Heijn et al. [320] demonstrated that

transport of GSSG in erythrocyte membrane vesicles is

inhibited by a very similar spectrum of organic anions

(including divalent sulfated and glucuronidated bile salts)

as the canalicular organic anion transport in hepatocytes

[246]. GS-DNP transport in erythrocytes from TR rats

was studied by two groups and it was found that the

activity was only slightly [320] or not [321] reduced as

compared to normal rats. Similarly, Board et al. [321]

detected no difference in the GS-DNP transport between

erythrocyte membranes from Dubin-Johnson patients and

normal subjects. Thus, it may be suggested that, although

the transporters in erythrocyte and liver canalicular mem-

branes resemble each other quite strongly functionally,

they may be derived from different genes. Nevertheless, it

is interesting to follow the achievements in the purification

of the erythrocyte transporter. A comprehensive review

concerning this subject has been published [218]. Unfortu-

nately, the results so far are rather confusing: the group of

Kondo et al. [322] reported the purification of a het-

erodimer consisting of a 62 kDa and 82 kDa protein which

has ATPase activity and is able to transport GSSG upon

reconstitution. On the other hand, the group of Awasthi et

al. has isolated a GS-DNP stimulated ATPase with a

postulated molecular mass of 38 kDa. Therefore, in con-

trast to what one would expect, GSSG and GS-DNP seem

to be transported by entirely different transport systems in

erythrocytes, although mutual inhibition of transport by

these substrates has been demonstrated in both liver [279]

and erythrocytes [320]. The interpretation of the putative

role of the GS-DNP-ATPase was further complicated by

the recent observation [323] that GS-DNP ATPase isolated

from human erythrocytes is also capable of transporting a

typical 'mdrl P-glycoprotein substrate', doxorubicin. It

was also shown that various other amphipathic substrates,

that are transported by mdrl P-glycoprotein, were able to

stimulate the ATPase activity of this isolated protein. From

these observations it was concluded that this putative

organic anion transporter actually translocates an even

broader spectrum of compounds including uncharged and

cationic amphipaths [323]. These data contrast with obser-

vations made with liver canalicular membrane vesicles

where no inhibition of ATP-dependent GS-DNP transport

is observed by compounds like doxorubicin [247,254].

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 235

Similarly, daunomycin does not inhibit canalicular GS-

DNP secretion by intact hepatocytes [246].

ATP-dependent transport of GSSG has also been re-

ported in the isolated perfused rat heart [324,325] and that

of GSSG and GS-conjugates, including leukotriene C 4, in

heart sarcolemma vesicles [326]. In a comparative study

with heart sarcolemma and liver plasma membrane vesi-

cles very similar K ms (150 nM and 250 nM respectively)

were found for LTC 4 transport in addition to a very similar

inhibition of LTC 4 transport by a spectrum of glutathione-

conjugates [326]. Since LTC 4 is synthesized by mast cells

(as well as several other blood cell types) Schaub et al.

[327] investigated transport in plasma membrane vesicles

of a mastocytoma cell line. They described a transport

system for LTC 4 that could be inhibited by GS-DNP and

that had a K m for LTC 4 similar as that in liver cell

membranes but with a much lower K m for ATP. More

recently, this group attempted to identify the polypeptide

that is responsible for LTC 4 transport in mastocytoma cells

by photoaffinity labeling with radioactive LTC 4. They

observed incorporation of radioactivity in a 190 kDa mem-

brane glycoprotein and speculated that the transporter might

be related to MRP (for multidrug resistance related pro-

tein; see Section 4) which has a similar molecular mass.

This contention was recently further substantiated by

showing that membrane vesicles from HL60 /ADR cells

which overexpress MRP display strongly increased trans-

port activity for LTC 4 and GS-DNP [328]. Very recently,

MiJller et al. [329] reported very similar results with an-

other MRP-overexpressing cell line as well as with MRP-

transfected cells. Zaman et al. [330] showed that MRP is

also ubiquitously expressed in normal human tissues, but

particularly low in liver. Thus endogenous MRP expres-

sion may be responsible for organic anion secretion in

many tissues, but it is unclear whether the low MRP

expression in liver could explain the very active organic

anion secretion into bile.

It was also reported that GS-DNP is secreted by CaCo 2

cells. This cell line represents a model for polarized intesti-

nal epithelium. It was observed that, in analogy with

hepatocytes, the secretion of the glutathione-conjugate was

highly assymetrical and dependent on the intracellular

ATP concentration. However, in contrast to the situation in

liver, secretion was detected both at the apical and basolat-

eral pole of the cell [331]. This is in accordance with a

study by de Vries et al. [332] who showed in the isolated

perfused rat small intestine that secretion of the organic

anion, 1-naphthol-/3-glucuronide, occurs both at the lumi-

nal and at the serosal side. Furthermore, they found that

the secretion of this compound was normal in the isolated

intestine of the TR rat whereas biliary secretion was

strongly reduced.

Taken together, it may be concluded from these obser-

vations that although highly similar organic anion secre-

tion systems are present in a wide variety of cell types, it is

not very likely that these transporters are derived from the

same gene as that encoding the putative canalicular multi-

specific organic anion transporter.

4. Hepatobiliary transport of organic cations

During evolution, organisms have been exposed to many

cationic food components. Examples are plant alkaloids of

which many are readily absorbed and can be extremely

toxic due to interaction with neural and hormonal receptor

systems and/or genetic material. In the evolutionary pro-

cesses, excretory organs such as kidney, gut and liver

became adequately equipped with detoxifying mecha-

nisms~ These include metabolic conversion processes and

secretory systems for cationic compounds to remove such

potential life-threatening agents from the body [333,334].

At the same time, selective cationic transport mecha-

nisms for essential nutrients such as choline, certain vita-

mins, and nucleosides are necessary to maintain hepatic

cell function and the production of vital physiological

substrates. Against this background, it is not surprising that

multiple carrier-mediated transport systems for cationic

compounds evolved at the hepatic uptake and excretion

level.

Apart from the many transport processes in the sinu-

soidal and canalicular domains of the hepatocyte plasma

membrane, intracellular transport sites have been identified

in organelles such as endosomes, lysosomes, mitochondria

and probably the cell nucleus [333,335].

Classical physiology postulates organic: cations as a

homogenous class of agents that are transported by single

transport systems at the uptake and secretory level [334]. It

became clear however in the past two decades that organic

cations encompass a very heterogenous group of agents. At

least five different uptake mechanisms have been postu-

lated to date for exogenous cationic compounds (mostly

drugs) and at least two different translocation processes are

supposed to be present at the excretion level [333,335,336].

In addition, multiple transport processes have been identi-

fied for the abovementioned endogenous cations. The latter

systems, in general, poorly accommodate the exogenous

organic cations, although overlapping substrate specificity

may exist. The relative contribution of these multiple

cation transporters to the overall uptake and secretion rate

of a given compound is depending on the (unbound)

concentration of the particular agent in relation to the

affinity for the various transporting modalities.

4.1. Sinusoidal uptake o[ organic cations

The overall picture emerging from many studies using a

variety of experimental systems is that multiple mecha-

nisms exist for the hepatic uptake of organic cations.

Widely overlapping substrate specificity may exist and,

depending on the concentration of the organic cation stud-

ied, often more than one system will contribute to the

236 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

Type TT compounds ~ N C H 3 0"'~''*~ 0 ~ :~c.~

- . . . - , A t . . c._c. . o ' y . . r , .o.3

i0 I . . . . 0 i ~ ! v HOH2C-CH CH I

; i N ~ CH?. OH 0 H3C-C-O H H3C-C-O H CH =CH-CH2- H ~ N ~

PANCURONIUM ORG 6368 2 I I H3C~ ALCURONIUM OCH3

T y p e I" c o m p o u n d s

CH3 Q CH~ N-CH-CH-CH-N-CH S / ~ N +1

2 Z 2 3 --CH-CH-N-CH CH 2 I I 3

N-METHYLJMIPRAMINE THIAZINAMIUM

0

CH 3 CH 3

d-TUSOCURARINE

~7"-~CN3 CH 3 ~ ' ~ I I

~,,L_yc% c%

HEXAFLUORENIUM N-METHYLDEPTROPINE

C2H 5 C'H7 C4H9 O / C 2 H ' ' +, +1 +1 __O._11 :N , -CzH 5 C2Hs-~-C2H5 C3H7-~ -CH3 C4H9-~ -CH3 H2N C-NH-CHz-CH2 i

CH 3 C3H7 C4H9 'CzH 5

TRIETHYLMETHYLAMMONIUM TRIPROPYLMETHYLAMMONIUM /RIBUTYLMETHYLAMMONIUM PROCAINAMIDE ETHOBROMIDE

Fig, 6. Chemical structure of cationic drugs that are taken up in the liver by two separate (type 1 and type 2) transport systems. Type 1 compounds are

either hydrophilic/aliphatic organic cations or aromatic compounds with a clear separation of ring strucures and the positively charged nitrogen group.

Type 2 compounds are generally more bulky molecules with one or two cationic groups enclosed in (or close to) ring structures.

overall uptake rate of given compound. Nevertheless, the

individual transport systems exhibit quite different features

with regard to maximal transport rate, energization and

ion-dependency as well as to the influence of inhibitors.

Non-carrier-mediated mechanisms in hepatic uptake

Although carrier mediated transport is the major mecha-

nism for uptake of organic cations, it should be empha-

sized that passive uptake can occur for very lipophilic

organic cations. Examples are the fluorescent probe rho-

damine B [337,338] and the anticholinergic drug dep-

tropine [339] (see for structure Figs.Fig. 6 and Fig. 7 7).

As the consequence of its passive transfer mechanism,

uptake rate of rhodamine B is not inhibited by other

cationic compounds [340]. Due to the efficient uptake

process of rhodamine B, sharp tissue concentration gradi-

ents are observed for this agent in whole liver. The initially

observed periportal to perivenous concentration gradient

reverses in time due to a time-dependent redistribution and

a higher binding capacity for dye in the pericentral (zone

3) hepatocytes [337,338]. In contrast, the hydrophilic fluo-

rescent compound lucigenin (see Fig. 7), that cannot pass

the sinusoidal plasma membrane by passive transport or

carrier-mediated mechanisms, is probably internalized via

a vesicular uptake process [341,342]. This was inferred

from fluorescence-microscopy studies showing a clustered

intracellular distribution of lucigenin with dots of fluores-

cence close to the plasma membrane. Partial inhibition by

microfilament inhibitors such as cytochalasin B and noco-

dazole, as well as by competitors for membrane adsorption

such as poly(L-lysine) [341] indicated a process of adsorp-

ORGANIC CATIONS

CH3 (~)JC2 H 5

. ~ C O O H Rhodamine B

~ (lipophilic)

CH3 I

Lucigenin

[hydrophilie)

i CH 3

O II

H3C-C-O c ®

II

O (lip o pl~ilic/hydr ophilic)

Fig. 7. Bulky organic cations that largely differ in lipophi]icity and are

taken up in the liver by entirely different mechanisms: rhodamine B is

sufficiently lipohilic to pass membranes by lipoid diffusion [337]. In

contrast, two quarternary nitrogen centers make lucigenin much more

hydrophilic and this compound can only be taken up by an endocytic

mechanism [341,342]. Vecuronium has an intermediate lipophilicity and

in the deprotonized form is an amphipathic organic cation that is trans-

ported via caiTier-mediated mechanisms [336].

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 237

Sinusoid

03 OC ~

Cl', HC~

0 OC~_~

© OC:---'

Bile canaliculus

© ©

~h Sinusoid . - . _ i _ ~ . _ . . ~

4Tp 7

®

OC ~

OC"

@ Pgp CI-

Fig. 8. Tentative scheme depicting carrier-mediated transport of organic

cations at the sinuoidal and the canalicular level of the hepatocyte. Five

uptake systems are indicated for exogenous cations: type 1 carrier system

(1), type 2 carrier system (2), absorptive endocytosis (3), passive lipoid

diffusion (4) and paracellular permeation (5). Transport processes for

choline and thiamine exhibit Na+-dependency in cells and for NMN a

H+-dependency in basolateral membranes was shown. Sinusoidal mem-

branes contain the N a + / K + pump and the Na+/H+-antiporter. The

canalicular membrane possesses a CI- /HCO~- antiporter and at least

two organic cation carriers: mdrl P-glycoprotein and a carrier for rela-

tively hydrophilic organic cations.

tive endocytosis. Such a phenomenon has also been de-

scribed for basic drugs at the renal tubular level [343].

Carrier-mediated uptake systems for organic cations With regard to saturable hepatic uptake of endogenous

cations (see Fig. 8Fig. 8), various carrier-mediated trans-

port processes have been described, in studies in whole

(isolated perfused) liver and isolated hepatocytes. In some

of these studies it is difficult to distinguish the membrane

transport process from further metabolic steps since many

of these compounds are rapidly converted within the hepa-

tocyte. Yet, in most of these studies evidence was obtained

that carrier-mediated uptake and accumulation of the par-

ent compounds in the cells can occur in the absence of

sequential metabolic steps [336]. Kinetic analysis often

demonstrated the presence of high and low affinity uptake

processes for the individual compounds. The outcome of

inhibition studies with metabolic inhibitors or structurally

related cationic compounds is therefore dependent on the

range of substrate concentrations tested.

Choline uptake in hepatocytes was claimed to be (at

least partly) Na + dependent, could not be inhibited by

tetra-ethyl-ammonium but was strongly inhibited by hemi-

cholinium-3 [344]. Although thiamine seemed to competi-

tively inhibit choline transport, choline inhibited thiamine

uptake in a non-competitive manner [345]. Three indepen-

dent studies with isolated hepatocytes clearly showed

Na+-dependency of (high affinity) thiamine uptake [346-

348]. This process that was poorly inhibited by tetrameth-

ylammonium but in contrast strongly decreased by diben-

zyl-dimethyl ammonium [345,349] as well as other more

lipophilic ammonium and phosphonium derivatives

[349,350]. A study with the thiamine derivative dimethal-

lium [351], in which replacement of the hydroxyethyl

group by a methyl group prevents phosphorylation, showed

carrier-mediatedneur uptake dissociable from the phospho-

rylation step.

In general, caution should be exercised with regard to

such ion-dependency studies: replacement of Na + by

potassium, lithium or choline can produce many other

effects. These may include a direct influence on the cell

membrane, plasma membrane depolarization (high K+),

and substrate competition (choline).

In a more recent study on thiamine transport with

hepatocyte basolateral membrane vesicles Na+-depend -

ency of thiamine uptake could not be dete, cted [352]. In

contrast an electroneutral thiamine/H + exchange system

was demonstrated. This process was inhibitable by ammo-

nium ions, choline and imipramine but not by the endoge-

nous cations N-methylnicotinamide (NMN) and exogenous

cations such as PAEB, TBuMA and vecuronium [352]. An

electroneutral H+-antiport system was also detected for

NMN in the same membrane preparation [353] (see Fig.

8). Choline, TBuMA, PAEB and vecuronium did not

significantly affect pH-dependent transport of NMN [354],

whereas acetylcholine, spermine, hamaline and imipramine

were clearly inhibitory [353].

With regard to cartier-mediated uptake of exogenous

compounds studies have been focused around a number of

model cations. Tetraethylammonium (TEA)uptake in ba-

solateral membrane vesicles was claimed to be pH-inde-

pendent in one study [352] and found to be stimulated by a

negative membrane potential [352]. The uptake rate of

TEA was cis-inhibited by PAEB, TBuMA and vecuronium

[354] but interestingly not by NMN [354,355]. Contrasting

observations were made in two uptake studies in basolat-

eral plasma membrane vesicles with TEA [355] and TBuMa

[356] from which involvement of an electroneutral proton-

antiport translocation mechanism was concluded.

Transstimulation of TEA transport was demonstrated for

thiamine while cimetidine and quinidine inhibited this

counter transport [357]. Since thiamine showed cis-inhibi-

tion and trans-stimulation, involvement of the earlier men-

tioned thiamine transporter might play a role. Alterna-

tively, cross-contamination of the basolateral plasma mem-

238 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

brane fraction with endosomal/lysosomal membranes may

have occurred. In these organelles H+-antiport of quater-

nary drugs has been reported [358].

It is important to note here that cis-inhibifion does not

necessarily prove that a common carrier system is in-

volved: cationic substrates could inhibit TEA transport

without being transported by the same carrier. Such a

phenomenon also explains that uptake of small (type I)

cations is strongly inhibited by bulky (type II) organic

cations but not vice versa [336]. Only cis inhibition com-

bined with trans stimulation may give a clue for a com-

mon transporter. The abovementioned vesicle studies indi-

cate that at least two cation transporters are involved in

uptake of TEA and TBuMa in basolateral plasma mem-

brane vesicles. For TEA an electrogenic cation/cation

exchanger is operating [354,355] along with an electroneu-

tral H+-antiport system [355].

The characteristics of TEA transport in basolateral

membrane vesicles are only partially in line with hepato-

cyte uptake studies for the cationic drugs, azido-PAEB

[357], TBuMa [359] and vecuronium [360]. For all of these

compounds at least two uptake systems could be kineti-

cally demonstrated. These studies showed no Na+-or pH-

dependency of uptake; neither was evidence for membrane

potential driven uptake found. For TBuMA uptake the

latter observations was corroborated by direct microelec-

trode studies [359]. In the abovementioned basolateral

membrane vesicle studies, a negative membrane (diffusion

potential) did not drive uphill TEA transport [354,357].

The stimulatory effect of an intravesicular negative mem-

brane potential was therefore interpreted with the involve-

ment of a conductive pathway [354,355]. Whether such a

mechanism does also operate for larger organic cations

remains to be established.

In a study in isolated hepatocytes it was shown that

various classical metabolic inhibitors clearly affect ATP

synthesis and largely lower cellular ATP content [358].

These agents significantly (but not completely) inhibited

initial uptake rate of TBuMA [358,359] and vecuronium

[360]. Part of the effects of metabolic inhibitors on uptake

rate may also be due to interference with cellular seques-

tration steps such as electrogenic uptake into mitochondria

and proton antiport in endosomes and lysosomes (see

Section 4.2). The lack of influence of ATP and ATP-gen-

erating systems on uptake of TBuMa in basolateral plasma

membrane vesicles suggests that this transport is not di-

rectly ATP-dependent [356].

In conclusion: hepatic uptake of cationic drugs seems to

occur by at least two facilitated diffusion carrier processes

that are Na+-independent and of which one may be stimu-

lated by proton gradients. The latter process is not identical

to thiamine and NMN proton antiporters. The electrogenic-

ity of the uptake systems for exogenous cations currently

remains unresolved. ATP dependency could not be demon-

strated and facilitated (carrier-mediated) diffusion mecha-

nisms are the most likely candidates.

Heterogeneity in uptake systems for cationic drugs

The cationic drugs as a whole also show clear hetero-

geneity in hepatocyte uptake mechanisms. The relatively

small cations such as TEA [354,355], TBuMA [359], PAEB

[336] and its azido analogue APM [357] are transported by

a system that is inhibited by choline and lipophilic (amphi-

pathic) cations but not by cardiac glycosides and bile salts

[336,361,362]. They were categorized as type 1 organic

cations and comprise either aliphatic quaternary ammo-

nium compounds or molecules in which the cationic amine

group is at some distance from the aromatic ring structure.

Larger cationic drugs with the positively charged group

included in or situated close to large, aromatic ring struc-

tures (called type II compounds) behave differently: hep-

atic uptake is not affected by a large excess of choline or

the type I compounds but can be largely blocked by

cardiac glycosides and bile salts. This is explained by

competitive inhibition for a multispecific system that rec-

ognized bulky amphipathic compounds independent on

their charge [361 ]. It is of interest that hepatocyte uptake of

both type I and type II organic cations is stimulated by

inorganic counter anions [359,360] such as HCO 3 and I-.

This effect was speculated to be due to an improved

presentation of the cations to the membrane carriers via

ion-pair formation rather than to facilitation through

cation/anion symport [359].

Photoaffinity labeling and expression cloning of potential

carriers for uptake of organic cations

Preliminary photo affinity labeling studies with azido-

procainamide methyliodide (APM) [357,363] showed in

plasma membrane fraction of rat liver predominant label-

ing of 50 kDa and 72 kDa proteins while labeling studies

with the bulky type II compound azo-N-pentyl-deoxy-

ajmalinium (APDA) showed labeling of 48 kDa and 50

kDa proteins [364]. Incorporation of label in the APM

studies was inhibited by type II compounds (an interaction

also found in isolated hepatocytes) while labeling of poten-

tial carrier proteins with APDA was strongly reduced by

cardiac glycosides and bile salts (see above). Prior photo-

affinity labeling of hepatocytes with APDA reduced the

Vmax without changing the K m of N-propyl-deoxy ajma-

linium (NPDA) uptake into hepatocytes indicating that part

of the carriers was inactivated through covalent labeling.

More definitive identification of the putative carrier pro-

teins for organic cations is awaited.

Expression cloning of organic cation uptake carrier

protein has recently be attempted independently by two

groups [365,366] in studies with TEA and TBuMA respec-

tively. A 1.8-3.0 KB mRNA fraction seems to encode for

proteins that mediate TEA uptake in Xenopus laevis

oocytes. However, enrichment compared with total m-RNA

is only 5-fold and the transport signal compared to controls

is quite weak. Consequently complete sequencing of the

genetic code was not yet feasible. Cloning, sequencing and

large scale synthesis of such proteins may enable reconsti-

R.P.J. Oude Elf'erink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 239

tution and further characterization at the molecular level.

Among others, ATP-dependency and potential regulation

mechanisms could be studied with such techniques.

4.2. Intracellular sequestration of organic cations

Following uptake at the sinusoidal domain of the hepa-

tocyte, organic cations can reach high concentrations in the

cytoplasmic compartments [333-336,367]. Transmem-

brane concentration gradients of 15-fold and more have

been calculated, both for monovalent [368] and bivalent

positively charged model compounds [369]. For many

compounds this value largely exceeds the gradient that

would occur by passive equilibration of cations according

to the membrane potential ( - 35 mV in hepatocytes). It is

important to note that cytoplasma/bloodplasma concentra-

tion gradients under steady state conditions, even corrected

for extracellular and intracellular protein binding, do not

merely reflect the uptake process since they are the resul-

tant of the relative rates of uptake, biliary excretion and

sequestration into organelles [369]. The latter processes

may provide sink conditions especially if hepatic uptake

rate is relatively low. For some organic cations the sinu-

soidal uptake process is rate limiting in overall hepatobil-

iary transport [368,369]. In such cases cytoplasma/blood

plasma concentration gradients below unity have indeed

been reported [368,369]. Real uphill transport against an

electrochemical gradient of the cationic species however,

is difficult to prove since many of these studies rely on

subcellular distribution studies after whole liver homoge-

nization. Since considerable accumulation of organic

cations in various types of organelles can occur, the cyto-

plasmic cytosol concentrations are overestimated, at least

to some extent, since redistribution from particulate frac-

tion to the cytosol fractions cannot entirely be prevented.

Uptake of cationic drugs in endosomes / lysosomes

Cell fractionation studies of livers taken from animals

that were injected with organic cations in vivo revealed

subce]lular distribution patterns that are entirely different

from the patterns obtained after addition of these com-

pounds to homogenized livers in vitro [368-372]. In par-

ticular, enrichment of the mitochondrial and lysosomal

fractions was detected after in vivo administration of the

cations [368,372]. Lysosomal/endosomal accumulation

was confirmed by electronmicroscopy of electron-dense

precipitates of such agents [373] as well as by more recent

uptake studies in acidified multivesicular membrane prepa-

ration [370]. It was shown that organic cation transport is

partly ATP-dependent (see Fig. 9Fig. 9) and sensitive to

proton-dissipating agents such as monensin as well as to

H+-ATPase inhibitors, These data suggest two possible

mechanisms of uptake. Similar to the accumulation of

chloroquine and other lysosomotropic agents, weakly basic

compounds can diffuse across the membrane, become

protonated in the acidic interior and thereby trapped in the

~ , ~ uptake by MVB

6 0 0

0 CYTOPLASM~'~'- ~ ~ wll h alp

, o o . . . . . , , o

400

"~ 3 0 0

I11 200

or , , l

,OOoo ,Q~ 0

= 0 20 40 60 80 100 120 140 160

Time (min) Fig. 9. Uptake of of the permanently charged tubocurarine in muhivesicular bodies isolated from rat liver, representing a model lot endosomal and

lysosomal sequestration. Vesicles were incubated with the radioactive ligand in the presence or absence of an ATP-regenerating system. The inset shows

intravesicular acidification through an ATP-driven proton pump while CI- cations follow passively. Organic cations (quarternary amines) can be taken up

by proton/cation antiport. In contrast, tertiary amines such as chloroquine pass the membrane in the uncharged form and accumulate through intravesicular

protonation (from Ref. [370]).

240 R.P.J. Oude E!ferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

organelle. Alternatively, compounds can be taken up by a

putative proton-cation antiporter. The first mechanism

could be operable for compounds with tertiary amine

moieties, like tubocurarine or vecuronium, if it is assumed

that the singly charged species is permeable. This mecha-

nism is unlikely to occur with the permanently charged

quarternary amine, TBuMA. Therefore, an ATP driven

proton-antiport or a cation/hydroxyl symporter was pro-

posed [370].

Storage of organic cations in acidified organelles can be

partly prevented by prior administration of basic drugs

such as primaquin both in vitro and in the intact organ

[372]. The resulting displacement of organic cations from

the acidified compartment, however, did not lead to an

increased biliary excretion rate [370,372]. A decrease in

biliary output was observed for d-tubocurarine [372] as

well as TBuMA and vecuronium [370] after administration

of primaquin. This lack of stimulation may be explained

by the relatively small functional size of this displaceable

organelle compartment [370]. In addition, competitive inhi-

bition of the protonized tertiary amine with the quaternary

agents at the level of carrier transport in the canalicular

membrane may well have occurred. An alternative expla-

nation for the lower biliary output would be a decreased

intracellular (vesicular) transport of the organic cations to

the canalicular pole of the cell [370]. In such a process

biliary excretion of the cations would occur through fusion

of cation-containing vesicles with the canalicular plasma

membrane followed by exocytosis. However, such a pro-

cess remains to be demonstrated. In this respect it is of

interest that a proton-antiport system for TEA was detected

in canalicular plasma membrane vesicles [354]. This trans-

port of TEA, however, was not ATP-dependent, in contrast

to that in endosomes.

Uptake and binding of organic cations in mitochondria

Mitochondrial sequestration of organic cations consti-

tutes another major storage compartment in the hepatocyte

[333,358,369]. Since the lysosomal sequestration process

has a limited capacity, especially at high doses of organic

cations, association with mitochondria may become a

quantitatively important process [369]. A number of factors

determine mitochondrial sequestration: (a) the considerable

intracellular volume of mitochondria; (b) the presence of

mitochondrial DNA, to which the cationic drugs can bind;

(c) the sizeable negative membrane potential of the or-

ganelle ( - 2 2 0 mV) that drives the uptake process

[358,374]. Fluorescent cationic probes are widely used for

measurement of the organelle membrane potential [375].

Initial uptake rate of various organic cations into isolated

rat liver mitochondria was shown to be highly sensitive to

agents that dissipate the mitochondrial membrane potential

such as lipophilic cationic dyes and to ionophores such as

CCCP and valinomycin [358]. Electrogenic transport into

the inner mitochondrial space was found to be (at least

partly) carrier-mediated and mutual competitive inhibition

was demonstrated for various model organic cations [376].

Apart from electrogenic uptake an electroneutral proton

antiport (or OH- symport) systems has been identified

[377].

Uptake rate into the mitochondria was positively corre-

lated with the lipophilicity of the cationic model com-

pounds [376]. This was supposed to be due to the relative

affinity for putative carriers in the mitochondrial mem-

branes. It is of interest that many cationic agents that

avidly enter mitochondria are also substrate for the MDR

(P-glycoprotein) system [378,379]. P-glycoproteins how-

ever, have not been detected in mitochondria and lyso-

somes. Yet, the apparent accumulation of cationic drugs in

these organelles in expense of the cytoplasmic concentra-

tion may contribute to drug resistance in non-steady state

conditions. On the other hand intercalation of some or-

ganic cations such as ethidium bromide with mitochondrial

DNA is a central aspect of cytotoxicity and efficient

cellular extrusion by P-glycoprotein may therefore lead to

drug resistance [378,380].

Carrier-mediated mitochondrial transport processes were

described for choline and quinidine [381] as well as sper-

midine [382]. Saturable and electrophoretic choline trans-

port was competitively inhibited by hemicholiniums, quini-

dine and quinine, the latter agent being a 10-fold stronger

inhibitor than the first [381]. The polyvalent polyamines

are also transported via a uniport system or channel that is

membrane potential dependent [336] and indirectly cou-

pled to transport of inorganic phosphate. Interestingly,

photoaffinity labeling studies with lipophilic (type 2) or-

ganic cations in hepatocytes consistently reveal labeling of

the mitochondrial protein carbamoyl synthetase that is

localized inside the mitochondrial matrix, suggesting an

intra-organelle localization of the probe [115,383].

As indicated above, the mitochondrial pool of organic

cations can be largely mobilized by addition of inhibitors

of mitochondrial function. In isolated perfused livers, addi-

tion of CCCP or valinomycin leads to massive efflux of

the hepatically stored organic cations back into the perfu-

sion medium. This confirms that mitochondria constitute a

major component of the hepatic storage compartment for

organic cations.

Various organic cations including d-tubocurarine [384],

adriamycin, spermine, and related anticancer agents as

well as alkyl guanidines inhibit respiratory chain activities

of isolated mitochondria [379,385]. Cytotoxic effects of

such agents have been, at least partly, attributed to inhibi-

tion of oxidative phosporylation [379,384]. Conclusion:

although the outer membrane of mitochondria provides

some barriers against the effects of especially hydrophilic

cationic agents [379], more lipophilic basic drugs may

easily reach these potential sites of toxicity [386]. Both

passive membrane potential driven uptake [358,374] and

carrier-mediated transport across both the outer and inner

mitochondrial membranes [374,376,381] may occur;

choline may play a role in this process.

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 241

Binding of cationic drugs to the cell nucleus

Apart from the lysosomal/endosomal and mitocbon-

drial sequestration of organic cations, the cell nuclear

material may bind considerable amounts of these agents.

Nuclear staining with fluorescent cations such as propid-

ium, ethidium and acridine dyes is generally used in

cytofluorometric analysis [387-389]. The relative amount

of cationic drugs in the cell nucleus is dependent on the

liver load [369]. The extent of nuclear binding of cationic

drugs therefore does not only depend on the inherent

affinity of the cationic drugs for the polyanionic DNA but

also on 'saturation' of the lysosomal and mitochondrial

pools. It is assumed that quaternary probes such as ethid-

ium passively pass pores in the nuclear envelope and bind

to the polyanionic polynucleotides through electrostatic

forces. The strong intercalation of some cytostatic drugs

with double-strain DNA is partly explained by their cationic

character.

4.3. Canalicular transport of organic cations

Excretion of organic cations into bile occurs under quite

different conditions compared with renal tubular secretion:

major H+-gradients as present in the renal tubuli probably

do not exist at the bile canaliculus. Proton-antiport, as

described for several cationic probes in the kidney [390], is

therefore less likely to occur at the hepatic excretory level.

Another major difference with urinary excretion is the

presence of biliary micelles and vesicles in bile that can

avidly bind hydrophilic and lipophilic organic cations [391 ].

Passive transport of cationic drug across the canaliculus

followed by binding to biliary micelles could in principle

lead to biliary 'sink' conditions and net output of cations

into bile. Excretion of organic cations into the canalicular

lumen occurs against the negative membrane potential.

However, the membrane potential 'hill' could be taken via

electroneutral exchange with organic or inorganic cations

or be facilitated by formation of ion-pairs with suitable

counter anions. Marked stimulation of transport of cationic

probes across liver plasma membranes by lipophilic anions

has been demonstrated [392]. This was explained by dissi-

pation of the intramembrane potential hill in the lipid

bilayer and/or ion-pair transport of the uncharged

cation/anion complex. In this respect it has been shown

that tetrapbenylborate completely blocks the biliary excre-

tion of TBuMA at an unchanged bile flow in isolated

perfused livers [393]. The lipophilic anion might have

favored electrogenic reabsorbtion of the organic cation or

could have facilitated intracellular binding/sequestration

to macromolecules and organelles.

Evidence was obtained for heterogeneity of organic

cation transport from studies with canalicular liver plasma

membrane vesicles (cLPM). Intravesicular accumulation

and initial uptake rate of TEA in inside-out canalicular

vesicles showed no apparent ATP-dependency. The pres-

ence of an electroneutral T E A / H + exchange mechanism

exhibiting trans-stimulation was shown. PAEB, vecuro-

nium and tributylmethyl ammonium cis-inhibited this pro-

cess [354]. The choice of the transport model compound

TEA seems less appropriate since it is unlikely to be

excreted in bile [394]. Similar results were found, however,

for the tributyl-methyl analogue, which is strongly secreted

into bile (Moseley, pers. communication). In contrast, the

accumulation of NMN in cLPM did not show a pH-gradi-

ent dependent mechanism [353]. Earlier observations

showed that inside-out cLPM manifest ATP-dependent

accumulation of daunomycin [395]. Accumulation in cLPM

was shown to be inhibited by other cytostatic (basic) drugs

such as vinblastine, vincristine and andriamycin as well as

by other lipohilic drugs with a tertiary amine function such

as verapamil and quinidine [395]. More recently, vinblas-

tine accumulation in cLPM was reported which exhibited

temperature dependency and osmotic sensitivity [396].

Daunomycin avidly binds to membranes ,~o that actual

transmembrane transport of the drug in vesicles is quite

difficult to prove. Therefore the relation of these in vitro

cation studies with the biliary excretion in the intact liver

remains to be clarified.

Evidence ~ r carrier-mediated transport in the biliarv

excretion of cationic antineoplastic drugs

In the isolated perfused organ and in the,, liver in vivo,

daunomycin is extensively metabolized and converted to

glucuronide conjugates among other metabolites [397].

Yet, abundant biliary excretion of unchanged doxorubicin

was also observed [398]. Elimination of the anthracyclin

into bile can be completely blocked by verapamil at an

unchanged bile flow (see Fig. 10Fig. 10). In this experi-

ment verapamil was added at the moment that most of the

added doxorubicin was taken up in the liver, excluding

potential interaction at the uptake level [398]. Two other

liver perfusion studies showed a profound inhibition of

colchicine excretion into bile by the lipoph~lic cyclic pep-

tide cyclosporin [399,400]. Biliary excretion of the basic

fluorescent dye acridine orange is also largely affected by

cyclosporin [401]. Finally, an inhibitory effect of vera-

pamil on the biliary excretion of 3H-vincristine was found

by Watanabe et al. [402]. Daunomycin, vincristine,

colchicine and acridine orange are known substrates for

P-glycoprotein and verapamil and cyclosporin are strong

inhibitors of this pump [403,404]. These studies therefore

are suggestive for the involvement of P-glycoprotein

and/or related ATP-dependent translocator(s) in the bil-

iary excretion process of organic cations. The ATP-depen-

dent vesicular accumulation of cationic drugs in isolated

canalicular membrane vesicles [395,396] as well as the

canalicular localization of P-glycoproteins (mdrla, mdrlb,

mdr2), that was demonstrated by immunochemistry and

photoaffinity labeling [405-408], may support such a hy-

pothesis.

In the light of the interesting hypothesis of involvement

of P-glycoprotein in biliary secretion of organic corn-

242 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

10

E O

E

t~

O

X LU

0 .1 ,

0.01 -

0.001

/ /

/ /

/

/

Addition

J i

/

/ ' 'o ' 0 20 4 6O

Time (min)

i i

80 1 O0

Fig. 10. Inhibition of biliary excretion of doxorubicin by verapamil in the isolated perfused rat liver. Doxorubicin was allowed to be taken up from the perfusate during 20 rain. Subsequently verapamil was administered to the perfusate in a single dose. Verapamil did not affect bile flow but reduced biliary excretion rate of unchanged (HPLC assayed) doxorubicin temporarily for more than 90%. At t = 40 rain biliary excretion of doxorubicin is restored due to metabolic inactivation of verapamil in the liver (from Ref. [398]).

pounds more detailed information on this peculiar trans-

port system will be provided and recent findings on its

possible function in organic cation transport will be inte-

grated.

4.4. The role o f P-glycoproteins

P-glycoproteins (Pgps) are members of the super family

of transporter proteins that contain ATP-binding cassettes,

also called ABC proteins or traffick ATPases [403,409-

413]. The proteins consist of about 1200 amino acids that

constitute two homologous halves each with one nu-

cleotide binding segment and 6 transmembrane domains.

The apparent MW varies from 130-190 kDa, dependent

on species and cellular origin. Murine species exhibit three

gene products: mdrla, mdr lb and mdr2 while in the

human two variants MDR1 and MDR3 were identified.

MDR1 and mdr la or b confer drug resistance whereas

MDR3 and the related murine protein mdr2 do not.

M D R 1 / m d r l a / l b compared with M D R 3 / m d r 2 have

very different tissue localization and are also differently

induced [405,414-418]. M D R 3 / m d r 2 is particularly ex-

pressed in the canalicular domain of the hepatocyte and

may function as a phospholipid translocator or flippase

[419,420] (see also Section 5.3). In contrast,

M D R I / m d r l a / l b display a more general body distribu-

tion, being highly expressed at the apical domains of cells

in colon, small intestine, liver, adrenal glands and to a

lesser extent in choroid plexus, placenta, testis and bone

marrow [403,409]. Recent studies showed its presence in

peripheral blood lymphocytes [421,422]. The strategic body

localization points to an active protective function towards

potentially toxic agents and/or an essential role in trans-

port of endogenous substrates.

The multidrug resistant tumor phenotype, characterized

by a decreased cellular drug accumulation of cationic

drugs, is directly related to overproduction of this naturally

occurring protein via gene-amplification [403,409,411,423].

It was recently demonstrated that this resistance is achieved

by ATP-dependent extrusion of drugs from cells by P-

glycoprotein [424-426]. The genes responsible for the

various MDR modalities have been cloned and sequenced

[423,427-429]. Although the transport function can be

reconstituted in artificial systems [424,425,429] or trans-

fected in M D R ( - ) cell lines, insufficient quantities are

available to date for structural analysis. Yet, mutational

analysis enabled further characterization at the molecular

level [403,409].

Mechanisms o f P-glycoprotein mediated membrane

translocation ~1 ~ drugs

At least two potential drug binding sites, one in each

half of the molecule, have been detected through photo-

affinity labeling/proteolytic digestion studies and by site

directed mutagenesis [403,409]. Subtle changes in amino

acid composition in the transmembrane domains 3 and 11

can change the relative binding affinity of various sub-

strates but can also differently affect the rate of dissocia-

tion of various drugs from the protein. It should be realized

that both association and dissociation are essential for

membrane translocation of organic compounds. The inter-

relation of the above mentioned binding sites in the func-

tional transport structure of Pgp is not exactly known, but

recent studies support a single channel model in which the

two binding sites somehow cooperate [403]. P-gp variants

have been proposed to function as volume regulated CI-

channels [430] and even as ATP channels [431]. However,

it is difficult to explain how all these functions would be

incorporated in a single transmembrane protein.

In addition, a proton pump activity has been claimed

since Pgp(+) cell lines exhibit an increased (cytoplasmic)

pH [432-437]. Chloride and proton (or cation) fluxes

could be somehow linked in the sense that ATP-hydrolysis

leads to binding of the cationic species that for vectorial

transport is either passively followed by C1 or exchanged

for H +. A recent study does not support the idea that

changes in intra-or extracellular pH mediate Pgp-depen-

dent multidrug resistance [438].

With regard to drug transport, it remains to be estab-

lished whether the P-glycoprotein really functions as a

R.P.J. Oude EIferink et al./ Biochimica et Biophysica Acta 1241 (1995) 215-268 243

conventional transporter or, alternatively, acts as a mem-

brane extruder that merely expels amphiphilic compounds

that entered the membrane by non-ionic lipoid diffusion

[439]. The latter concept has been called the 'hydrophobic

vacuum cleaner' model and may bear resemblance with

phospholipid flippase-type of translocation [419]. The lat-

ter modality implies that access of substrate to the trans-

port protein occurs directly from the lipid phase and that

the drug is flipped from the inner leaflet of the bilayer to

the outer. This model also predicts that the P-glycoproteins

could decrease the initial uptake rate of the substrate in

cells in vitro since it would not differentiate between drug

entering the membrane from the inner or outer cellular

space [440]. Homolya et al. [441] have tried to substantiate

this concept by looking at the uptake of acetoxymethyl

esters of fluorescent (anionic) dyes. Due to their hydropho--

bic character these compounds diffuse through the plasma

membrane and are extremely rapidly hydrolysed by abun-

dant intracellular hydrolases. It was shown that cells over.-

expressing mdrl Pgp accumulate the hydrolysed com-

pounds much less than control cells, suggesting that the

acetoxymethyl esters are extruded by Pgp from the mem-

brane before they can reach the intracellular space.

Stein et al. [442] studied the behavior of a variant of

MDR1 in which a glycine at position 185 is replaced by

valine. On basis of the differential effect of this mutation

on the initial uptake vs. efflux rates of various substrates

they proposed that Pgp can extract drugs from the internal

as well as from the external leaflet of the membrane by

different paths.

In an alternative model, Gottesman and Pastan [443]

have tried to combine the different reported transport

functions of Pgp for protons, chloride and amphipathic

compounds. In this hypothetical model ATP hydrolysis is

linked to the transport of protons into the transporter, with

chloride following passively. Once within the transporter

these ions will draw water osmotically into the pump out

of the membrane. Amphipathic drugs within the membrane

then: "should follow the water, and drugs will be removed

from the membrane just as a scrubber on a smokestack

removes water-soluble materials from smoke". Although

this model (like the vacuum-cleaner model) would explain

the broad specificity for amphipathic compounds it is not

in line with the observation that the chloride channel

activity, if indeed an inherent feature of Pgp, could be

dissociated from its drug transporter activity [430]

Substrate specifici~ o f P-glycoprotein

A central feature of multidrug resistance is the phe-

nomenon of cross-resistance: persistent exposure to an

amphiphilic cytostatic compound leads to overexpression

of the 170 kDa protein. The overproduced protein accom-

modates such a broad range of substrates that resistance

occurs for many related antineoplastic drugs [403]. How-

ever, it is highly unlikely that evolution has resulted in a

P-glycoprotein that is specific for anti-tumor agents alone.

Indeed, it can bind and may transport a wealth of naturally

occurring as well as therapeutic and diagnostic agents

[404 ,~ ,445] . Many of these agents stimulate ATP hydro-

lysis by purified and reconstituted Pgp [424-426] and are

likely candidates for vectorial transport. Some compounds

may only bind to Pgp but do so with high affinity. Such

inhibitors may have value as so called 'reversal agents'

that can antagonize the MDR phenotype.

Structure-transport relation o f P-glycoprotein

The wide variety of organic compounds that seems to

be accommodated by P-glycoprotein (at least the

MDR 1 / m d r l a / mdr lb Pgps) raises the question what the

common physicochemical and structural features of such

substrates are. Many systematic structure-kinetic studies

have been performed, including three-dimensional struc-

tural analysis [404,444-452]. Most of these studies con-

firm a general picture that comprises a number of struc-

tural features:

(a) The presence of at least one but preferentially several

planar aromatic ring structures enabling interaction

with a supposedly 'flat ' hydrophobic structure in the

P-glycoprotein drug-binding domains

(b) In relation to this, a partition coefficient (octanol-aque-

ous buffer) higher than 1 but preferentially exceeding

2

(c) Preferentially a cationic charge at physiological pH,

for instance through a protonable nitrogen center in the

molecule

(d) A proper molar refractivity reflecting the 'space fill-

ing' molecular volume, indicating that a minimal

'bulky' structure is a necessary prerequisite for a

proper MDR substrate.

The fact that the molecular mass of MDR substrates

mostly exceeds 400 Da is in line with such a structure. The

latter characteristic is a necessary element since a combi-

nation of aromatic rings and protonatable nitrogens in

itself is not sufficient to modulate pump activity [404,453].

The importance of a nitrogen centre in MDR

modulators/substrates was demonstrated for the acetamide

group of colchicine [453] and is also supported by the

finding that removal of the basic center from doxorubicin

partially destroys its feature of a MDR substrate [454].

Nevertheless some steroidal compounds without a basic

center have been shown to be transported by the MDRI

gene product. These include certain steroidal hormones

and cardiac glycosides such as digoxin.

At least two drug binding sites seem to be present in

P-glycoprotein [403,411,455]. A model was, proposed in

which both halves of the transporter, each with one bind-

ing site, come together to form a single transport channel

[403,456]. This in spite of the fact that the binding sites of

verapamil and vinblastin are not identical [45'7]. The differ-

ing abilities of various drugs that can reverse MDR to

inhibit other drug or drug analogue binding also indicates

244 R.P.Z Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

that P-glycoprotein has multiple drug-binding sites (see

[423,458]. Recent evidence indicates a dihydropyridine

selective acceptor site that is allosterically coupled to a

vinca-alkaloid-selective binding site [459]. Aliosterically

coupled drug binding domains have also been reported for

other channel/pump and receptor proteins [459]. The im-

plication of the latter aspect is that both negative and

positive allosteric effects (synergism) may occur in the

presence of more than one MDR-substrates/inhibitor. It

also means that Pgp-directed bifunctional modulators could

be developed for counteraction of drug resistance. Other

types of 'bifunctional' MDR-modulators that are in the

stage of development comprise compounds that both in-

hibit nucleoside transport and MDR1 [460] or consist of

agents that simultaneously inhibit P-glycoprotein and non-

Pgp carrier system(s) (see below).

Non P-glycoprotein related MDR phenomena

A number of laboratories have independently observed

that reduced accumulation of drugs in tumor cell lines can

occur without mdrl gene overexpression while multidrug

resistance in such cells is poorly reversed by chemosensi-

tizers that are effective in cells overexpressing MDR1

[330,461-465]. Recently, a potential carrier channel pro-

tein has been cloned: the MDR-related protein (MRP) that

like P-glycoprotein belongs to the ABC protein superfam-

ily [462]. MRP overexpression results in high levels of a

190 kDa ATP-binding plasma membrane protein

[330,462,464]. However, MRP is only distantly related to

the characterized members of the superfamily. The amino

acid identity with human MDRI is only 14% and predomi-

nantly confined to the ATP binding site [462]. The MRP-

gene is ubiquitously expressed in human tissues, but its

expression in liver is relatively low [330]. Although the

protein is mainly localized on the plasma membrane, sub-

stantial amounts may be present in the endoplasmic reticu-

lum and other intracellular organelles [464]. Cell-type spe-

cific variation in MRP distribution between plasma mem-

brane and intracellular compartments may explain why

decreases in cytoplasmic drug accumulation are in some

cases due to drug extrusion and in others related to intra-

cellular drug sequestration [462]. Recent studies [465] indi-

cate that the specific tyrosine kinase inhibitor genestein

highly influences intracellular distribution of cationic drugs

in cells with MRP overexpression but fails to do so in Pgp

(MDR1) cell lines. Tyrosine kinase may be involved in

deactivation or down regulation of the MRP pump. In

contrast, P-glycoproteins seem to be regulated via protein

kinase C (see 4.4.6).

In a recent study by Jedlitschky et al. [328] it was

demonstrated that plasma membrane vesicles of MRP-

overexpressing cells display increased ATP-dependent

transport of leukotriene C 4 and dinitrophenyl-

glutathione,which both are typical organic anions. These

experiments suggest that MRP is a primary active organic

anion pump. In order to explain the (relatively low) resis-

tance of MRP-overexpressing cells towards cationic cyto-

toxic drugs like daunomycin, one has to assume that this

transporter accommodates both organic anions and neutral

or cationic cytotoxic drugs. Alternatively, it may be that

metabolism of daunomycin and other drugs into anionic

species, like glutathione conjugates, is sufficient in these

cells to give rise to increased efflux via an organic anion

transporter. Very recently, Miiller et al. [329] reported very

similar results with another MRP-overexpressing cell line

as well as with MRP-transfected cells.

Interestingly, another protein of 110 kDa is also overex-

pressed in some non-Pgp MDR cell lines. In normal tissues

with a secretory/excretory function such as the liver, this

protein is highly expressed and predominantly present in

intracellular vesicles [466]. From a recent study [330] it

was indeed concluded that MRP overexpression can not

account for all lbrms of non-Pgp MDR. In non-steady state

conditions the activity of such an intracellular pump could

lead to MDR via lowering of the cytoplasmic drug concen-

tration.

Regulation o f P-glycoprotein expression and activ~v in the

liver

In a number of pathologic and experimental conditions

major fluctuations in Pgp activity have been demonstrated.

Increased MDRI RNA levels have been demonstrated in

hepatocellular carcinoma (HCC) [418], in cirrhotic non-

tumorous tissue in HCC patients but not in other patients

with liver cirrhosis [418]. In contrast Teeter et al. [467,468]

found no consistent elevation or even a lowered expression

of MDRI gene in metastatic colon tumors and HCC while

MDR2 was clearly underexpressed in metastatic tumors.

In rodents mdr genes are induced in preneoplastic and

especially neoplastic liver nodules, in the later stages of

carcinogenesis. However, the underlying molecular mecha-

nisms are unresolved [414,469,470]. During regeneration

of mouse liver both mdr2 and mdrla are induced whereas

in mouse HCC only mdr la is elevated [414,417]. In-

creased levels of mdrla and l b have also been reported in

hepatoma derived cell lines and hepatocytes treated with

acetyl aminofluorene (AAF) [417].

P-glycoprotein is also overexpressed in canalicular

structures that are formed between adult rat hepatocytes in

primary culture [471-473]. During regeneration of rat

liver, after partial hepatectomy, and following tetrachloride

[416] and colchicine treatment [474], hepatic mdrl and

mdr2 are largely but differently induced. After CC14 and

colchicine, increased levels of mdrl Pgp are already found

at 3 h after treatment, remaining increased for up to 5

days. A 5-to 10-fold increase in transcription rate is found.

Mdr2 expression after CCL4 increased only after 48 h and

is back to normal after 72 h. Increased expression of mdr2

was found especially in the canalicular domain, initially in

the periportal zone and subsequently spreading to pericen-

tral cells. Both obstructive-and a-naphthyl isothiocyanate-

induced cholestasis in rats and monkeys increase mdrla

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 245

and mdrlb gene expression in liver [415]. In the latter

study a possible correlation with biliary excretory function

was found: treatment of rats with AAF led both to in-

creased mdrl mRNA levels and vinblastine excretion into

bile. However, the causal relationship of these phenomena

remains to be established. In a recent study of Lee et al.

[473], the differential expression of P-glycoprotein genes

in primary rat hepatocytes cultures was studied; mdrlb

was only moderately elevated. Expression of mdr2 Pgp,

which is most abundant in rat liver even decreased and a

reciprocal relation step between mdrlb and mdr2 seems to

be present. The level of mdrlb was strongly correlated

with actin mRNA, suggesting a common regulation of

gene expression. Microtubuli and microfilament inhibitors

(colchicine, cytochalasin B) inhibited the normal mdrlb

overexpression. A similar differential activation was ob-

served during liver regeneration and hepatocarcinogenesis:

mdrlb expression was much more increased than that of

the mdrla and mdr2 gene [414].

Apart from variations in P-glycoprotein levels, also the

intrinsic activity of the pump seems to be regulated (see

Fig. l lFig. 11) [475]. There is abundant evidence that

basal activity of P-glycoprotein is modulated through phos-

phorylation/dephosphorylation reactions mediated through

protein kinase C (PKC) and probably other kinases

[465,476]. In most cell lines stimulation of PKC by phor-

bol esters leads to an increased phosphorylation of Pgp and

increased drug effiux from cell lines with overexpression

of P-glycoprotein [477-484]. Analysis of the primary se-

quence of Pgp indicates 37 potential phosporylation sites

of which 14 are probably accommodated by PKC. Other

kinases such as cAMP dependent kinases as well as an

unidentified kinase that is neither identical to PKC nor to

protein kinase A [485] mediates phosphorylation of Pgp.

PKC is a family of at least 10 closely related phospholipid

and thiol dependent isoenzymes of which some are also

Ca2+-dependent. Most of the isotorms are activated by

diacylglycerol (DAG) (see Fig. 11) [486]. Among the most

selective PKC inhibitors are synthetic peptide substrate

analogs [487]. It is of interest that many PgP substrates can

activate phospholipase-C and thereby can stimulate the

pump activity via DAG release. The positive regulation by

PKC of mdr-encoded PgP was elegantly demonstrated in

cotransfection studies using the human MDR1 gene and

PKCa cDNA [488]. Staurosporin and its more selective

7-hydroxy derivative are inhibitors of PKC that bind to its

catalytic domains [489] and reverse the phorbol ester ef-

fects [490]. However, staurosporins are cationic drugs and

bind to P-glycoprotein directly [491]. Indeed, it has been

shown that staurosporin and some derivatives can reverse

drug resistance irrespective of their PKC inhibitory activity

[492]. However, two other more specific inhibitors of PKC

exhibited similar effects on phosphorylation of P-glyco-

protein and its pumping activity: sodium butyrate [485]

that decreases Pgp phosphorylation and calphostin C that

binds to the regulatory domain of PKC [493,494]. The dual

SINUSOID

÷ ¢X::-- bile acids q

c a r d d ~

El>

CANALJCULUS

I

5 3 .'z.kq

0¢ HCO~

CI-

Fig. 1 I. Intracellular sequestrat ion sites o f organic cat ions and potential

regulat ion mechan i sms for transport . O C * = organic cations: 1 = type I

uptake system; 2 = type 2 uptake system: M = multispecific uptake sys-

tem; VP = vasopressin; R = receptor: G = GTP-b ind ing protein: D A G =

diacylglycerol ; P K C = protein kinase C,

inhibition of Pgp by staurosporins through direct binding

and lowering its phosphorylation is probably true for many

'reversal agents' and have been demonstrated for vera-

pamil, phenothiazines, cyclosporins and tamoxifen (see

[476,495]) among others. Interestingly, treatment of some

colon cancer cell lines with butyrate leads to decreased

transport of vinblastine and adriamycin but not of

colchicine [485]. This indicates that the PKC influence on

MDR-mediated translocation may be drug-class specific.

PKC mediated phosphorylation of P-glycoprotein also

modifies the efficacy of reversal agents [494]. How drug

transport via P-glycoprotein is changed by phosphorylation

is presently unclear. In general, decreased phosphorylation

does not affect binding of substrates to the MDR1 proteins

[494] although azidopine is a remarkable exception [494].

In relation to P-glycoprotein, activity of PKC plays a

dual role: it modulates the intrinsic activity of the pump

but also regulates its cellular expression. PKC agonists

activate P-glycoprotein gene expression whereas stau-

246 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

rosporin and other PKC antagonists lower its cellular

expression [496,497].

Protein kinase C and biliary excretion of drugs

PKC may also be involved in the regulation of canalicu-

lar drug transport. Organic anion efflux from isolated

hepatocytes, that likely reflects the bile canalicular excre-

tion process via cMOAT, is increased by vasopressin and

phorbol esters and inhibited by staurosporin [268]. A re-

cent study demonstrates that such agents also affect the

biliary excretion rate of the monovalent organic cation

TBuMA [498]. A dose-dependent stimulation was found

for PMA. Vasopressin also increased biliary output of the

quaternary amine whereas staurosporin blocked these ef-

fects and also lowered the basal excretion rate. These

effects were observed in the intact (isolated perfused) liver

and were unrelated to moderate changes in bile flow and

uptake rate of the organic cation [498]. Whether the effects

on cation transport are related to P-glycoprotein activity

remains to be demonstrated. Preliminary studies indicate

that high TBuMA concentrations fail to inhibit dauno-

mycin transport in canalicular membrane vesicles and do

not affect the biliary excretion of classical MDR substrates

in the perfused liver [398]. Yet a number of reversal agents

such as verapamil and quinidine strongly inhibit TBuMA

excretion into bile at an unchanged bile flow [398]. These

observations point to multiplicity in hepatic excretory

mechanisms for organic cations and also indicate that the

effects of verapamil may not be specific for MDR-media-

ted transport.

Current evidence for the involvement of P-glycoprotein in

hepatobiliary cation transport

Recent studies in highly purified right-side out canalicu-

lar membrane vesicles confirm the earlier observation

[395,396] of ATP-dependent cation transport. In these

studies the quaternary amines N-pentyl-quinidinium (NPG)

and N-pentyl-deoxyajmalinum (NPDA) were used [499].

Especially the quinidium derivative has the advantage of

low aspecific binding to the plasma membranes, a phe-

nomenon that endangers proper experimentation with

daunomycin and vincristine. The pentyl analogue of quini-

dine with a permanent positive charge exhibited a 10-fold

higher uptake compared with the tertiary compound.

Photoaffinity labeling of the canalicular membranes with

N-azo-pentyl-deoxyajmalinum revealed polypeptides of

143 and 108 kDa [499]. Transport in vesicles was in-

hibitable by verapamil and daunomycin but not by tauro-

cholate. N-pentyl-quinidinum transport could also be ele-

gantly demonstrated in Sf9 insect cells transfected with the

mdrlb gene product via a recombinant baculovirus vector

[500]. ATP-dependent transport of NPG, NPDA and

daunomycin as well as quinidine was clearly demonstrated

and interestingly the differences in K m of transport of

these agents in the mdr-transfected cells was almost identi-

cal to the ones observed in the canalicular membrane

vesicle experiments. Photoaffinity labeling of the Sf9

membrane vesicles again revealed a 140 kDa protein that

probably represents a non-glycosylated form of P-glyco-

protein [500].

In summary, a number of indirect observations support

the hypothesis that P-glycoprotein is one of the putative

cation carriers responsible for biliary excretion of bulky

(amphiphilic) organic cations:

(a) ATP-dependent transport in canalicular membrane

vesicles and in plasma membrane vesicles of MDR

transfected cells of various amphiphilic basic drugs

such as daunomycin [395,396] as well as quinidine,

deoxyajmalinum and their quaternary N-pentyl deriva-

tives [499,500]

(b) Excretion of cationic antineoplastic agents in bile in

the intact organ is highly inhibited by verapamil and

other so-called reversal agents [398,400-402]

(c) Induction of mdrla through partial hepatectomy and

regeneration leads to an increased biliary clearance of

vinblastine [415]

(d) Excretion rate into bile of unchanged doxorubicin is

increased by phorbol esters suggesting PKC regulation

[3981.

Some, more hydrophilic, quaternary ammonium com-

pounds such as TBuMA, PAEB and d-tubocurarine do not

affect doxorubicin excretion into bile even when used at

extremely high doses [398]. Nevertheless TBuMA excre-

tion is highly influenced by PKC modulators [398,498].

These observations point to at least one additional organic

cation transport system at the canalicular level that might

be ATP-dependent, PKC stimulated but not identical to

mdrla or mdrlb. A non-P glycoprotein (MRP) type of

carrier could be responsible or alternatively transport activ-

ity might be related to a 110 kDa protein that apart from

the abovementioned 140 kDa polypeptide is consistently

photolabeled in these canalicular membrane preparations

[364].

The supposed cation transport systems in the canalicu-

lus likely display overlapping substrate specificity depend-

ing on the hydrophobicity/hydrophilicity balance in the

particular molecules. Hydrophobic bile salts and other

lipophilic organic anions such as estradiol glucuronide

[501,502] may inhibit both transport systems [503].

Whether hormonal steroids, cardiac glycosides and their

metabolites are also accommodated by these ATP-depen-

dent carrier systems in the canaliculus is presently un-

known. It is of interest that evidence is accumulating for a

role of P-glycoprotein in transmembrane transport of

steroidal compounds such as corticosterone [504], cortisol

[505] and digoxin [506-508] in renal tubular monolayers

[509], adrenal gland, kidney in vivo [510], multidrug resis-

tant tumor cell lines [505,510] and perfused rat kidney

[507,508].

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 247

5. Hepatobiliary transport of lipids

Hepatocytes secrete substantial amounts of lipids, both

at their basolateral (sinusoidal) side into the blood and at

their apical (canalicular) pole into the bile. The first pro-

cess involves the packaging of triglycerides, cholesterol

(ester) and phospholipids together with apolipoprotein B

into nascent Very Low Density Lipoproteins (VLDL) in

the endoplasmic reticulum, as well as processing of the

particles in the Golgi complex and secretion by exocytotic

mechanFig. 12Fig. 12isms [511]. Nascent High Density

Lipoproteins (HDL), in the form of discoidal apolipo-

protein-phospholipid complexes containing apolipoprotein

A-I, A-II and/or A-IV, are also secreted into the blood by

the liver cells [512]. The apical route, i.e., the movement

of free cholesterol and phospholipids from hepatocytes

into the bile, occurs via clearly different, but as yet

incompletely understood mechanisms. Data from a large

amount of studies published during the past two decades,

suggest that the secretion of cholesterol and phospholipid

is coupled. Thus, in general, fluctuations in bile secretion

of phospholipid induced by experimental means, e.g., by

feeding choline-deficient diets to rats [513], result in paral-

lel alterations in the cholesterol secretion pattern. It has,

however, been recognized that under certain metabolic

conditions deviations can occur [514-516]. It is also evi-

dent that the secretion of cholesterol and phospholipids is,

at least in part, governed by bile salts. How and where in

the sequence of events leading to the secretion of lipids the

bile salts exert their regulatory function is still a matter of

debate. Basically, the relationship between bile salt secre-

tion and that of cholesterol and phospholipids is described

by a hyperbolic function, i.e., lipid secretion plateaus at

high bile salt secretion, in all animal species studied so far.

However, in different species marked quantitative differ-

ences exist in the amounts of lipids secreted per amount of

bile salt [517] as well as in the maximal secretion rates of

phospholipids and cholesterol [517]. This may, in part, be

related to interspecies differences in the bile formation

process, or, according to the model of Mazer and Carey

%

-4

~, o o o ° ° ~ ~ , o o o o ° °

o ~ \Theory o o ° o o o o o o o o O O O o o o o ° ~ o o o o o o o O O o o ° ° ° °

- - I I I ] I

50 ioo 15o 20o 25o 300

BS~ (#mol.kg -I. hr -I)

Fig. 12. The relation between biliary bile salt secretion (BS~e c) and biliary

phospholipid secretion (L~e c). The data represent a compilation of several

studies with different animal species. These data were used by Mazer and

Carey for a mathematical model of bile salt-induced phospholipid secre-

tion. (from Ref. [518]).

[518], by interspecies differences in hepatic phospholipid

synthesis rate (see Section 5.2 and Fig. 12). It is also

well-established that the capacity of the individual bile

salts to induce lipid secretion correlates with its relative

hydropbobicity [519-521] as reflected in its so-called criti-

cal micellar concentration (CMC). However, intracanalicu-

lar bile salt micelle formation alone can not be responsible

for regulation of biliary lipid secretion. This would not be

in line with the observed presence of cholesterol-phospho-

lipid vesicles in native bile samples [522-524], with obser-

vations such as the isolated hypo-or hypersecretion of

phospholipids [516] and of cholesterol [514,515] under

certain experimental conditions and with the highly

species-specific coupling of lipid-to-bile salts [517] There-

tore, alternative modes of regulation must exist. Recent

work indicates the presence of (an) additional regulatory

mechanism(s) involving canalicular membrane proteins,

specifically the mdr2 P-glycoprotein [125]. The issues that

will be addressed in this Section concern the origin of the

biliary lipids and the factors involved in the regulation of

their secretion. In addition, some considerations on the

potential physiological function(s) of biliary lipids will be

presented. In view of the scope of this review, emphasis

will be on recent developments in this field of research; for

a more complete overview of available literature the reader

is referred to a number of excellent reviews on this subject

[5,517,525,526].

5.1. Composition and physical.h~rm o[ bilia~ lipids

Cholesterol and phospholipids comprise a major part of

the organic fraction of bile in quantitative terms; their

secretion into the bile amounts up to 0.8-1.2 g /day and

3-4 g/day, respectively, in humans [517] and to 15 m g /

day and 200 rag/day, respectively, in the rat [527]. Con-

sidering the large differences in lipid composition that

exist between liver and bile, it is evident that secretion

does not occur in a random fashion but rather involves an

advanced level of 'sorting'. Bile phospholipids consist

almost exclusively of phosphatidylcholine (PC) of specific,

relatively hydrophilic, fatty acid composition in all animal

species studied. The sn 1 position of biliary PC usually is

occupied by the saturated fatty acid species palmitate

(16:0) whereas the sn 2 position invariably contains an

unsaturated species, predominantly oleate (18:1) or

linoleate (18:2). This contrasts with PC of the canalicular

membrane which has a high content of arachidonate (20:4)

in the sn 2 position. Using state-of-the-art analytical tech-

niques, however, approximately 25 additional PC species

could be identified in normal human [528] and rat [529]

bile, albeit in small proportions. Fatty acid composition

can be influenced by diet, e.g., diets rich in n - 3 fatty

acid-containing fish oils lead to enrichment of biliary PC

species with these specific fatty acids [530] and, in the

perfused rat liver, by acute administration of hydrophilic

fatty acids such as 16:1 [529]. Only minor amounts of

248 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

sphingomyelin and phosphatidylethanolamine, both major

constituents of hepatic membranes, are present in the bile

under normal conditions. Thus, the hepatocyte selectively

recruits specific PC species for secretion into bile and/or

effectively excludes other phospholipids from entering this

pathway. Cholesterol appears to be present in bile exclu-

sively in its unesterified form under any condition.

It is well established that lipids are present in bile in the

form of aggregates, of which the size, shape and composi-

tion are primarily determined by the concentration of these

lipids relative to that of the biliary bile salts as well as by

the physicochemical characteristics of the phospholipids

and bile salts present. A large body of evidence is avail-

able to indicate that simple micelles, consisting of bile

salts and cholesterol, mixed micelles, containing bile salts,

cholesterol and phospholipids, and unilamellar vesicles,

consisting of cholesterol and phospholipids and trace

amounts of bile salts, can co-exist as biliary lipid carriers

in a dynamic, i.e., interchangeable, fashion [531]. In addi-

tion, multilamellar vesicles (liquid crystals) may be present

in lithogenic-supersaturated bile [53 l]. The vesicular struc-

tures contain a characteristic and constant pattern of (glyco)

proteins, with the canalicular transmembrane glycoprotein

aminopeptidase N as an important representative [532].

The term 'bile lipoprotein complexes' has been proposed

for vesicular structures containing a 7 kDa apoprotein

designated as anionic polypeptide fraction (APF) [533,534].

Whether these structures represent a distinct entity within

the population of vesicular carriers and whether they have

specific physiological importance remains unclear. It has

also been claimed that 'lamellae', composed of stretched

membrane-like bilayers, comprise the major cholesterol-

carrying particles in human bile [535,536]. However, as

recently discussed in great detail [531], these structures

may represent artifacts of the microscopic technique em-

ployed.

A key question concerning the regulation of biliary lipid

secretion relates to the form in which the lipids actually

enter the bile canaliculus. This question is difficult to

approach directly because of the inaccesibility of the

canalicular lumen for direct bile sampling at the site of its

formation. In the 'traditional' view (see further below),

based on the pioneering studies of Kay and Entenman

[537], Hardison and Apter [538] and Wheeler and King

[539], solubilization of bile-specific lipids from intra-

cellular or canalicular membranes into mixed micelles

reflects the initial event in the secretory process. More

recent concepts, generated by the physicochemical analysis

of freshly collected bile by quasi-elastic light scattering

and electronmicroscopic techniques [522-524], postulate a

form of vesicular secretion. This form of secretion has

been proposed to result from an exocytotic process or from

membrane-shedding or budding, and is, under physio-

logical conditions, followed by micellar dissolution of the

vesicles. As bile salts rapidly dissolve the PC-cholesterol

vesicles in unsaturated bile, only simple and mixed mi-

celles are found after equilibration, i.e., the situation that

usually occurs under experimental conditions when bile is

collected for prolonged periods of time after creation of a

bile fistula. Clearly, the different mechanisms of secretion

are by nature not mutually exclusive and may even occur

simultaneously, perhaps contributing to a variable extent to

total lipid secretion under different conditions.

5.2. The origin and precursor pools of biliary lipids

Considerable efforts have been made to define the

origin of biliary lipids. Concerning the question to what

extent de novo synthesis contributes to biliary cholesterol

secretion, recent studies employing tritiated water as a

precursor yielded values in the order of 7-16% [540,541 ]

under control conditions in the rat. The observation that

the specific activity of cholesterol in bile under these

experimental conditions is much higher than in the liver

[540,542] has led to the speculation that newly synthetized

cholesterol is preferentially secreted from the liver into the

bile or, alternatively, is derived from a specific hepatic

pool. As elegantly shown by Robins et al. [543], this

observation may in fact reflect the secretion of (newly

synthetized) cholesterol by a specific subpopulation of

hepatocytes. It is known that HMG-CoA reductase, which

is considered to be the rate-limiting enzyme in cholesterol

biosynthesis, is sharply localized to the periportal region of

the liver acinus [544]. From other studies [34] it is known

that this region of hepatocytes is primarily involved in bile

salt transport during normal enterohepatic cycling. The

relative amount of newly synthetized cholesterol can ex-

perimentally be altered to a considerable degree by manip-

ulation of hepatic cholesterogenesis [540,545]. Recent

studies [546,547] indicate that, under certain circum-

stances, the rate of hepatic cholesterol synthesis may co-

regulate the total amount of cholesterol secreted into the

bile. Nevertheless, the major part of biliary cholesterol is

derived from preformed sources, including cholesterol in

membranes of liver cells, cholesterol released from hepatic

cholesterylester pools and cholesterol originating from

lipoproteins taken up by the hepatocytes from the blood

compartment. With respect to the latter, numerous studies

on the contribution of lipoprotein cholesterol(ester) em-

ploying radiolabeled tracers have suggested the existence

of a metabolically active cholesterol pool that is replen-

ished from lipoprotein cholesterol, with HDL cholesterol

possibly as an important contributor, but interpretation of

these studies is hampered by the rapidity and extent of

cholesterol exchange that occurs in vivo. Recent work by

the group of Robins [548] indicates that all of the pre-

formed free cholesterol in the liver should be considered as

a single kinetic pool, potentially available for biliary secre-

tion. From a functional point of view, there seems to be an

interrelationship between pools available for VLDL pro-

duction and bile secretion or, alternatively, there may be a

single pool present for both secretory pathways; in general,

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 249

experimental procedures resulting in increased or de-

creased secretion of VLDL cholesterol have an opposing

effect on bile cholesterol secretion [514,515,530].

A number of studies from different laboratories have

shown that biliary PC is mainly derived from preformed

hepatic and extrahepatic pools. From recent work by Portal

et al. [549] it was concluded that in rats, under basal

conditions, choline-labeled HDL-PC contributes 38% of

total biliary PC as compared to 44% from a preformed

hepatic microsomal pool. It should be stressed, however,

that the relatively large contribution of HDL-PC reported

by these authors may reflect an unique feature of hepatic

HDL metabolism: i.e., of the way that HDL surface lipid is

handled by the liver. It has been shown that liposomal

bile-type PC, that enters the endosomal-lysosomal pathway

after uptake by hepatocytes, is extensively degraded and

remodeled and only to a limited extent secreted into the

bile [550]. De novo synthesis, mainly via the CDP-choline

pathway, contributes only to a limited extent to biliary PC:

values in the order of 3-14% have been reported for rats

using [J4C]choline as a precursor [551]. This relatively

small proportion is surprising considering the fact that

bile-salt induced PC secretion was dramatically reduced in

rats when PC synthesis was impaired by feeding choline-

deficient diets, while total hepatic PC content was not

altered by the experimental diets [513]. In this respect,

there seems to be an analogy to VLDL secretion, a process

that is also critically dependent upon the synthesis of PC

[552] and therefore is inhibited by feeding choline-defi-

cient diets to rats [552,553]. It thus seems that specific de

novo synthetized phospholipids, selected on the basis of

their biosynthetic routes, are required for both secretory

processes. Coenzyme labeling with [l,l-2H]ethanol, that

does not discriminate between both synthetic routes for

PC, suggested that liver, bile and plasma are supplied from

a common pool [554]. The high specific activity found in

bile compared to liver found in [14C]choline-labeling stud-

ies, on the other hand, were suggested to indicate that

newly synthetized biliary PC originates from a small hep-

atic pool with high turnover [555] although secretion by a

subset of hepatocytes should also be considered. In this

respect it is of interest to note that HDL particles, newly

synthetized and secreted by the perfused rat liver, were

found to have a PC composition very similar to that of

bile, whereas the PC composition of VLDL exactly mir-

rored that of whole liver [556]. In addition, secretion of

bile-type PC in HDL was stimulated by bile salts, suggest-

ing that PC of bile and HDL originates from the same pool

with both pathways of secretion susceptible to regulation

by bile salts [556].

5.3. Intracellular trafficking of biliary lipids

As can be inferred from the data discussed, the

'anatomical' localization of the putative precursor pool(s)

for biliary lipids within the liver cells has not been defined,

but probably includes components of the plasma mem-

brane, endoplasmic reticulum and Golgi network. Attempts

to identify these pools by extensive subcellular fractiona-

tion of livers during increased phospholipid flux, followed

by comparison of phospholipid composition and choles-

terol/phospholipid content of the fractions with that of

bile, have been unsuccessful [516]. Based on the knowl-

edge of lipid trafficking obtained with non-hepatic cell-

types [557], on studies demonstrating increased numbers of

vesicles in the pericanalicular region during high lipid

secretion [558,559] and on studies employing inhibitors of

intracellular vesicular transport [ 182,186], it has been in-

ferred that biliary lipids derived from (specific) areas of

the endoplasmic reticulum are transported towards the

canaliculus by vesicle movement, as discussed in detail in

a recent review by Coleman and Rahman [5]. The Golgi

apparatus and the trans Golgi network have been suggested

to play an important role in the sorting of the lipids,

analogous to their function in sorting and processing of

proteins. This is mainly based on the observation that it

shows transient localisation of fluorescent lipid precursors

[560] and of bile salts [193]. In 1975 Gregory et ai. [184]

suggested that phospholipid transport, probably together

with cholesterol, is vesicular. In the presence of micro-

tubule poisons phospholipid transport is inhibited by about

40-60%. Complete inhibition has not been reported. In

some studies cholesterol transport seems to be more sus-

ceptible to the inhibitory effect of these compounds than

lecithin [183]. Studies in which the effects of monensin, a

N a+ /K + ionophore interfering with Golgi function, on

lipid secretion were evaluated yielded conflicting results

[561,562], indicating that the role of the Golgi apparatus is

by no means clear at the moment. Nevertheless, in the

general view, endosomal (and Golgi-derived) vesicles are

transferred to the bile canaliculus in a microtubule-depen-

dent, i.e., colchicine and vinblastine-sensitive, fashion. The

nature of the vesicles, e.g., whether they transport bile-de-

stined lipids only or also carry other bile-destined com-

pounds to the canaliculus, whether they are remodelled

during their transport to the canaliculus by the action of

phospholipid transfer proteins and, most importantly, by

what mechanism they are targeted to the canaliculus, re-

mains unclear at the moment and awaits their isolation and

further characterization. Several groups have attempted to

isolate the vesicles involved taking advantage of the fact

that these vesicles should be enriched in phosphatidyl-

choline (PC) of the typical biliary PC type; 16:0, 18:2 and

16:0, 18:1. Graham et al. [563] reported the isolation of

these vesicles. They fractionated cells on gradients and

found biliary type vesicles in a fraction enriched in

canalicular markers. However, others (R. Coleman, per-

sonal communication) have not been able to confirm these

results and further studies are required to characterize the

putative biliary lipid transport vesicles. The question arises

why it is so difficult to isolate these vesicles. There are

250 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

two rather obvious reasons. Firstly, it can not be excluded

that they simply do not exist. Secondly, the vesicles may

be so short-lived that during the isolation procedures they

fuse rapidly with intracellular membranes. At high bile salt

fluxes accumulation of vesicles in the pericanalicular re-

gion of the hepatocyte has been observed [564]. It is not

yet clear whether these vesicles do indeed consist of lipids

destined for secretion into bile. Under normal, physio-

logical, conditions it appears to be difficult to visualize

such vesicles.

The vesicular pathway is most probably not the only

pathway mediating lipid secretion. Stein et al. [565] could

not demonstrate the inhibitory effect of microtubule poi-

sons at low rates of lipid secretion. Although these results

have to be confirmed they suggest that alternative path-

ways are operating as well. Recently Cohen et al. [566]

suggested a role for PC transfer protein, which is abun-

dantly present in the hepatocyte [567], in transport of PC

from ER to canalicular membrane. In an in vitro model,

employing the naturally fluorescent PC (i.e., 1-palmitoyl,

2-parinaroyl PC), they demonstrated a rapid transfer of this

phospholipid from small unilamellar vesicles modelling

endoplasmic reticulum to model canalicular membranes

(devoid of PC) which could be induced by submicellar

concentrations of bile salts (BS). Though submicellar,

these concentrations were still at least one order of magni-

tude higher than the estimated intracellular concentration

of bile salts. Furthermore, support for this hypothesis from

in vivo studies is, as yet, still lacking.

Non-vesicular transport to the bile canaliculus by lateral

diffusion has also been considered for delivery of lipids to

the bile canaliculus. In line with the first option, Robins et

al. [542] reported marked kinetic differences when hepatic

newly synthetized cholesterol and preformed cholesterol

were separately labeled and traced into the bile. The

specific activity of preformed cholesterol in the canalicular

membrane was closer to that in bile than in any other

organelle immediately after isotope administration, which

they explained by direct transport of these cholesterol

molecules from the blood through the plasma membrane to

the canaliculus, without entering the interior of the cell.

Although other interpretations of the data are possible, this

is a very intruiging possibility. Lateral diffusion of PC

through the inner leaflet of the plasma membrane may also

supply PC to the canalicular membrane. This could explain

the higher specific activity of canalicular membrane PC

compared to microsomal PC after injection of [3H]PC-

labeled HDL in rats. It could also explain the very marked

increase in canalicular membrane PC specific activity rela-

tive to that of the microsomes when phospolipid secretion

was inhibited by dehydrocholate under the above-men-

tioned conditions [549]. Lateral diffusion of (fluorescently

labeled) phospholipids from the basolateral membrane to

the apical pole, passing the tight junctional barrier via the

inner leaflet of the plasma membrane, has been reported to

occur in cultured polarized MDCK cells [557].

5.4. Regulation of bilia~ lipid secretion: the role of bile

salts

Although, as mentioned earlier, the stimulatory effects

of bile salts on lipid secretion have been known for more

than three decades, the fundamental issues of where and

how these compounds exert their effects still are controver-

sial. Concerning the localization of the bile salt action, the

discussion has been focused on the question whether it is

exerted at the intracellular as opposed to the canalicular

level, or, in other words, whether it is a matter of 'push-

ing' or of 'pulling'. It is evident that the processes of

supply and of actual secretion must occur coordinately to

maintain ongoing secretion, but, as will be discussed in

this section, available evidence strongly indicates that the

quantitative regulation takes place at the level of the bile

canaliculus. With respect to the intracellular bile salt ef-

fects, it has been suggested that intracellular mixed micelle

formation may occur by interaction of bile salts, during

their passage through the hepatocyte, with intracellular

membranes [568]. However, this assumption probably is

incorrect, not in the least because the intracellular bile salt

concentration is far below CMC. A large number of stud-

ies have been devoted to the interactions of bile salts with

the endoplasmic reticulum and the Golgi complex to evalu-

ate the role of these organelles in transcellular bile salt

transport and to provide a background for the intracellular

flow of lipids required to maintain bile salt-induced biliary

lipid secretion. As indicated in Section 5.2, and discussed

in detail by Coleman and Rahman [5], a number of these

studies have provided evidence that, en route from the

sinusoidal to the canalicular membrane, bile salts mobilize

the bile-destined lipids from their intracellular sources for

transport towards the canaliculus. The findings that bile

salt administration leads to (i) an increased abundance and

altered morphology of vesicles in the pericanalicular re-

gion in vivo [558,559], (ii) acceleration of the translocation

of Golgi-derived phospholipid analogues to the peri-

canalicular region in hepatocyte couplets and stimulation

of their biliary secretion in isolated perfused rat liver [560],

and (iii) an increase in vesicular transcytosis of compounds

such as horseradish peroxidase [569] and inulin [570], have

been interpreted to suggest that bile salts are able to direct

vesicular transport towards the bile canaliculus. The in-

volvement of microtubuli in the movement of bile-destined

vesicles was revealed in studies showing impaired lipid

secretion after colchicine or vinblastine treatment, although

the role of microtubuli-dependent (vesicular) transport was

only appreciable under the condition of high bile salt load

[186]. Furthermore, recent studies by Katagiri et al. [571]

demonstrated that colchicine does not inhibit lipid secre-

tion as induced by /3-muricholate and tauroursodeoxy-

cholate, implying that the cholchicine-effect may be spe-

cific for more hydrophobic bile salts, such as taurocholate.

The importance of proper functioning of another part of

the cytoskeleton, i.e., the actin filaments surrounding the

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 251

canaliculus, for lipid secretion was suggested by the

demonstration of reduced biliary lipid secretion after phal-

loidin treatment [572]. However, the question how bile

salts actually affect cytoskeleton-dependent vesicular trans-

port has remained unanswered until now. It is clear, on the

other hand, that bile salts are secreted into the bile unac-

companied by lipids: (i) bile salts are secreted in the form

of monomers (see Section 3), (ii) 'kinetic' studies in the

isolated perfused rat liver with very low endogenous bile

salt output showed that the secretion of (radiolabeled)

taurocholate preceeded the induced peak in lipid secretion

by minutes [573], and (iii) it is well known that several

hydrophilic organic anions are able to inhibit biliary lipid

secretion without affecting that of bile salts [574-576].

This latter phenomenon, usually referred to as 'uncou-

pling', has recently been used to provide evidence that,

under physiological conditions, quantitative regulation of

bile salt-induced lipid secretion is exerted at the level of

the canaliculus and not at some site within the hepatocyte.

In these studies, reviewed by Verkade et al. [577], it was

first shown that hydrophilic organic anions, that avidly

interact with biliary bile salts, are unable to uncouple

lipid-from-bile salt secretion in mutant ( G Y / T R - ) rats

[221,244]. These rats lack canalicular ATP-dependent or-

ganic anion transport and therefore show impaired bile

secretion and increased intracellular concentrations of the

uncoupling agents [259,275]. This observation demon-

strates that uncoupling takes place within the canaliculus

and this strongly suggests that the quantitative regulation

of lipid secretion by bile salts is exerted after secretion of

the bile salts across the bile canalicular membrane

[259,275].

As discussed in Section 3, bile salts build up extremely

high concentrations in the canalicular lumen, largely ex-

ceeding their CMC values. These bile salts subsequently

are able to induce the secretion from the canalicular mem-

brane of cholesterol and specific PC as a function of their

concentration and their hydrophobicity. It was recently

demonstrated that the residence time of the bile salts inside

the canaliculus represents an additional parameter in the

quantitative regulation of bile salt mediated lipid secretion

[578]. The specificity of the process at canalicular level is

documented very clearly by the release of bile-specific PC

species from isolated rat liver canalicular membranes in

vitro [579] as well as by retrograde, intrabiliary injection of

taurocholate and taurochenodeoxycholate in rats in vivo

[580]. In these latter studies, the detergent CHAPS induced

the instantaneous release of lipid with a general membrane

profile [580]. It has been suggested that the hydrophilic-hy-

drophobic balance of biliary bile salts, at least in part,

determines the selective composition of bile PC. More

hydrophobic bile salts stimulate the secretion of arachi-

donic-and stearic-rich PC in humans, implying a physical-

chemical mechanism for selective secretion [581-583].

Recent work [584], on the other hand, showed that in rats

selectivity was greatest at unphysiologically low secretion

rates and that selectivity was already partially lost when

PC secretion was in the physiological range. In these

experiments, bile salt hydrophobicity had no acute effects

on the selection of biliary PC species. The results of the

latter studies are in line with the well-known change in

biliary phospholipid composition to a more aspecific pat-

tern that occurs during the course of administration of bile

salts to rats at supraphysiological rates, eventually leading

to cholestasis [585,586]. The occurence of this type of

cholestasis has been explained by inadequate lipid supply

during excessive secretion leading to a depleted, function-

ally impaired, canalicular membrane. Under normal condi-

tions, the canalicular domain is the most rigid domain of

the hepatocytic plasma membrane, relatively enriched in

cholesterol and sphingomyelin. It has been suggested that

the canalicular membrane contains microdomains of higher

fluidity containing the bile-destined lipids, possibly repre-

senting fusion points of the supplying vesicles, that would

allow for a more easy interaction with the bile salts.

Although never demonstrated in canalicular membranes up

to now, such microdomains have been observed in other.

in vitro systems [587]. Bile salts have been shown to

induce the pinching-off of membrane lipids from red blood

cell membranes in the form of vesicles [588]. Accordingly,

it has been suggested that intercalation of bile salts into the

canalicular membrane (microdomains) would cause out-

ward curving and budding of the membrane at this point,

ultimately causing it to be pinched off as a vesicle of

specific, relatively hydrophilic lipid composition. Although

attractive, direct proof for such a mechanism of secretion

has not been provided.

5.5. Regulation of biliar3, lipid secretion: the role of mdr2

P-glycoprotein

Recent studies demonstrate that the process of lipid

secretion in the mouse is critically depending upon the

presence of the canalicular mdr2 P-glycoproteins

[125,672]. This dependency was revealed by the observa-

tion that bile of mice in which the gene encoding this

protein was disrupted by homologous recombination did

not contain measurable amounts of phospholipids in the

presence of normal bile salt output [125]. As phospholipid

secretion was the only function that was completely abol-

ished in the homozygous ( - / - ) mice and also the only

one that was substantially reduced, by 40%, in the het-

erozygous ( + / - ) mice, it is very likely that an inability

to deliver phospholipids to the bile represents the primary

defect in the mdr2 P-glycoprotein knock-out mice. Thus,

not included in any of the proposed mechanisms of biliary

lipid secretion sofar, it turns out that the delivery of PC

from the hepatocytes into the bile requires the active

participation of a canalicular membrane protein, which, of

course, is fully in line with a quantitative regulation of the

process at this level. Apparently, mdr2 P-glycoprotein

activity and canalicular bile salts cooperate functionally in

252 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

the lipid secretion process. Cholesterol secretion was re-

duced by about 90% in the ( - / - ) mice but virtually

normal in the heterozygotes, again underscoring the depen-

dency of cholesterol secretion on (a minimal level) of

phospholipid secretion. Bile salt composition in mice is

very hydrophilic ( + 60% muricholate with the remainder

taurocholate). When a more hydrophobic bile salt, tauro-

cholate, was infused into ( - / - ) mice cholesterol secre-

tion could be partly restored [672]. Thus, cholesterol and

phospholipid secretion can be partly uncoupled.

The mechanism of the mdr2 P-glycoprotein action re-

mains speculative at the moment. As discussed by Smit et

al. [125], it may facilitate the entry of (bile-specific)

phospholipids into the membrane, the translocation of

these phospholipids through the membrane as well as their

actual exit into the bile. An attractive model, in view of

recently published work on phospholipid translocation in

rat liver canalicular membranes by Berr et al. [589] as well

as the proposed mode of action of other P-glycoproteins

[590], involves the action of the protein as a 'flippase', i.e.,

translocating bile-specific PC from the inner to the outer

leaflet of the membrane. This process could result in a

local protrusion of PC-enriched buds into the canalicular

lumen that subsequently, under the influence of bile salts,

pinch off to form biliary vesicles. The data of Berr et al.

[589] suggest the presence of a flippase activity in the

canalicular membrane of rat liver with kinetic properties

similar to those of the system identified previously in

microsomes by Bishop and Bell [591]. In these studies the

transport of the water-soluble short chain PC analogue

L-a-dibutyroyl-glycero-3-PC was used as a measure of

flippase activity. Saturable uptake of the radiolabeled com-

pound was observed in microsomes as well as in canalicu-

lar membrane vesicles, with similar K m values but two-fold

higher Vma x values in the ¢analicular membranes compared

to microsomal membranes. This transport was not affected

by taurocholate and ATP-independent, and its relation to

bile-specific (long-chain) PC translocation remains to be

established. How this activity relates to mdr2 P-glyco-

protein, which on the basis of its relationship to other

MDR proteins would be expected to act in an ATP-depen-

dent manner, remains to be established. Very recently,

cytosol ~ canaliculus

~ -- -- - - ~ ~ biliary

~ - - - - " ~ ' ~ vesicles

E ~ m dr 2 ~ bile salt J ~ (rnicelles)

PHOSPHATIDYLCHOLINE CHOLESTEROL

"APICAL OUTER LEAFLET PHOSPHOLIPIDS" ~ "INNER LEAFLET PHOSPHOLIPIDS"

GLYCOSPHINGOLIPIDS PHOSPHATIDYLETHANOLAMINE SPHINGOMYELIN PHOSPHATIDYLSERINE

Fig. 13. Hypothetical model of the mechanism of mdr2 Pgp-mediated lipid secretion. Phosphatidylcboline is mainly supplied to the membrane via

phosphatidylcholine-transfer protein which inserts PC into the inner leaflet, mdr2 Pgp translocates PC to the outer leaflet into PC-rich microdomains and

this ATP-dependent process leads to phospholipid imbalance in the membrane. Luminal bile salt micelles or monomers further destabilize these domains

and this leads (via an unknown mechanism) to the formation and release of vesicular structures. In the absence of PC-rich domains luminal bile salt

micelles are unable to extract phospholipids from the (outer leaflet of the) membrane [125]. This may be caused by a high proportion of

glycosphingolipids, sphingomyelin and cholesterol in the outer leaflet. Luminal bile salts can, however, extract cholesterol directly from the outer leaflet

[672]. This extraction is more efficient when mixed micelles of bile salts and PC are present because these have a higher affinity for cholesterol.

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 253

Ruetz and Gros [420] provided experimental data which

support the model of mdr2 Pgp as a primary active

flippase. They transfected the mdr2 gene into the yeast

sec6-4 mutant which has a defect in the secretory appara-

tus. This mutation leads to the accumulation of plasma

membrane proteins in secretory vesicles. After isolation,

these vesicles were used to demonstrate translocation of

fluorescent NBD-PC from the inner to the outer leaflet.

The translocation was strictly ATP-dependent and specific,

based on the observation that vesicles from yeast trans-

fected with the mdrla gene were inactive. Interestingly,

verapamil inhibited this translocation while colchicine and

vinblastine were without effect. Further evidence support-

ing the flippase function of mdr2 Pgp was provided by

Smith et al. [673] who used fibroblasts from transgenic

mice that express MDR3, the human homologue of mdr2

Pgp. After metabolic labeling of intracellular phosphatidyl-

choline with radioactive choline, translocation from the

inner to the outer leaflet was assayed and this was com-

pared with normal mouse fibroblasts which do not express

mdr2. Translocation of radioactive phosphphatidylcholine

to the outer leaflet was measured by the possibility to

exchange with phosphatidylcholine-transfer protein and li-

posomes in the medium. In MDR3 expressing fibroblasts a

more rapid translocation of PC was observed than in

control fibroblasts.

In conclusion, these data suggest that mdr2 Pgp may

indeed be a primary active PC translocator. The passive

flippase described by Berr et al. [589] probably represents

another function. Since it was demonstrated that conven-

tionally purified canalicular membrane fractions contain

significant amounts of intracellular membranes [105] this

flippase activity may also have an intracellular localiza-

tion.

The inclusion of PC-flipping in a model describing the

mechanism of biliary PC secretion assumes that lipid

asymmetry of the canalicular membrane is an important

parameter and clearly imposes restraints on the mechanism

of lipid supply to the membrane. If PC is supplied via

phosphatidylcholine transfer protein to the inner leaflet

there is an obvious role for mdr2 Pgp in the translocation

of PC from the inner to the outer leaflet of the membrane.

If, however, vesicular supply is the major mechanism, a

substantial amount of PC molecules will already be present

in the outer leaflet of the membrane because PC may be

expected to be present in both leaflets of such vesicles. In

that case a substantial amount of PC is already present in

the outer leaflet of the canalicular membrane and therefore

available for bile salt-mediated secretion without the need

for a flippase. By virtue of its primary active transport of

PC, mdr2 Pgp may induce a surplus of phospholipid

molecules in the outer leaflet compared to the inner leaflet

of the canalicular membrane. Such a lipid imbalance can

create localized instable structures in the membrane (mi-

crodomains) and these strucures may be especially vulner-

able to bile salt-mediated extraction. Under the influence

of bile salts these instable structures may pinch off as

vesicles, since it has been observed that lipids in primary

bile exist as vesicles [592] (see Fig. 13Fig. 13). How this

happens is, however, as yet, completely unclear. Perhaps

new emerging rapid-fixation techniques will solve this

urgent problem. Recently, Crawford et al. [593] employing

an ultra-rapid freezing technique have succeeded in visual-

izing vesicular structures attached to the canalicular mem-

brane. Whether these vesicles represent intermediate struc-

tures in the biliary lipid secretion process is not yet clear.

5.6. Functional aspects o( biliar 3, lipid secretion

A number of physiological functions for the biliary

lipids can be discerned, some of them well-recognized and

some appreciated to a lesser extent. Apart from conversion

into bile salts, nature has not provided an enzymic machin-

ery for the breakdown of cholesterol molecules. Therefore,

the biliary pathway represents first and foremost the major

route for the removal of cholesterol from the body and

functions as a crucial factor in the maintenance of choles-

terol homeostasis [594]. Somewhat paradoxically, it ap-

pears that a large part of the secreted molecules are

reabsorbed from the intestine [595], which may, however,

reflect the presence of a control mechanism in cholesterol

metabolism at the intestinal level.

In bile, the presence of cholesterol and phospholipids

may protect the hepatocytes and the cell lining of the bile

ductules from the cytotoxic effects of the (detergent) bile

salts. This function is probably best examplified by the

development of liver disease in the mdr2 P-glycoprotein

knock-out mice produced by Smit et al. [125] which is

mild directly after birth but progresses to severe liver

disease during lifetime and ends with the development of

hepatocellular carcinoma. Liver histology showed degener-

ative features (of hepatocytes) throughout the lobule, irreg-

ular size with nuclear polymorphism, and focal necrosis

with formation of eosinophilic bodies. Upon electronmi-

croscopic examination, a prominent widening and turtuos-

ity of the canaliculi with Loss of microvilli was observed.

In these mice, the bile ducts were affected as well, with

extensive portal expansion owing to ductular proliferation.

This pathology may be the result of prolonged exposure of

apical membranes along the biliary tree to the detergent

action of lipid-free bile salt micelles [596].

Another (potential) function of the biliary lipids may

involve the hepatobiliary transport of certain hydrophobic

organic anions. As shown initially by Tazuma et al.

[597,598], several hydrophobic organic anions show a very

high affinity for the (vesicular) lipid fraction in bile and it

has been suggested that these vesicular structures act as

carriers of these compounds during their transit from the

liver to the intestine. To what extent the presence/absence

of biliary lipids actually affects the transport kinetics of the

organic anions remains to be established.

254 R.P.J. Oude Elferink et aL / Biochimica et Biophysica Acta 1241 (1995) 215-268

Finally, biliary phospholipids play an important role in

the absorption of dietary lipids from the intestine, addi-

tional to the well-established function of bile salts in this

process. Studies by Davidson et al. [599] have provided

evidence for a physiologic role of biliary phospholipids in

the regulation of intestinal apolipoprotein B48 expression,

an apolipoprotein essential for proper assembly and secre-

tion of chylomicrons containing the absorbed dietary fats.

6. Regulation of canalicular transport and its conse-

quence for development of cholestasis

As outlined in the introductory section, the generation

of bile flow critically depends upon (active) transport of a

number of organic solutes from hepatocytes into the

canalicular lumen. Consequently, disturbances of these

transport processes may lead to impairment of bile forma-

tion; i.e., cholestasis.

6.1. Regulation o f transport

Bile flow is not a continuous process. It is subject to

diurnal variations and depends on the delivery of bile salts

and non-bile salt organic anions and cations to the canalic-

ular lumen. For example, in many species, but not in the

rat, the bile salt flux to the liver is greatly enhanced after

gallbladder contraction. The gallbladder contains 60% of

the total bile salt pool. It contracts after each meal and

several times per day during the spontaneous phases of

gastrointestinal motitility. These fluxes in substrate supply

need to be modulated in order to dampen intracellular bile

salt concentration changes. Enhanced secretion at high

concentration and decreased efflux during situations of low

concentration could be of importance here. The liver as

part of the digestive system is under the control of gastro-

intestinal hormones. For example, secretin not only stimu-

lates bicarbonate secretion [600] but also enhances biliru-

bin secretion [601-603]. Gastrointestinal hormones may be

candidates for the fine tuning of the secretory processes in

the liver as part of the digestive process. Against this

background regulation of canalicular transport processes

may be relevant and it is even possible that some forms of

cholestasis may result from a disordered regulation of bile

secretion.

Regulation by recruitment o f transporters f rom intra-

cellular stores

Canalicular cartier proteins that are involved in hepato-

biliary transport are synthesized in the endoplasmic reticu-

lum and are then transported through the Golgi network to

the basolateral membranes of the hepatocyte [604]. From

there they are directed to the canalicular domains by way

of a microtubule-dependent transcytotic pathway [199].

Canalicular transport activity depends on synthesis and

processing of the carder proteins in the ER and Golgi, and

on transport and removal of the carrier proteins to and

from the canalicular membranes. Regulation at the level of

transcription and translation represents long-term regula-

tion. For example, in primary rat hepatocyte cultures, the

sodium-dependent taurocholate cotransporting polypeptide

is down-regulated at the mRNA level [159]. In addition to

this type of regulation, canalicular transport activity can be

decreased during cholestasis by a redistribution of the bile

salt and organic anion transporters from the canalicular

either to intracellular compartments or to the basolateral

domains [605,606]. Whether, and if so to what extent, such

processes actually contribute to the etiology of (certain

forms of) cholestasis remains to be established.

Intracellular storage of transporter proteins represents a

flexible type of regulation, whereby a cell can rapidly

respond to fluctuations of hormones. The existence of this

type of regulation has been extensively described for the

glucose transporter GLUT4. Adipocytes and muscle cells

respond to insulin by recruitment of glucose transporters

from intracellular vesicles to the plasma membrane. In this

way the hormone prepares the cell for increased glucose

transport [607]. Also the hepatic receptor for epithelial

growth factor responds in this way [608]. There are indi-

rect indications that canalicular carrier proteins are also

present in intracellular vesicles [105,609]. For example,

evidence was found for MOAT activity in small intra-

cellular organelles in isolated hepatocytes. This activity

was absent in hepatocytes of TR rat livers [609]. Using

free flow electrophoresis, Kast et al. isolated three subcel-

lular fractions: endoplasmic reticulum with electrogenic,

ATP-independent, taurocholate transport; canalicular mem-

branes with both ATP-dependent taurocholate transport

and ecto-ATPase activity; and an as yet unidentified frac-

tion with ATP-dependent taurocholate transport but devoid

of ecto-ATPase [105]. These results provide circumstantial

evidence for the presence of a pool of subcellular or-

ganelles with canalicular ATP-dependent transport activity.

Apparently these organelles do not simply represent re-

trieved or endocytosed canalicular membranes since they

probably contain a selective repertoire of canalicular mark-

ers and, for example, lack canalicular ecto-ATPase. This

subfraction may represent a reservoir of canalicular bile

salt transporters, waiting for a signal to be inserted into the

canaliculus. This potential intracellular pool of transporters

may be recruited by certain stimuli and this would allow

short-term regulation of canalicular transport. Whether this

is indeed the case needs to be proven. However, it is

evident that hormones and intracellular messengers can

modulate canalicular transport activity. For example,

canalicular and ductular C1- /HCO 3 exchange activity

mediates biliary bicarbonate secretion. The canalicular ac-

tivity can be stimulated by cyclic-AMP [99]. This increase

can be blocked by colchicine. The colchicine effect sug-

gests that a C1 / H C O 3 exchanger is targeted from a

pre-formed pool of exchanger proteins to the canalicular

domain via a microtubule-dependent pathway.

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 255

Both the group of Haussinger [610-612] and of Boyer

[613] demonstrated rapid upregulation of maximal bile salt

transport capacity by cell swelling, although the observed

effects differed quantitatively. Since the effect was acute

and maximal transport capacity increased it is reasonable

to assume that in some way cell swelling induces recruit-

ment of intracellular transport protein stores to the canalic-

ular membrane. Several, more physiological signals may

have the same effect as cell swelling through a common

pathway of second messengers (Fig. 14Fig. 14). In hepato-

cyte couplets, cAMP causes the microtubule-dependent

targeting of bile salt transporters to the canalicular do-

mains with a concomittant increase in canalicular secretion

of bile salts [614]. These observations indicate that intra-

cellular storage of active transporters is possible and this

may be a mechanism that regulates transport activity on

the canalicular membrane.

Regu la t ion by s e c o n d m e s s e n g e r s

Intracellular Ca 2+ can be modulated by mobilizing

Ca 2+ from intracellular inositol-triphosphate (IP3)-sensi-

tive pools or by opening hormone-sensitive Ca2+-channels

to allow entry of extracellular Ca 2+ [615]. Vasopressin

allows Ca-'" entry from the intracellular pool and from

extracellular Ca z~ [616]. Vasopressin and the Ca 2+

ionophore A23187 both inhibit bile flow. 2,5-di(tert-

butyl)-l,4-benzohydroquinone mobilizes C a 2+ from the

intracellu[ar stores only and also inhibits bile flow [616]. In

contrast, tauroursodeoxycholic acid causes an increase in

intracellular Ca 2 ~ through mobilization of Ca 2 + from both

the intracellular stores as well as extracellular Ca 2+. Yet,

tauroursodeoxycholic acid stimulates bile flow and re.-

verses the cho[estatic action of other agents [616]. En-

dothelin-1, also increases intracellular Ca 2+. In low dose it

enhances bile flow and in high dose it causes cholestasis

[617,618]. This seems contradictory and suggests a com-

plex relation between intracellular Ca 2+ and bile flow.

This complexity also follows from the work of Hamada et

al. who have shown that glucagon and vasopressin, hor-

mones that act via cAMP and intracellular Ca 2÷ respec-

tively, both increase bile flow in the isolated perfused rat

liver and, when given together, act synergistically

[619,620]. Little is known about the regulation of individ-

ual transporters by phosphorylation because only few

transporters are characterized at the polypeptide level. The

effect of phosphorylation on the activity of P-glycoprotein

has been extensively discussed in Section 4. The canalicu-

lar transporter for non-bile salt organic anions is either

directly or indirectly stimulated by activation of protein

kinase C [621]. cAMP has been shown to stimulate vesicu-

lar transport of horse radish peroxidase [622] and to stimu-

late the bile salt-independent bile flow [623].

6.2. Choles tas i s

Normal bile flow results from the concerted action of a

series of transport carriers in the liver: transporters in the

basolateral membrane that mediate the Na%dependent and

Na+-independent uptake of bile salt and non-bile salt

organic anions and cations, microtubuli-dependent

vesicle-mediated transcytosis and intracellular transport by

as yet undefined routes, and the various ATP-dependent

canalicular transporters. Dysfunction of any of these trans-

porters or pathways or disorders of energy supply, will

result in cholestasis. Clinical cholestasis is characterized

by either 'cholate stasis' or 'bilirubinostasis', terms coined

by Desmet [624]. Early primary bi[iary cirrhosis, when

there is no jaundice yet, is characterized by "cholate stasis'

hypotonic swelling cAMP

ursodeoxycholate

0 BS

BS

hypertonic shrinking

Fig. 14. Potential regulation of canalicular transport activity by recruitment of transporters from intracellular stores. A traction of the canalicular transporter proteins may be stored in intracellular compartments, probably endosomes. Upon proper signals, like regulatory volume decrease, cAMP and ursodeoxycholate these membranes may be inserted in the canalicular membrane, which leads to an increased transport maximum. Adapted from Ref. [668].

256 R.P.J. Oude Elferink et al./Biochimica et Biophysica Acta 1241 (1995) 215-268

Table 2 Cholestatic drugs and agents

Cholestatic agent Animal Reference

Taurocholate rat [656] Taurolithocholate rat,hamster [657,658] Glycolithocholate-3 a-O-sulfate rat [659.660] Lithocholate-3 ot-O-glucuronide rat [661] Ethinylestradiol rat [636,662] Estradiol- 17B-glucuronide rat [641 ] Chlorpromazine rat [663] Cyclosporin rat [664,665] Manganese-bilirubin rat [643,666] a-Naphthylisothiocyanate rat,mouse [645] Phalloidin rat [629]

with elevated serum bile salt levels and increased alkaline

phosphatase. Byler's disease may be the best example of

'cholate stasis' marked by high serum bile salt levels and

only very mild jaundice. Examples of 'bilirubinostasis' are

leptospirosis, with greatly elevated serum bilirubin levels

caused by a combination of decreased biliary secretion and

increased supply because of hemolysis, Dubin-Johnson and

Rotor syndromes with elevated serum levels of conjugated

bilirubin and normal alkaline phosphatase activity and

normal bile salt levels. Most cases of cholestasis are

mixtures of 'cholate stasis' and 'bilirubinostasis'.

Experimental models of cholestasis include drug-induced

cholestasis

Table 2 gives an overview of various drugs and chemi-

cals that cause cholestasis in animals. Major actions of

some of these agents include: a decrease of N a + / K +

ATPase activity and an increase of membrane viscosity of

predominantly the sinusoidal membrane [625]; an increase

of tight-junctional permeability [626]; an increase of intra-

cellular calcium [627]; changes of the microcirculation

[628]; effects on the cytoskeleton with irreversible poly-

merization of microfilaments, which has consequences for

both transcellular transport as well as junctional permeabil-

ity [629]. Apart from reduced transport activities cholesta-

sis can be induced or aggravated by the disappearance of

the canalicular contractions. These contractions are Ca 2 +-

dependent and agents that block Ca 2+ release from IP 3-

sensitive stores, also inhibit canalicular contractions, cGMP

generated by nitric oxide is such an agent [614]. Canalicu-

lar contractions represent a major propulsive force for the

bile flow. The pathophysiological importance of these

contractions is not well understood but one could imagine

a sort of canalicular 'ileus', similar to intestinal ileus, in

their absence. Phalloidin-induced cholestasis is most prob-

ably caused by its irreversible polymerisation of actin into

microfilaments [629].

Chronic administration of cyclosporin A reduces the

hepatic excretory function in man and causes cholestasis in

the rat [630,631]. Cyclosporin is an important immuno-

suppressive drug, used e.g., after organ transplantation. It

is metabolized and secreted by the liver. Cyclosporin A

inhibits ATP-dependent uptake of bile salts, leukotriene

C 4, and daunomycin in human and rat liver canalicular

membrane vesicles [632,633]. In addition, the drug inhibits

Na+-dependent taurocholate uptake in sinusoidal mem-

brane vesicles [634]. Also in cultured rat hepatocytes,

cyclosporin reduces both taurocholate uptake and effiux

[635]. However, ATP-dependent canalicular transport of

bile salts shows the most potent inhibition suggesting that

inhibition of biliary bile salt secretion is the main cause of

cyclosporin-induced cholestasis [632,633].

Ethinyl estradiol is also a cholestatic agent which in the

rat reduces bile flow [636]. In isolated rat hepatocytes it

reduces the uptake of taurocholate [637]. Further proof that

ethinyl estradiol causes alterations in the sinusoidal mem-

brane and in sinusoidal membrane function came from

Berr et al. and Rosario et al. [625,638]. They showed that

the drug decreases Na+-K + ATPase activity and tauro-

cholate uptake in isolated rat hepatocytes and that alter-

ations of sinusoidal membrane fluidity may play a role.

Bossard et al. confirmed this effect by showing a reduction

of taurocholate uptake in sinusoidal membrane vesicles but

they also showed that the most dramatic change is a

decrease of ATP-dependent canalicular bile salt transport,

as measured in canalicular rat liver membrane vesicles

[605]. In addition the ATP-dependent transport of the

glutathione-conjugate, GS-DNP, in canalicular membranes

from ethinyl estradiol-treated rats is reduced [605]. As was

seen with cyclosporin, ethinyl estradiol-induced cholestasis

mainly results from an inhibition of the main concentrative

step in hepatobiliary secretion, the ATP-dependent canalic-

utar secretion of bile salts.

Estradiol- 17/3(/3-D-glucuronide) causes a reversible type

of cholestasis by inhibiting the hepatobiliary secretion of

taurocholate and non-bile salt organic anions [639,640].

This steroid inhibits the biliary excretion not the hepatic

uptake of taurocholate [639]. In addition, it increases the

tight-junctional permeability [626]. Estradiol-3/3(/3-D-

glucuronide) and estriol-3/3(/3-o-glucuronide) are not

cholestatic [641]. Estriol 17/3(/3-D-glucuronide), estriol-

16a( /3-D-glucuronide) and testosterone-17/3(/3-D-

glucuronide) show intermediate cholestatic potency [642].

Infusion of manganese together with bilirubin induces

cholestasis probably by precipitation of manganese-biliru-

bin in the biliary tree [643,644]. Similarly, c~-naphthyli-

sothiocyanate (ANIT) [645] causes severe necrosis of bil-

iary epithelial cells which leads to obstruction of the

biliary tree and subsequent cholestasis. In addition it in-

creases tight junction permeability.

Expression and localization of transporters during extra-

hepatic cholestasis

Bile duct ligation, as an experimental model for extra-

hepatic cholestasis, leads to dramatic changes in the ex-

pression and domain-specific localization of plasma mem-

R.P.Z Oude Elferink et al. /Biochimica et Biophysica Acta 1241 (1995) 215-268 257

brane proteins. In general it appears to induce a shift of

canalicular membrane proteins from their specific apical to

a basolateral or random localization. This was shown in

several immunocytochemical studies using antibodies

against canalicular antigens [646,647], but also by means

of functional studies [606]. It was also shown that transcy-

tosis, both fluid phase and receptor-mediated, is abrogated

during complete cholestasis [199,647,648] and impaired

during reduced bile flow [649]. Several mechanisms may

be proposed for the altered localization of canalicular

membrane proteins. Firstly, the impaired transcytosis can

be due to a compromised fusion process of transcytotic

vesicles with the canalicular membrane. This may also

hold for the fusion of vesicles containing de novo synthe-

sized canalicular membrane proteins. Indeed, accumulation

of canalicular antigens in pericanalicular vesicles was ob-

served during bile duct ligation [647] very similar to what

was demonstrated for the transcytotic marker, polymeric

IgA [199]. If fusion of such vesicles with the basolateral

plasma membrane is not impaired this will lead to a net

redistribution of proteins. It was shown that after release of

the bile duct ligation the pericanalicular vesicles rapidly

disappear, probably by fusion with the canalicular mem-

brane [647]. Secondly, canalicular membrane proteins may

be retro-endocytosed during bile duct obstruction and redi-

rected to the basolateral domain. Thirdly, the increase in

paracellular permeability that was reported to occur during

bile duct ligation [647] indicates that the barrier function of

the tight junction is drastically decreased. This may lead to

lateral redistribution of canalicular proteins over the entire

plasma membrane including the basolateral domain.

Bile duct ligation also affects the expression of trans-

port proteins. In a preliminary report Gartung et al. [650]

showed that it leads to down-regulation of the basolateral

Na+-dependent bile salt co-transporter. Similarly, Kothe et

al. [651] found a strong decrease in the activity of the

canalicular organic anion transporter as well as a complete

abrogation of the canalicular GSH secretion. The latter

may, however, be due to increased gamma-glutamyltrans-

peptidase activity. Reconstruction of bile flow led to a

slow recovery of cMOAT activity, which may reflect the

rate of synthesis of this transporter.

7. Perspectives

In the past decades a vast amount of research has been

invested by many groups in unraveling the mechanisms of

hepatobiliary transport. Although this taught us much about

the physiology of several transport processes, this work

had not led to the unequivocal assignment of transport

functions to specific membrane proteins. In the past few

years, however, this situation is dramatically changing and

several candidate proteins have now been cloned which

play a function in uptake of organic compounds by the

liver. The introduction of molecular biology and more

specifically that of expression cloning has caused a break-

through in this deadlock situation. It may be expected that

in the near future many more transporters, including those

of the canaliculus, will be recognized. This represents an

important advance in hepatology because the use of cDNA

probes will allow the analysis of expression of these

transporters under different physiological and pathological

conditions. This opens possibilities to analyse the molecu-

lar mechanism of action of these transporters and the

definition of the multiplicity of transport proteins involved

in the secretion of very diverse compounds. Only then one

can start to explain the different conditions of intrahepatic

cholestasis that are presently not understood. The exact

molecular description of transport mechanisms will also be

of importance to pharmacology; it will give a clue to the

diverse forms of drug-induced liver disease, and it will

eventually also open the way to a rational design of drugs

that are specifically delivered to parenchymal and non-

parenchymal cells, to hepatocellular carcinomas and to

cells of the biliary tree. It may be that eventually a

similarly impressive collection of gene products for trans-

port will be known as now already exists for drug

metabolism.

Because of the high ultrastructural complexity of the

liver as an organ it will always remain essential to study

hepatobiliary transport processes in the intact organ and

animal. Unfortunately, until now no model system has

been developed that can properly mimick biliary transport.

The canaliculus is unique in its membrane composition

and localization in the cell and in the organ. Even the

hepatocyte couplet that partially retains its polarity and is

generally seen as an attractive model system, has no free

flowing primary bile. It may therefore under many condi-

tions share more properties with a cholestatic hepatocyte

than with a normal hepatocyte. The need for intact animals

is well exemplified by the discovery of the function of

mdr2 P-glycoprotein. Although this protein and its gene

were well characterized it has taken many years before a

(putative) function in phospholipid secretion could be pro-

posed; no other model system than the intact liver exists in

which bile salt mediated phospholipid secretion can be

studied. It was therefore not until a transgenic knockout

animal was available that the involvement of mdr2 P-

glycoprotein in this complex process could be observed.

Mutant and transgenic animals will therefore remain in-

valuable for studies into hepatobiliary transport.

The complexity of the hepatocyte also explains why cell

biology of the hepatocyte as an epithelial cell lags behind

that of other epithelia like enterocytes and proximal tubule

cells. Relatively little is known about the exact biosyn-

thetic routing of canalicular (transport) proteins, let alone

about its possible defects. Nevertheless, the use of polariz-

ing hepatocyte cell lines will open a new, exciting area of

research.

258 R.P.Z Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

Acknowledgements

The authors acknowledge the support from the EEC

program BIOMED I, project #BMHI-CT93-1436; 'Design

and development of therapeutic and diagnostic agents for

the treatment of diseases of the liver and biliary tract'.

References

[1] Klaassen, C.D. and Watkins III, J.B. (1984) Pharmacol. Rev. 36,

1-67.

[2] Arias, I.M., Che, M.X., Gatmaitan, Z., Leveille, C., Nishida, T. and

Stpierre, M. (1993)Hepatology 17, 318-329.

[3] Boyer, J.L., Graf, J. and Meier, P.J. (1992) Annu. Rev. Physiol. 54,

415-438.

[4] Nathanson, M.H. and Boyer, J.L. (1991) Hepatology 14, 551-566.

[5] Coleman, R. and Rahman, K. (1992) Biochim. Biophys. Acta 1125,

113-133.

[6] Coleman, R. (1987) Biochem. J. 244, 249-261.

[7] Erlinger, S. (1987) in Physiology of the Gastrointestinal Tract

(Johnson, L.R., ed.), pp. 1557-1580, Raven Press, New York.

[8] Anwer, M.S., Berk, P.D., Suchy, F.J. and Wolkoff, A.W. (1992)

Hepatology 15, 1179-1193.

[9] Crawford, J.M., Strahs, D.C.J., Crawford, A.R. and Barnes, S.

(1994) J. Lipid Res. 35, 1738-1748.

[10] Petzinger, E. (1994) Rev. Physiol. Biochem. Pharmacol. 123, 47-

211.

[11] Tiribelli, C. (1992) J. Hepatol. 14, 385-390.

[12] Kuipers, F., Dijkstra, T., Havinga, R., Van Asselt, W. and Vonk,

R.J. (1985) Biochem. Pharmacol. 34, 1731 - 1736.

[13] Kuipers, F., Havinga, R., Bossenieter, H., Toorop, G.P., Hindriks,

F.R. and Vonk, R.J. (1985) Gastroenterology 88, 403-411.

[14] Remie, R., Rensema, J.W. and Kuipers, F. (1991) in Biliary

Excretion of Drugs and Other Chemicals (Siegers, C.P. and

Watkins, J.B., eds.), pp. 127-145, G. Fisher Verlag, Stuttgart.

[15] Mulder, G.J., Scholtens, E. and Meijer, D.K.F. (1981) in Methods

in Enzymology, Vol. 77 (Jakoby, W.B., ed.), pp. 21-30, Academic

Press, New York.

[16] Accatino, L., Hono, J., Koenig, C., Pizarro, M. and Rodriguez, L.

(1993) J. Hepatol. 19, 95-104.

[17] Meijer, D,K.F., Weitering, J.G. and Vermeer, G.A. (1983) Eur. J.

Clin. Pharmacol. 24, 549-556.

[18] Herman, R.H., Redinger, R.N. and Small, D.M. (1971) Surgery

Forum 22, 378-380.

[19] Askin, J.R., Lyon, D.I., Shull, S.D., Wagner, C.I. and Soloway,

R.D. (1978) Gastroenterology 74, 560-565.

[20] Dujovne, C.A., Gustafson, J.H. and Dickey, R.A. (1982) Clin.

Pharmacol. Ther. 31, 187-194.

[21] Vonk, R.J., Kneepkens, C.M.F., Havinga, R., Kuipers, F. and

Bijleveld, C.M.A. (1986) J. Lipid. Res. 27, 901-904.

[22] Bijleveld, C.M.A., Vonk, R.J., Kuipers, F., Havinga, R., Boverhof,

R., Koopman, B.J., Wolthers, B.G. and Fernandes, J. (1989)

Gastroenterology 97, 427-432.

[23] van der Meer, R., Welberg, J.W.M., Kuipers, F., Kleibeuker, J.H.,

Mulder, N.H., Termont, D.S.M.L., Vonk, R.J., De Vries, H.T. and

De Vries, E.G.E. (1990) Gastroenterology 99, 1653-1659.

[24] Angelin, B., Arvidsson, A., Dahlqvist, R., Hedman, A. and

Schenk-Gustafsson, K. (1987) Eur. J. Clin. Invest. 17, 262-265.

[25] Hedman, A., Angelin, B., Arvidsson, A., Beck, O., Dahlqvist, R.,

Nilsson, B., Olsson, M. and Schenck-Gustafsson, K. (1991) Clin.

Pharmacol. Ther. 49, 256-262.

[26] Hedman, A., Angelin, B., Arvidsson, A., Dahlqvist, R. and Nils-

son, B. (1990) Clin. Pharmacol. Ther. 47, 20-26.

[27] Meijer, D.K.F. and Nijssen, H.M.J. (1991) in Perfused Liver:

Clinical and Basic Applications (Ballet, F. and Thurman, R.G.,

eds.), pp. 165-208, John Libbey, London.

[28] Nijssen, H.M.J., Pijning, T., Meijer, D.K.F. and Groothuis, G.M.M.

(1992) Hepatology 15, 302-309.

[29] Meijer, D.K.F., Keulemans, K. and Mulder, G.J. (1981) Methods

Enzymol. 77, 81-94.

[30] Reichen, J. (1991) in Research in Perfused Liver (Ballet, F. and

Thurman, R.G., eds.), pp. 149-163, John Libbey, London.

[31] Pang, K.S., Goresky, C.A. and Schwab, A.J. (1991) in Research in

Perfused Liver (Ballet, F. and Thurman, R.G., eds.), pp. 259-302,

John Libbey, London.

[32] Meijer, D.K.F., Vonk, R.J., Keulemans, K. and Weitering, J.G.

(1977) J. Pharmacol. Exp. Ther. 202, 8-21.

[33] Proost, J.H., Nijssen, H.M.J.. Strating, C.B., Meijer, D.K.F. and

Groothuis, G.M.M. (1993)J. Pharmacokinet. Biopharm. 21, 375-

394.

[34] Groothuis, G.M.M., Hardonk, M.J., Keulemans, K.T.P., Nieuwen-

huis, P. and Meijer, D.K.F. (1982) Am. J. Physiol. 243. G455-

G462.

[35] Groothuis, G.M.M., Hardonk, M.J. and Meijer, D.K.F. (1985)

Trends Pharmacol. Sci. 6, 322-327.

[36] Groothuis, G.M.M. and Meijer, D.K.F. (1992) Enzyme 46, 94-138.

[37] Pang, K.S., Lee, W.F., Cherry, W.F., Yuen, V., Accaputo, J.,

Schab, A.J. and Goresky, C.A. (1988) J. Pharmacokinet. Biopharm.

16, 595-632.

[38] Nijssen, H.MJ., Pijning, T., Meijer, D.K.F. and Groothuis, G.MM.

(1991 ) Biochem. Pharmacol. 42, 1997-2002.

[39] Blom, A., Scaf, A.H.J. and Meijer, D.K.F. (1982) Biochem. Phar-

macol. 31, 1553-1565.

[40] Goreski, C.A., Ziegler, W.H. and Bach, G.G. (1970) Circ. Res. 27,

739-764.

[41] Mischinger, HJ., Walsh, T.R., Liu, T., Rao, P.N., Rubin, R..

Nakamura, K., Todo, S. and Starzl, T.E. (1992) J. Surg. Res. 53,

158-165.

[42] Riedel, G.L., Scholle, J.L. and Shepherd, A.P. (1983) Am. J.

Physiol. 245, 769-774.

[43] Palmen, N.G.M., Evelo, C.T.A., Borm, PJ.A. and Henderson, P.T.

(1992) Toxicol. In Vitro 6, 357-365.

[44] Pang, K.S. (1919) in Drug Metabolism and Drug Toxicity (Mitchell,

J.R. and Homing, M.G., eds.), pp. 331-352, Raven Press. New

York.

[45] Ullrich, V. and Diehl, H. (1971) Eur. J. Biochem. 20, 509-512.

[46] Evans, A.M., Hussein, Z. and Rowland, M. (1993) J. Pharm. Sci.

82, 421-428. [47] Pang, K.S., Cherry, W.F., Accaputo, J., Schwab, A. and Goresky,

C.A. (1988)J. Pharmacol. Exp. Ther. 247, 690-700.

[48] Reichen, J. (1988) J. Clin. Invest. 81, 1462-1469.

[49] Berry, M.N. and Friend, D.S. (1969)J. Cell. Biol. 43. 506-520.

[50] Krumdieck, C.L., Dos Santon, J.E. and Ho, K.J. (1980) Anal.

Biochem. 104, 118-123.

[51] Connors, S., Ramlin, D., Gandolfi, A.J., Krumdieck, C.L., Koep,

L.J. and Brendel, K. (1990) Toxicology 61, 171-183.

[52] Dogterom. P. and Rothuizen, J. (1993) Drug. Metab. Dispos. 21,

705-709.

[53] Smith, P.F., Krack, G. and McKee, R.L. (1986) Biol. Cell. 22,

706-712.

[54] Saadker, G.W., Slooff, MJH. and Groothuis, G.MM. (1992)

Biochern. Pharmacol. 43, 1479-1485.

[55] Sandker, G.W., Weert, B., Olinga, P., Wolters, H., Slooff, MJ.H.,

Meijer, D.K.F. and Groothuis, G.M.M. (1994) Biochem. Pharma-

col. accepted,

[56] Brendel, K., Gandolfi, A.J., Krumdieck, C.L. and Smith, P.F.

(1987) Science 8, 11-15.

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 259

[57] Powis, G. (1989) Drug Metab. Rev. 20, 379-394.

[58] Thompson, M.B., Davis, D.G. and Morris, R.W. (1993) J. Lipid.

Res. 34, 553-561.

[59] Olinga, P., Merema, M.T., Meijer, D.K.F., Slooff, MJ.H. and

Groothuis, G.M.M. (1993) ATLA-Altern. Lab. Anim. 21,466-468.

[60] Schwenk, M. (1980) Arch, Toxicol. 44, 113-126.

[61] Berry, M.N., Halls, H.J. and Grivell, M.B. (1992) Life, Sci. 51,

1-16. [62] Miyanchi, S., Sawada, Y., Iga, T., Hanano, M. and Sugiyama, Y.

(1993) Pharm. Res. 10, 434-440.

[63] Klaassen, C.D. and Watkins, LB.,Ill (1984) Pharmacol. Rev. 36,

1-67.

[64] Berry, M.N., Edwards, A.M. and Berritt, G.J. (1991) in Isolated

Hepatocytes Preparation, Properties and Applications (Burdon R.H.,

ed.), pp. 182-184, Elsevier, Amsterdam.

[65] Wolkoff, A.W., Samuelson, A.C., Johansen, K.L., Nakata, R.,

Withers, D.M. and Sosiak, A. (1987) J. Clin. Invest. 79, 1259-1268.

[66] Groothuis, G.M.M., Hulstaert, C.E., Kalicharan, D. and Hardonk,

M.J. (1981) Eur. J. Cell. Biol. 26, 43-51.

[67] Gebhardt, R. (1986) in Research in Isolated and Cultured Hepato-

cytes (Guillouzo, A. and Guguen-Guillouzo, C., eds.), pp. 353-376,

John Libbey Eurotext Ltd./INSERM,

[68] Oude Elferink, R.P.J., Ottenhoff, R., Liefting, W.G.M., Schoen-

maker, B., Groen, A.K. and Jansen, P.L.M. (1990) Am. J. Physiol.

258, G699-G706.

[69] Tarao, K., Clinger, E.J., Ostrow, D. and Balistreri, W.F. (1982)

Am. J. Physiol. 243, G253-G258.

[70] Kwekkeboom, J., Princen, H.M.G., Van Voorthuizen, E.M., Mei-

jer, P. and Kempen, H.J.M. (1989) Biochem. Biophys. Res. Com-

mun. 162, 619-625.

[71] Petzinger, E. and Frimmer, M. (1988) Biochim. Biophys. Acta 937,

135-144.

[72] Liang, D., Hagenbuch, B., Stieger, B. and Meijer, P.J. (1993)

Hepatology 18, 1163-1166.

[73] F~511mann, W., Petzinger, E. and Kinne, R.K.H. (1990) Am..1.

Physiol. 258, C700-C712.

[74] Foliot, A.. Glaise, D., Erlinger, S. and Guguen-Guillonzo, C.

(1985) Hepatology 5, 215-219.

[75] Galivan, J. (1982) Arch. Biochem. Biophys. 214, 850-852.

[76] Chesn6, C., Guyomard, C., Grislain, L., Clerc, C., Fautrel, A. and

Guillouzo, A. (1991) Toxicol. In Vitro 5, 479-482.

[77] Quistorff, B. and Grunnet, N. (1985) Biochem. J. 229, 221-226.

[78] Lindros, K.O. and Penttil~i, K.E. (1985) Biochem. J. 228, 757-760.

[79] Braakman, I., Keij, J., Hardonk, M.J., Meijer, D.K.F. and Groothuis,

G.M.M. (1991) Hepatology 13, 73-82.

[80] Buscher, H.P., Fricker, G., Gerok, W,, Kurz, G., Miiller, M.,

Schneider, S., Schramm, U. and Schreyer, A. (1987) in Bile Acids

and the Liver (Paumgartner, G., Stiehl, A. and Gerok, W., eds.),

pp. 95-110, MTP Press, Lancaster, England.

[81] Rogiers, V. (1992) in Human Cells in In Vitro Pharmaco-toxicol-

ogy (Rogiers, V., Sonck, W., Shephard, E. and Vercruysse, A.,

eds.), pp. 77-115, VUB Press. Brussels.

[82] Sandker, G.W., Vos, R.M.E., Delbressine, L.P.C., Slooff, M.J.H.,

Meijer, D.K.F. and Groothuis, G.M.M. (1994) Xenobiotica 24.

143-155.

[83] Azer, S.A. and Stacey, NH. (1993) Biochem. Pharmacol. 46,

813-819.

[84] Day, J.F., Thorpe, S.R. and Baynes, J.W. (1979) J. Biol. Chem.

254, 595-597.

[85] Chalmers, D.K., Scholz, G.H., Topliss, D.J., Kolliniatis, E., Munro,

S.L.A., Craik, D.J., Iskander, M.N. and Stockigt, J.R. (1993) J.

Med. Chem. 36, 1272-1277.

[86] Ihrke, G., Neufeld, E.B., Meads, T., Shanks, M.R., Cassio, D.,

Laurent, M., Schroer, T.A., Pagano, R.E. and Hubbard, A.L.

(1993) J. Cell Biol. 123, 1761-1775.

[87] Petzinger, E., Follmann, W., Blumrich, M., Walther, P., Hentschel,

J., Bette, P., Maurice, M. and Feldmann, G. (1994) Eur. J. Cell.

Biol. 64, 328-338. [88] Blumrich, M., Zeyenblumrich, U., Pagels, P. and Petzinger, E.

(1994) Eur. J. Cell. Biol. 64, 339-347.

[89] Bayer, J.L., Philips, J.M. and Graf, J. (1990) Methods Enzymol.

192, 501-516.

[90] Wilton, J.C., Williams, D.E., Strain, A.J., Parslow, R.A., Chipman,

J.K. and Coleman, R. (1991) Hepatology 14, 180-183.

[91] Verkade, H.J., Zaal, K.J., Derksen, J.T., Vonk, R.J., Hoekstra, D.,

Kuipers, F. and ScherphoL G.L. (1992)Biochem. J. 284, 259-265.

[92] Fentem, J.H., Foster, B., Mills, C.O., Coleman, R. and Chipman,

J.K. (1990) Toxicol. In Vitro 4, 452-457.

[93] Nathanson, M.H. and Burgstahler, A.D. (1992) Cell. Calcium. 13,

89-98.

[94] GraL J. and Boyer, J.L. (1990) J. Hepatol. 10, 387-394.

[95] Watanabe, N., Tsukada, N., Smith, C.R. and Phillips, M.J. (1991)

J. Cell. Biol. 113, 1069-1080.

[96] Watanabe, N., Tsukada, N., Smith, C.R., Edwards, V. and Phillips,

M.J. (1991) Lab. Invest. 65, 203-213.

[97] Thibault, N., Claude, J.R. and Ballet, F. (1992) Toxicology 73,

269-279.

[98] Wilton, J.C., Chipman, J.K., Lawson, C.J., Strain, A.J. and Cole-

man, R. (1993) Biochem. J. 292, 773-779.

[99] Benedetti, A., Strazzabosco, M., Ng, O.C. and Boyer, J.L. (1994)

Proc. Natl. Acad. Sci. USA 91,792-796.

[100] Bruck, R., Nathanson, M.H., Roelofsen, H. and Boyer, J.L. (1994)

Hepatology 20, 1032-1040.

[101] Blitzer, B.L. and Donovan, C.B. (1984) J. Biol. Chem. 259,

9295-9301. [102] Meier, P.J., Meier-Abt, A.S., Barrett, C. and Boyer. ,I.L. (1984) J.

Biol. Chem. 259, 10614-10622.

[103] Hardikar, W. and Suchy, F.J. (1993) Hepatology 18, 1278-1279.

[104] Meier, P.J., Sztul, E.S., Reuben, A. and Boyer, J.L. (1984) J. Cell.

Biol. 98, 991-1000.

[105] Kast, C., Stieger, B., Winterhalter, K.H. and Meier, P.J. (1994)J.

Biol. Chem. 269, 5179-5186.

[106] Adachi, Y., Kobayashi, H., Kurumi, Y., Shouji, M., Kitano, M. and

Yamamoto, T. (1991) Hepatology 14, 655-659.

[107] Novak, D.A., Ryckman, F.C. and Suchy, F.J. (1989) Hepatology

10, 447-453.

[108] Wolters, H., Kuipers, F., Slooff, M.J.H. and Vonk, R.J. (1992) J.

Clin. Invest. 90, 2321-2326.

[109] Wolters, H., Spiering, M., Gerding, A., Slooff, M.JH., Kuipers, F.,

Hardonk, M.J. and Vonk, R.J. (1991) Biochirn. Biophys. Acta

1069, 61-69.

[110] lnoue, M., Kinne, R., Tran, T. and Arias, I.M. (1982) Hepatology

2, 572-579.

[111] Bear, C.E., Davison, J.S. and Shaffer, E.A. (1987) Biochim. Bio-

phys. Acta 903, 388-394.

[112] Wolkoff, A.W. and Chung. C.T. (1980) J. Clin. ]Invest. 65, 1152-

1169.

[113] Stremmel, W., Gerber, M.A., Glezerov, V., Thung, S.N., Kochwa,

S. and Berk, P.D. (1983) J. Clin. Invest. 71, 1796-1805.

[114] yon Dippe, P., Ananthanarayanan, M. and Levy, D. (1986) Biochim.

Biophys. Acta 862, 352-360.

[115] Buscher, H.-P., Fricker, G., Gerok, W., Kramer, W., Kurz, G.,

Miiller, M. and Schneider, S. (1986) in Receptor-Mediated Uptake

in the Liver (Greten, H., Windier, E. and Beisiegel, U., eds.), pp.

189-199, Springer Verlag, Berlin Heidelberg.

[116] Ruetz, S., Fricker, G., Hugentobler, G., Winterhalter, K., Kurz, G.

and Meier, P.J. (1987) J. Biol. Chem. 262, 11324-11330.

[117] Soreq, H. and Seidman, S. (1992) Methods Enzymol. 207,225-231.

[118] Mori, M., Izumi, T. and Shimizu, T. (1993) Biochim. Biophys.

Acta 1168, 23-29.

[119] Shneider, B.L. and Moyer, M.S. (1993) J. Biol. Chem. 268,

6985-6988.

260 R.P.J. Oude Elferink et a l . / Biochimica et Biophysica Acta 1241 (1995) 215-268

[120] Hortensteiner, S., Vogt, E., Hagenbuch, B., Meier, P.J., Amrhein,

N. and Martinoia, E. (1993) J. Biol. Chem. 268, 18446-18449.

[121] Martinoia, E., Grill, E., Tommasini, R., Kreuz, K. and Amrhein, N.

(1993) Nature 364, 247-249.

[122] Molenaar, D., Bolhuis, H., Abee, T., Poolman, B. and Konings,

W.N. (1992)J. Bacteriol. 174, 3118-3124.

[123] Boyer, J.L., Ng, O.C., Ananthanarayanan, M., Hofmann, A.F.,

Schteingart, C.D., Hagenbuch, B., Stieger, B. and Meier, P.J.

(1994) Am, J. Physiol. 266, G382-G387.

[124] Kitamura, T., Jansen, P.L.M., Hardenbrook, C., Kamimoto, Y.,

Gatmaitan, Z. and Arias, I.M. (1990) Proc. Natl. Acad. Sci. USA

87, 3557-3561.

[125] Smit, J.J.M., Schinkel, A.H., Oude Elferink, R.P.J., Groen, A.K.,

Wagenaar, E., Van Deemter, L., Mol, C.A.A.M., Ottenhoft, R.,

Van der Lugt, N.M.T., van Roon, M.A., Van der Valk, M.A.,

Offerhaus, G.J.A., Berns, A.J.M. and Borst, P. (1993) Cell 75,

451-462.

[126] Capecchi, M.R. (1994) Scientific American March, 34-41.

[127] Schinkel, A.H., Smit, J.J.M., Vantellingen, O., Beijnen, J.H., Wa-

genaar, E., Vandeemter, L., Mol, C.A.A.M., Vandervalk, M.A.,

Robanusmaandag, E.C., Teriele, H.P.J., Berns, A.J.M. and Borst,

P. (1994) Cell 77, 491-502.

[128] Hammerton, R.W., Krzeminski, K.A., Mays, R.W., Ryan, T.A.,

Wollner, D.A. and Nelson, W.J. (1991) Science 254, 847-850.

[129] Anwer, M.S. and Hegner, D. (1978) Hoppe-Seyler. 's. Z. Physiol.

Chem. 359, 181-192.

[130] Meier, P.J. (1989) J, Hepatol. 9, 124-129.

[131] Zimmerli, B., Valantinas, J. and Meier, P.J. (1989) J. Pharmacol.

Exp. Ther. 250, 301-308.

[132] Petzinger, E., Muller, N., Follmann, W., Dentscher, J. and Kinne,

R.K.H. (1989) Am. J. Physiol. 256, G78-G86.

[133] Anwer, M.S., O'Maille, E.RL., Hofmann, A.F., DiPietro, R.A. and

Michelotti, E. (1985) Am. J. Physiol. 249, G479-G488.

[134] Hardison, W.G.M., Bellentani, S., Heasley, V. and Sbelhamer, D.

(1984) Am. J. Physiol. 246, G477-G483.

[135] Bellentani, S., Hardison, W.G.M., Marchegiano, P., Zanasi, G. and

Manenti, F. (1987) Am. J. Physiol. 252, G339-G344.

[136] Stacey, N. and Koteckex, B. (1988) Gastroenterology 95,780-786.

[137] Scharschmidt, B.F. and Stephens, J.E. (1981) Proc. Natl. Acad. Sci.

USA 78, 986-990.

[138] Meier, P.J., Meier-Abt A, S.T., Barrett, C. and Boyer, J.L. (1984)

J. Biol. Chem. 259, 10614-10622.

[139] Weinman, S., Graf, J. and Boyer, J.L. (1989) Am. J. Physiol. 256,

G826-G832.

[140] Nunez, L., Amigo, L., Rigotti, A., Puglielli, L., Mingrone, G.,

Greco, A.V. and Nervi, F. (1993) FEBS Lett. 329, 84-88.

[141] Reference deleted.

[142] Lidofsky, S.D., Fitz, J.G., Weisiger, R.A. and Scharschmidt, B.F.

(1993) Am. J. Physiol. 264, G478-G485.

[143] Weinman, S.A. and Weeks, R.P. (1993) Am. J. Physiol. 265,

G73-G80.

[144] Wehner, F. (1993) Pflugers. Arch-Eur. J. Physiol. 424, 145-151.

[145] Kramer, W., Bickel, U., Buscher, H.P., Gerok, W. and Kurz, G.

(1982) Eur. J. Biochem. 129, 13-24.

[146] Von Dippe, P. and Levy, D. (1983) J. Biol. Chem. 258, 8896-8901.

[147] Wieland, T., Nassal, M., Kramer, W., Fricker, G., Bickel, U. and

Kurz, G. (1984) Proc. Natl. Acad. Sci. USA 81, 5232-5236.

[148] Ananthanarayanan, M., Von Dippe, P. and Levy, D. (1988) J. Biol.

Chem. 263, 8338-8343.

[149] Von Dippe, P. and Levy, D. (1990) J. Biol. Chem. 265, 14812-

14816.

[150] Von Dippe, P., Amoui, M., Alves, C. and Levy, D. (1993) Am. J.

Physiol. 264, G528-G534.

[151] Porter, T.D., Beck, T.W. and Kasper, C.B. (1986) Arch. Biochem.

Biophys. 248, 121-129,

[152] Hagenbuch, B., Lubbert, H., Stieger, B. and Meier, P.J. (1990) J.

Biol. Chem. 265, 5357-5360,

[153] Hagenbuch, B,, Stieger, B., Foguet, M., Lubbert, H. and Meier,

P.J. (1991) Proc. Natl. Acad. Sci. USA 88, 10629-10633.

[154] Reichen, J. and Berk, P.D. (1979) Biochem. Biophys. Res. Com-

mun. 91,484-489.

[155] Honscha, W., Schulz, K., Muller, D. and Petzinger, E. (1993) Ear.

J. Pharmacol. Mol. Pharmacol. 246, 227-232.

[156] Hagenbuch, B. and Meier, P.J. (1994) J. Clin. Invest. 93, 1326-

1331.

[157] Boyer, J.L., Hagenbuch, B., Ananthanarayanan, M., Suchy, F.,

Stieger, B. and Meier, P.J. (1993) Proc. Natl. Acad. Sci. USA 90,

435-438.

[158] Ananthanarayanan, M., Bucuvalas, J.C., Shneider, L., Sippel, C.J.

and Suchy, F.J, (1991) Am. J. Physiol. 261, G810-G817.

[159] Liang, D., Hagenbuch, B., Stieger, B. and Meier, P.J. (1993)

Hepatology 18, 1162-1166.

[160] Hsu, S.M., Waldron, J.W., Hsu, P.L. and Hough, A.J. (1993) Hum.

Pathol. 24, 1040-1057.

[161] Yamazaki, M., Suzuki, H., Hanano, M., Tokui, T., Komai, T. and

Sugiyama, Y. (1993) Am. J. Physiol. 264, G36-G44.

[162] Hinchman, C.A., Truong, A.T. and Ballatori, N. (1993) Am. J.

Physiol. 265, G547-G554.

[163] Baldini, G., Passamonti, S., Lunazzi, G.C., Tiribelli, C. and Sotto-

casa, G.L. (1986) Biochim. Biophys. Acta 856, 1-10.

[164] Persico, M. and Sottocasa, G.L. (1987) Biochim. Biophys. Acta

930, 129-134.

[165] Stremmel, W. and Berk, P.D. (1986) J. Clin. Invest. 78, 822-826.

[166] Stremmel, W. and Diede, H.E. (1990) J. Hepatol. 10, 99-104.

[167] WolkofL A.W.. Sosiak, A., Greenblatt, H.C., Van Renswoude, J.

and Stockert, R.J. (1985) J. Clin. Invest. 76, 454-459.

[168] Goeser, T., Nakata, R., Braly, L.F., Sosiak, A., Campbell, C.G.,

Dermietzel, R., Novikoff, P.M., Stockert, R.J., Burk, R.D. and

Wolkoff, A.W. (1990) J. Clin. Invest. 86, 220-227.

[169] Min, A.D.. Goeser, T., Liu, R., Campbell, C.G., Novikoff, P.M.

and Wolkoff, A.W. (1991) Hepatology 14, 1217-1223.

[170] Wolkoff, A.W., Samuelson, A,C., Johansen, K.L., Nakata, R.,

Withers, D.M. and Sosiak, A. (1987) J. Clin. Invest. 79, 1259-1268.

[171] Wolkoff, A.W. (1989) in Hepatic transport of organic substances

(Petzinger, E., Kinne, R.K.H. and Sies, H., eds.), pp. 221-232,

Springer Verlag, Berlin.

[172] Min, A.D., Johansen, K.L., Campbell, C.G. and Wolkoff, A.W,

(1991) J. Clin. Invest. 87, 1496-1502.

[173] Campbell, C.G., Spray, D.C. and Wolkoff, A.W. (1993) J. Biol.

Chem. 268. 15399-15404.

[174] Sottocasa, G.L., Baldini, G., Sandi, G., Lunazzi, G. and Tribelli, C.

(1982) Biochim. Biophys. Acta 685, 123-128.

[175] Marchegiano, P., Carubbi, F., Tiribelli, C., Amarri, S., Stebel, M.,

Lunazzi, G.C., Levy, D. and Bellentani, S. (1992) Biochem. Bio-

phys. Res. Commun. 183, 1203-1208.

[176] Torres, A.M., Lunazzi, G.C., Stremmel, W. and Tiribelli, C. (1993)

Proc. Natl. Acad. Sci. USA 90, 8136-8139.

[177] Jacquemin, E., Hagenbuch, B., Stieger, B., Wolkofl-; A.W. and

Meier, P.J. (1991) J. Clin. Invest. 88, 2146-2149.

[178] Jacquemin, E., Hagenbnch, B., Stieger, B., Wolkoff, A.W. and

Meier, P.J. (1994) Proc. Natl. Acad. Sci. USA 91, 133-137.

[179] Meijer, D.K.F. (1989) in Handbook of Physiology Section 6, vol

III (pp. 717-758,

[180] Stolz, A., Takikawa, H., Ookhtens, M. and Kaplowitz, N. (1989)

Annu. Rev. Physiol. 51, 161-176.

[181] Zucker, S.D., Storch, J., Zeidel, M.L. and Gollan, J.L. (1992)

Biochemistry. 31, 3184-3192.

[182] Barnwell, S.G., Lowe, P.J. and Coleman, R. (1984) Biochem. J.

220, 723-731.

[183] Rao, C.V. and Mehendale, H.M. (1991) Biochem. Pharmacol. 42,

2323-2332.

[184] Gregory, D.H., Vlahcevic, Z.R., Prugh, M.F. and Swell, L. (1978)

Gastroenterology 74, 93-100.

R.P.J. Oude Elferink et a l . / Biochimica et Biophysica Acta 1241 (1995) 215-268 261

[185] Crawford, J.M. and Gollan, J.L. (1988) Am. J. Physiol. 255,

GI21-GI31.

[186] Crawford, J.M., Berken, C.A. and Gollan, J.L. (1988) J. Lipid Res.

29, 144-156.

[187] Nakai, T., Katagiri, K., Hoshino, M., Hayakawa, T, and Ohiwa, T.

(1992) Biochem. J. 288, 613-617.

[188] Crawford, J.M., Crawford, A.R. and Strahs, D.CJ. (1993) Hepatol-

ogy 18, 903-911.

[189] Crawford, J.M. and Gollan, J.L. (1991) Hepatology 14, 192-197.

[190] Crawford, J. and Gollan, J.L. (1988) Am. J. Physiol. 255, GI21-

GI31,

[191] Mori, M., Oyamada, M., Sakauchi, F. and Ogawa, K. (1987)

Virchows. Arch. 53, 37-43.

[192] Dumont, M,, Dhont, C., Durandschneider, A.M., Legranddefretin,

V., Feldmann, G. and Erlinger, S. (1991) Hepatology 14, 10-15.

[193] Lamri, Y., Roda, A., Dumont, M., Feldmann, G. and Erlinger, S.

(1988) J. Clin. Invest. 82, 1173-1182.

[194] Lamri, Y., Erlinger, S., Dumont, M., Roda, A. and Feldmann, G.

(1992) Liver 12, 351-354.

[195] Aoyama, N., Ohya, T., Chandler, K., Gresky, S. and Holzbach.

R.T. (1991)Hepatology 14, I-9.

[196] Aoyama, N., Tokumo, H., Ohya, T., Chandler, K. and Holzbach,

R.T. (1991) Am. J. Physiol. 261, G305-G311.

[197] Melby, E.L,, Prydz, K., Olsnes, S. and Sandvig, K. (1991) J. Cell

Biochem. 47, 251-260.

[198] Oude Elferink. R.P.J., Bakker, C.T.M,, Roelofsen, H., Middelkoop,

E., Ottenhoff, R.. Heijn, M. and Jansen, P.L.M. (1993) Hepatology

17, 434-444.

[199] Barr, V.A. and Hubbard, A.L. (1993) Gastroenterology 105, 554-

571.

[200] Scott, L.J. and Hubbard, A.L. (1993) J. Biol. Chem. 268, 19160.

[201] Weisz, O.A., Machamer, C.E. and Hubbard. A.L, (1992) J. Biol.

Chem. 267, 22282-22288.

[202] Schell, M.J., Maurice, M., Stieger, B. and Hubbard, A.L. (1992) J.

Cell Biol. 119, 1173-1182.

[203] Bortles, J.R., Feracci, H.M., Stieger, B. and Hubbard, A.L. (1987)

J. Biol. Chem. 105, 1241-1251.

[204] de Vries. M.H., Groothuis, G.M.M., Mulder, G.J., Nguyen, H. and

Meijer, D.K.F. (1985) Biochem. Pharmacol. 34, 2129-2135.

[205] Proost, J.H., Nijssen, H.MJ.. Strating, C.B., Meijer, D.K.F. and

Groothuis. G.MM. (1993)J. Pharmacokinet. Biopharm. 21, 375-

394.

[206] Mizuma, T., Komori, M., Ueno, M. and Horikoshi, I. (1991) J.

Pharm. Pharmacol. 43, 446-448.

[207] Lauterburg, B.H., Adams, J.D. and Mitchell, J.R. (1984) Hepatol-

ogy 4, 586-590.

[208] Deleve, L.D. and Kaplowitz, N. (1991) Pharmacol. Ther. 52,

287-305.

[209] Ookhtens, M., Lyon, L, Fernandez-Checa, J. and Kaplowitz, N.

(1988) J. Clin. Invest. 82, 608-616.

[210] Garciaruiz, C., Fernandezcheca, J.C. and Kaplowitz, N, (1992) J.

Biol. Chem. 267, 22256-22264.

[211] Aw, T.Y., Ookhtens. M., Kuhlenkamp, J.F. and Kaplowitz, N.

(1987) Biochem. Biophys. Res. Commun. 143, 377-382.

[212] Fernandez-Checa, J.C., Ren, C., Yee aw, T., Ookhtens, M. and

Kaplowitz, N. (1988) Am. J. Physiol. 255, g403-g408.

[213] Fernandez-Checa, J.C., Yi, J.R., Garciaruiz, C., Knezic, Z., Tahara,

S.M. and Kaplowitz, N, (1993) J. Biol. Chem. 268, 2324-2328.

[214] Meier. P.J., Meier-Abt A, S.T. and Boyer, J.L. (1987) Biochem. J.

242, 465-469.

[215] Inoue, M., Akerboom, T.P.M., Sies, H., Kinne, R., Thao, T. and

Arias, I.M. (1984) J. Biol. Chem. 259, 4998-5002.

[216] Akerboom, T., lnoue, M., Sies, H., Kinne, R. and Arias, I.M.

(19831Eur. J. Biochem. 141,211-215.

[217] Scharschmidt, B.F. and Schmid, R. (1978) J. Clin. Invest. 62,

1121-1131.

[218] Zimniak, P. and Awasthi, Y.C. (1993) Hepatology 17, 330-339.

[219] Larkin, J.M. (1993) Gastroenterology 105, 594-597.

[220] Inoue, M., Kinne, R., Tran, T. and Arias, I.M. (1983) Eur. J.

Biochem. 134, 467-471.

[221] Jansen, P.L.M., Peters. W.H. and Lamers, W.H. (1985) Hepatology

5, 573-579.

[222] Dubin, I.N. and Johnson, F.B. (1954)Medicine 33, 155-197.

[223] Sprintz, H. and Nelson, R.S. (1954) Ann. Intern. Med. 41,952-960.

[224] Arias, I.M. (1961) Am. J. Med. 31, 510-518.

[225] Javitt, N.B., Kondo, T. and Kuchiba, K. (1978) Gastroenterology

75, 931-932.

[226] Cornelius, C.E., Arias, I.M. and Osburn, B.I (1965) J. Am. Vet.

Med. Ass. 146, 709-713.

[227] Alpert, S., Mosher, M., Shanske, A. and Arias. I.M. (1969) J. Gen.

Physiol. 53, 238-247.

[228] Jansen, P.L.M., Vanklinken, J.W., Vangelder, M., Ottenhoff, R.

and Oude Elferink, R.PJ. (1993) Am. J. Physiol. 265, G445-G452.

[229] Jansen, P.L.M., Groothuis, G.MM., Peters. W,HM. and Meijer,

D.K.F. (1987) Hepatology 7, 71-76.

[230] Jansen, P,L.M. and Oude Elferink, R.P.J. (1993) in Hepatic Trans-

port and Bile Secretion: Physiology and Pathophysiology (Tavoloni,

N. and Berk, P.D., eds.), pp. 721-731, Raven Press, New York.

[231] Ruetz, S., Hugentobler, G. and Meier, P.J. (1988) Proc. Natl. Acad.

Sci. USA 85, 6147-6151.

[232] Muller, M., Ishikawa, T., Berger, U., Klunemann, C., Lucka, L.,

Schreyer, A., Kannicht, C., Reutter, W., Kurz, G. and Keppler, D.

(1991) J. Biol. Chem. 266, 18920-18926.

[233] Nishida, T., Gatmaitan, Z., Che, M.X. and Arias, I.M. (1991) Proc.

Natl. Acad. Sci. USA 88, 6590-6594.

[234] Stieger, B., Oneill, B. and Meier, P.J. (1992) Biochem. J. 284,

67-74.

[235] Becker, A., Lucka, L., Kiliam C., Kannicht, C. and Reutter, W.

(1993) Eur. J. Biochem. 214, 539-548.

[236] Lin. S.H. and Guidotti, G. (1989) J. Biol. Chem. 264, 14408-14414.

[237] Bartles, J.R.. Braiterman, L.T. and Hubbard, A.L. (1985) J. Biol.

Chem. 260, 12792-12802.

[238] Culic, O,, Huang, Q.H., Flanagan, D., Hixson. D. and Lin, S.H.

(1992) Biochem. J. 285, 47-53.

[239] Petell, J.K., Diamond, M., Hong, W., Bujanover, Y., Amarri, S.,

Pittschieler, K. and Doyle, D. (1987) J. Biol. Chem. 262, 14753-

14765.

[240] Sippel, C.J., Suchy, FJ., Ananthanarayanan, M. and Perlmutter,

D.H. (1993)J. Biol. Chem. 268, 2083-2091.

[241] Sippel, C.J., Mccollum. M.J. and Perlmutter, D.tt. (1994)J. Biol.

Chem. 269. 2820-2826.

[242] Sippel, CJ., Fallon, R.J. and Perlmutter, D.H. (1994) J. Biol.

Chem. 269, 19539-19545.

[243] Abraham, E.H., Prat, A.G.. Gerweck, L., Seneverame, T., Arceci,

R.J., Kramer, R., Guidotti, G. and Cantiello, H.F. (1993) Proc.

Natl. Acad. Sci. USA 90, 312-316.

[244] Kuipers, F., Enserink, M., Havinga, R., Van de, r Steen, A.B.M.,

Hardonk, M.J., Fevery, J. and Vonk, R.J. (1988) J. Clin. Invest. 81,

1593-1599.

[245] Takikawa, H., Sano, N., Narita, T., Uchida, Y., Yamanaka, M.,

Horie, T., Mikami, T. and Tagaya, O. (1991~ Hepatology 14,

352-360.

[246] Oude Elferink, R.P.J., Ottenhoff, R., Radominska, A., Hofmann,

A.F., Kuipers, F. and Jansen, P,L.M. (1991) Biochem. J. 274,

281-286.

[247] Nishida, T., Hardenbrook, C., Gatmaitan, Z. and Arias, I.M. (1992)

Am. J. Physiol. 262. G629-G635.

[248] Huber, M., Guhlmann, A., Jansen, P.L.M. and Keppler, D. (1987)

Hepatology 7, 224-228.

[249] Oude Elferink, R.P.J., Haan de, J., Lambert, K., Hofmann, A.F.

and Jansen, P.L.M. (1987) Hepatology 7, 1109.

[250] Oude Elferink, R.PJ., Ottenhoff, R., Liefting, W., De haan. J. and

Jansen, P.L.M. (1989) J. Clin. Invest. 84, 476-483.

[251] Oude Elferink, R.P.J., Ottenhoff, R., Liefting, W., Schoemaker, B.,

262 R,P.Z Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

Groen, A.K. and Jansen, P.L.M. (1990) Am. J. Physiol. 258,

G699-G706.

[252] Kitamura, T., Alroy, J., Gatmaitan, Z., Inoue, M., Mikami, T.,

Jansen, P.L.M. and Arias, I.M. (1992) Hepatology 15, 1154-1159.

[253] Ishikawa, T., Muller, M., Klunemann, C., Schaub, T. and Keppler,

D. (1990) J. Biol. Chem. 265, 19279-19286.

[254] Nishida, T., Gatmaitan, Z., Roychowdhry, J. and Arias, I.M. (1992)

J. Clin. Invest. 90, 2130-2135.

[255] Kobayashi, K., Sogame, Y., Hayashi, K., Nicotera, P. and Orre-

nius, S. (1988) FEBS Lett. 240, 55-58.

[256] Akerboom, T.P.M., Narayanaswami, V., Kunst, M. and Sies, H.

(1991) J. Biol. Chem. 266, 13147-13152.

[257] Kobayashi, K., Komatsu, S., Nishi, T., Hara, H. and Hayashi, K.

(1991) Biochem. Biophys. Rese. Commun. 176, 622-626.

[258] Kobayashi, K., Sogame, Y., Hara, H. and Hayashi, K. (1990) J.

Biol. Chem. 265, 7737-7741.

[259] Verkade, H.J., Havinga, R., Gerding, A., Vonk, R.J. and Kuipers,

F. (1993) Am. J. Physiol. 264, G462-G469.

[260] Adachi, Y., Kobayashi, H., Kurumi, Y., Shouji, M., Kitano, M. and

Yamamoto, T. (1991) Hepatology 14, 1251-1258.

[261] Adachi, Y., Kobayashi, H., Shouji, M., Kitano, M., Okuyama, Y.

and Yamamoto, T. (1993) Life. Sci. 52, 777-784.

[262] Nicotera, P., Moore, M., Bellomo, G., Mirabelli, F. and Orrenius,

S. (1985) J. Biol. Chem. 260, 1999-2002.

[263] Nicotera, P., Moore, M., Bellomo, G., Mirabelli, F. and Orrenius,

S. (1985) FEBS Lett. 187, 121-125.

[264] Zimniak, P., Ziller, S.A, Panfil, I., Radominska, A., Wolters, H.,

Kuipers, F., Sharma, R., Saxena, M., Moslen, M.T., Vore, M..

Vonk, R., Awasthi, Y.C. and Lester, R. (1992) Arch. Biochem.

Biophys. 292, 534-538.

[265] Sharma, R., Gupta, S., Singh, S.V., Medh, R.D., Ahmad, H.,

Labelle, E.F. and Awasthi, Y.C. (1990) Biochem. Biophys. Res.

Commun. 171, 155-161.

[266] Pikula, S., Hayden, J.B., Awasthi, S., Awasthi, Y.C. and Zimniak,

P. (1994) J. Biol. Chem. 269, 27566-27573.

[267] Pikula, S., Hayden, J.B., Awasthi, S., Awasthi, Y.C. and Zimniak,

P. (1994) J. Biol. Chem. 269, 27574-27579.

[268] Roelofsen, H., Ottenhoff, R., Oude Elferink, R.P.J. and Jansen,

P.L.M. (1991) Biochem. J. 278, 637-641.

[269] Okuda, H., Tavoloni, N., Blaschke, T.F., Kiang, C.L., Jones,

M.JT., Waggoner, J.G., Sardana, M.K., Sassa, S., Shrager, R.I. and

Berk, P.D. (1991) Hepatology 14, 1153-1160.

[270] Kuipers, F., Radominska, A., Zimniak, P., Little, J.M., Havinga,

R., Vonk, R.J. and Lester, R. (1989) J. Lipid Res. 30, 1835-1845.

[271] Maudgal, D.P., Maxwell, J.D., Lees, L.J. and Wild, R.N. (1982)

Br. J. Clin. Pharrnacol. 14, 213-217.

[272] Arvidsson, A., Leijd, B., Nord, C.E. and Angelin, B. (1988) Eur. J.

Clin. Invest. 18, 261-266.

[273] Xia, Y., Lambert, K. and Hofmann, A.F. (1988) Gastroenterology

94, A606.

[274] Oude Elferink, R.P.J. (1989) in Bile Duct and Bile Duct Abnormal-

ities (Tytgat, G.N.J. and Huibregtse, K., eds.), pp. 9-14, Thieme

Verlag, New York.

[275] Verkade, H.J., Wolbers, M.J., Havinga, R., Uges, D.R., Vonk, R.J.

and Kuipers, F. (1990) Gastroenterology 99, 1485-1492.

[276] Levine, W.G. (1978)Ann. Rev. Pharmacol. Toxicol. 18, 81-96.

[277] Bergan, T. (1987) Drugs 34, 89-104.

[278] Akerboom, T.P.M. and Sies, H. (1989) Methods Enzymol. 173,

523-535.

[279] Akerboom, T.P.M., Bilzer, M. and Sies, H. (1982) FEBS Lett. 140,

73-76.

[280] Silva, J.M., Mcgirr, L. and Obrien, P.J. (1991) Arch. Biochem.

Biophys. 289, 313-318.

[281] Gilbert, H.F. (1982) J. Biol. Chem. 2, 12086-12091.

[282] Chance, B., Sies, H. and Boveris, A. (1979) Physiol. Rev. 59,

527-605.

[283] Smith, P.F., Alberts, D.W. and Rush, G.F. (1987) Toxicol. Appl.

Pharmacol. 89, 190-201.

[284] Chung, P.M, Cappel, R.E. and Gilbert, H.F. (1991) Arch. Biochem.

Biphys. 288, 48-53.

[285] Stein, A.F., Gregus, Z. and Klaassen, C.D. (1988) Toxicol. Appl.

Pharmacol. 93, 351-359.

[286] Ballatori, N. and Truong, A.T. (1989) Am. J. Physiol. 256, G22-

G30.

[287] Ballatori, N. and Truong, A.T. (1992) Am. J. Physiol. 263, G617-

G624.

[288] Ballatoil, N., Jacob, R. and Boyer, J.L. (1986) J. Biol. Chem. 261,

7860-7865.

[289] Fernandezcheca, J.C., Takikawa, H., Horie, T., Ookhtens, M. and

Kaplowitz, N. (1992) J. Biol. Chem. 267, 1667-1673.

[290] Ballatori, N. and Dutczak, W.J. (1994)J. Biol. Chem. 269, 19731- 19737.

[291] Yi, J.R., Lu, S.L., Fernandezcheca, J. and Kaplowitz, N. (1994)J.

Clin. Invest. 93, 1841-1845.

[292] Fernandezcheca, J.C., Ookhtens, M. and Kaplowitz, N. (1993) J.

Biol. Chem. 268, 10836-10841.

[293] Ballatori, N. (1991)Drug Metab. Rev. 23, 83-132,

[294] Alexander, J. and Aaseth, J. (1980) Biochem. Pharmacol. 29, 2129-2133.

[295] Houwen, R., Dijkstra, M., Kuipers, F., Smit, E., Havinga, R. and

Vonk, R.J. (1990) Biochem. Pharmacol. 39, 1039-1044.

[296] Dijkstra, M., Kuipers, F., Havinga, R., Smit, E.P. and Vonk, R.J.

(1990) Pediatr. Res. 28, 339-343.

[297] Gross, J.B., Myers, B.., Kost, L.., Kuntz, S.. and La Russo, N.

(1989) J. Clin. Invest. 83, 30-39.

[298] Kressner, M.S., Stockert, R.J., Morell, A.G. and Sternlieb, I.

(1984) Hepatology 4, 867-870.

[299] Moil, M., Hattori, A., Sawaki, M., Tsuzuki, N., Sawada, N.,

Oyamada, M., Sugawara, N. and Enomoto, K. (1994) Am. J. Pathol. 144, 200-204.

[300] Suzuki, M. and Aoki, T. (1994) Pediatr. Res. 35, 598-601.

[301] Muramatsu, Y., Yamada, T., Miura, M., Sakai, T., Suzuki, Y.,

Serikawa, T., Tanzi, R.E. and Matsumoto, K. (1994)Gastroenterol-

ogy 107, 1189-1192.

[302] Alexander, J., Aaseth, J. and Refsvik, T. (1981) Acta Pharmacol.

Toxicol, 49, 190-194.

[303] Ballatori, N. and Clarkson, T.W. (1982) Science 216, 61-62.

[304] Ballatori, N. and Clarkson, T.W. (1983) Am. J. Physiol. 244, G435-G441.

[305] Ballatori, N. and Clarkson, T.W. (1984) Biochem. Pharmacol. 33,

1087-1092.

[306] Dutczak, W.J. and Ballatori, N. (1994) J. Biol. Chem. 269, 9746- 9751.

[307] Alexander, J., Aaseth, J. and Mikalsen, A. (1986) Acta Pharmacol.

Toxicol. 59, 486-489.

[308] Gregus, Z. and Varga, F. (1985) Acta Pharmacol. Toxicol. 56,

398-403.

[309] Klaassen, C.D. (1974) Toxicol. Appl. Pharmacol. 29, 447-457.

[310] Gyurasics, A., Varga, F. and Gregus, Z. (1991) Biochem. Pharma-

col. 42, 465-468.

[311] Gyurasics, A., Varga, F. and Gregus, Z. (1991) Biochem. Pharma- col. 41,937-944.

[312] Anundi, I., Hogberg, J. and Vahter, M. (1982) FEBS Lett. 145, 285-288.

[313] Delnomdedieu, M., Basti, M.M., Otvos, J.D. and Thomas, D.J.

(1994) Chem. Biol Interact. 90, 139-155,

[314] Gyurasics, A., Koszorus, L., Varga, F. and Gregus, Z. (1992)

Biochem. Pharrnacol. 44, 1275-1281.

[315] Srivastava, S.K. and Beutler, E. (1969) J. Biol. Chem. 244, 9-16. [316] Board, P.G. (1981) FEBS Lett. 124, 163-165.

[317] Prchal, J., Srivastava, S.K. and Beutler, E. (1975) Blood 46, 111-117.

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 263

[318] Kondo, T., Murao, M. and Taniguchi, N. (1982) Eur. J. Biochem.

125, 551-554.

[319] Kondo, T., Kawakami, Y., Taniguchi, N. and Beutler, E. (1987)

Proc. Natl. Acad. Sci. USA 84, 7373-7377.

[320] Heijn, M., Oude Elferink, R.P.J. and Jansen, P.L.M. (1992) Am. J.

Physiol. 262, C104-C110.

[321] Board, P., Nishida, T., Gatmaitan, Z., Che, M.X. and Arias, I.M.

(1992) Hepatology 15, 722-725.

[322] Kondo, T., Miyamoto, K., Gasa, S., Taniguchi, N. and Kawakami,

Y. (1989) Biochem. Biophys. Res. Commun. 162, 1-8.

[323] Awasthi, S., Singhal, S.S., Srivastava, S.K, Zimniak, P., Bajpai.

K.K., Saxena, M., Sharma, R., Ziller, S.A., Frenkel, E.P., Singh,

S.V., He, N.G. and Awasthi, Y.C. (1994) J. Clin. Invest. 93,

958-965.

[324] Ishikawa, T., Zimmer, M. and Sies, H. (1986) FEBS Lett. 200,

128-132.

[325] Ishikawa, T. and Sies, H. (1984) J. Biol. Chem. 259, 3838-3843.

[326] Ishikawa, T. (1989) J. Biol. Chem. 264, 17343-17348.

[327] Schaub, T., lshikawa, T. and Keppler, D. (1991) FEBS Lett. 279,

83-86.

[328] Jedlitschky, G., Leier, 1., Buchholz, U., Center, M. and Keppler, D.

(1994) Cancer Res. 54, 4833-4836.

[329] Muller, M., Meijer, C., Zaman, G.J.R., Borst, P,, Scheper, R.J.,

Mulder, N.H., De Vries, E.G.E. and Jansen, P.L.M. (1994) Proc.

Natl. Acad. Sci. USA 91, 13033-13037

[330] Zaman, G.JR., Versantw)ort, C.HM., Smit, J.JM., Eijdems,

E.WHM., de Haas, M., Smith, A.J., Broxterman, H.J., Mulder,

N.H., De Vries, E.GE., Baas. F. and Borst, P. (1993) Cancer Res.

53. 1747-1750.

[331] Oude Elferink, R.P.J., Bakker, C.T.M. and Jansen, P.L.M. (1993)

Biochem. J. 290, 759-764.

[332] de Vries, M.H., Redegeld, F.A.M., Koster, A.S., Noordhoek, J., de

Haan, J.G., Oude Elferink, R.P.J. and Jansen, P.L.M. (1989)

Naunyn-Schmied. Arch. Pharmacol. 340, 588-592.

[333] Meijer, D.K.F. and Groothuis, G.M.M. (1991) in Oxford Textbook

of Clinical Hepatology, Volume 1 (Mclntyre, N., Benhamou, J.P.,

Bircher, J.. Rizzetto, M. and Rodes, J., eds.), pp. 40-78, Oxford

University Press, Oxford, New York, Tokyo.

[334] Meijer. D.K.F. (1989) in Handbook of Physiology, Sect. 6, Vol. III

(Schuhz, S.G., Forte, J.G. and Rauner, B.B., eds.), pp. 717-758.

Oxford University Press, New York.

[335] Meijer. D.K.F., Mol, W.EM., Miiller, M. and Kurz, G. (1990) J.

Pharmacokinet. Biopharm. 18, 35-70.

[336] Steen, H. and Meijer, D.K.F. (1991) Prog. Pharmacol. Clin. Phar-

macol. 8. 239-272.

[337] Braakman, I.. Groothuis, G.MM. and Meijer, D.K.F. (1987) Hepa-

tology 7, 849-855.

[338] Braakman, I., Groothuis, G.MM. and Meijer, D.K.F. (1989) J.

Pharmacol. Exp. Ther. 249. 869-873.

[339] Van der, Sluijs P., Spanjer, H.H. and Meijer, D.K.F. (1987) J.

Pharmacol. Exp. Ther. 240, 668-673.

[340] Braakman, 1. (1988) Hepatocyte heterogeneity in organic cation

transport, Thesis, Rijksuniversiteit Groningen.

[341] Braakman, I.. Verest, O., Pijning, T., Meijer, D.K.F. and Groothuis,

G.MM. (1990)Mol. Pharmacol, 36, 532-536.

[342] Braakman, 1., Pijning, T., Verest, O., Weert, B., Meijer, D.K.F. and

Groothuis, G.MM. (1989) Mol. Pharmacol. 36, 537-542.

[343] Tulkens, P.M. (1991) Eur. J. Clin. Microbiol. 10, 100-106.

[344] Yoshioka, K. and Nishimura, H. (1988) Experientia 44, 889-892.

[345] Yoshioka, K., Nishimura, H., Himukai, M. and Iwashima, A.

(1985) Biochim. Biophys. Acta 815, 499-504.

[346] Lumeng, L., Edmondson, J.W., Schenker, S. and Li, T.-K. (1979)

J. Biol. Chem. 254, 7265-7268.

[347] Chen. C.P. (1978)J. Nutr. Sci. Vitaminol. 24, 351-362.

[348] Yoshioka, K. (1984) Biochim. Biophys. Acta 778, 201-209.

[349] Yoshioka, K., Nishimura, H. and Hasegawa, T. (1985) Biochim.

Biophys. Acta 819, 263-266.

[350] Yoshioka, K. and Nishimura, H. (1986) Experientia 42, 1022-1023.

[351] Yoshioka, K., Nishimura, H. and Iwashima, A. (1983) Biochim.

Biophys. Acta 732, 308-311.

[352] Moseley, R.H., Vashi, P.G., Jarose, S.M., Dickinson, C.J. and

Permoad, P.A. (1992) Gastroenterology 103, 1056-1065.

[353] Moseley, R.H., Morrissette. J. and Johnson, T.R. (1990) Am. J.

Physiol. 259, G973-G982.

[354] Moseley, R.H., Jarose, S.M. and Permoad, P. (1992) Am. J.

Physiol. 263, G775-G785.

[355] Mckinney, T.D. and Hosford, M.A. (1992) Am. J. Physiol. 263,

G939-G946.

[356] King, P.D. and Cai, Z.-S. (1993) Gastroenterology 104, A927.

[357] Mol, W.EM., Miiller, M., Kurz, G. and Meijer, D.K.F. (1992)

Biochem. Pharrnacol. 43, 2217-2226.

[358] Steen, H., Maring, J.G. and Meijer, D.K.F. (1993) Biochem.

Pharmacol. 45, 809-818.

[359] Steen, H., Oosting, R. and Meijer, D.K.F. (1991) J. Pharmacol.

Exp. Ther. 258, 537-543.

[360] Mol, W.EM., Fokkema, G.N., Weert, B. and Meijer, D.K.F. (1988)

J. Pharmacol. Exp. Ther. 244, 268-275.

[361] Steen, H., Merema, M. and Meijer, D.K.F. (1992) Biochem. Phar-

macol, 44, 2323-2331.

[362] Faraji-Shadan, F., Stubbs, J.D. and Bowman, P.D. (1990) Med.

Hypotheses 32, 81-84.

[363] Mol, W.EM., Miiller, M. and Meijer, D.K.F. (1989) Arch. Pharm.

322, 613-615.

[364] Miiller, M., Mol, W.EM., Meijer, D.K.F. and Kurz, G. (1994)

submitted.

[365] Xie, M.-H., Cochran, M., Lidofsky, S.D. and Scharschmidt, B.F.

(1992) Hepatology 16, 147A.

[366] Moseley, R.H., Smit, H., van Solkema, B.G.H., Wang, W. and

Meijer, D.K.F. (1994) Hepatology 20, 261A.(Abstract)

[367] Meijer, D.K.F. (1976) in Intestinal permeation (Kramer, M. and

Lauterbach, F., eds.), pp. 196-207, Excerpta Medica, Amsterdam.

[368] Meijer, D.K.F., Wester, J. and Gunnik, M. (1972) Naunyn.

Schmiedebergs. Arch. Pharmacol. 273, 179-192.

[369] Mol, W.EM. and Meijer, D.K.F. (1990) Biochem. Pharmacol. 39,

383-390.

[370] Van Dyke, R.W., Faber. E.D. and Meijer, D.K.F. (1992) J. Phar-

macol. Exp. Ther. 261, 1-11.

[371] Meijer, D.K.F., Weltering, J.G. and Vonk. R.J. (1976) J. Pharma-

col. Exp. Ther. 198, 229-239.

[372] Weltering, J.G., Lammers, W., Meijer, D.K.F. and Mulder, G.J.

(1977) Naunyn. Schmiedebergs. Arch. Pharmacol. 299, 277-281.

[373] Weltering, J.G., Mulder, G.J., Meijer, D.K.F., Lammers, W., Veen-

huis, M. and Wendelaar Bonga, S.E. (1975)Naunyn. Schmiede-

bergs, Arch. Pharmacol. 289, 251-256.

[374] Woolley, G.A., Kapral, M.K. and Deber, C.M. (1987)FEBS Lett.

224, 337-342.

[375] Bunting, J.R., Phan, T.V., Kamali, E. and Dowben, R.M. (1989)

Biophys. J. 56, 979-993.

[376] Steen, H. (1992) Mechanisms for the hepatic disposition of organic

cations. Molecular aspects of membrane transport in the hepato-

cyte, Thesis, Rijksuniversiteit Groningen.

[377] Kr~imer, R., Mayr, U., Heberger, C. and Tsompanidou, S. (1986)

Biochim. Biophys. Acta 855, 201-210.

[378] Singh, G., Sharkey, S.M. and Moorehead, R. (1992) Pharmacol.

Ther. 54, 217-230.

[379] Diwan, J.J., Yune, H.H., Bawa, R., Haley, T. and Mannella, C.A.

(1988) Biochem. Pharmacol. 37, 957-961.

[380] Berman, J.J. (1988) Cell Biol. Toxicol. 4, 325-332.

[381] Porter, R.K., Scott, J.M. and Brand, M.D. (1992) J. Biol, Chem.

267, 14637-14646.

264 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

[382] Toninello, A., Dalla Via, L., Siliprandi, D. and Garlid, K.D. (1992)

J. Biol. Chem. 267, 18393-18397.

[383] Miiller, M. (1988) Membrantransport von organischen kationen in

hepatocyten. Ein beitrag zur aufkl~rung von hepatischen trans-

portvorg~ingen, Dissertation, Universitiit von Freiburg.

[384] Koch, J.K. and Gallagher, C.H. (1960) Biochem. Pharrnacol. 3,

231-243.

[385] Meier, P.J., Knickelbein, R., Moseley, R.H., Dobbins, J.W. and

Boyer, J.L. (1985) J. Clin. Invest. 75, 1256-1263.

[386] Mannella, C.A. and Wang, Q. (1989) Biochim. Biophys. Acta 981,

363-366.

[387] Zimmermann, H.W. (1986) Angew. Chem. [Int Ed.] 25, 115-130.

[388] Yeh, C.G., Hsi, B. and Faulk, W.P. (1981) J. Immunol. 43,

269-275. [389] Ronot, X., Benel, L., Adolphe, M. and Mounolou, J.C. (1986) Biol

Cell. 57, 1-8.

[390] Ullrich, K.J. and Rumrich, G. (1993) Clin. Investig. 71, 843-848.

[391] Vonk, R.J., Scholtens, E., Keulemans, G.TP. and Meijer, D.K.F,

(1978) Naunyn. Schmiedebergs. Arch. Pharmacol. 302, 1-9.

[392] Ruifrok, P.G. (1982) Biochem. Pharmacol. 31, 1431-1435.

[393] Neef, C., Keulemans, K.TP. and Meijer, D.K.F. (1984) Biochem.

Pharmacol. 33, 3991-4002.

[394] Neef, C., Keulemans, K.TP. and Meijer, D.K.F. (1984) Biochem.

Pharmacol. 33, 3977-3990.

[395] Kamimoto, Y.. Gatmaitan, Z., Hsu, J. and Arias, I.M. (1989) J.

Biol. Chem. 264, 11693-11698.

[396] Sinicrope, F.A., Dudeja, P.K., Bissonnette, B.M., Safa, A.R. and

Brasitus, T.A. (1992) J. Biol. Chem. 267, 24995-25002.

[397] Ballet, F., Vrignaud, P., Robert, J., Rey, C. and Poupon, R. (1987)

Cancer Chemother. Pharmacol. 19, 240-245.

[398] Smit, J.W., Steen, H. and Meijer, D.K.F. (1994) in Transport in the

Liver (Keppler, D. and Jungermann, K., eds.), pp. Abstract-Abno.

27, Falk Symposium No. 74, Heidelberg.

[399] Speeg, K.V. and Maldonado, A.L. (1993) Cancer Chemother.

Pharmacol. 32, 434-436.

[400] Speeg, K.V., Maldonado, A.L., Liaci, J. and Muirhead, D. (1992)

Hepatology 15, 899-903.

[401] Thalhammer, T., Stapf, V., Gajdzik, L. and Graf, J. (1994) Eur. J.

Pharmacol. 270, 213-220.

[402] Watanabe, T., Miyauchi, S., Sawada, Y., Iga, T., Hanano, M.,

Inaba, M. and Sugiyarna, Y. (1992) J. Hepatol. 16, 77-88.

[403] Gottesman, M.M. and Pastan, I. (1993) Annu. Rev. Biochem. 62,

385-427. [404] Zamora, J.M., Pearce, H.L. and Beck, W.T. (1988) Mol. Pharma-

col. 33, 454-462.

[405] Buschman, E., Arceci, R.J., Croop, J.M., Che, M.X., Arias, 1.M.,

Housman, D.E. and Gros, P. (1992) J. Biol. Chem. 267, 18093-

18099.

[406] Thiebaut, F., Tsuruo, T., Hamada, H., Gottesman, M.M., Pastan, I.

and WiUingham, M.C. (1987) Proc. Natl. Acad. Sci. USA 84.

7735-7738. [407] Bradley, G., Georges, E. and Ling, V. (1990) J. Cell. Physiol. 145,

398-408.

[408] Fojo, A.T., Ueda, K., Slamon, D.J., Poplack, D.G., Gottesman,

M.M. and Pastan, I. (1987) Proc. Natl. Acad. Sci. USA 84,

265 -269. [409] Germann, U.A., Pastan, I. and Gottesman, M.M. (1993) Semin.

Cell Biol. 4, 63-76.

[410] Doige, C.A. and Ames, G.FL. (1993) Annu. Rev. Microbiol. 47,

291-319.

[411] Roninson, I.B. (1992) Biochem. Pharmacol. 43, 95-102.

[412] Schinkel, A.H. and Borst, P. (1991) Semin. Cancer Biol. 2, 213-

226.

[413] Borst, P. (1991) Acta Oncol. 30, 87-105. [414] Teeter, L.D., Estes, M., Chan, J.Y., Atassi, H., Sell, S., Becker,

F.F. and Kuo, M.T. (1993) Mol. Carcinog. 8, 67-73.

[415] Schrenk, D., Gant, T,W., Preisegger, K.-H., Silverman, J.A.,

Marino, P.A. and Thorgeirsson, S.S. (1993) Hepatology 17, 854-

860.

[416] Nakatsukasa, H., Silverman, J.A., Gant, T.W., Evarts, R.P. and

Thorgeirsson, S.S. (1993) Hepatology 18, 1202-1207.

[417] Kuo, M.T. (1993) Mol. Carcinog. 7, 73-75.

[418] Chenivesse, X., Franco, D. and Br&hot, C. (1993) J. Hepatol. 18,

168-172.

[419] Smit, 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., Van der Valk, M.A.,

Offerhaus, G.J.A., Berns, A.J.M. and Borst, P. (1993) Cell 75,

451-462.

[420] Ruetz, S. and Gros, P. (1994) Cell 77, 1071-1082.

[421] Chaudhary, P.M., Mechetner, E.B. and Roninson, I.B. (1992)

Blood 80, 2735-2739.

[422] Gupta, S. and Gollapudi, S. (1993) J. Clin. Immunol. 13, 289-301.

[423] Endicott, J.A. and Ling, V. (1989) Annu. Rev. Biochem. 58,

137-171.

[424] Ambudkar, S.V., Lelong, I.H., Zhang, J.P., Cardarelli, C.O.,

Gottesman, M.M. and Pastan, I. (1992) Proc. Natl. Acad. Sci. USA

89, 8472-8476.

[425] Sharom, F.J., Yu, X.H. and Doige, C.A. (1993) J. Biol. Chem. 268,

24197-24202.

[426] Garrigos, M., Belehradek, J.,Jr., Mir, L.M. and Orlowski, S. (1993)

Biochem. Biophys. Res. Commun. 196, 1034-1041.

[427] Hammond, J.R., Johnstone, R.M. and Gros, P. (1989) Cancer Res.

49, 3867-3871.

[428] Gros, P., Croop, J. and Housman, D, (1986) Cell 47, 371-380.

[429] Ueda, K., Cardarelli, C., Gottesman, M.M. and Pastan. I. (1987)

Proc. Natl. Acad. Sci. USA 84, 3004-3008.

[430] Gill, D.R., Hyde, S.C., Higgins, C.F., Valverde, M.A., Mintenig,

G.M. and Sepfilveda, F.V. (1992) Cell 71, 23-32.

[431] Abraham, E.H., Prat, A.G., Gerweck, L., Seneveratne, T., Arceci,

R.J., Kramer, R.. Guidotti, G. and Cantiello, H.F. (1993) Proc.

Natl. Acad. Sci. USA 90, 312-316.

[432] Roepe, P.D., Wei, L.Y., Cruz, J. and Carlson, D. (1993) Biochem-

istry. 32, 11042-11056.

[433] Boscoboinik, D., Gupta, R.S. and Epand, R.M. (1990) Br. J.

Cancer. 61,568-572.

[434] Keizer, H.G. and Joenje, H. (1989) J. Natl. Cancer Inst. 81,

706-709.

[435] Thiebaut, F.. Currier, S.J., Whitaker, J., Haugland, R.P., Gottes-

man, M.M., Pastan, I. and Willingham, M.C. (1990) J. Histochem.

Cytochem. 38, 685-690.

[436] Speeg, K.V.,Jr., Deleon, C. and Mcguire, W.L. (1992) Cancer Res.

52, 3539-3546.

[437] Skovsgaard, T. (1989) in Anticancer drugs (Tapiero, H., Robert, J.

and Lampidis, T.J., eds.), pp. 233-244, Colloque INSERM/John

Libbey Eurotext Ltd.

[438] Altenberg, G.A., Young, G., Horton, J.K., Glass, D., Belli, J.A.

and Reuss, L. (1993) Proc. Natl. Acad. Sci. USA 90, 9735-9738.

[439] Higgins, C.F. and Gottesman, M.M. (1992) Trends Biochem. Sci.

17, 18-21.

[440] Shalinsky, D.R., Jekunen, A.P., Alcaraz, J.E., Christen, R.D., Kim,

S., Khatibi, S. and Howell, S.B. (1993) Br. J. Cancer. 67, 30-36.

[441] Homolya, L., Hollo, Z., Germann, U.A., Pastan, I., Gottesman,

M.M. and Sarkadi, B. (1993) J. Biol. Chem. 268, 21493-21496.

[442] Stein, W.D., Cardarelli, C., Pastan, I. and Gottesman, M.M. (1994)

Mol. Pharmacol 45, 763-772.

[443] Gottesman, M.M. and Pastan, I. (1993) Annu. Rev. Biochem. 284,

278-284.

[444] Klopman, G., Srivastava, S., Kolossvary, I., Epand, R.F., Ahmed,

N. and Epand, R.M. (1992) Cancer Res. 52, 4121-4129,

[445] Selassie, C.D., Hansch, C. and Khwaja, T.A. (1990) J. Med. Chem.

33, 1914-1919. [446] Pearce, H.L., Safa, A.R., Bach, N.J., Winter, M.A., Cirtain, M.C.

and Beck, W.T. (1989) Proc. Natl. Acad. Sci. USA 86, 5128-5132.

R.P.Z Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 265

[447] Ford, J.M., Bruggemann, E.P., Pastan, I., Gottesman, M.M. and

Hait, W.N. (1990) Cancer Res. 50, 1748-1756.

[448] Nogae, I., Kohno, K., Kikuchi, J., Kuwano, M., Akiyama, S.I.,

Kiue, A., Suzuki, K.I., Yoshida, Y., Cornwell, M.M., Pastan, I. and

Gottesman, M.M. (1989) Biochem. Pharmacol. 38, 519-527.

[449] Hofsli, E. and Nissen-Meyer, J. (1990) Cancer Res. 50, 3997-4002.

[450] Dellinger, M., Pressman, B.C., Calderon-Higginson, C., Savarai,

N., Tapiero, H., Kolonias, D. and Lampidis, T.J. (1992) Cancer

Res. 52, 6385-6389.

[451] Ramu, A. and Ramu, N. (1992) Cancer Chemother. Pharmacol. 30,

165-173.

[452] Piwnica-Worms, D., Chiu, M.L., Budding, M., Kronauge, J.F.,

Kramer, R.A. and Croop, J.M. (1993) Cancer Res. 53, 977-984.

[453] Tang-Wai, D.F., Brossi, A., Arnold, L.D. and Gros, P. (1993)

Biochemistry 32, 6470-6476.

[454] Priebe, W., Van, N.T., Burke, T.G. and Perez-Soler, R. (1993)

Anti-Cancer. Drugs 4, 37-48.

[455] Sehested, M., Skoosgaard, T., Van Deurs, B. and Winther-Nielsen,

H. (1987) Br. J. Cancer. 56, 747-751.

[456] Bruggemann, E.P., Currier, S.J., Gottesman, M.M. and Pastan, I.

(1992) J. Biol. Chem. 267, 21020-21026.

[457] Tamai, I. and Sara, A.R. (1991) J. Biol. Chem. 266, 16796-16800.

[458] Loo, T.W. and Clarke, D.M. (1993) J. Biol. Chem. 268, 19965-

19972.

[459] Ferry, D.R., Russell, M.A. and Cullen, M.H. (1992) Biochem.

Biophys. Res. Commun. 188, 440-445.

[460] Chen, H.-X., Bamberger, U., Heckel, A., Guo, X. and Cheng,

Y.-C. (1993) Cancer Res. 53, 1974-1977.

[461] Gruber, A., Briese, B., Arestrtim, I., Vitols, S., BjiSrkholm, M. and

Peterson, C. (1993) Leuk. Res. 17, 353-358.

[462] Grant, C.E., Valdimarsson, G., Hipfner, D.R., Almquist, K.C.,

Cole, S.PC. and Deeley, R.G. (1994)Cancer Res. 54, 357-361.

[463] Miilder, H.S., van Grondelle, R., Westerhoff, H.V. and Lankelma,

J. (1993) Eur. J. Biochem. 218, 871-882.

[464] Krishnamachary, N. and Center, M.S. (1993) Cancer Res. 53,

3658-3661.

[465] Versantvoort, C.HM., Schuurhuis, G.J., Pinedo, H.M., Eekman,

C.A., Kuiper, C.M., Lankelma, J. and Broxterman, H.J. (1993) Br.

J. Cancer 68, 939-946.

[466] Scheper, R.J., Broxterman, H.J., Scheffer, G.L., Kaaijk, P., Dalton,

W.S., Van Heijningen, T.HM., Van Kalken, C.K., Slovak, M.L.,

De Vries, E.GE., Van der, V.alk P., Meijer, C.JLM. and Pinedo,

H.M. (1993) Cancer Res. 53, 1475-1479.

[467] Teeter, L.D., Hsu, H.-C., Curley, S.A., Tong, M.J. and Kuo, M.T.

(1993) Int. J. Oncol. 2, 73-80.

[468] Kuo, M.T., Zhao, J.Y., Teeter, L.D., Ikeguchi, M. and Chisari,

F.V. (1992) Cell Growth Differ. 3, 531-540.

[469] Nakatsukasa, H., Evarts, R.P., Burt, R.K., Nagy, P. and Thorgeirs-

son, S.S. (1992) Mol. Carcinog. 6, 190-198.

[470] Bradley, G.. Sharma, R., Rajalakshmi, S. and Ling, V. (1992)

Cancer Res. 52, 5154-5161.

[471] Fardel, O., Ratanasavanh, D., Loyer, P., Ketterer, B. and Guil-

louzo, A. (1992) Eur. J. Biochem. 205, 847-852.

[472] Fardel, O., Morel, F. and Guillouzo, A. (1993) Carcinogenesis 14,

781-783.

[473] Lee, C.H.. Bradley, G., Zhang, J.-T. and Ling, V. (1993) J. Cell.

Physiol. 157, 392-402.

[474] Acufia, C., Vollrath, V., Wielandt, A., Andrade, L., Duarte, I. and

Chianale, J. (1994) submitted.

[475] Hait, W.N. and Aftab, D.T. (1992) Biochem. Pharmacol. 43, 103-107.

[476] Grunicke, H. and Hofmann, J. (1992) Pharmacol. Ther. 55, 1-30.

[477] Chambers, T.C., Chalikonda, I. and Eilon, G. (1990) Biochem.

Biophys. Res. Commun. 169, 253-259.

[478] Chambers, T.C., McAvoy, E.M., Jacobs, J.W. and Eilon, G. (1990) J. Biol. Chem. 265, 7679-7686.

[479] Ford, J.M. and Hait, W.N. (1990) Pharmacol. Rev. 42, 155-199.

[480] Fine, R.L., Patel, J., Hamilton, T.C., Cowan, K., Curt, G.A.,

Friedman, M.A. and Chabner, B.A. (1986) Proc. Am. Assoc.

Cancer Res. 27, 271. [481] Hamada, H., Hagiwara, K., Jakajima, T. and Tsuruo, T. (1987)

Cancer Res. 47, 2860-2865.

[482] Mellado, W. and Horwitz, S.B. (1987) Biochemistry. 26, 6900-

6904.

[483] Center, M.S. (1983) Biochem. Biophys. Res. Commun. 115, 159-

166. [484] Mickley, L.A., Bates, S.E., Richert, N.D., Currier, S., Tanaka, S.,

Foss, F., Rosen, N. and Fojo, A.T. (1989) J. Biol. Chem. 264,

18031-18040.

[485] Bates, S.E., Currier, S.J., Alvarez, M. and Fojo, A.T. (1992)

Biochemistry 31, 6366-6372.

[486] Nishizuka, Y. (1988)Nature 334, 661-665.

[487] Ward, N.E. and O'Brian, C.A. (1993) Biochemistry 32, 11903-

11909.

[488] Yu, G., Ahmad, S., Aquino, A., Fairchild, C.R., Trepel, J.B., Ohno,

S., Suzuki, K., Tsuruo, T., Cowan, K.H. and Glazer, R.I. (1991)

Cancer Commun. 3, 181 - 189.

[489] Tamaoki, T., Nomoto, H., Takahashi, I., Kato, Y., Morimoto, M.

and Tomita, F. (1986) Biochem. Biophys. Res. Commun. 135,

397-402.

[490] Hidaka, H. and Kobayashi, R. (1992) Annu. Rev. Pharmacol.

Toxicol. 32, 377-397.

[491] Sato, W., Yusa, K., Naito, M. and Tsuruo, T. (1990) Biochem.

Biophys. Res. Commun. 173, 1252-1257.

[492] Wakusawa, S., lnoko, K., Miyamoto, K., Kajita, S., Hasegawa, T.,

Harimaya, K. and Koyama, M. (1993) J. Antibiot. 46, 353-355.

[493] Kobayashi, E., Nakano, H., Morimoto, M. and Tamaoki, T. (1989)

Biochem. Biophys. Res. Commun. 159, 548-553.

[494] Bates, S.E., Lee, J.S., Dickstein, B., Spolyar, M. and Fojo, A.T.

(1993) Biochemistry 32, 9156-9164.

[495] Brunton, V.G. and Workman, P. (1993) Cancer Chemother. Phar-

macol. 32, 1-19.

[496] Chaudhary, P.M. and Roninson, I.B. (1993)J. Natl. Cancer Inst.

85, 632-639.

[497] Chaudhary, P.M. and Roninson, I.B. (1992) Oncol. Res. 4, 281-

290.

[498] Steen, H., Smit, H., Nijholt, A., Merema, M. and Meijer, D.K.F.

(1993) Hepatology 18, 1208-1215.

[499] Mtiller, M., Mayer, R., Hero, U. and Keppler, D. (1994) FEBS

Lett.

[500] Mittenbiihler, K., Miiller, M., Wallstab, A.. Biichter, M,, Hero, U.,

Spiess, E., Silverman, J.A., Thorgeirsson, S.S. and Keppler, D.

(1994) J. Biol. Chem.

[501] Gosland, M., Tsuboi, C., Hoffman, T., Goodin, S. and Vore, M.

(1993) Cancer Res. 53, 5382-5385.

[502] Vore, M. (1993)Toxicol. Appl. Pharmacol. 118, 2-7.

[503] Arias, I.M. (1993) Hepatology 18, 216-217.

[504] Wolf, D.C. and Horwitz, S.B. (1992) Int J. Cancer 52, 141-146.

[505] Van Kalken, C,K., Broxterman, H.J., Pinedo, H.M., Feller, N.,

Dekker, H., Lankelma, J. and Giaccone, G. (1993) Br. J. Cancer

67, 284-289.

[506] De Lannoy, I.AM. and Silverman, M. (1992) Biochem. Biophys.

Res. Commun. 189, 551-557.

[507] Hori, R., Okamura, N., Aiba, T. and Tanigawara, Y. (1993) J.

Pharmacol. Exp. Ther. 266, 1620-1625.

[508] Okamura, N., Hirai, M., Tanigawara, Y., Tanaka, K., Yasuhara,

M., Ueda, K., Komano, T. and Hori, R. (1993) J. Pharmacol. Exp.

Ther. 266, 1614-1619.

[509] Ito, S., Koren, G., Harper, P.A. and Silverman, M. (1993) Can. J.

Physiol. Pharmacol. 71, 40-47.

[510] Becker, K.F., Allmeier, H. and HiSllt, V. (1992) lqiorm. Metab. Res.

24, 210-213.

266 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

[511] Gibbons, G.F. (1990)Biochem. J. 268, 1-13.

[512] Eisenberg, S. (1984) J. Lipid. Res. 25, 1017-1058.

[513] Robins, S.J. and Armstrong, M.J. (1976) Gastroenterology 70,

397-402.

[514] Marzolo, M.P., Amigo, L. and Nervi, F. (1993) J. Lipid Res. 34,

807-814.

[515] Smit, M.J., Temmerman, A.M., Wolters, H., Kuipers, F., Beynen,

A.C. and Vonk, R.J. (1991) J. Clin. Invest. 88, 943-951.

[516] Hardison, W.G.M. and Evans, W.H. (1986) J. Hepatol. 3, 318-327.

[517] Lanzini, A. and Northfield, T.C. (1991) Eur. J. Clin. Invest. 21,

259-272.

[518] Mazer, N.A. and Carey, M.C. (1984) J. Lipid Res. 25, 932-953.

[519] Gurantz, D. and Hofmann, A.F. (1984) Am. J. Physiol. 247,

G736-G748.

[520] Carulli, N., Loria, P., Bertolotti, M., Ponz de Leon, M., Menozzi,

D., Medici, G. and Piccagli, I. (1984) J. Clin. Invest. 74, 614-624.

[521] Alvaro, D., Cantafora, A., Attili, A.F., Ginanni Corradini, S., De

Luca, C., Minervini, G., Di Biase, A. and Angelico, A. (1986)

Comp. Biochem. Physiol. [B]. 83, 551-554.

[522] Somjen, G.J. and Gilat, T. (1983) FEBS Lett. 156, 265-268.

[523] Ulloa, N., Garrido, J. and Nervi, F. (1987) Hepatology 7, 235-244.

[524] Cohen, D.E., Angelico, M. and Carey, M.C. (1989) Am. J. Physiol.

257, G1-G8.

[525] Hofmann, A.F. (1990) Hepatology 12, 17S-25S.

[526] Hay, D.W. and Carey, M.C. (1990) Hepatology 12, 6S-14S.

[527] Smit, M.J., Temmerman, A.M., Havinga, R., Kuipers, F. and

Vonk, R.J. (1990) Biochem. J. 269, 781-788.

[528] Hay, D.W., Cahalane, M.J., Timofeyeva, N. and Carey, M.C.

(1993) J. Lipid Res. 34, 759-768.

[529] Robins, S.J., Fasulo, J.M., Robins, V.F. and Patton, G.M. (1991) J.

Lipid Res. 32, 985-992.

[530] Smit, M.J., Verkade, H.J., Havinga, R., Vonk, R.J., Scherphof,

G.L., Tveld, G.I. and Kuipers, F. (1994)J. Lipid. Res. 35, 301-310.

[531] Cohen, D.E., Kaler, E.W. and Carey, M.C. (1993) Hepatology 18,

1522-1531.

[532] Rigotti, A., Nunez, L., Amigo, L., Puglielli, L., Garrido, J., Santos,

M., Gonzalez, S., Mingrone, G., Greco, A. and Nervi, F. (1993) J.

Lipid Res. 34, 1883-1894.

[533] Domingo, N., Chanussot, F., Botta, D., Reynier, M.O., Crotte, C.,

Hauton, J. and Lafont, H. (1993) Lipids 28, 883-887.

[534] Chanussot, F., Domingo, N., Tuchweber, B., Lafont, H. and Yousef,

I. (1992) Scand. J. Gastroenterol. 27, 238-242.

[535] S~imjen, G.J., Marikovsky, Y., Wachtel, E., Harvey, P.R.C., Rosen-

berg, R., Strasberg, S.M. and Gilat, T. (1990) Biochim. Biophys.

Acta 1042, 28-35.

[536] Somjen, G.J., Coleman, R., Koch, M.HJ., Wachtel, E., Billington,

D., Townsandrews, E. and Gilat, T. (1991) FEBS Lett. 289,

163-166. [537] Kay, R.E. and Entenman, C. (1961) Am. J. Physiol. 200, 855-859.

[538] Hardison, W.G.M. and Apter, J.T. (1972) Am. J. Physiol. 222,

61-67.

[539] Wheeler, H.O. and King, K.K. (1972) J. Clin. Invest. 51, 1337-

1350.

[540] Turley, S.D. and Dietschy, J.M. (1981) J. Biol. Chem. 256, 2438-

2446. [541] Scheibner, J., Fuchs, M., Schiemann, M., Tauber, G., H6rmann, E.

and Stange, E.F. (1993) Hepatology 17, 1095-1102.

[542] Robins, S.J., Fasulo, J.M., Collins, M.A. and Patton, G.M. (1985)

J. Biol. Chem. 260, 6511-6513.

[543] Robins, S.J., Fasulo, J.M., Lessard, P.D. and Patton, G.M. (1993)

Biochem. J. 289, 41-44.

[544] Singer, I.I., Kawka, D.W., Kazazis, D.M., Alberts, A.W., Chen,

J.S., Huff, J.W. and Ness, G.C. (1984) Proc. Natl. Acad. Sci. USA

81, 5556-5560.

[545] Stone, B.G., Eridison, S.K., Graig, W.Y. and Copper, A.D. (1985)

J. Clin. Invest. 76, 1773-1781.

[546] Robins, S.J., Fasulo, J.M., Lessard, P.D. and Patton, G.M. (1993) J.

Lipid Res. 34, 1445-1450.

[547] Bilhartz, L.E., Spady, D.K. and Dietschy, J.M. (1989) J. Clin.

Invest. 84, 1181-1187.

[548] Robins, S.J., Fasulo, J.M., Leduc, R. and Patton, G.M. (1989)

Biochim. Biophys. Acta 1004, 327-331.

[549] Portal, I., Clerc, T., Sbarra, V., Portugal, H., Pauli, A.M., Lafont,

H., Tuchweber, B., Yousef, I. and Chanussot, F. (1993) Am. J.

Physiol. 264, GI052-GI056.

[550] Verkade, H.J., Derksen, J.T.P., Gerding, A., ScherphoE G.L.,

Vonk, R.J. and Kuipers, F. (1991) Biochem. J. 275, 139-144.

[551 ] Robins, S.J. and Brunengraber, H. (1982) J. Lipid Res. 23,604-608.

[552] Yao, Z.M. and Vance, D.E. (1988)J. Biol. Chem. 263, 2998-3004.

[553] Verkade, H.J., Fast, D.G., Rusinol, A.E., Scraba, D.G. and Vance,

D.E. (1993)J. Biol. Chem. 268, 24990-24996.

[554] Cronholm, T., Curstedt, T. and Sj6vall, J. (1983) Biochim. Bio-

phys. Acta 753. 276-279.

[555] Kawamoto, T., Akino, T., Nakamura, M. and Mori, M. (1980)

Biochim. Biophys. Acta 619, 35-47.

[556] Robins, S.J., Fasulo, J.M. and Patton, G.M. (1990) Am. J. Physiol.

259, G205-G211.

[557] Simons, K. and Van Meer, G. (1988)Biochemistry. 27, 6197-6202.

[558] Goldsmith, M.A., Huling, S. and Jones, A.L. (1983) Gastroenterol-

ogy 84, 978-986.

[559] Jones, A.L., Schmucker, D.L., Mooney, J.S., Ockner, R.K. and

Adler, R.D. (1979) Lab. Invest. 40, 512-517.

[560] Crawford, J.M., Vinter, D.W. and Gollan, J.L. (1991) Am. J.

Physiol. 260, G119-G132.

[561] Casu, A. and Camogliano, L. (1990) Biochim. Biophys. Acta 1043,

113-115.

[562] Reynier, M.O., Abouhashieh, I., Crotte, C., Carbuccia, N., Richard,

B. and Gerolami, A. (1992) Gastroenterology 102, 2024-2032.

[563] Graham, J., Ahmed, H. and Northfield, T.C. (1989) in Trends in

Bile Acid Research (Paumgartner, G., Stiehl, A. and Gerok, W.,

eds.), pp. 177-188, Kluwer Academic Publishers, Lancaster.

[564] Sakisaka, S., Cheng, N.OI. and Boyer, J. (1988) Gastroenterology

95, 793-804.

[565] Stein, O., Sanger, L. and Stein, Y. (1974) J. Cell Biol. 62,

390-405.

[566] Cohen, D.E., Leonard, M.R. and Carey, M.C. (1994) Biochemistry

33, 9975-9980.

[567] Wirtz, K.WA. (1991) Annu. Rev. Biochem. 60, 73-99.

[568] Reuben, A, and Allen, R.M. (1986) Biochim. Biophys. Acta 876,

1-12.

[569] Hayakawa, T., Cheng, O., Ma, A. and Boyer, J.L. (1990) Gastro-

enterology 99, 216-228.

[570] Lorenzini, I., Sakisaka, S., Meier, P.J. and Boyer, J.L (1986)

Gastroenterology 91, 1278-1288.

[571] Katagiri, K., Nakai, T. and Hoshino, M. (1992) Gastroenterology

102, 1660-1667.

[572] Dubin, M. and Erlinger, S. (1980) Clin. Sci. 58, 545-548.

[573] Lowe, P.J., Barnwell, S.G. and Coleman, R. (1984) Biochem. J.

222, 631-637.

[574] Kuipers, F., Derksen, J.PT., Gerding, A., ScherphoL G.L. and

Vonk, R.J. (1987)Biochim. Biophys. Acta 922, 136-144.

[575] Apstein, M.D. (1984) Gastroenterology 87, 634-638.

[576] Pattinson, N.R., Willis, K.E. and Chapman, B.A. (1987) Dig. Dis.

Sci. 32, 615-619. [577] Verkade, H.J., Vonk, R.J. and Kuipers, F. (1994) Hepatology 17,

1074-1080. [578] Verkade, H.J., Wolters, H., Gerding, A., Havinga, R., Fidler, V.,

Vonk, R.J. and Kuipers, F. (1993) Hepatology 17, 1074-1080.

[579] Yousef, I.M. and Fisher, M.M. (1976) Can. J. Biochem. 54,

1040-1046.

[580] Coleman, R., Rahman, K., Kan, K.S. and Parslow, R.A. (1989)

Biochem. J. 258, 17-22.

R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268 267

[581] Van-Berge-Henegouwen, G.P., VanderWerf, S.DJ. and Ruben, A.T.

(1987) Clin. Chim. Acta 165, 27-37.

[582] Salvioli, G., Romani, M., Loria, P., Carulli, N. and Pradelli, J.M.

(1985) J. Hepatol. 1, 291-300.

[583] Ben-, F., Schreiber, E. and Frick, U. (1992) Hepatology 16, 71-81.

[584] Shamburek, R.D. and Schwartz, C.C. (1993) J. Lipid Res. 34,

1833-1842.

[585] Yousef, I.M., BarnwelL S., Gratton, F., Tuchweber, B., Weber, A.

and Roy, C.C. (1987) Am. J. Physiol. 252, G84-G91.

[586] Hardison, W.G.M., Hatoff, D.E., Miyai, K. and Weiner, R.G.

(1981) Am. J. Physiol. 241, G337-G343.

[587] Coleman, R., Rahman, K., Bellringer, M.E., Kan, K. and Hamlin,

S. (1989) in Trends in Bile Acid Research (Paumgartner, G., Stiehl,

A. and Gerok, W., eds.), pp. 161-176, Kluwer Academic Publish-

ers, Lancaster.

[588] Billington, D. and Coleman, R. (1978) Biochim. Biophys. Acta

509, 33-47.

[589] Berr, F., Meier, P.J. and Stieger, B. (1993) J. Biol. Chem. 268,

3976-3979.

[590] Higgins. C.F. and Gottesman, M.M. (1992)Trends Biochem. Sci.

17, 18-21.

[591] Bishop, W.R. and Bell, R.M. (1985) Cell 42, 51-60.

[592] Ulloa, N., Garrido, J. and Nervi, F. (1987) Hepatology 7, 235-244.

[593] Crawford, J.M., Barnes, S., Stearns, R.C., Hastings, C.L. and

Godleski, J.L. (1994) Lab. Invest.

[594] Dietschy, J.M., Turley, S.D. and Spady, D.K. (1993) J. Lipid Res.

34, 1637- 1659.

[595] Grundy, S.M. (1983) Ann. Rev. Nutr. 3, 71-96.

[596] Mauad, T.H., van Nieuwkerk, C.M.J., Dingemans, K.P., Smit,

J.J.M.. Schinkel, A.H., Notenboom, R.G.E., van den Bergh Weer-

man, M.A., Verkruisen, R.P., Groen, A.K., Oude Elferink, R.P.J.,

Van der Valk, M.A., Borst, P. and Offerhaus, G.J.A. (1994) Am J.

Pathol. 145, 1237-1245.

[597] Tazuma, S., Barnhart, R.L., Reeve, L.E., Tokumo, H. and Holzbach,

R.T. (1988) Am. J. Physiol. 255,

[598] Tazuma, S. and Holzbach, R.T. (1987) Proc. Natl. Acad. Sci. USA

84, 2052- 2056.

[599] Davidson. N.O., Drewek, M.J., Gordon, J.I. and Elovson, J. (1988)

J. Clin. Invest. 82, 300-308.

[600] Wheeler, H.O. and Mancusi-Ungaro, P.L. (1966) Am. J. Physiol.

210, 1153-1159.

[601] Kountouras, J., McKavanagh, S., Burmicky, M. and Billing, B.H.

(1987) J. Hepatol. 4, 198-205.

[602] Ricci, G.L. and Fevery, J. (1989) Gut 30, 1266-1269.

[603] Ricci, G.L.. Michiels, R. and Fevery, J. (1984) Hepatology 4,

651-657.

[604] Bartles, JR., Feracci, H.M., Stieger, B. and Hubbard, A.H. (1987)

J. Cell. Biol. 105. 1241-1251.

[605] Bossard, R., Stieger, B., Oneill, B., Fricker, G. and Meier, P.J.

(1993) J. Clin. Invest. 91, 2714-2720.

[606] Fricker, G., Landmann, L. and Meier, P.J. (1989) J. Clin. Invest.

84, 876-885.

[607] Silverman, M. (1991) Annu. Rev. Biochem. 60, 757-794.

[608] Gladhaug, I.P. and Christoffersen, T. (1988) J. Biol. Chem. 263,

12199-12203.

[609] Oude Elferink, R.P.J., Bakker, C.T.M., Roelofsen, H., Middelkoop,

E., Ottenhoff, R., Heijn, M. and Jansen, P.L.M. (1993) Hepatology

17,434-444.

[610] Hallbrucker, C., Lang, F., Gerok, W. and Haussinger, D. (1992)

Biochem. J. 281,593-595.

[611] Haussinger, D., Saha, N., Hallbrucker, C., Lang, F. and Gerok, W.

(1993) Biochem. J. 291, 355-360.

[612] Haussinger, D., Hallbrucker, C., Saha, N., Lang, F. and Gerok, W.

(1992) Biochem. J. 288, 681-689.

[613] Bruck, R., Haddad, P., Graf, J. and Boyer, J.L. (1992) Am. J.

Physiol. 262, G806-G812.

[614] Boyer, J.L., McGrath, J.W. and Ng, O.C. (1993) Hepatology 18,

107A.(Abstract)

[615] Rasmussen, H. (1990) Hoppe-Seyler's. Z. Physiol. Chem. 371,

191-206.

[616] Nathanson, M.H., Gautam, A.. Bruck, R., Isales, C.M. and Boyer,

J.L. (1992) Hepatology 15, 107-116.

[617] Tanaka, A., Katagiri, K., Hoshino, M., Hayakawa, T., Tsukada, K.

and Takeuchi, T. (1994) Am. J. Physiol. 266, G324-G329.

[618] Isales, C.M., Nathanson, M.H. and Bruck. R. (1993) Biochem.

Biophys. Res. Commun. 191, 1244-1251.

[619] Hamada, Y., Karjalainen, A., Setchell, B., Millard, J. and Bygrave,

F.L. (1992) Biochem. J. 281, 387-392.

[620] Hamada, Y., Karjalainen, A., Setchell, B., Millard. J. and Bygrave,

F.L. (1992) Biochem. J. 283, 575-581.

[621] Roelofsen, H., Ottenhoff, R., Oude Elferink, R.P.J. and Jansen,

P.L.M. (1991) Biochem. J. 278, 637-641.

[622] Hayakawa, T., Bruck, R., Ng, O.C. and Boyer, J.L. (1990) Am. J.

Physiol. 259, G727-G735.

[623] Lenzen, R., Hruby, V,J. and Tavoloni, N. (1990) Am. J. Physiol.

259, G736-G745.

[624] Desmet, V.J. (1987) in Pathology of the Liver (MacSween, R.N.M.,

Anthony, P.P. and Scheuer, P.J., eds.), pp. 364-423, Churchill

Livingstone, Edinburgh.

[625] Rosario, J., Sutherland, E., Zaccaro, L. and Simon, F.R. (1987)

Biochemistry 27, 3939-3946.

[626] Kan, K.S., Monte, M.J., Parslow, R.A. and Coleman, R. (1989)

Biochem. J. 261,297-300.

[627] Combettes, L., Dumont, M., Berthon, B., Erlinger, S. and Claret,

M. (1988)J. Biol. Chem. 263. 2299-2303.

[628] Akerboom, T.P.M., Schneider, I.. Vom Dahl, S. and Sies, H.

(1991 ) Hepatology 13, 216-221.

[629] Dubin, M., Maurice, M., Feldmann, G. and Erlinger, S. (1978)

Gastroenterology 75,450-455.

[630] Cadranel, J.F., Erlinger, S., Desruenne, M., Luciani, J., Lunel, F.,

Grippon, P.. Cabrol. A. and Opolon, P. (1992) Dig. Dis. Sci. 37,

1473-1476.

[631] Stone, B.G., Udani, M., Sanghvi, A., Warty, V., Plocki, K.,

Bedetti, C.D. and Van Thiel, D.H. (1987) Gastroenterology 93,

344-351.

[632] Bohme, M., Buchler, M., Muller, M. and Keppler, D. (1993) FEBS

Lett. 333, 193-196.

[633] Kadmon, M., Klunemann, C., Bohme, M., Ishikawa, T., Gorgas,

K., Otto, G., Herfarth, C. and Keppler, D. (1993) Gastroenterology

104, 1507-1514.

[634] Moseley, R.H., Johnson, T.R. and Morisette, J.M. (1990) J. Phar-

macol. Exp. Ther. 253, 974-980.

[635] Kukongviriyapan, V.. and Stacey, N.H. (1988) J. Pharmacol. Exp.

Ther. 247, 685-689.

[636] Heikel, T.A.J. and Lathe. G.H. (1970) Br. J. Pharmacol. 38,

593-601.

[637] Stacey, N.H. (1986) Biochem. Pharmacol. 35, 2495-2500.

[638] Ben-, F., Simon, F.R. and Reichen, J. (1984) Am. J. Physiol. 247,

G437-G443.

[639] Vore, M., Durham, S., Yeh. S. and Ganguly, T. (1991) Biochem.

Pharmacol. 41, 431-437.

[640] Kitamura, T., Brauneis, U., Gatmaitan, Z. and Arias, I.M. (1991)

Hepatology 14, 640-647.

[641] Cohen, D.E., Leonard, M.R. and Carey, M.C. (1992) Hepatology

16, 90a.

[642] Meyers, M., Slikker, W. and Vore, M. (1981) J. Pharmacol. Exp.

Ther. 218, 63-73.

[643] Boyce, W. and Witzleben, C.L. (1973) Am. J. Pathol. 72, 427-432.

[644] Ayotte, P. and Plaa, G.L. (1988) Hepatology 8, 1069-1078.

[645] Becker, B.A. and Plaa, G.L. (1965) Toxicol. Appl. Pharmacol. 7,

708-718.

[646] Durand-Schneider, A.M., Maurice. M., Dumont, M. and Feldmann,

G. (1987) Hepatology 7, 1239-1248.

[647] Stieger, B., Meier, P.J. and Landmann, L. (1994) Hepatology 20,

201-212.

[648] Larkin, J.M. and Palade, G.E. (1991) J. Cell Sci. 98, 205-216.

268 R.P.J. Oude Elferink et al. / Biochimica et Biophysica Acta 1241 (1995) 215-268

[649] Rank, J. and Wilson, I.D. (1983) Hepatology 3, 241-247.

[650] Gartung, C., Ananthanarayanan, M., Rahman, M.A., Stolz, A,,

Suchy, F.J. and Boyer, J.L. (1993) Hepatology 18, 139A.(Abstract)

[651] Kothe, M.J.C., Bakker, C.T.M., De haan, J., Maas, A., Jansen,

P.L.M. and Oude Elferink, R.P.J. (1993) Hepatology 18,

138A.(Abstract)

[652] De Herder, W.W., Van der Hul, R.L., Bonthuis, F., Jansen, P.L.M.,

Oude Elferink, R.P.J. and Visser, T.J. (1989) Ann. Endocrinol. 50,

98. [653] Oude Elferink, R.P.J. and Jansen, P.L.M. (1994) Pharmacol. Ther.

64, 77-97.

[654] Polhuis, M., Kuipers, F., Vonk, R.J. and Mulder, G.J. (1989) J.

Pharmacol. Exp. Ther. 249, 874-878.

[655] Muhler, A., Oude Elferink, R.P.J. and Weinmann, H.J. (1993)

MAGMA 1, 134-139.

[656] Herz, R., Paumgartner, G. and Preisig, R. (1976) Scand. J. Gas-

troenterol. 11,741-746.

[657] Javitt, N.B. (1966) Nature 210, 1262-1263.

[658] King, J.E. and Schoenfield, L.J. (1971) J. Clin. Invest. 50, 2305-

2312.

[659] Yousef, I.M., Tuchweber, B., Vonk, R.J., Masse, D., Audet, M.

and Roy, C.C. (1981) Gastroenterology 80, 233-241.

[660] Van der Meer, R., Vonk, R.J. and Kuipers, F. (1988) Am. J.

Physiol. 254, G644-G649.

[661] Oelberg, D.G., Chari, M.V., Little, J.M., Adcock, E.W. and Lester,

R. (1984) J. Clin. Invest. 73, 1507-1514.

[662] Gumucio, J.J. and Valdivieso, V.D. (1971) Gastroenterology 61,

339-344.

[663] Tavoloni, N., Reed, J.S. and Boyer, J.L. (1982) Proc. Soc. Exp.

Biol. Med. 170, 486-492.

[664] Farthing, M.J.G. and Clark, M.L. (1981) Biochem. Pharmacol. 30,

3311-3316.

[665] Rotolo, F.S., Branum, G.D., Browers, B.A. and Meyers, W.C.

(1986) Am. J. Surg. 151, 35-40.

[666] Witzleben, C.L. (1971) Am. J. Pathol. 62, 181-194.

[667] Heijn, M. (1994) Characterization of ATP-dependent organic anion

transport, PhD. Thesis, Amsterdam.

[668] Haussinger, D. (1994) in Transport in the Liver (Keppler, D. and

Jungermann, K., eds.), pp. 97-116, Kluwer Academic Publ., Dor-

drecht/Boston/London.

[669] Stieger, B., Hagenbuch, B., Landmann, L., HSchli, M., Schroeder,

A. and Meier, P.J. (1994) Gastroenterology 107, 1781-1787.

[670] Kullak-Ublick, G.A., Hagenbuch, B., Stieger, B., Wolkoff, A.W.

and Meier, P.J. (1994) Hepatology 20, 411-416.

[671] Boyer, J.L., Ng, O.C., Ananthanarayanan, M., Hofmann, A.F.,

Schteingart, C.D., Hagenbuch, B., Stieger, B., and Meier, P.J.

(1994) Am. J. Physiol. 266: G382-G387.

[672] Oude Elferink, R.P.J., Ottenhoff, R. van Wijland, M., Smit, J.J.M.,

Schinkel, A.H. and Groen, A.K. (1995) J. Clin. Invest. 95, 31-38.

[673] Smith, A.J., Timmermans-Hereijgers, J.L.P.M., Roelofsen, B.,

Wirtz, K.W.A., van Blitterswijk, W.J., Smit, J.J.M., Schinkel, A.H.

and Borst, P. (1994) FEBS Lett. 354, 263-266.