ology 214 (2006) 199–208www.elsevier.com/locate/ytaap

Toxicology and Applied Pharmac

3α-6α-Dihydroxy-7α-fluoro-5β-cholanoate (UPF-680), physicochemicaland physiological properties of a new fluorinated bile acid that prevents

17α-ethynyl-estradiol-induced cholestasis in rats

Carlo Clerici a,⁎, Danilo Castellani a, Stefania Asciutti a, Roberto Pellicciari b,Kenneth D.R. Setchell c,d, Nancy C. O'Connell c,d, Bahman Sadeghpour b,Emidio Camaioni b, Stefano Fiorucci a, Barbara Renga a, Elisabetta Nardi a,Giuseppe Sabatino a, Mattia Clementi a, Vittorio Giuliano a, Monia Baldoni a,Stefano Orlandi a, Alessandro Mazzocchi a, Antonio Morelli a, Olivia Morelli a

a Clinica di Gastroenterologia ed Epatologia, Dipartimento di Medicina Clinica e Sperimentale Università degli Studi di Perugia, 06122 Perugia, Italyb Istituto di Chimica e Tecnologia del Farmaco, Università degli Studi di Perugia, via del Liceo, 06123 Perugia, Italy

c Division of Pathology, Clinical Mass Spectrometry, Cincinnati Children's Hospital Medical Center, Cincinnati, OH 45229, USAd Department of Pediatrics, University of Cincinnati, College of Medicine, Cincinnati, OH 45229, USA

Received 9 November 2005; revised 13 December 2005; accepted 14 December 2005Available online 17 February 2006

Abstract

3α-6α-Dihydroxy-7α-fluoro-5β-cholanoate (UPF-680), the 7α-fluorine analog of hyodeoxycholic acid (HDCA), was synthesized to improvebioavailability and stability of ursodeoxycholic acid (UDCA). Acute rat biliary fistula and chronic cholestasis induced by 17α-ethynyl-estradiol(17EE) models were used to study and compare the effects of UPF-680 (dose range 0.6–6.0 μmol/kg min) with UDCA on bile flow, biliarybicarbonate (HCO3

−), lipid output, biliary bile acid composition, hepatic enzymes and organic anion pumps. In acute infusion, UPF-680 increasedbile flow in a dose-related manner, by up to 40.9%. Biliary HCO3

− output was similarly increased. Changes were observed in phospholipidsecretion only at the highest doses. Treatment with UDCA and UPF-680 reversed chronic cholestasis induced by 17EE; in this model, UDCA hadno effect on bile flow in contrast to UPF-680, which significantly increased bile flow. With acute administration of UPF-680, the biliary bile acidpool became enriched with unconjugated and conjugated UPF-680 (71.7%) at the expense of endogenous cholic acid and muricholic isomers.With chronic administration of UPF-680 or UDCA, the main biliary bile acids were tauro conjugates, but modification of biliary bile acid pool wasgreater with UPF-680. UPF-680 increased the mRNA for cytochrome P450 7A1 (CYP7A1) and cytochrome P450 8B (CYP8B). Both UDCA andUPF-680 increased the mRNA for Na+ taurocholate co-transporting polypeptide (NCTP). In conclusion, UPF-680 prevented 17EE-inducedcholestasis and enriched the biliary bile acid pool with less detergent and cytotoxic bile acids. This novel fluorinated bile acid may have potentialin the treatment of cholestatic liver disease.© 2005 Elsevier Inc. All rights reserved.

Keywords: Cholestasis; UPF-680; Bile acids; FXR; UDCA

Introduction

The clinical utility of bile acid therapy for a number ofcommon hepatic and gastrointestinal conditions is well estab-lished (Poupon et al., 2003; Paumgartner and Beuers, 2004).Chenodeoxycholic acid (CDCA) was the first bile acid to be

⁎ Corresponding author. Fax: +39 075 578 3687.E-mail address: clerici@unipg.it (C. Clerici).

0041-008X/$ - see front matter © 2005 Elsevier Inc. All rights reserved.doi:10.1016/j.taap.2005.12.017

commercially developed for gallstone dissolution (Danzinger etal., 1972). Due to the intrinsic hepatotoxicity of this ratherhydrophobic molecule, its clinical use was superceded afterintroduction of the highly hydrophilic bile acid, UDCA (Ward etal., 1984). Clinical applications of UDCA have since beenextended beyond conditions of cholelithiasis (Bachrach andHofmann, 1982), and this bile acid is now universally used forthe treatment of a variety of cholestatic liver diseases (Riely andBacq, 2004; Kaplan et al., 2004; Paumgartner and Beuers, 2004).

Fig. 1. Structure of 3α,6α-dihydroxy-7α-fluoro-5β-cholanoic acid (UPF-680).

200 C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

While the exact mechanism of action of UDCA therapy inimproving biochemical and histological indices of liver diseaseis not fully established, alterations in the composition of thebiliary bile acid pool are considered to play a significant role.Following oral administration, UDCA is absorbed passivelyalong the small intestine and transported to the liver where itundergoes first-pass metabolism by conjugation with the aminoacids, glycine and taurine, before being secreted into bile(Hofmann, 1989). The biliary pool consequently becomes morehydrophilic, and associated improvements in liver diseasemarkers are commonly observed (Hofmann, 1994; Bellentaniet al., 1989). The beneficial effects of UDCA are not limited toits physicochemical properties but may also be explained byseveral other actions, including its choleretic activity, its effectsat the intestinal level in inhibiting cholesterol and bile acidabsorption and its immunomodulatory effects (Van Milligen deWit et al., 1999). Furthermore, UDCA binds weakly to farnesoidX receptor (FXR) which contrasts primary bile acids that inhibitcholesterol 7α-hydrolase activity, the rate-limiting enzyme inbile acid synthesis (Makishima et al., 1999; Parks et al., 1999).

Despite the clinical effectiveness of UDCA, its overallefficacy may be limited by its relatively poor solubility,absorption and hepatic extraction, thus limiting its biliaryenrichment and by its extensive intestinal bacterial biotransfor-mation to CDCA and lithocholic acid (LCA), bile acids with alow critical micellar concentration (CMC) and high detergencyand hepatotoxicity. The search for alternative novel bile acidsthat exhibit the selective benefits of UDCA but without thesedrawbacks continues. Two patterns of structural modificationare possible in bile acids, and these target either the side-chainor the steroid nucleus.

We have synthesized a novel bile acid, namely, 3α,6α-dihydroxy-7α-fluoro-5β-cholanoate (UPF-680), in which thesteroid nucleus has been modified by the incorporation of afluorine atom at carbon C-7. The fluorine atom is as small as thehydrogen atom and has a much stronger electronic effect. It canonly accept, and cannot donate electrons, and is thereforeendowed with a unique hydrophilic/hydrophobic balancecreating different physicochemical properties to that of ahydroxyl group. The properties of a fluorinated derivative ofUDCA have been previously described (Roda et al., 1995), butwe now describe for the first time the physicochemical,physiological and metabolic behavior of the fluorinatedderivative of HDCA (UPF-680) in the rat model of acute andchronic cholestasis induced by 17α-ethynyl-estradiol (17EE).

Methods

Description of the new bile acid

Chemistry. All chemicals were purchased from Sigma-Aldrich (St. Louis,MO). UPF-680 (Fig. 1) was prepared using HDCA as a starting material.HDCA was converted in two steps to methyl 3α-hydroxy-6-oxo-5β-cholanoate according to established procedure (Iida et al., 1988). Subsequentconversion of the 6-oxo group, by using lithium diisopropyl amide andtrimethylsylil chloride, into the corresponding silyl enolether followed byelectrophilic fluorination with Selectfluor resulted in the correspondingmethyl 3α-hydroxy-7α-fluoro-6-keto-5β-cholanoate. Reduction of the lattercompound with sodium borohydride afforded the two isomers methyl 3α,6α-

dihydroxy-7α-fluoro-5β-cholanoate and methyl 3α,6β-dihydroxy-7α-fluoro-5β-cholanoate, respectively (63/37 ratio). Hydrolysis with alkali gave thecorresponding desired UPF-680. The structures of both compounds wereconfirmed by 1H NMR and 13C NMR with a Bruker AC 200 MHzspectrometer. All derivatives displayed a high pressure liquid chromatographypurity Nx95%, as detected with a Shimadzu LC-10 workstation (Shimadzu,Kyoto, Japan) equipped with an SPD-10A UV–Vis detector and using aLichrospher 100 RP-18 (250 × 5 mm, 5 μm) as a column. The flow rate was1 ml/min, the detection was carried out at 205 and 220 nm, and the mobilephase was a mixture phosphate buffer (3 mM Na2HPO4, pH 3.8)/acetonitrile70/30, v/v) or a mixture of H2O/acetonitrile + 0.1% TFA 40/60.

Evaluation of critical micellar concentration (CMC). The CMC of UPF-680, UDCA, CA and LCA was measured in water by ORANGE OT method(Roda et al., 1983).

Acute experiments in biliary fistula rat. Male Sprague–Dawley rats (225–300 g), housed in the University of Perugia Animal Facility, weremaintained on a standard diet with water ad libitum and housed in atemperature (21 °C–23 °C)- and humidity (45–50%)-controlled room undera constant 12-h light, 12-h dark cycle. All animals received humane careaccording to the criteria outlined in the Guide for the Care and Use ofLaboratory Animals prepared by the National Academy of Sciences andpublished by the NIH (publication 86-23, revised 1985).

The rats (fasted 24 h) were anesthetized with ketamine (50 mg/kg) andphenobarbital (100 mg/kg) (ip). Body temperature was maintained at constant37.0 °C with a temperature controller lamp (Yellow Springs Instrument Co.,Yellow Springs, Ohio) to prevent hypothermic alterations of bile flow. Aftercatheterization of the right jugular vein, the abdomen was opened through amidline incision and the common bile duct was isolated and cannulated (PE-10,Intramedic, Clay Adams).

UPF-680 was dissolved in saline solution with 2% bovine serum albumin,pH 7.4. Fifteen rats were divided in 3 groups (I, II and III) of 5 animals each.Saline solution was infused via the right external jugular vein by a Harvardmicroliter syringe pump for 60 min. Using the same jugular vein, UPF-680 wasinfused at a different dose (Group I: 0.6 μmol/kg min, Group II: 2.6 μmol/kgmin, Group III: 6.0 μmol/kg min) for 90 min. UPF-680 was then replaced againwith saline solution for a further 60 min. Bile samples were collected by theexternal biliary fistula every 15 min in order to determine the bile flow, HCO3

−

secretion, biliary phospholipids and cholesterol secretion.A microgasmetric procedure was adopted to evaluate biliary HCO3

−

concentration (Micro CO2 system, Harleco, American Hospital Corporation,Gibbstorm, NJ).

The pooled bile samples collected in the Group III were analyzed forbile acid composition by fast atom bombardment ionization massspectrometry (FAB-MS) and, in one animal of Group III, by gaschromatography-mass spectrometry (GC-MS). Bile acids were extractedfrom the pooled bile samples (80–120 μl volumes), after the addition of aninternal standard nordeoxycholic acid (1–10 μg), by liquid–solid extractionand fractionated into distinct groups of unconjugated bile acids andconjugated bile acids using lipophilic anion exchange chromatography onLipidex-DEAP (Setchell and Worthington, 1982; Rodrigues and Setchell,1996). The conjugated bile acids were solvolyzed and enzymaticallyhydrolyzed prior to conversion to the methyl ester-trimethylysilyl etherderivatives for analysis by GC-MS. Bile acids were identified from theirretention indices and mass spectra as compared to authentic standards and

201C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

quantified by GC-MS from peak height responses relative to the peak heightresponse for the added internal standard and assuming a unity responsefactor for each bile acid.

Biliary phospholipid concentration was measured by an enzymatic methodbased on choline oxidase (Wako Pure Chemical industries, MI, Italy).

Biliary cholesterol concentration was determined by an enzymatic procedurebased on cholesterol oxidase (Boehringer Mannheim, MI, Italy).

Chronic cholestasis in the rats. Twenty-eight male Wistar rats (225–300 g)housed in the University of Perugia animal facility were maintained on astandard diet with water ad libitum for 7 days and housed in a temperature (21°C–23 °C)- and humidity (45–50%)-controlled room under a constant 12-h light, 12-h dark cycle. Protocols were approved by the animal researchcommittee. The cholestasis was induced by an ip injection of 17EE at a dosageof 5 mg/kg administered for 7 days.

Group I (N = 7): without treatment (control group); Group II (N = 7): 17EEalone (5 mg/kg ip) for 7 days; Group III (N = 7): 17EE (5 mg/kg ip) plus UDCAat a dosage of 30 mg/kg ip for 7 days; Group IV (N = 7): 17EE (5 mg/kg ip) plusUPF-680 at the equivalent dosage of 30 mg/kg of UDCA ip for 7 days.

On the last day after anesthesia with ketamine (50 mg/kg) andphenobarbital (100 mg/kg) ip, a midline incision was preformed on the rats'abdomen and the common bile duct was isolated and cannulated (PE-10,Intramedic, Clay Adams, Parsippany, NJ). The bile samples were collectedby the external biliary fistula every 15 min for 90 min and then weighed inorder to determine the bile flow.

The bile acid composition was analyzed by high pressure liquidchromatography and GC-MS (Setchell and Worthington, 1982).

At the end of experiment in 16 of 28 rats, the liver was removed in order toquantify, by PCR, the mRNA relative abundance of the principal enzymes andpumps involved in synthesis and transport of bile acids, namely, CYP7A1,CYP8B, BSEP, NTCP and SHP.

Total RNA was isolated from the rat liver by using the TRIzol reagentaccording to the specifications of the manufacturer (Invitrogen, Carlsbad,CA). Total RNA was processed directly to cDNA by reverse transcriptionwith Superscript III (Invitrogen, Carlsbad, CA). Two micrograms of RNAwas added to mixture which contains DnaseI reaction buffer 10× and 1 UDNaseI. The mix was incubated 15 min at room temperature then 1 μlEDTA 25 mM solution was added and the mixture was incubated at 95 °Cfor 5 min. Then, 4 μl of first strand buffer 5× (250 mM Tris–HClpH = 8,3; 375 mM KCl; 15 mM MgCl2), 2 μl of DTT 0.1 M, 1 μl ofdNTP's mix 10 mM, 0.5 μl of random primers 300 ng/μl, 1 μl of RNaseout and 1 μl of Super Script III were added. The mixture was incubated atroom temperature for 10 min and at 42 °C for 50 min and then heated at95 °C for 5 min to inactivate the enzyme and cooled at 4 °C. All PCRprimers were designed using software PRIMER3-NEW using publishedsequence data from the NCBI database. Primers were synthesized by MWGBIOTECH. For rat β-ACTIN, the sense primer was 5′-TAC CAC TGGCAT TGT GAT GG-3′ and antisense 5′-TTA ATG TCA CGC ACG ATTTC-3′; for rat NTCP, the sense primer was: 5′-GCA TGA TGC CAC TCCTCT TAT AC-3′ and the antisense 5′-TAC ATA GTG TGG CCT TTTGGA CT-3′; for rat BSEP, the sense primer was: 5′-AAG GCA AGA ACTCGA GAT ACC AG-3′ and the antisense 5′-TTT CAC TTT CAA TGTCCA CCA AC-3′; for rat SHP, the sense primer was: 5′-CCT GGA GCAGCC CTC GT-3′ and the antisense 5′-AAC ACT GTA TGC AAA CCGAGG-3′; for rat CYP7A1, the sense primer was: 5′-CTG CAG CGA GCTTTA TCC AC-3′ and the antisense 5′-CCT GGG TTG CTA AGG GACTC-3′; and for rat CYP8B, the sense primer was: 5′-CCC CTA TCT CTCAGT ACA CAT GG-3′ and the antisense 5′-GAC CAT AAG GAG GACAAA GGT CT-3′. In control experiments with 3 replicates, no falsepositives were detected. Amplification reactions contained 2 μl cDNA, 12.5μl of the 2× Quanti tect SYBR Green RT-PCR Master Mix (Qiagen,Milano, Italy) and 0.75 μl of each of the specific primers. Primerconcentrations in the final volume of 25 μl were 300 nM. All reactionswere performed in triplicate in an iCycler iQ system (Biorad, Hercules,CA), and the thermal cycling conditions were: 2 min at 95 °C followed by50 cycles of 95 °C for 10 s and 60 °C for 30 s. We used the expression ofβ-ACTIN to normalize the expression data of FXR target genes. β-ACTINwas used to correct for differences in the amount of total RNA added to a

reaction and to compensate for different levels of inhibition during reversetranscription of RNA and during PCR.

Statistical analysis

Data reported are the mean ± SEM of the number of experiments indicated.The Student's t test for continuous variables was performed for paired data.

Results

Critical micellar concentration (CMC)

The CMC of UPF-680 was 25 mM, H2O. By comparison,the CMC values for UDCA, CA and LCAwere 19.0, 13.0 and2.5 mM, H2O, respectively.

Effects on bile flowThe average bile flow at baseline (the pre-infusion period)was

102.95 μl/kg min ± 1.7 SEM. A dose-dependent increase in bileflow was observed with UPF-680 infusion. Mean (±SEM) bileflow increased to 113.5 ± 4.3 μl/kg min, 132.5 ± 1.8 μl/kg minand 145.1 ± 7.0μl/kgmin, respectively, with doses of 0.6, 2.6 and6.0 μmol/kg min of UPF-680. These represented statisticallysignificant increments of 10.2%, 28.7% and 40.9% over basalbile flow. When the UPF-680 infusion was interrupted, the bileflow returned to baseline values after 30 min (Fig. 2A).

Effects on HCO3− secretion

The increase in bile flow observed with UPF-680 wasaccompanied by an associated increase in biliary HCO3

− output.The average HCO3

− output in the pre-infusion period was 2.62μmol/kg min ± 0.13 SEM. In the group of animals in whichUPF-680 was administered at a dose of 0.6 μmol/kg min, HCO3

−

output increased to 3.01 μmol/kg min ± 0.06 SEM, representingan increment of 14.9% (P = 0.01) as compared to the pre-infusion period. With doses of 2.6 and 6.0 μmol/kg min, HCO3

−

output increased to 4.91 ± 0.18 and 7.03 ± 0.69 μmol/kg min,respectively, which accounted for an increase of 87.4% andN100% (both values significant P b 0.003) compared to the pre-infusion period. At all doses when the UPF-680 infusion wasinterrupted, the HCO3

− output returned to baseline values after30 min (Fig. 2B).

Effects on lipid outputDuring infusion of UPF-680, there was no change in biliary

cholesterol secretion at any doses administered (Fig. 2C).The average biliary phospholipid output in the pre-infusion

period was 0.081 μmol/kg min ± 0.012 SEM. During infusionof UPF-680 at a dose of 0.6 μmol/kg min, there was nosignificant increased of biliary phospholipid secretion, whilewith doses of 2.6 and 6.0 μmol/kg min, biliary phospholipidsecretion increased respectively to 0.11 ± 0.003 and0.27 ± 0.062 μmol/kg min ± SEM. Dose-dependent increasesin biliary phospholipids were observed with UPF-680 infusionwith an increment of 35.8% (P b 0.05) and N100% (P b 0.05)over baseline when UPF-680 was administrated respectively atdoses of 2.6 μmol/kg min and 6.0 μmol/kg min (Fig. 2D).

Fig. 2. Effect of acute intraperitoneal administration of UPF-680 in different doses in rats on bile flow (A), on HCO3− output (B), on cholesterol output (C) and on

phospholipids output (D). Fifteen male Sprague–Dawley rats (3 groups of 5 animals each) were infused via the right external jugular vein for 60 min with salinesolution and then treated with UPF-680 at a different dose (Group I: 0.6 μmol/kg min, Group II: 2.6 μmol/kg min, Group III: 6.0 μmol/kg min) for 90 min. UPF-680was then replaced again with saline solution for a further 60 min. Bile samples were collected by the external biliary fistula every 15 min. Bile flow showed asignificant increase over basal bile flow at the three administrated doses (°P b 0.05 at 0.6 μmol/kg min, xP b 0.05 at 2.6 μmol/kg min, *P b 0.05 at 6.0 μmol/kg min).HCO3

− output showed a significant increase compared to the pre-infusion period at the three administrated doses (°P = 0.01 at 0.6 μmol/kg min, xP b 0.003 at 2.6 μmol/kg min, *P b 0.003 at 6.0 μmol/kg min). Biliary cholesterol secretion was unchanged. Biliary phospholipids output showed a significant increase compared to the pre-infusion period at the following doses: 2.6 μmol/kg min (xP b 0.05), 6.0 μmol/kg min (*P b 0.05).

202 C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

Hepatic metabolism of UPF-680 and effects on biliary bileacid composition

Negative ion fast atom bombardment ionization massspectrometry analysis of the bile samples collected in thegroup C during infusion of each bile acid yielded specific ionsin the mass spectra corresponding in molecular weight to the(M-H)-ions of the parent molecule UPF-680 and its taurineconjugate (data not shown). The biliary pool became enrichedin this bile acid at the expense of the endogenous bile acids,CA and the muricholic acid isomers. Detailed GC-MS analysisof the bile from one of the animals confirmed UPF-680 as themajor peak in both the unconjugated and conjugated bile acidfractions (Fig. 3), and, quantitatively, the total bile acidconcentration increased in this animal from 4502 μmol/l atbaseline to 7003 μmol/l during infusion of UPF-680. Theincrease in bile acid secretion in bile was sustained even aftertermination of UPF-680 infusion, suggesting that the fluori-nated bile acids were retained within the hepatocyte andcontinued to drive bile flow.

The biliary bile acid composition (Fig. 3) was dramaticallyaltered by the administration of UPF-680. Conjugated andunconjugated UPF-680 accounted for 71.7% of the total bileacids during intravenous infusion with the highest dose of UPF-680, and the proportion of CA and the muricholate isomers,

which at baseline accounted for most of the biliary bile acids,was significantly reduced (Table 1).

Effect of UPF-680 in chronic cholestasis induced by 17EECholestasis induced by 17EE was completely prevented by

UPF-680 and also by UDCA (Fig. 4). The average (±SEM) bileflow of the control group was 70.5 ± 3.5 μl/kg min, while forthose animals treated with 17EE average, bile flow was45.8 ± 4.6 μl/kg min, reflecting a reduction in bile flow of35.0% (P b 0.05). Average bile flow in the animals treated with17EE and UDCAwas 78.9 ± 3.6 μl/kg min SEM, which was notstatistically significant when compared with baseline values.The group of animals treated with 17EE and UPF-680 had anaverage bile flow of 92.8 ± 1.0 μl/kg min. This represented a31.6% increase (P b 0.05) when compared with the controlgroup and a 17.6% increase (P b 0.05) compared with animaltreated with 17EE and UDCA.

HPLC and GC-MS analysis of bile confirmed absorption,hepatic extraction and canalicular secretion of the UPF-680. Thebile acid composition as determined by HPLC and GC-MS ofthe individual groups of animals is shown in Table 2.

PCR analysis of the hepatic parenchyma established theeffects of UPF-680 on mRNA levels of the principal enzymesand pumps involved in bile acid synthesis and transport.

Fig. 3. Gas chromatography-mass spectrometry (GC-MS) profiles of the methyl ester-trimethylsilyl ether derivatives of bile acids isolated in the unconjugated andconjugated fractions of Sprague–Dawley rat bile collected at baseline and during infusion with 0.6 μmol/kg min of UPF-680. The major bile acids identified by massspectrometry are indicated, and the quantitative concentrations detailed in Table 1.

203C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

CYP7A1 mRNA level increased after administration of bothUDCA (P b 0.05) and UPF-680 (P b 0.05), although the latterbile acid stimulated expression more than UDCA in this 17EE-induced cholestasis model (Fig. 5).

CYP8B mRNA level was not significantly reduced during17EE-induced cholestasis, while UDCA (P b 0.05) and UPF-680 (P b 0.05) both stimulated mRNA for CYP8B synthesis in17EE-induced cholestasis (Fig. 5).

The bile salt transporter mRNA level of BSEP was notsignificantly changed by any of the treatments (Fig. 6), while

Table 1GC-MS quantitative and qualitative evaluations of bile acids composition at baselin

Biliary bile acid Baseline (μmol/l)

Conjugated Unconjugated Total (%

Deoxycholic 0.501 0.002 0.503 (Chenodeoxycholic 0.092 – 0.092 (α-Muricholic 2.011 0.011 2.022 (Cholic 0.486 0.003 0.489 (UPF-680 – – –β-Muricholic 0.364 0.004 0.368 (Δ22-β-muricholic 0.097 0.001 0.098 (ω-Muricholic 0.114 0.004 0.118 (Other bile acids 0.811 0.001 0.812 (Total 4.476 0.026 4.502 (

Bile samples were collected by the external biliary fistula.

NTCPmRNA level was increased by both UDCA (P b 0.05) andUPF-680 (P b 0.05) during 17EE-induced cholestasis (Fig. 6).Neither UDCAnorUPF-680 had any effect on themRNA for theexpression of SHP, even though SHPmRNA level was increasedafter administration of 17EE (P b 0.05) (Fig. 6).

Discussion

We describe for the first time the metabolism and hepaticeffects of a novel bile acid that is a 7α-fluoro derivative of

e and during infusion UPF-680 (6.0 μmol/kg min) in a Sprague–Dawley rat

During UPF-680 infusion (μmol/l)

) Conjugated Unconjugated Total (%)

11.2%) 0.237 0.001 0.238 (3.4%)2.0%) 0.070 0.005 0.075 (1.1%)44.9%) 0.941 0.020 0.961 (13.7%)10.9%) 0.222 0.010 0.231 (3.3%)

4.110 0.910 5.020 (71.7%)8.2%) 0.177 0.020 0.197 (2.8%)2.2%) 0.167 0.001 0.168 (2.4%)2.6%) 0.041 0.015 0.056 (0.8%)18.0%) 0.015 0.041 0.056 (0.8%)100%) 5.980 1.023 7.003 (100%)

Fig. 4. Effect of chronic administration of UPF-680 and UDCA on cholestasisinduced by 17EE. 28 male Wistar rats (4 groups of 7 animals each) were treateddaily for 7 days with either 17EE (5 mg/kg ip), 17EE + UDCA (30 mg/kg ip),17EE + UPF-680 (at the equivalent dosage of 30 mg/kg of UDCA ip) or notreatment (control). Bile flow was determined collecting bile samples by theexternal biliary fistula every 15 min for 90 min. Bile flow (average ± SEM) wassignificantly reduced (P b 0.05) in 17EE-treated animals (45.8 ± 4.6 μl/kg min)versus control group (70.5 ± 3.5 μl/kg min); in the animals treated with 17EEand UDCA (78.9 ± 3.6 μl/kg min), bile flow was not statistically significantwhen compared with control group; in the animals treated with 17EE and UPF-680 (92.8 ± 1.0 μl/kg min), bile flow showed a significant increase (P b 0.05)when compared with the control and UDCA groups.

204 C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

hyodoeoxycholic acid (UPF-680). The introduction of afluorine atom in the 7α-position markedly decreases thedetergency of the bile acid molecule relative to the propertiesof other bile acids such as UDCA, CA and LCA. This isreflected by differences in the measured CMC of UPF-680 andis responsible for the low toxicity of UPF-680 on cell

Table 2High pressure liquid chromatography (HPLC) and GC-MS quantitative and qualitativUPF-680 and UDCA on cholestasis induced by 17EE in male Wistar rats

Total concentration Group I Group II

CTRLμmol/l ± SD

CTR % 17EEμmol/l ± SD

17EE

UPF680 Tauro 0 0 0 0Free 0 0 0 0

UDCA Tauro 0 0 0 0Free 0 0 0 0

murCA Tauro 1461.9 ± 41.6 77.2 1299.8 ± 14.2 85.8Free 0 0 0 0

CA Tauro 253.2 ± 56.0 13.4 143.1 ± 1.9 9.4Free 78.2 ± 23.7 4.1 28.6 ± 7.2 1.9

CDCA Tauro 60.3 ± 14.4 3.2 27.0 ± 6.9 1.8Free 0 0 1.7 ± 0.8 0.1

DCA Tauro 40.9 ± 11.7 2.2 12.7 ± 2.5 0.8Free 0 0 1.7 ± 2.6 0.1

LCA Tauro 0 0 0 0Free 0 0 0 0

Total 1894.5 ± 147.4 1514.6 ± 36.1 2587

Bile was collected from rats by external biliary fistula.Group I (control,N = 7); Group II (17EE 5 mg/kg ip, N = 7); Group III (17EE + UDCAmg/kg of UDCA ip, N = 7).

membranes. Insertion of a fluorine atom also renders theanalogue metabolically stable, especially with respect tointestinal bacterial degradation.

Acute administration of UPF-680 induced dose-dependenteffects on bile flow and HCO3

− output, and moreover UPF-680was more effective than UDCA in generating a choleresis(Clerici et al., 1996; Piazza et al., 2000). This increasedcholeresis is likely to reduce the hepatotoxic effect caused bybile acid retention during cholestasis. The choleretic propertyof UPF-680 in the acute experimental model of cholestasisused in our studies may be explained by a greater proportionof unchanged UPF-680 being secreted in the bile ducts,particularly at the highest dosage (6.0 μmol/kg min). UPF-680significantly increased HCO3

− secretion possibly as a result ofthe cholehepatic shunt pathway (Hofmann, 1984) or throughhepatocyte HCO3

− pump stimulation (Erlinger, 1996). Bycontrast, UPF-680 had no effect on biliary lipid secretion,except for increasing biliary phospholipids secretion at thehighest doses administered.

The metabolism of UPF-680 revealed that, with acuteadministration, its unconjugated and conjugated forms weresecreted in bile at the expense of a reduction in the proportions ofthe primary bile acids, CA and β-muricholic acids, accounting forthe 71.7% of the total biliary bile acid pool. On the other hand, withchronic administration of UPF-680, this novel bile acid appeared inbile mainly in the conjugated form, and, in this regard, itsmetabolism was similar to that of UDCA, but modification ofbiliary bile acid pool was greater with UPF-680. UPF-680 enrichedthe biliary bile acid pool of 67.3%as compared to 55.6% induced byUDCA. A further difference in the metabolism of UPF-680 whencompared with UDCAwas the apparent lack of formation of LCA,indicating that the presence of the fluorine atom completelyprevented intestinal bacterial biotransformation, yet it did not limit

e evaluations of bile acids composition after chronic administration for 7 days of

Group III Group IV

% UDCA +17EEμmol/l ± SD

UDCA % UPF-680 +17EE μmol/l ± SD

UPF-680

0 0 2050.0 ± 130.0 67.20 0 3.3 ± 1.2 0.11435.5 ± 57.2 55.5 0 01.6 ± 0.4 0.1 0 0952.3 ± 60.1 36.8 810.7 ± 51.0 26.60 0 0 0116.8 ± 27.0 4.5 109.8 ± 19.2 3.635.8 ± 5.7 1.4 35.7 ± 16.2 1.222.7 ± 6.5 0.9 22.6 ± 7.8 0.71.3 ± 0.3 0.05 0.9 ± 0.3 0.0316.3 ± 1.2 0.6 17.3 ± 12.5 0.61.0 ± 0.2 0.04 0 04.0 ± 1.8 0.2 0 00 0 0 0

.3 ± 160.4 3050.3 ± 238.2

30 mg/kg ip, N = 7); Group IV (17EE + UPF-680 at the equivalent dosage of 30

Fig. 5. Effects of chronic administration of UPF-680 and UDCA on cholestasis induced by 17EE for 7 days on cytochrome P450 7A1 (CYP7A1) and cytochromeP450 8B (CYP8B) mRNA level in liver from 16 male Wistar rats (4 groups of 4 animals each). The mRNA levels were determined by PCR, using β-ACTIN asinternal control. Values are expressed as mean ± SEM of 4 rats per group, each assay was carried out in triplicate. Group I: control; Group II: 17EE (5 mg/kg ip);Group III: 17EE + UDCA (30 mg/kg ip); Group IV: 17EE + UPF-680 (at the equivalent dosage of 30 mg/kg of UDCA ip). CYP7A1 mRNA level decreased afteradministration of 17EE compared to control (*P b 0.05) while increased after administration of both UDCA and UPF-680 compared to control (*P b 0.05) and to17EE-induced cholestasis group (**P b 0.05). CYP8B mRNA level increased after administration of UDCA compared both to control (*P b 0.05) and to 17EE-induced cholestasis group (**P b 0.05), while administration of UPF-680 increased CYP8B mRNA level only compared to 17EE-induced cholestasis group(**P b 0.05).

205C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

the ability of the liver to conjugate the molecule in the side-chainwith taurine and glycine.

FXR, an endogenous sensor for bile acids, regulates bileacid synthesis and excretion (Parks et al., 1999). FXRinhibits bile acid synthesis by inducing the expression ofSHP, which represses the transcription of CYP7A1. FXR alsoinhibits the expression of NTCP and activates BSEP (Chawlaet al., 2001). Estrogens could induce directly the expressionof SHP (Lai et al., 2003) or indirectly by activation of FXRby the endogenous bile acids that accumulate in thehepatocyte in cholestasis (Parks et al., 1999). Cholestasisresults in systemic and intrahepatic retention of potentiallyhydrophobic and toxic bile acids that exacerbate liver injury,ultimately leading to biliary fibrosis and cirrhosis. Down-regulation of hepatocellular transport systems contribute toimpaired hepatobiliary excretion of bile acids and otherbiliary constituents during cholestasis. Estrogens reduce bileflow by modulating a number of hepatocellular transporters(Mottino et al., 2002; Huang et al., 2000; Stieger et al., 2000;Bossard et al., 1993; Simon et al., 1996). Thus, cholestasisinduced by administration of estrogens to rodents has been

linked to a reduction of activity and/or expression of theATP-BSEP (Stieger et al., 2000; Bossard et al., 1993). ATP-dependent BSEP is important for transport of monovalentanions into bile and stimulates bile-salt-dependent bile flow.It has been shown previously that estrogen-induced chole-stasis is associated with a reduced activity of the NTCP, theprimary carrier for sodium-dependent conjugated bile saltsuptake from portal blood (Bossard et al., 1993; Simon et al.,1996), and a suppression of endogenous bile acid synthesis(Simon et al., 1996), factors that further contribute to thereduction of bile-acid-dependent bile flow observed in the rat.The most predictable mechanism responsible for primaryimpairment of bile flow in estrogen-treated rats (Mottino etal., 2002; Huang et al., 2000; Koopen et al., 1999; Accatinoet al., 1998; Fickert et al., 2001; Crocenzi et al., 2001;Stieger et al., 1994) may be activation of FXR, which, in itsturn, regulates a chain of genes involved in the feedbackinhibition of bile acid synthesis and transport (Kast et al.,2002; Goodwin et al., 2000; Chiang, 2002). For this reason,we examined the influence of UPF-680 on this orphanreceptor.

Fig. 6. Effects of chronic administration of UPF-680 and UDCA on cholestasis induced by 17EE for 7 days on BSEP, NTCP and SHP mRNA level in liver from 16male Wistar rats (4 groups of 4 animals each). The mRNA levels were determined by PCR, using β-ACTIN as internal control. Values are expressed as mean ± SEM of4 rats per group, each assay was carried out in triplicate. Group I: control; Group II: 17EE (5 mg/kg ip); Group III: 17EE + UDCA (30 mg/kg ip); Group IV:17EE + UPF-680 (at the equivalent dosage of 30 mg/kg of UDCA ip). BSEP mRNA level was not significantly changed in any group. NTCP mRNA level wasdecreased by 17EE administration compared to control (*P b 0.05) and increased after administration of both UDCA and UPF-680 compared to control (*P b 0.05)and to 17EE-induced cholestasis group (**P b 0.05). SHP mRNA level increased after administration of 17EE compared to control (*P b 0.05), while any change wasnoticed after administration of both UDCA and UPF-680.

206 C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

FXR is activated by endogenous bile acids that accumulate incholestasis with the rank order of potency CDCA N deoxycholicacid N LCA N CA.

In 17EE cholestasis, an increment in mRNA SHP expressionwas observed that may be due to the accumulation ofendogenous bile acids such as CDCA in the hepatocytes.Cholestasis was completely reversed by the administration ofUDCA and even more effectively by infusion of UPF-680,lower doses of the latter being required for efficacy.

The mRNA for the bile acid synthetic enzymes CYP7A1and CYP8B was also increased, showing a change in thebiliary bile acid pool, to one with less detergent bile acids that

increase the mRNA for CYP7A1 and CYP8B. The greaterincrease of mRNA for CYP7A1 induced by UPF-680 ascompared to UDCA could be responsible for the higherconcentration of less detergent bile acids in the bile acid pool.Moreover, an increased synthesis of bile acids could beresponsible for an enhancement on bile-acid-dependent bileflow thereby decreasing cholestasis. As previously reported,the co-transporter NTCP mRNA level was reduced byadministration of 17EE. This transporter is uniformly down-regulated in all experimental models of cholestasis and incholestatic liver diseases in humans (Lee and Boyer, 2000).Administration of both UDCA and UPF-680 increased NTCP

207C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

mRNA level, showing a possible contribution to reverse17EE-induced cholestasis. BSEP mRNA level is onlymodestly impaired in 17EE cholestasis, and bile saltsexcretion continues, albeit at reduced levels, even in theface of complete bile duct obstruction (Lee and Boyer, 2000).Our experiments are in agreement with others that show that17EE does not change the mRNA for ATP-dependenttransporter BSEP, and it was not significantly changed byany treatment.

In conclusion, substitution of a fluorine atom at position C-7α of HDCA to produce the novel bile acid UPF-680 preventsbiotransformation to more hydrophobic metabolites such asLCA but does not inhibit its hepatic conjugation. UPF-680 hada lower detergency, gave a greater enrichment of the biliary bileacid pool and induced a greater choleresis than UDCA. UPF-680 prevented cholestasis induced by 17EE, and theseproperties warrant consideration as a therapeutic bile acid forthe cholestatic liver diseases.

References

Accatino, L., Pizarro, M., Solis, N., Koenig, C.S., 1998 (Jul.). Effects ofdiosgenin, a plant-derived steroid, on bile secretion and hepatocellularcholestasis induced by estrogens in the rat. Hepatology 28 (1),129–140.

Bachrach, W.H., Hofmann, A.F., 1982. Ursodeoxycholic acid in the treatment ofcholesterol cholelithiasis. Part 1. Dig. Dis. Sci. 27, 737–761.

Bellentani, S., Tabarroni, G., Barchi, T., Ferretti, I., Fratti, N., Villa, E., Manenti,F., 1989. Effect of ursodeoxycholic acid treatment on alanine aminotrans-ferase and gamma-glutamyltranspeptidase serum levels in patients withhypertransaminasemia. Results from a double-blind controlled trial. J.Hepatol. 8 (1), 7–12.

Bossard, R., Stieger, B., O'Neill, B., Fricker, G., Meier, P.J., 1993.Ethinylestradiol treatment induces multiple canalicular membrane transportalterations in rat liver. J. Clin. Invest. 91, 2714–2720.

Chawla, A., Repa, J.J., Evans, R.M., Mangelsdorf, D.J., 2001. Nuclear receptorsand lipid physiology: opening the X-files. Science 294, 1866–1870.

Chiang, J.Y., 2002. Bile acid regulation of gene expression: roles of nuclearhormone receptors. Endocr. Rev. 23, 443–463.

Clerici, C., Gentili, G., Errico, A., Cammilleri, F., Brasacchio, F., Morelli, A.,Pellicciari, R., Costantino, G., Mattoli, L., and Annibali, D., 1996.Regulation of biliary bicarbonate secretion by bile salts. MTP PressLimited-1996-Kluwer Academic Publishers Group. pp. 82–89.

Crocenzi, F.A., Sanchez Pozzi, E.J., Pellegrino, J.M., Favre, C.O.,Rodriguez Garay, E.A., Mottino, A.D., Coleman, R., Roma, M.G.,2001. Beneficial effects of silymarin on estrogen-induced cholestasis inthe rat: a study in vivo and in isolated hepatocyte couplets. Hepatology34 (2), 329–339.

Danzinger, R.G., Hofmann, A.F., Schoenfield, L.J., Thistle, J.L., 1972.Dissolution of cholesterol gallstones by chenodeoxycholic acid. N. Eng. J.Med. 286, 1–8.

Erlinger, S., 1996. Mechanisms of hepatic transport and bile secretion. ActaGastro-Enterol. Belg. 59 (2), 159–162.

Fickert, P., Zollner, G., Fuchsbichler, A., Stumptner, C., Pojer, C., Zenz, R.,Lammert, F., Stieger, B., Meier, P.J., Zatloukal, K., Denk, H., Trauner, M.,2001. Effects of ursodeoxycholic and cholic acid feeding on hepatocel-lular transporter expression in mouse liver. Gastroenterology 121,170–183.

Goodwin, B., Jones, S.A., Price, R.R., Watson, M.A., McKee, D.D., Moore,L.B., Galardi, C., Wilson, J.G., Lewis, M.C., Roth, M.E., 2000. A regu-latory cascade of the nuclear receptors FXR, SHP-1, and LRH-1 repressesbile acid biosynthesis. Mol. Cell 6, 517–526.

Hofmann, A.F., 1984. Chemistry and enterohepatic circulation of bile acids.Hepatology 4 (5 Suppl.), 4S–14S.

Hofmann, A.F., 1989. Enterohepatic circulation of bile acids. In: Schultz, S.G.(Ed.), Handbook of Physiology. Section on the Gastrointestinal System,pp. 567–596.

Hofmann, A.F., 1994. Pharmacology of ursodeoxycholic acid, an enterohepaticdrug. Scand. J. Gastroenterol., Suppl. 204, 1–15.

Huang, L., Smit, J.W., Meijer, D.K., Vore, M., 2000. Mrp2 is essential forestradiol-17beta (beta-D-glucuronide)-induced cholestasis in rats. Hepatol-ogy 32, 66–72.

Iida, T., Momose, T., Tamara, T., Matsumoto, T., Chang, F.C., Goto, J., Nambara,T., 1988. Potential bile acid metabolites. 13. Improved route to 3β,6β- and3β,6α-dihydroxy-5β-cholanoic acids. J. Lipid Res. 29, 165–171.

Kaplan, M.M., Cheng, S., Price, L.L., Bonis, P.A., 2004. A randomizedcontrolled trial of colchicine plus ursodiol versus methotrexate plusursodiol in primary biliary cirrhosis: ten-year results. Hepatology 39 (4),915–923.

Kast, H.R., Goodwin, B., Tarr, P.T., Jones, S.A., Anisfeld, A.M., Stoltz,C.M., Tontonoz, P., Kliewer, S., Willson, T.M., Edwards, P.A., 2002.Regulation of multidrug resistance-associated protein 2 (ABCC2) bythe nuclear receptors pregnane X receptor, farnesoid X-activatedreceptor, and constitutive androstane receptor. J. Biol. Chem. 277,2908–2915.

Koopen, N.R., Post, S.M., Wolters, H., Havinga, R., Stellaard, F., Boverhof, R.,Kuipers, F., 1999. Differential effects of 17α-ethinylestradiol on the neutraland acidic pathways of bile salt synthesis in the rat. J. Lipid Res. 40,100–108.

Lai, K., Harnish, D.C., Evans, M.J., 2003. Estrogen receptor alpha regulatesexpression of the orphan receptor small heterodimer partner. J. Biol. Chem.278 (38), 36418–36429.

Lee, J., Boyer, J., 2000. Molecular alterations in hepatocyte transportmechanism in acquired cholestasis liver disorders. Semin. Liver Dis. 20,373–384.

Makishima, M., Okamoto, A.Y., Repa, J.J., Tu, H., Learned, R.M., Luk, A.,Hull, M.V., Lustig, K.D., Mangelsdorf, D.J., Shan, B., 1999. Identificationof a nuclear receptor for bile acids. Science 284, 1362–1365.

Mottino, A.D., Cao, J., Veggi, L.M., Crocenzi, F., Roma, M.G., Vore, M., 2002.Altered localization and activity of canalicular Mrp2 in estradiol-17beta-D-glucuronide-induced holestasis. Hepatology 35, 1409–1419.

Parks, D.J., Blanchard, S.G., Bledsoe, R.K., Chandra, G., Consler, T.G.,Kliewer, S.A., Stimmel, J.B., Willson, T.M., Zavacki, A.M., Moore, D.D.,Lehmann, J.M., 1999. Bile acids: natural ligands for an orphan nuclearreceptor. Science 284, 1365–1368.

Paumgartner, G., Beuers, U., 2004. Mechanisms of action and therapeuticefficacy of ursodeoxycholic acid in cholestatic liver disease. Clin. Liver Dis.8 (1), 67–81.

Piazza, F., Montagnani, M., Russo, C., Azzaroli, F., Aldini, R., Roda, E., Roda,A., 2000. Competition in liver transport between chenodeoxycholic acid andursodeoxycholic acid as a mechanism for ursodeoxycholic acid and itsamidates' protection of liver damage induced by chenodeoxycholic acid.Dig. Liver Dis. 32 (4), 318–328.

Poupon, R.E., Lindor, K.D., Pares, A., Chazouilleres, O., Poupon, R.,Heathcote, E.J., 2003. Combined analysis of the effect of treatment withursodeoxycholic acid on histologic progression in primary biliary cirrhosis.J. Hepatol. 39 (1), 12–16.

Riely, C.A., Bacq, Y., 2004. Intrahepatic cholestasis of pregnancy. Clin. LiverDis. 8 (1), 167–176.

Roda, A., Hofmann, A.F., Mysels, K.J., 1983. The influence of bile salt structureon self association in aqueous solution. J. Biol. Chem. 258, 6362–6370.

Roda, A., Pellicciari, R., Polimeni, C., Cerre, C., Forti, G.C., Sadeghpour,B., Sapigni, E., Gioacchini, A.M., Natalini, B., 1995. Metabolism,pharmacokinetics, and activity of a new 6-fluoro analogue ofursodeoxycholic acid in rats and hamsters. Gastroenterology 108 (4),1204–1214.

Rodrigues, C.M.P., Setchell, K.D.R., 1996. Performance characteristics ofreverse-phase bonded silica cartridges for serum bile acid extraction.Biomed. Chromatogr. 10, 1–5.

Simon, F.R., Fortune, J., Iwahashi, M., Gartung, C., Wolkoff, A., Sutherland, E.,1996. Ethinylestradiol cholestasis involves alterations in expression of liversinusoidal transporters. Am. J. Physiol. 271, G1043–G1052.

208 C. Clerici et al. / Toxicology and Applied Pharmacology 214 (2006) 199–208

Stieger, B., Hagenbuch, B., Landmann, L., Hochli, M., Schroeder, A.,Meier, P.J., 1994. In situ localization of the hepatocytic Na+/taurocholate cotransporting polypeptide in rat liver. Gastroenterology107, 1781–1787.

Stieger, B., Fattinger, K., Madon, J., Kullak-Ublick, G.A., Meier, P.J., 2000.Drug- and estrogen-induced cholestasis through inhibition of the hepato-cellular bile salt export pump (Bsep) of rat liver. Gastroenterology 118,422–430.

Van Milligen de Wit, A.W., Kuiper, H., Camoglio, L., Van Bracht, J.,Jones, E.A., Tytgat, G.N., Van Deventer, S.J., 1999. Does ursodeoxy-

cholic acid mediate immunomodulatory and anti-inflammatory effectsin patients with primary sclerosing cholangitis? Eur. J. Gastroenterol.Hepatol. 11 (2), 129–136.

Ward, A., Brogden, R.N., Hell, R.C., Speight, T.M., Avery, G.S., 1984.Ursodeoxycholic acid: a review of its pharmacological properties andtherapeutic efficacy. Drugs 27, 95–131.

Setchell, K.D.R., Worthington, J., 1982. A rapid method for the quantitativeextraction of bile acids and their conjugates from serum using commerciallyavailable reverse-phase octadecylsilane bonded silica cartridges. Clin. Chim.Acta 125, 135–144.