Hepatobiliary transport of hepatic 3-hydroxy-3-methylglutaryl coenzyme a reductase inhibitors conjugated with bile acids

Hepatobiliary Transport of Hepatic 3-Hydroxy-3methylglutaryl Coenzyme A Reductase Inhibitors Conjugated With Bile Acids ERNST PETZINGER,1 LUTZ NICKAU,1 J~RGEN A. HORZ,1 SIEGFRIED SCHULZ, 1 GO'NTHER WESS, 2 ALFONS ENHSEN, 2 EUGEN FALK,2 KARL-HEINZ BARINGHAUS,2 HEINER GLOMBIK,2 AXEL HOFFMANN,2 STEFAN MULLNER,2 GEORG NECKERMANN,2 AND WERNER KRAMER2

To obtain prodrugs with affinity to liver parenchymal cells, the hepatic 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors HR 780 and lovastatin (syn. mevinolin) were conjugated with the bile acids cholic acid, taurocholic acid, and glycocholic acid. Hepatic uptake and biliary excretion of the coupled drugs were investigated and compared with the noncoupied drugs. Studies were performed with livers of normal Wistar rats, and TR-/GT- Wistar rats with deficient drug excretion. The experiments showed that the parent drug HR 780 was slowly excreted into bile. In contrast, the excretion of the bile acid-conjugated Hit 780 derivatives S 3554 (conjugated with cholate), S 3898 (conjugated with glycocholate), and S 4193 (conjugated with taurocholate) was rapid and very efficient in both groups of rat strains. The bile acid-conjugated HMGCoA reductase inhibitors showed a 10 to 20 times higher affinity for the uptake systems of bile acids than the noncoupled parent drug compounds, and even higher affinities than the bile acids themselves. The cholate conjugate of HR 780 (compound S 3554) was shown to be a noncompetitive inhibitor of taurocholate uptake and a competitive inhibitor of soditun-independent cholate uptake (Ki = 1 pmol/L). Uptake of radiolabeled S 3554 into isolated rat hepatocytes was observed to be rapid, cell specific, saturable, energy dependent, and carrier mediated. However, the carrier for S 3554 uptake was found not to be the cloned Na+-dependent taurocholate cotransporting polypeptide Ntcp. Expression of this carrier cRNA in Xenopus laevis oocytes did not stimulate S 3554 uptake. (HEPATOLOGY1995;22:1801-1811.)

Abbreviation: HMG CoA, 3-hydroxy-3-methylglutaryl coenzyme A. From the 1Institute of Pharmacology and Toxicology,Justus-Liebig-University Giessen, Giessen, and 2Hoechst Aktiengesellschaft, Leitung Pharma Forschung Stoffwechsel, Postfach, Germany. Received August 18, 1994; accepted August 3, 1995. Supported by the Deutsche Forschungsgemeinschaft, SFB 249, project B3 and by grant PE 250/6-1. Address reprint requests to: Prof Dr Ernst Petzinger, Institute of Pharmacology and Toxicology, Justus-Liebig-University Giessen, Frankfurter Strafle 107, D-35392 Giessen, Germany. Copyright © 1995 by the American Association for the Study of Liver Diseases. 0270-9139/95/2206-002853.00/0

Bile acid transport systems in hepatocytes are presumed to transport drugs. 1-6 To achieve an improvement of drug uptake into hepatocytes, drugs were coupled to bile acids. 7's No detailed study has been published to define this uptake of a d r u g - b i l e acid conjugate in hepatocytes. Cell-selective uptake of the conjugates is the first step in targeting them to the liver. We have studied the transport into liver parenchymal cells and biliary excretion of a radiolabeled drug-bile acid conjugate, compound S 3554. This compound is a cholate conjugate of the cholesterol-lowering drug HR 780, 9 which is related to the 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reductase inhibitor lovastatin (formerly named mevinolin). S 3554 is a prodrug that, to inhibit cholesterol biosynthesis, requires splitting offthe open lactone form of HR 780.1° We have compared the cellular uptake and canalicular secretion of radiolabeled S 3554 with radiolabeled HR 780. We have also characterized the inhibition of bile acid uptake by bile a c i d - H R 780 conjugates and bile acid-lovastatin conjugates, respectively. It was found that by coupling the drug to bile acids, the uptake properties are shifted from passive physical diffusion to a selective carrier-mediated cell uptake. As a result, a more efficient clearance by the liver of bile acid-conjugated drugs occurred. MATERIALS AND METHODS [3H]Tauroch01ic acid, specific radioactivity 77.7 GBq/mm01, was bought from New England Nuclear, Dreieich, Germany. [14C]Cholic acid, specific radioactivity 1.99 GBq/mm0l, was obtained from Amersham, Braunschweig, Germany. Two different charges of 14C-radiolabeled S 3554 were available from the Department of Radiochemicals, H0echst AG, Frankfurt. One charge was labeled in the cholate molecule. The specific radioactivity of this [14C]S 3554 was 547 MBq/g (483 MBq/ mmol). A second charge was radiolabeled in the HR 780 molecule; the specific radioactivity of this compound was 811 MBq/ g (717 MBq/mmol). The specific radioactivity of the two [ltC]HR 780 charges were 1628 MBq/g (702.5 MBq/mmol) and 2064 MBq/g (890 MBq/mmol). Unlabeled taurocholic acid was from Carl Roth KG, Karlsruhe, Germany; cholic acid and glycocholic acid were purchased from Sigma GMBH, Deisenhofen, Germany. Collagenase was from Boehringer, Mann-

1801

1802

PETZINGER ET AL

heim, Germany, and the Bio-Rad Protein Assay was from Bio-Rad Laboratories , Munich, Germany. The drugs lovastatin and HR 780 were from Hoechst AG, Frankfurt/M, Germany. The synthesis of drug-bile acid conjugates has been described elsewhere, s~2 All other chemicals were from Merck, Darmstadt, Germany, or from Serva, Heidelberg, Germany. All animals received humane care in compliance with the institution's guidelines (§9, abs. 2 and 3, TierSchG, FRG). Whole-Liver Experiments; Biliary Excretion Studies. Male Wistar rats were maintained on a standard diet (A]tromin, Altrogge, Lage/Lippe, Germany) and given free access to water. For in situ liver perfusion experiments, the rats were anesthetized with urethane (20% in solution, 5 mL/kg intraperitoneally) and received either 500 #L of a solution of the respective nonlabeled compounds (concentration, 0.1 and 1 mmol/L), or a mixture of unlabeled plus radiolabeled S 3554 (0.2 #Ci). The final concentration of 0.1 mmol/L was applied through a jejunal vein as bolus injection. The compounds were dissolved in 10 #L ethanol and subsequently diluted with 490 #L of 0.9% phosphate-buffered NaC1. The final ethanol concentration was no more than 2%, v/v. After injection of the drug compounds, bile was sampled every 5 minutes for 80 minutes and then every 10 minutes subsequently for 2 hours. Each experiment was repeated at least three times. A series of in situ perfusion experiments was performed with Wistar TR /GT- rats obtained from Drs P. L. M. Jansen and R. P. J. Oude-Elferink, Amsterdam. These rats lack the adenosine triphosphate-driven canalicular transporter for organic anions, MOAT (indicated by "TR-" = transport negative), but sustain the excretion of bile acids.13'l* The rats were double mutants, because of their deficiency in this transport as well as a 95% defect in glucuronyl transferase activity (indicated by "GT-"). 15The experiments were also carried out as described above for normal Wistar rats. For in vitro liver perfusion experiments, the liver was isolated and transferred into a perfusion box that was temperature controlled. The liver was perfused through a portal vein catheter with 37°C warm buffer A (containing 118 mmol/L NaC1, 25 mmol/L NaHCO~, 2.28 mmol/L K2SOt, 2.12 mmol/ L MgSOt, 2.5 mmol/L CaC12, 10 mmol/L glucose, and 2% (w/ v) dextran DT 70, pH 7.4), in a recirculating system for 30 minutes under carbon dioxide/oxygen gassing (5%/95% vol/ vol) to reach a bile flow of approximately 2.5 #L/g liver weight. During this time, bile was sampled for control analysis. The compounds were then added in a bolus of 50 #L into the perfusion buffer, and bile was sampled at 5-minute intervals (4×), and later at 10-minute intervals, in preweighed tubes. Bile samples were immediately frozen, stored in the dark, and subjected to thin-layer chromatography or high-pressure liquid chromatography analysis after appropriate sample preparation. Preparation of Isolated Rat Hepatocytes. Male Wistar rats of 250 to 300 g body weight received a standard diet and water ad libitum. Hepatocytes were isolated by perfusion of the liver with 0.05% collagenase in Krebs-Henseleit buffer according to the method described by Berry and Friend '6 with some modifications. 17 The cell suspension was diluted to a concentration of 2 × 105 cells (3.5 mg protein) per milliliter. Viability of the cells was 85% to 95% as tested by trypan blue staining. For the uptake experiments, the suspension was kept in shaking Erlenmeyer flasks at 37°C, under constant O2/CO2 gassing. Protein concentration was determined according to the Bradford method is using the Bio-Rad protein reagent.

HEPATOLOGYDecember 1995 Transport Experiments With Bile Acids, HR 780, and S 3554. To determine inhibition of bile acid uptake by d r u g bile acid conjugates, 2 × 106 hepatocytes in 1 mL Tyrode buffer were incubated for 30 seconds, with the conjugated drugs in a concentration of 100 #mol/L. The radiolabeled bile acids were added in final concentrations of 40 nmol/L [3H]taurocholate/10 #mol/L taurocholate or 0.55 pmol/L [14C]cholate/10 #mol/L cholate. For comparison, the transport of 3 #mol/L [3H]serine/10 #mol/L serine was measured. Inhibition concentrationso values were determined for bile acid uptake in the presence of increasing concentrations of each inhibitor. Controls received dimethyl sulfoxide buffer or other appropriate solvents used for the solution of the compounds. The final solvent concentration was less than 1% and had no detectable effect on the uptake. Uptake studies with isolated hepatocytes were also performed with the radiolabeled HR 780 and S 3554. Here 0.1 #Ci of each compound was added. From each cell suspension, samples of 100 #L were withdrawn 15, 45, 75, 105, and 135 seconds after the start of the experiment and at depicted intervals during the next 20 minutes. The samples were transferred to small tubes and immediately centrifuged at 10,000g through a silicone oil layer. 19 Radioactivity was measured in the cell pellet and in the supernatant with Rotiszint Eco Plus, in a Packard 2660 fluid scintillation counter (Rath, Karlsruhe, Germany). Contamination of the cell pellet with adhering supernatant was measured by adding extracellular [3H]inulin to the cells. The inulin space in the cell pellet was less than 0.2% of the radioactivity added. Expression of the Na÷-dependent Taurocholate Cotransporting Polypeptide in X e n o p u s l a e v i s Oocytes. The Ntcpcomplementary DNA was a gift of Prof Dr P. Meier-Abt and Drs Bruno Hagenbuch and Bruno Stieger, Zfirich. The complementary RNA was obtained from the prLNaBA clone (which contained the Ntcp-complementary DNA) by standard techniques as described by Sambrook et al 2° and Jacquemin et al. 42Xenopus laevis oocytes were obtained from the ovaries of anesthetized female frogs by incubation of the tissue in collagenase (2 mg/mL) containing modified Barth's Solution buffer. Anesthesia was achieved with 0.15% Tricaine. The oocytes were incubated in MBS medium at 17°C and then 0.5 ng of the specific cRNA/oocyte was injected by means of a Bachhofer nanoliter micro pump. Two and 3 days after the expression of the Ntcp, the uptake of radiolabeled HR 780, S 3554, and taurocholate were measured in the transport buffer medium either in the presence or absence of sodium ions. Phalloidin-Response Test. Phalloidin was applied at a concentration of 10 pmol/L to isolated rat hepatocytes. After 20 minutes, cells developed membrane protrusions caused by the uptake of phalloidin by a bile acid transport system. 21 The percentage of protrusion-bearing cells was determined in a Btirker-Tfirk chamber, under a light microscope. More than 90% of the hepatocytes developed membrane protrusions within that time. For protection, the test compounds were added in an equimolar concentration (10 pmol/L) 30 seconds before phalloidin. Protection was expressed as percent decrease of protrusion-bearing cells versus controls. Calculation and Statistics. The kinetic data of the transport experiments were from at least three experiments obtained with "n" different cell preparations. Mean values a r e given by x _+ SD. Significance was tested by the bifactorial analysis program BMDP 2V. The initial velocity (vi) of uptake was calculated by linear regression from the first five uptake

HEPATOLOGYVol. 22, No. 6, 1995

PETZINGER ET AL

drug - bile acid conjugates

,

.o !.~3.R

£ ~ O N O N#

OH

HR780

H=~

No

ONON

~'R

ON

1803

appeared earlier in bile than HR 780, and the concentration reached a peak after 15 minutes. The peak radioactivity of S 3554 was 8 to 10 times higher than that of HR 780. The excretion of HR 780-derived radioactivity was found to be at a low level and very delayed. Bile flow was not reduced by HR 780 or S 3554 under these conditions in situ. For comparison, the excretion of radiolabeled [3H]taurocholic acid into bile was measured. The bile acid was excreted much faster than S 3554. A peak concentration was reached within 5 minutes, and 96% to 98% of the injected dose recovered after 20 minutes. The occurrence of S 3554 in bile raised the question whether bile acid-conjugated drugs are secreted by canalicular bile acid carriers or by drug carriers. Canalicular bile acid secretion is performed by an adenosine

NIF" 2 ~ c

mevinolin 6

bile acid conjugates with: HR780 mevinolin R

S 2887

excretion [nmol/50/JI bile]

TC-excretion [pmolF-rj0/JIbile]

2000

5 1500 4

o.

8 3554

.,^coo. H

S 3898

2-

,~.~o~. H

8 4193

1

3-

1000

5OO 0.

FIG. 1. Structures of HMG-CoA reductase inhibitors HR 780 and lovastatin (syn. mevinolin) coupled with bile acids cholate, taurocholate, and glycocholate.

20

values (those measured between 15 and 135 seconds). IC5ovalues were determined by logarithmic regression.

6 - excretion [nmol/50/JI bile]

RESULTS Drug-Bile Acid Conjugates. Figure 1 depicts the structures of the compounds tested. The natural bile acid cholate and its glycine and taurine conjugates, taurocholate and glycocholate, were modified to 3/~-(2aminoethoxy)-derivatives,z1'12 This was to obtain the spacer for the coupling with the open ring forms of lovastatin (syn. mevinolin) and HR 780. The modified bile acids retained two alpha-hydroxy groups at C7 and C12 and a carboxy or sulfonic acid terminus, corresponding to nonamidated or amidated bile acids. Thus, essential recognition sites for transport by hepatic bile acid carrier proteins still remained in the drug-bile acid molecules. Excretion of HMG-CoA R e d u e t a s e l n h i b i t o r s a n d Their Bile A c i d Conjugates Into Bile In Situ a n d In Vitro. The conjugation with cholate increased markedly the elimination of HR 780 into bile (Fig. 2A). In normal Wistar rats in situ, approximately 88% of the radioactivity from injected [14C]S 3554, but only 38% of [14C]HR 780, was recovered in bile sampled over a period of 120 minutes. The S 3554-derived radioactivity

40

60

80

100

120

A

t [min] -eHR 780 "*-S 3554 ~TC

54 3 2 1 0

o

--

I

20

,'o

;o t (min]

8'0

' 100

' 120

a

II.HR 780 ~ S 3554

FIG. 2. Biliary excretion of HR 780 and its cholate conjugate S 3554 in normal Wistar rats (A) and TR-/GT- Wistar rats (B). From each compound, 50 nmol (12.5 nmol radiolabeled S 3554/HR 780 corresponding to 0.25 ~Ci/rat plus 37.5 nmol unlabeled compound/ rat) were injected in the jejunal vein of anesthetized rats from which bile was collected. About 90% of S 3554 but 40% of HR 780 derived radioactivity appeared in bile during 2 hours. Whereas compound S 3554 peaked from 15 to 35 minutes after injection, the excretion of HR 780 was constant over the whole sampling period. For comparison, 5 nmol taurocholate/0.25 #Ci [aH]taurocholate was applied to normal Wistar rats (A).

1804

PETZINGER ET AL

HEPATOLOGYDecember 1995

A

uptake [pmol/mg prot.] 30002S00 t T

B

a000.uptake [pmoi/mgprot.]

] ] 0

6

control

10

18

t[min] ~

HR780

~

8:1884

20

.

o

5

.

.

1o t[min]

.

16

20

--~- 8 4193 --

1000

0

8 Ilia

--

control

o ,, mevinolin

- ~ - 8 2887

U

800

6O0

400

200

0

I

0

I

b

I

i

i

100200300400500600 t [sec] Dcontrol ~HR 780 ~S 3554

triphosphate-dependent transporter that was recently described in several l a b o r a t o r i e s . 22"26 It is not yet known whether this transporter accepts drugs. Anionic drugs are, however, secreted after metabolic modifica-

FIG. 3. Inhibition of taurocholate (A, B) and cholate (C) uptake by HMG-CoA reductase inhibitors and their bile acid conjugates in isolated rat hepatocytes. One hundred micromolars of each compound was added to the cells 30 seconds before (z4C)taurocholate/ taurocholate (40 nmol/L/10 #tool/L) or (14C)cholate/cholate (0.55 #tool/L/10 #mol/L), respectively. (A) Inhibition of taurocholate uptake by bile acid conjugates of HR 780; (B) inhibition of taurocholate uptake by bile acid conjugates of lovastatin (syn. mevinolin); (C) inhibition of cholate uptake by HR 780 and S 3554.

tions by a second transporter, which again is adenosine triphosphate dependent. This system is a canalicular multispecific organic anion carrier, which transports glutathione-S-drug conjugates, glucuronidated drugs,

HEPATOLOGYVol. 22, No. 6, 1995

lOO-

PETZINGER ET AL 1805

Inhibition of vl [%]

A

76

shown). This indicated t h a t each bile acid conjugate of HR 780 was t r a n s p o r t e d by a canalicular t r a n s p o r t system as if it were a bile acid. Structure.Activity Relationship of HR 780 and HR 780.BUe Acid Conjugates on Bile Acid Uptake. Bile acid conjugates of HR 780 interfere with hepatic bile acid

80 26 ¸

700

vi ipmol/(mg prot.* rain)] a~t

0

i

i

i

i rllll

0,1

f

r

i

IBJ~H

1

r

i

I

100

"*'- 8 4193

w~)/m~l

•

lo00

1000

Cone. of Inhibitor [IJM]

--~- 8 3554

pint.

800

IllllJ

10

I(~

500

-e- HR780 400

lOO,

Inhibition of vi [%]

B

/

7660-

300 200 100

0

26-

A

0

5 10 15 20 vi/s [pmol/(mg * min * tJM)] control

0

l

0,1

,

,

ilill,

i

1

,

i

llllil

i

i

i

llllll

10 100 Cone. of inhibitor hJM]

8 8884

~

8 4198

i

t

i

~

$

8884

"8-

8

3 8 6 4 SuM

25

I~M

llllll ~ -

1000

--D- HR780

FIG. 4. Concentration-dependent inhibition of taurocholic acid (A) and cholic acid (B) uptake by bile acid conjugates of HMG-CoA reductase inhibitors. Initial uptake velocitywas determined from the uptake during 15 to 135 secondsin either the presenceor the absence of the inhibitors. ICsovalues were determined graphically from the plots.

2000"

8

3 8 5 4 SuM

vi [pmol/(mg prot., rnin)]

e.

e.

1500

6" lie

1000

and bilirubin glucuronides. 23'27-29 This carrier was reported to be probably identical 30"32 with the multidrug related protein t r a n s p o r t e r described by Cole et al. 88 In a genetically a b e r r a n t Wistar r a t strain, TRrats, the multispecific organic anion system is absent, 13z5'34'3~ and these rats secrete bile acids but not conjugated drugs into bile. We have used T R - / G T - rats to test the elimination of radiolabeled S 3554 after injection in situ (Fig. 2B). As with normal Wistar rats, S 3554-derived radioactivity appeared in bile samples much earlier, and at a m a r k e d l y higher concentration t h a n HR 780-derived radioactivity. T R - / G T - rats excreted 75% of the injected a m o u n t of [z4C]S 3554 and 18% of [~4C]HR 780 in bile over a period of 120 minutes. Again bile flow was not impaired by the compounds, but the total bile production in T R - / G T - rats was 25% lower t h a n in normal rats. Finally, isolated perfused livers obtained from the T R - / G T - rats excreted S 3554, S 3898, and S 4193 (not

6000 ~ ' ~ , "

|

B

0

20 40 80 80 100 v i l e [pmol/(mg • min • IJM)] control 8 $$84 SpM

120

~ 8 8684 IuM -4-. 8 8684 SuM

Fro. 5. Woolf-Hofsteeplots of the inhibition kinetics by bile acidHMG-CoA reductase inhibitors on (A) taurocholate uptake (in normal Tyrode buffer) and (B) cholate uptake in sodium-free Tyrode buffer (sodium substituted by choline; see Materials and Methods). Insets showLineweaver-Burkplots from which K~values were determined.

1806

PETZINGER ET AL

uptake 1200 1

HEPATOLOGYDecember 1995

[pmol/mg prot.]

1000

800

600

400

200

0 0

100

200

A

300

400

500

600

t [sec] I¢ontrol ~ S 3554

800

uptake

[pmol/mg prot.]

uptake. Figure 3A and 3C depicts the inhibition of taurocholic acid uptake and cholic acid uptake by HR 780 and three of its bile acid conjugates. All conjugates completely prevented taurocholate uptake when applied at 100 #mol/L final concentration, whereas inhibition by HR 780 was significantly less. Irrespective of whether cholic acid (S 3554), taurocholic acid (S 4193), or glycocholic acid (S 3898) was conjugated to the drug, the ICso values were identical: 5 #mol/L for S 3554, and 7 #mol/L for S 3898 and S 4193 (Fig. 4A). In contrast, the ICso value for HR 780 was 100 #mol/L. Comparable results were obtained, if, instead of taurocholic acid uptake, the uptake of cholic acid was inhibited (Fig. 4B). Regarding cholic acid uptake, the compounds had even lower ICso values: 3 #mol/L for S 3554 and 2 ~mol/ L for S 4193, whereas for HR 780 an ICso of 31 #mol/ L was determined. However, the ratios between the IC~o values of the bile acid-HR 780 conjugates versus HR 780 remained identical. Kinetic experiments showed a competitive inhibition by S 3554 (the cholate conjugate; Fig. 5B) and S 4193 (the taurocholate conjugate; not shown) under sodium-free conditions versus cholate uptake. The K~ values were 1 #mol/L for both compounds. In contrast, S 3554 noncompetitively inhibited taurocholate uptake (Fig. 5A). Because of the low Ki values, the binding of HR 780 and its conjugates was strong, and washing the cells only partly abolished the noted inhibition. To prove whether the inhibition is a result of nonspecific disturbance of the sodium gradient, the sodiumdependent serine uptake was measured for comparison. Whereas serine uptake into hepatocytes was not inhibited by 100 #mol/L HR 780, 100 #mol/L S 3554 inhibited approximately 34% of the initial Vi of serine uptake. Investigation With L o v a s t a t i n a n d L o v a s t a t i n Bile

Acid-Conjugates. Lovastatin (syn. mevinolin) is used

600

for clinical treatment of hypercholesterolemia.86'37The inhibition of taurocholate uptake by 100 #mol/L lovastatin and its conjugate with cholic acid, substance S 2887, is given in Fig. 3B. The conjugate again was a much stronger inhibitor than the free drug. S 2887 inhibited 95% of initial taurocholate uptake, whereas lovastatin inhibited 47%. The ICso value of S 2887 was 9 #mol/L, but was 100 #mol/L for lovastatin. These values are almost identical to the corresponding ICso data of HR 780 and its cholate conjugate.

400

Transport o f R a d i o l a b e l e d HR 780 a n d S 3554 Into Isolated R a t Hepatocytes. Our studies so far indicated

200

high-affinity binding of each drug-bile acid conjugate

0

0

B

100

200

300

400

t [sec] ~control -~-HR 780

500

600

FIG. 6. Uptake of (14C)S 3554 and (14C)HR 780 in isolated rat hepatocytes. Isolated rat hepatocytes were incubated at 37°C in Tyrode buffer containing 3% bovine serum albumin when 100 #mol/L nonlabeled HR 780 and S 3554 was added 30 seconds before 5 #mol/ L of the labeled compounds. (A) (14C)S 3554 uptake is decreased by 100 #mol/L S 3554; (B) 100 #mol/L HR 780 does not inhibit uptake of (14C)HR 780.

PETZINGER ET AL

HEPATOLOGYVol. 22, No. 6, 1995

uptake [pmol/mg prot.]

1200

1000

1807

uptake [pmol/mg cell protein]

1000 800 800

see 600 400 400

200

200

0

100

200

A

300

400

500

600

t [sec]

!

o

500 t [$ec]

e 3 7 °C ~ 4 °C

800

0

FIG. 8. Inhibition of S 3554 uptake by oligomycin (A) and decreased uptake into permeabilized (×) hepatecytes. Oligomycin, 10 ~g/mL cell suspension, was present during 10 minutes in an albuminfree Tyrode buffer. The cells were washed with albumin containing Tyrode buffer to remove oligomycin and [z4c]s 3554 was added (5 #mol/L final concentration). Permeabilization of hepatocytes was achieved by freezing and thawing the cells. Those cells were incubated at 37°C in albumin containing Tyrode buffer.

uptake [pmol/rng prot.]

600 to bile acid carrier proteins. Radiolabeled conjugate [14C]S 3554 was used to measure whether the conjugates are indeed transported into hepatocytes by saturable carrier-mediated pathways. Figure 6 shows the uptake of [14C]S 3554 and [z4C]HR 780 into hepatocytes in the presence of an excess of each unlabeled, cold compound. Whereas cold S 3554 inhibits, HR 780 does not. Mutual uptake inhibition was not observed between HR 780 and S 3554 (not shown). A strong temperature dependence for the uptake of S 3554 was observed, but not for that of HR 780 (Fig. 7). Only S 3554 uptake but not HR 780 uptake was totally blocked by the metabolic inhibitor oligomycin (Fig. 8). The resid-

400

200

0 0

B

100

200

300

400

t [sec] t 3 7 °C .*-4 °C

500

600

FIG. 7. Temperature dependence of the uptake of S 3554 and HR 780 into isolated rat hepatocytes. Uptake of 5 #mol/L (14C)S 3554 (A) and (z4C)HR 780 (B) was measured in Tyrode buffer containing 3% albumin at 37°C and 4°C. Whereas S 3554 uptake is completely blocked at 4°C, the uptake of HR 780 is only delayed.

1808

PETZINGER ET AL

HEPATOLOGYDecember 1995

uptake [pmol/rng prot.]

1400

Sub. t(mln]

s ss~Ktso) control(120) S S~KS0} control(SO) s ~4(s)

1200

control(6) HR 780{120) control(120) HR ~0(S0)

1000

I I i m l m 4 l

I ! I I t

600

800 Sub. t[min] s 3684(t20) control(120) S S~SO) control(60) S S~S) control(E) HR 780(t20) control(t20) HR 780(60) control(S0)

600

400

200

I

almmmmN

control(6) ,

1000 1600 uptake [dpm/oocyto]

I

2000

2500

I

I

B

I

I I i i I

I I

I

I

HR7S0{6) i control(S)

I

600

Z

0 0

5

10

A

15

20

25

3O

t [rain]

• hepatocytes -*- hepatoma cells

1400 uptake [pmol/mg prot.]

1200

1000

1600

2000

2600

3o0o

umb Hm/oo~)

FIG. 10. Lack of specific uptake of S 3554 and HR 780 intoXenop u s laevis oocytes after 3 days expression of 0.5 ng prLNaBA derived cRNA (containing the Ntcp clone) per oocyte. Neither in the presence (A) or absence (B) of sodium ions was the uptake of [I*C]S 3554/ [14C]HR 780 (each with 5 #mol/L final concentration) increased during 120 minutes' incubation of the oocytes. For transport measurements the uptake buffer (see reference 50) contained 3% bovine serum albumin. As a control the uptake of [3H]taurocholate was measured with the same oocyte batch. Taurocholato uptake was stimulated 60- to 80-fold on injection of 0.5 ng cRNA within 3 days of expression.

ual uptake under these conditions was the same as uptake by permeabilized hepatocytes (Fig. 8). The uptake of S 3554 was not sodium dependent and required albumin in the incubation buffer. In the presence of albumin, a Km of 26 #mol/L and a Vm~ of 1,010 pmol/ mg protein/min was determined for this uptake of S 3,554. So far the results showed that S 3554 is transported carrier-mediated whereas HR 780 enters hepatocytes by nonionic diffusion. This was further verified by uptake experiments with hepatoma cells. These are cells that lack saturable bile acid transporters. 38'zs Uptake of [14C]S 3554 was lacking in hepatoma cells, whereas, in contrast, [14C]HR 780 was taken up by these cells, too (Fig. 9). Transport Studies With Xenopus laevis Oocytes After Expression of the Na+-Dependent Taurocholate Cotransporting Polypeptide. The uptake of bile acids into

1000

800

4O0

200

0

S

I

contro~SO)

HR7~s)

A

I

5

10

15

20

25

t [min] hepatocytes -*-hepatoma cells

30

FiG. 9. Lack of uptake of S 3554 into hepatoma cells. Uptake of 5 #mol/L (14C)S 3554 (A) and 5 #mol/L (14C)HR 780 (B) was measured in isolated rat hepatocytes and in Fao Reuber H 35 hepatoma cells at 37°C in Tyrode buffer containing 3% albumin. Hepatoma cells do not take up S 3554.

HEPATOLOGYVol. 22, No. 6, 1995

100

PETZINGER ET AL

inhibition l%]

76 ¸

50-

Mev.

HR780

to

c

8 2887 8 3664

FIG. 11. Inhibition of phalloidin response in isolated rat hepatocytes by bile acids, by HMG-CoA reductase inhibitors, and by bile acid conjugates of HMG-CoA reductase inhibitors. Phalloidin response was determined from the percentage of hepatocytes that showed membrane protrusions a f a r incubation with 10 pmol/L phalloidin. Mev: mevinolin; tc: taurocholic acid; c: cholic acid; S 2887: cholate conjugate of mevinolin; S 3554: cholate conjugate of HR 780.

hepatocytes occurs by different transport systems that overlap in substrate recognition and by sodium dependency. 4°42 Up to now with one cloned bile acid carrier, the Ntcp, we could directly prove the uptake of S 3554. However, expression of the Ntcp in Xenopus laevis oocytes did not improve S 3554 uptake, neither in the absence (Fig. 10B) nor in the presence (Fig. 10A) of sodium ions. When the uptake of HR 780 in the oocytes was studied too, it was found that this uptake was also not increased on Ntcp-cRNA expression (Fig. 10). Inhibition of Phalloidin Response. Phalloidin is a cyclopeptide, which is assumed to be taken up into hepatocytes by a bile acid transport system. 21'4a-45 This system was named multispecific bile acid transporter. Bile acids inhibit phalloidin uptake into hepatocytes, thereby preventing the development of membrane protrusions, which otherwise are seen during incubation of hepatocytes with phalloidin. 4e To test whether the phalloidin-sensitive bile acid transporter of the basolateral membrane is targeted by the drug-bile acid conjugates, the effect of the compounds on phalloidinpoisoned hepatocytes was investigated. Equimolar concentrations (10 #tool/L) of S 2887 or S 3554 prevented more than 90% of the cell response (development of membrane protrusions) that was normally observed after incubation in the presence of 10 #mol/L phalloidin. Conversely, lovastatin and HR 780 had almost no effect. The bile acids taurocholate and cholate at 10 pmol/L exercised an intermediate effect of 28% and 40% inhibition (Fig. 11). DISCUSSION

In previous studies it was shown that drugs that are coupled to bile acids showed marked improvement of

1809

hepatoceUular uptake and bfliary excretion. 7'1° Our conclusion was that drugs can be directed specifically to the liver on conjugation with bile acids. The drugs tested in this study, HR 780 and lovastatin, are HMGCoA reductase inhibitors, which are clinically used to lower cholesterol content in blood. The major target organ of these drugs is the liver. When the compounds were coupled to bile acids, the distribution throughout the body changed markedly, and much lower body burden by the drug was achieved; e.g., in the heart and the testicles the drug concentration was ~oth1° because of a hepatic first-pass elimination. In this investigation, we have studied whether the liver-specific pharmacokinetic properties of bile acid-derived prodrugs of HMG-CoA reductase inhibitors are attributable to a specific uptake, and excretion of the bile acid conjugates, by bile acid carriers. Conjugation of the drugs with bile acids exerted a dramatic change in their liver clearance. This was a result of a much higher affinity of the tested drug-bile acid conjugates toward bile acid uptake carriers. The conjugated compound S 3554 was efficiently taken up by a saturable, carrier-mediated pathway, whereas the drug alone, HR 780, was taken up slowly by passive physical diffusion. The conjugated drugs clearly exerted carrier-targeting specificity toward hepatocytes; e.g., other cells investigated exhibited no uptake of S 3554 but showed comparable small uptake of HR 780, similar to hepatocytes. We assumed that a bile acid transporter is involved in the uptake of this drug-bile acid conjugate, S 3554. Bile acid carriers have been cloned from rat liver by two groups: the Ntcp and the oatp carrier from P. Meie~s group in Z u r i c h , 41'42 and an isoform of microsomal epoxide hydrolase cloned by D. Levy's group, Los Angeles. 47'48 It was suggested that further bile acid/ drug carriers exist in the basolatoral membrane of hepatocytes. 5'49'5° Our studies with the Ntcp indicated that this bile acid carrier is not the transporter for S 3554. We suggest that another bile acid carrier is involved in the uptake of S 3554. This bile acid carrier might be a cholate transporter (cholate was a competitive inhibitor of S 3554 uptake whereas taurocholate inhibited noncompetitively) and should transport even in the absence of sodium ions. An experiment with Xenopus laevis oocytes that had received 0.5 ng oatp-cRNA also could not show uptake of S 3554 by this transport system (data unpublished). Our working hypothesis is that the phalloidin-sensitive multispecific bile acid transporter could be a hepatic transport system for the drug-bile acid conjugates. This would explain why phalloidin poisoning was almost completely prohibited by two bile acid-drug compounds. The affinity of the drug-bile acid conjugates for bile acid carriers was very much determined by the bile acids and less by the drug. The ICso values obtained with these drug-bile acid conjugates were of the same order of magnitude as the ICso values of taurocholate-

1810

PETZINGER ET AL

conjugated chlorambucil, v Because the structures of chlorambucil and the HMG-CoA reductase inhibitors are totally different, it appears that it is the bile acid moiety that is responsible for the interaction of the compounds with the bile acid carriers. Further, because the IC~o values were very similar irrespective of the bile acid (cholate, taurocholate, or glycocholate) that was conjugated to the compounds, the type of bile acid in the conjugates appears to have no crucial effect on their affinity. However, the drug-bile acid conjugates should not be considered simply as bulky bile acids. It is concluded that the drugs themselves have some additional effects on the membrane. This was indicated by the observation of a weak inhibition of serine uptake by the conjugates, although serine uptake was not inhibited by the unconjugated drugs nor by bile acids. It is very likely that part of this inhibition is attributable to the lipophilicity of the conjugates. The pharmacokinetics of S 3554 mimicked those of bile acids, not only for uptake but also for secretion. This was observed when the biliary excretion of the synthetic conjugates was investigated by use of TR-/ GT- rats, which have preserved bile acid secretion but significantly lack excretion of negatively charged drugs. The results clearly document that canalicular output of S 3554 in the bile of these animals occurs in a similar way to that in the bile of normal rats. When the appearance curve in bile of S 3554 was compared with the appearance curves of taurocholate and HR 780, respectively, the S 3554 curve is seen between the HR 780 and taurocholate curves. This also demonstrates that drug excretion kinetics were shifted on conjugation with a bile acid to bile acid excretion kinetics. The current study favors the concept of hepatocellular uptake and biliary excretion of drug-bile acid conjugates by bile acid carriers. This mechanism elegantly explains the changes in the clearance kinetics we have found. Further studies on cloned bile acid and drug transporter are, however, required to verify this concept and to identify in molecular terms the transport systems involved.

Acknowledgment: We are very much indebted to Drs Peter Jansen and Ronald Oude-Elferink, Amsterdam, for the delivery of TR-/GT- rats. We gratefully acknowledge the gift of the Ntcp clone by Drs Peter Meier-Abt, Bruno Hagenbuch, and Bruno Stieger. We also thank Dr Bruce Boschek for his support in preparation of the manuscript. REFERENCES 1. Petzinger E, Frimmer M. Comparative studies on the uptake of 14C-bile acids and 3H-demethylphalloin in isolated rat liver cells. Arch Toxicol 1980;44:127-135. 2. Frimmer M. What we have learned from phalloidin. Toxicol Lett 1987;35:169-182. 3. Petzinger E, Ziegler K, FrimmerM. Occurrenceofamultispecific transporter for the hepatocellular accumulation of bile acids and various cyclopeptides. In: Paumgartner G, Stiehl A, Gerok W, eds. Bile acids and the liver. Lancaster: MTP Press, Ltd, 1987:111-124.

HEPATOLOGYDecember 1995 4. Zimmerli B, Valantinas J, Meier PJ. Multispecificity of Na ÷dependent taurocholate uptake in basolateral (sinusoidal) rat liver plasma membrane vesicles. J Pharmacol Exp Ther 1989; 250:301-308. 5. Petzinger E, FSllmann W, Blumrich M, Schermuly R, Schulz S, Hahnen J, Feit PW. Interaction of bumetanide derivatives with hepatocellular bile acid uptake. Am J Physiol 1993;265:G942G954. 6. Petzinger E. Transport of organic anions by the liver. Rev Physiol Biochem Pharmacol 1994; 123:49-211. 7. Kramer W, Wess G, Schubert G, Bickel M, Girbig F, Gutjahr U, Kowalewski S, et al. Liver-specific drug targeting by coupling to bile acids. J Biol Chem 1992;267:18598-18604. 8. Kramer W, Wess G. European Patent Application; EP 0417 725 A2. 1989. 9. Wess G, Kesseler K, Baader E, Bartmann W, Beck G, Bergmann A, Jendralla H, et al. Stereosetective synthesis of HR780: a new highly potent HMG-CoA reductase inhibitor. Tetrahedron Lett 1990;31:2545-2548. 10. Kramer W, Wess G, Ehnsen A, Bock K, Falk E, Hoffmann A, Neckermann G, et al. Bile acid derived HMG-CoA reductase inhibitors. Biochim Biophys Acta 1994; 1227:137-154. 11. Wess G, Kramer W, Ehnsen A, Glombik H, Baringhaus KH, Bock K, Kleine H, et al. Preparation of 3a- and 3fl-(w-aminoalkoxy)7a,12a-dihydroxy-5/~-cholanoic acid esters: versatile shuttles for drug targeting. Tetrahedron Lett 1993;34:817-818. 12. Wess G, Kramer W, Schubert G, Ehnsen A, Baringhaus KH, Glombik, H, Miillner S, et al. Synthesis of bile acid-drug conjugates: potential drug-shuttles for liver specific targeting. Tetrahedron Lett 1993;34:819-822. 13. Jansen PLM, Peters WH, Lamers WH. Hereditary chronic conjugated hyperbilirubinemia in mutant rats caused by defective hepatic anion transport. HEPATOLOGY1985;5:573-579. 14. Jansen PLM, Groothuis GMM, Peters WHM, Meijer DKF. Selective hepatobiliary transport defect for organic anions and neutral steroids in mutant rats with hereditary conjugated hyperbilirubinemia. HEPATOLOGY1987;7:71-76. 15. Jansen PLM, Peters WHM, Meijer DKF. Hepatobiliary excretion of organic anions in double-mutant rats with a combination of defective canalicular transport and uridine-5'diphosphate-glucuronosyltransferase deficiency. Gastroenterology 1987; 93:10951103. 16. Berry MN, Friend DS. High-yield preparation of isolated rat liver parenchymal cells. J Cell Biol 1969;43:496-520. 17. Petzinger E, Seeger R. Scanning electron microscopic studies on the cytolytic effect of phallolysin on isolated rat hepatocytes and AS-30 D hepatoma cells. Naunyn-Schmiedebergs Arch Pharmacol 1976;295:211-213. 18. Bradford MM. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976; 72:248-254. 19. Klingenberg M, Pfaff E. Means of terminating reactions. Methods Enzymol 1967; 10:680-684. 20. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual. Ed 2. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1989. 21. Petzinger E. Competitive inhibition of the uptake of demethylphalloin by cholic acid in isolated hepatocytes: evidence for a transport competition rather than a binding competition. Naunyn-Schmiedebergs Arch Pharmacol 1981;316:345-349. 22. Adachi Y, Kobayashi H, Kurumi Y, Shouji M, Kitano M, Yamamoto T. ATP-dependent taurocholate transport by rat liver canalicular membrane vesicles. HEPATOLOGY1991; 14:655-659. 23. Akerboom TPM, Narayanaswami V, Kunst M, Sies H. ATP-dependent S-(2,4-dinitrophenyl)glutathionetransport in canalicular plasma membrane vesicles from rat liver. J Biol Chem 1991;266:13147-13152. 24. Mailer M, Ishikawa T, Berger U, Kliinemann C, Lucka L, Schreyer A, Kannich C, et al. ATP-dependent transport of taurocholate across the hepatocyte canalicular membrane mediated by a ll0-kDa glycoprotein binding ATP and bile salts. J Biol Chem 1991;266:18929-18926.

HEPATOLOGYVol. 22, No. 6, 1995 25. Nishida T, Gatmaitan Z, Che M, Arias IM. Rat liver canalicular membrane vesicles contain an ATP-dependent bile acid transport system. Proc Natl Acad Sci USA 1991;88:6590-6594. 26. Stieger B, O'Neill B, Meier PJ. ATP-dependent bile-salts transport in canalicular rat liver plasma-membrane vesicles. Biochem J 1992;284:67-74. 27. Akerboom T, Inoue M, Sies H, Kinne R, Arias M. Biliary transport of glutathione disulfide studied with isolated rat-liver canalicular-membrane vesicles. Eur J Biochem 1984; 141:211-215. 28. Kobayashi K, Sogame Y, Hara H, Hayashi K. Mechanism of glutathione S-conjugate transport in canalicular and basolateral rat liver plasma membranes. J Biol Chem 1990;265:7737-7741. 29. Kobayashi K, Komatsu S, Nishi T, Hara H, Hayashi IL ATPdependent transport for glucuronides in canalicular plasma membrane vesicles. Biochem Biophys Res Commun 1991; 176: 622-626. 30. Leier I, Jedlitschky G, Buchholz U, Cole SPC, Deeley RG, Keppler D. The mrp gene encodes an ATP-dependent export pump for leukotriene C4 and structurally related conjugates. J Biol Chem 1994;269:27807-27810. 31. Jedlitschky G, Leier I, Buchholz U, Center M, Keppler D. ATPdependent transport of glutathione S-conjugates by the multidrug-resistance-associated protein. Cancer Res 1994;54:48334836. 32. Miiller M, Meijer C, Zaman GJR, Borst P, Scheper RJ, Mulder NM, de Vries EGE, et al. Overexpression of the gene encoding the multidrug resistance-associated protein results in increased ATP-dependent glutathione S-conjugate transport. Proc Natl Acad Sci USA 1994;91:13033-13037. 33. Cole SPC, Bhardwaj G, Gerlach JH, Mackie JE, Grant CE, Almquist KC, Stewart AJ, et al. Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line. Science 1992;258:1650-1654. 34. Oude-Elferink RPJ, OttenhoffR, Liefting W, De Haan J, Jansen PLM. Hepatobiliary transport of glutathione and glutathione conjugate in rats with hereditary hyperbilirubinemia. J Clin Invest 1989;84:476-483. 35. Oude-Elferink RPJ, OttenhoffR, Liefting WGM, Schoemaker B, Groen AK, Jansen PLM. ATP-dependent efflux of GSSG and GSconjugate from isolated rat hepatocytes. Am J Physiol 1990; 258(Gastrointest Liver Physiol 21):G699-G709. 36. Brown MS, Goldstein JL. Recepter-mediated control of cholesterol metabolism. Science 1976;191:150-154. 37. Endo A. The discovery and development of HMG-CoA reductase inhibitors. J Lipid Res 1992;33:1569-1582.

PETZINGER ET AL

1811

38. Von Dippe P, Levy D. Expression of the bile acid transport system during liver development and in hepatoma cells. J Biol Chem 1990;265:5942-5945. 39. Blumrich M, Zeyen-Blumrich U, Pagels P, Petzinger E. Immortalization of rat hepatocytes by fusion with hepatoma cells. II. Studies on the transport and synthesis of bile acids in hepatocytoma (HPCT) cells. Eur J Cell Biol 1994; 64:339-347. 40. Hagenbuch B, Jacquemin E, Meier PJ. Na+-dependent and Na ÷independent bile acid uptake systems in the liver. Cell Physiol Biochem 1994;4:198-205. 41. Hagenbuch B, Stieger B, Foquet M, Liibbert H, Meier PJ. Functional expression cloning and characterization of the hepatocyte Na+/bile acid cotransport system. Proc Natl Acad Sci U S A 1991;88:10629-10633. 42. Jacquemin E, Hagenbuch B, Stieger B, Wolkoff AW, Meier PJ. Expression cloning of a rat liver Na+-independent organic anion transporter. Proc Natl Acad Sci USA 1994;91:133-137. 43. Frimmer M, Petzinger E, Rufeger H, Veil LB. The role of bile acids in phalloidin poisoning. Naunyn-Schmiedeberg's Arch Pharmacol 1977;301:145-147. 44. Petzinger E, Frimmer M. Comparative studies on the uptake of 14C-bile acids and 3H-Demethylphalloin in isolated rat liver cells. Arch Toxicol 1980;44:127-135. 45. Frimmer M. What we have learned from phalloidin. Toxicol Lett 1987;35:169-182. 46. Frimmer M, Petzinger E. Mechanism of phalloidin intoxication. I. Cell membrane alterations. In: Poppe H, Bianchi L, Reutter W, eds. Membrane alterations as basis of liver injury. Lancaster Ltd, England: MTP Press, 1976:293-299. 47. Alves C, yon Dippe P, Amoui M, Levy D. Bile acid transport into hepatecyte smooth endoplasmic reticulum vesicles is mediated by microsomal epoxide hydrolase, a membrane protein exhibiting two distinct topological orientations. J Biol Chem 1993;268: 20148-20155. 48. Von Dippe P, Amoni M, Alves C, Levy D. Na÷-dependent bile acid transport by hepatocytes is mediated by a protein similar to microsomal epoxide hydrolase. Am J Physiol 1993;264:G528G534. 49. Kullak-Ublick G-A, Hagenbuch B, Stieger B, WolkoffAW, Meier PJ. Functional chracterization of the basolateral rat liver organic anion transporting polypeptide. HEPATOLOGY1994;20:411-416. 50. Honscha W, Schulz K, Miiller D, Petzinger E. Two different mRNA's from rat liver code for the transport ofbumetanide and taurocholate in Xenopns laevis oocytes. Eur J Pharmacol Mol Pharmacol Section 1993;246:227-232.