Toxicology in Vitro 28 (2014) 218–230
Contents lists available at ScienceDirect
Toxicology in Vitro journal homepage: www.elsevier.com/locate/toxinvit
Toxicity and intracellular accumulation of bile acids in sandwich-cultured rat hepatocytes: Role of glycine conjugates Sagnik Chatterjee a, Ingrid T.G.W. Bijsmans b, Saskia W.C. van Mil b, Patrick Augustijns a, Pieter Annaert a,⇑ a b
KU Leuven Department of Pharmaceutical and Pharmacological Sciences, Drug Delivery and Disposition, O&N2, Herestraat 49 – Bus 921, 3000 Leuven, Belgium Department of Metabolic Diseases, University Medical Center Utrecht, Utrecht, The Netherlands
a r t i c l e
i n f o
Article history: Received 10 March 2013 Accepted 30 October 2013 Available online 7 November 2013 Keywords: Bile acids Sandwich-cultured rat hepatocytes Hepatotoxicity Bile acid conjugation Chenodeoxycholic acid Cholestasis
a b s t r a c t Excessive intrahepatic accumulation of bile acids (BAs) is a key mechanism underlying cholestasis. The aim of this study was to quantitatively explore the relationship between cytotoxicity of BAs and their intracellular accumulation in sandwich-cultured rat hepatocytes (SCRH). Following exposure of SCRH (on day-1 after seeding) to various BAs for 24 h, glycine-conjugated BAs were most potent in exerting toxicity. Moreover, unconjugated BAs showed signiﬁcantly higher toxicity in day-1 compared to day-3 SCRH. When day-1/-3 SCRH were exposed (0.5–4 h) to 5–100 lM (C)DCA, intracellular levels of unconjugated (C)DCA were similar, while intracellular levels of glycine conjugates were up to 4-fold lower in day3 compared to day-1 SCRH. Sinusoidal efﬂux was by far the predominant efﬂux pathway of conjugated BAs both in day-1 and day-3 SCRH, while canalicular BA efﬂux showed substantial interbatch variability. After 4 h exposure to (C)DCA, intracellular glycine conjugate levels were at least 10-fold higher than taurine conjugate levels. Taken together, reduced BA conjugate formation in day-3 SCRH results in lower intracellular glycine conjugate concentrations, explaining decreased toxicity of (C)DCA in day-3 versus day-1 SCRH. Our data provide for the ﬁrst time a direct link between BA toxicity and glycine conjugate exposure in SCRH. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Bile acids (BAs) are the major organic solutes of bile. They are amphiphilic molecules facilitating the intestinal absorption of fat-soluble compounds such as lipophilic vitamins and dietary lipids. Under normal physiological conditions, BAs are actively taken up from portal blood into the hepatocytes in a sodium-dependent manner, by the sodium taurocholate cotransporting polypeptide [NTCP (human) or Ntcp (rat); SLC10A1/Slc10a1]. In addition, sodium-independent uptake of BAs by hepatic isoforms of the organic anion transporting polypeptide (OATP/Oatp; SLCO/Slco) family Abbreviations: BAs, bile acids; BAAT, bile acid CoA: amino acid N-acyltransferase; BEI, biliary excretion index; BSEP/Bsep, bile salt export pump (human/rat); DMSO, dimethyl sulfoxide; CDCA, chenodeoxycholic acid; CA, cholic acid; FBS, Fetal Bovine Serum GCDCA, glycochenodeoxycholic acid; GDCA, glycodeoxycholic acid; HBSS, Hanks’ Balanced Salt Solution; HEPES, 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid; NOAEL, no observed adverse effect level; NTCP/Ntcp, sodium taurocholate cotransporting polypeptide (human/rat); PBS, Phosphate Buffered Saline; SCH, sandwich-cultured hepatocytes; SCRH, sandwich-cultured rat hepatocytes; TCDCA, taurochenodeoxycholic acid; TDCA, taurodeoxycholic acid. ⇑ Corresponding author. Tel.: +32 16 33 03 03; fax: +32 16 33 03 05. E-mail addresses: [email protected]
(S. Chatterjee), [email protected]
(I.T.G.W. Bijsmans), [email protected]
(S.W.C. van Mil), [email protected]
(P. Augustijns), Pieter.Annaert@ pharm.kuleuven.be (P. Annaert). 0887-2333/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tiv.2013.10.020
also plays a role. Subsequently, BAs are conjugated in the liver with glycine and taurine (at the C24 carboxylic acid moiety), and may also be metabolized by different liver enzymes such as cytochrome P450s, glucuronosyl- and sulfo-transferases. They ﬁnally undergo efﬂux into the bile canaliculi by the bile salt export pump (BSEP/ Bsep, an ATP binding cassette (ABC) transporter; ABCB11/Abcb11) (Kullak-ublick et al., 2004). Multidrug resistance-associated protein-2 (MRP2/Mrp2; ABCC2/Abcc2) is another canalicular ABC transporter that mediates the canalicular efﬂux of divalent BAs such as the sulfated glycine or taurine conjugates. MRP3/Mrp3 (ABCC3/Abcc3) and MRP4/Mrp4 (ABCC4/Abcc4), localized at the basolateral side of the hepatocytes, are responsible for the efﬂux of BAs back into the sinusoidal blood (Alrefai and Gill, 2007). BSEP is a transport protein primarily responsible for the elimination of monovalent BAs at the canalicular membrane. Inhibition of BSEP has been implicated in the pathogenesis of estrogen-induced intrahepatic cholestasis of pregnancy (Pauli-Magnus et al., 2010). Genetic defects in BSEP can lead to progressive familial intrahepatic cholestasis type 2 and benign recurrent intrahepatic cholestasis type 2 (van Mil et al., 2004; Dawson et al., 2009). BSEP/Bsep function can also be affected by endo- (e.g., estradiol17b-D-glucuronide) and xenobiotics leading to impaired canalicular BA excretion and subsequent excessive intracellular accumulation of BAs (Stieger et al., 2000). These clinical adverse events have
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
been suggested to be associated with disturbed BA homeostasis due to altered BA elimination as a key underlying cause. Serum BA concentrations of about 3–7 lM are considered as physiological (Scherer et al., 2009; Tribe et al., 2009). When BAs accumulate in hepatocytes above their physiological concentration range (which may or may not be associated with increased serum BA concentrations), clinical hepatotoxicity may be observed. Higher BA concentrations (in the mM range) are thought to exert cytotoxicity by causing necrosis. However, in the lM range of supra-physiological BA concentrations, BA-induced apoptosis is supposed to be the predominant mechanism (Perez and Briz, 2009). Several mechanisms have been suggested to explain BA-induced apoptosis, including: (i) direct activation of the death receptors Fas and TRAIL-R2, (ii) mitochondrial permeability transition induction and subsequent cytochrome c release and, (iii) mitochondrial and endoplasmic reticulummediated oxidative stress (reviewed in Perez and Briz (2009)). The toxicity associated with excessive BA accumulation results in impaired liver functions as observed in case of cholestasis. Higher plasma and hepatic BA concentrations have indeed been reported in case of cholestasis in both humans and rodents. In chronic cholestatic patients, hepatic BA concentrations as high as 430 lM have been measured (Fischer et al., 1996; Rolo et al., 2003). To provide more direct evidence for toxicity associated with such high BA levels, in vitro work directly linking intracellular BA exposure to toxicity is of interest. Various experimental animal models are available for studying the pathophysiological alterations during intrahepatic cholestasis (Rodríguez-Garay, 2003). However, it is not possible to accurately control and measure (hepatic) BA concentrations in vivo. In contrast, cell-based in vitro systems can provide insight into the role of individual BA species (and/or their metabolites) in mediating hepatotoxicity following elevated concentrations. BA-mediated toxicity has been studied in different in vitro systems including isolated rat liver mitochondria, cell lines (e.g., HepG2), and isolated hepatocytes in suspension and in conventional primary cultures of rat hepatocytes (Spivey et al., 1993; Hillaire et al., 1995; Martinez-Diez et al., 2000; Rolo et al., 2002, 2004). However, these in vitro systems only partly reﬂect the intricate interplay between (uptake and efﬂux) transporters and metabolizing enzymes that takes place inside the liver to regulate the intracellular BA concentration. Sandwich-cultured hepatocytes (SCH) represent an in vitro model that supports expression of relevant basolateral and canalicular transporters (De Bruyn et al., 2013). In addition, cultured hepatocytes have been reported to contain the necessary enzymes such as BAAT (bile acid CoA: amino acid N-acyltransferase), which mediates conjugation of unconjugated BAs with either glycine or taurine (Falany et al., 1994; Rembacz et al., 2010). SCH are therefore an excellent model to further explore the link between BA-mediated toxicity and intracellular hepatic exposure to these BAs or their conjugates. In sandwich-cultured rat hepatocytes (SCRH), the expression/ activity level of the different enzymes and transporters changes with culture time (De Bruyn et al., 2013). More speciﬁcally, Ntcp expression in SCRH gradually decreases (Tchaparian et al., 2011). Functional bile canaliculi are fully developed in SCRH by day-3 after seeding (Liu et al., 1998). Hence, it is expected that Bsep-mediated BA efﬂux will constitute an elimination mechanism in day-3 but not in day-1 SCRH. In the present study we have shown that, BA toxicity in SCRH [as determined with an assay reﬂecting (loss of) hepatocyte-speciﬁc function] is linked directly to the in vitro hepatobiliary disposition proﬁles of these BAs (and their metabolites/conjugates). In vitro incubations in day-1 and day-3 SCRH revealed culture-time dependent cytotoxicity and hepatic disposition of BAs, thus providing direct evidence for the putative link between toxicity of certain
BA species and the intracellular levels they reach in (cultured) hepatocytes. 2. Materials and methods 2.1. Materials Williams’ E Medium (WEM), L-glutamine, penicillin–streptomycin mixture (contains 10,000 IU/ml potassium penicillin and 10,000 lg/ml of streptomycin sulfate), Fetal Bovine Serum (FBS), Hanks’ Balanced Salt Solution (HBSS), Ca++/Mg++-free HBSS, Phosphate Buffered Saline (PBS; 1 and 10), Trypan blue solution (0.4%) were purchased from Lonza Verviers SPRL (Verviers, Belgium). ITS™ + Premix (contains insulin 6.25 mg/l, transferrin 6.25 mg/l, selenous acid 6.25 mg/l, bovine serum albumin 1.25 g/l and linoleic acid 5.35 mg/l) was purchased from BD Biosciences (Erembodegem, Belgium). Sulfuric acid (95–97%) was purchased from Chem-Lab NV (Zedelgem, Belgium). All the BAs, collagenase type IV (from Clostridium histolyticum), recombinant human insulin, Dulbecco’s modiﬁed Eagle’s Medium 10 (10 DMEM), Triton X-100, dexamethasone, urea, diacetyl monoxime, thiosemicarbazide, iron (III) chloride hexahydrate, ortho-phosphoric acid, ethylene glycol-bis(2-aminoethylether)-N,N,N0 , N0 -tetraacetic acid (EGTA) were purchased from Sigma–Aldrich (Schnelldorf, Germany). HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) was purchased from MP Biochemical (Illkirch, France). TRIzolÒ reagent was purchased from Life Technologies (Gent, Belgium). The anti-BAAT antibody raised in rabbit was purchased from Proteintech™ (local distributor Bio Connect, The Netherlands). Thermostable 96-well plates were kindly provided by Greiner Bio-One BVBA (Wemmel, Belgium). Collagen was prepared inhouse from rat tails according to established procedures. For incubations with sandwich-cultured hepatocytes described below, ‘standard buffer’ consisted of HBSS adjusted to pH 7.4 and ‘Ca++/ Mg++-free buffer’ consisted of Ca++/Mg++-free HBSS containing 1 mM EGTA (adjusted to pH 7.4). 2.2. Animals Rats were housed according to the Belgian and European laws, guidelines and policies for animal experiments, housing and care in the central animal facilities of the university. Approval for this project was granted by the Institutional Ethical Committee for Animal Experimentation. 2.3. Isolation of rat hepatocytes Hepatocytes were isolated from male Wistar rats (170–200 g) based on a two-step collagenase perfusion method, as described previously (Annaert et al., 2001), without adding trypsin inhibitor. After isolation, cells were centrifuged (50g) for 3 min at 4 °C and the pellet was re-suspended in WEM containing 5% FBS, 2 mM Lglutamine, 100 IU/ml penicillin, 100 lg/ml streptomycin, 4 lg/ml insulin, and 1 lM dexamethasone (seeding medium). Hepatocytes were counted using a hemocytometer and cell viability was determined using Trypan blue. Cells were diluted to a ﬁnal concentration of 1 106 cells/ml, in seeding medium. Freshly-isolated hepatocytes used in experiments had a viability of at least 85%. 2.4. Sandwich-cultured rat hepatocytes 24-Well/6-well plates were coated with ice-cold collagen solution (1.5 mg/ml ﬁnal concentration: 50 ll/well for 24-well and 150 ll/well for 6-well plates), prepared by neutralizing a mixture of 4 parts of rat-tail collagen, 4 parts of deionized water, and 1 part
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
of 10 DMEM with 1 part of 0.2 N NaOH (ﬁnal pH 7.4), and placed overnight at 37 °C in a humidiﬁed incubator. Three hours before seeding, each well was hydrated with 500 ll (2 ml/well for 6-well plates) of PBS at 37 °C. Hepatocytes were seeded at a density of 0.5 106 cells/well in 500 ll or 1.75 106 cells/well in 1.75 ml seeding medium for 24-well and 6-well plates, respectively. After incubating the cells for 1–2 h at 37 °C in a humidiﬁed atmosphere with 5% CO2 (Binder CO2 incubator, Binder GmbH), unattached cells were removed by swirling the plate and immediately aspirating the medium. To obtain a ‘‘sandwich’’ conﬁguration, the cells were overlaid with 50 ll or 200 ll of rat tail collagen solution (1.5 mg/ml, pH 7.4) for 24- and 6-well plates, respectively. One hour later, pre-warmed seeding medium was added onto the cultures that were kept in a humidiﬁed atmosphere with 5% CO2. The medium was changed every day with culture medium consisting of WEM supplemented with 1% (v/v) ITS™ + Premix, 100 IU/ml penicillin, 2 mM L-glutamine, 100 lg/ml streptomycin, and 0.1 lM dexamethasone.
(20 °C) 7:3 mixture of methanol:water, containing 1 lM d4-cholic acid (d4-CA) as internal standard, and centrifuged at 14,000 rpm for 15 min. After centrifugation, supernatants were collected and kept at 20 °C till analysis by LC–MS(/MS). In an alternative incubation protocol, aimed at the determination of the total formation and sinusoidal efﬂux of GCDCA and TCDCA, the 10-min efﬂux phase was omitted, i.e., culture medium was collected for analysis immediately after incubation of the day-1 and day-3 SCRH with CDCA for 30 min or 4 h. Subsequently, the cells were rinsed four times with standard buffer at 4 °C, followed by harvesting the SCRH according to the above-mentioned procedure for determination of intracellular contents. The supernatants collected after 10 min incubation with either standard or Ca++/Mg++-free buffer and the collected culture medium samples (obtained after the alternative incubation protocol) were diluted with two volumes of a 7:3 mixture of methanol:water, containing 1 lM d4-CA as internal standard, and centrifuged at 14,000 rpm for 15 min. After centrifugation, supernatants were collected and kept at -20 °C till analysis by LC–MS(/MS).
2.5. Determination of urea formation in SCRH To determine the urea synthesis capacity in SCRH, culture medium was aspirated and the hepatocytes were rinsed twice with HBSS at 37 °C. The cells were incubated with HBSS containing 10 mM HEPES, 2 mM L-glutamine, 10 mM ammonium chloride and 3 mM ornithine (250 ll/well for 24-well plates) for 1 h at 37 °C in a humidiﬁed atmosphere with 5% CO2. After 1 h of incubation, 60 ll of the incubation medium/well was mixed with 240 ll of color reagent (see below) in a 96-well thermostable plate. The mixture of color reagent and incubate was heated at 85 °C for 20 min in a water bath and subsequently cooled down by keeping the plate at 4 °C for 10 min. Absorbance was measured at 525 nm using a Tecan Inﬁnite M200 plate reader (Austria). The color reagent was prepared by mixing the following solutions A and B in 2:1 ratio (v/v) just before the measurement. Solution A: 30 ml of concentrated sulfuric acid and 10 ml orthophosphoric acid added slowly to 60 ml of ferric chloride solution (160 mg/l FeCl36H2O) and mixed gently; the solution was kept on ice. Solution B: a 1:1 mixture of 60 mg thiosemicarbazide in 100 ml of ELGA water and 1200 mg diacetyl monoxime dissolved in 100 ml ELGA water. 2.6. Comparison of BA toxicity between day-1 and day-3 SCRH BA toxicity was compared between day-1 and day-3 SCRH in two batches of rat hepatocytes. For this purpose, 5, 50, 100, 150, 200, 250, and 500 lM of CDCA or DCA and 5, 50, 100, 200, 350, 500 and 1000 lM of UDCA or GCDCA were incubated with day-1 and day-3 SCRH. After 24 h of incubation with the BAs, urea formation was measured at day-2 and day-4. DMSO content was standardized to 0.3% for each condition. 2.7. Measurement of accumulation of CDCA, DCA and their taurine and glycine conjugates To reveal a possible relation between BA toxicity and intracellular exposure of hepatocytes to BAs, day-1 and day-3 SCRH were incubated for 30 min (CDCA only) or 4 h with CDCA and DCA. In a ﬁrst incubation protocol, hepatocytes were rinsed twice with either standard or Ca++/Mg++-free buffer at 37 °C, immediately after the 30 min/4 h incubation. Subsequently, hepatocytes were incubated for 10 min (37 °C) with either standard or Ca++/Mg++-free buffer (B-CLEARÒ procedure, Qualyst Transporter Solutions, Durham, NC). Next, incubation buffer was aspirated [or collected to determine BA efﬂux; see below for efﬂux buffer sample processing] and cells were immediately rinsed four times with ice-cold standard buffer. Hepatocytes were then harvested with a chilled
2.8. LC–MS analysis BA analysis was performed either with a Waters LC–MS system (Milford, MA, USA) or a Thermo Fisher Scientiﬁc LC-MS/MS system (Thermo Fischer, Breda, The Netherlands). The Waters system consisted of an HT alliance HPLC component and Micromass ZQ MS component with ESI source. The Thermo Fisher Scientiﬁc system consisted of a TSQ Quantum Access™ mass spectrometer coupled with Accela™ U-HPLC system (Thermo Fischer, Breda, The Netherlands). For analysis with the Waters LC-MS system, separation was achieved with a Phenomenex Gemini C18 column (3 l, 150 4.6 mm) with matching guard and pre-column, at a ﬂow rate of 400 ll/min. The injection volume was 25 ll. The mobile phase was 15% methanol, 15% ammonium formate with formic acid (pH adjusted to 3.5) and 70% acetonitrile. The run time was 12 min. For analysis with the Thermo Fisher Scientiﬁc LC-MS/MS system, a Kinetex C18 column (1.7 lm, 100 A, 50 3 mm) with a KrudKatcher ultra HPLC in-line ﬁlter (PhenomenexR, The Netherlands) was used to achieve optimum separation. The injection volume was 10 ll. The chromatographic conditions were: 300 ll/ min ﬂow rate for 3 min, with mobile phase consisting of 80:20 methanol:5 mM ammonium formate buffer (pH adjusted to 3.5 with formic acid). Analysis was done in the negative electrospray ionization mode. The monitored mass/charge ratios and transitions are included in Tables 1a and 1b, respectively. BA concentrations in samples were determined by calculating the area ratio of analyte to internal standard. Analytical methods were validated for reproducible quantiﬁcation of (i) CDCA and its conjugates, and (ii) DCA and its conjugates, with nominal QC sample concentrations of 0.05, 0.5 and 2 lM. The interday inaccuracy varied from +0.31% to 5.97% and from +1.96% to +4.78% for analysis of CDCA (+ conjugates) and DCA (+ conjugates), respectively. The reproducibility expressed by % CV was 0.55–9.2% for CDCA (+ conjugates), and
Table 1a Molecular ions monitored for LC–MS analysis of CDCA, DCA and corresponding glycine and taurine conjugates in cell lysate from SCRH. d4-CA is used as internal standard. Analyte
Retention time (min)
CDCA DCA GCDCA GDCA TCDCA d4-CA
391 391 448 448 498 411
8.9 9.3 7.2 7.6 8.3 6.8
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230 Table 1b Transitions monitored for UPLC–MS/MS analysis of GCDCA, GDCA, TCDCA, TDCA, DCA along with internal standard d4-CA. CDCA was analyzed unfragmented. Analyte
Retention time (min)
CDCA GCDCA TCDCA d4-CA DCA GDCA TDCA
391.400 448.276 498.263 411.256 391.257 448.276 498.248
391.400 74.267 80.244, 124.083 347.171 345.274 74.303 80.244, 124.132
2.0 1.3 1.0 1.4 2.1 1.3 1.0
3.4–5.2% for DCA (+ conjugates). For analysis with the Thermo LC– MS/MS system, the interday inaccuracy for the CDCA (+ conjugates) group varied from 4.9% to +7.9%, and from +1.2% to 7.1% for DCA (+ conjugates). The reproducibility (% CV) for analysis at different days was 3.8–13.4% for BAs in the CDCA group, while it was 2.0–10% for BA in the DCA group. 2.9. BAAT mRNA levels measurement
(+10 min) incubation periods. Intracellular BA concentrations were estimated by dividing amounts of intracellularly accumulated BAs by total hepatocyte volume. The hepatocyte cellular volume used to calculate intracellular BA concentrations (6.34 pl/cell) was based on the mean of 2 previously reported values for rat hepatocyte cellular volume (Uhal and Roehrig, 1982; Brouwer, 2010). 2.11. Statistics ANOVA (F-test in MS Excel version 2007) was used to evaluate statistical signiﬁcance of differences between day-1 and day-3 toxicity (Fig. 1) or accumulation (Fig. 2) proﬁles, by comparing separately ﬁt proﬁles for each culture day versus the simultaneous ﬁt obtained with day-1 and day-3 measurements combined (Gibson et al., 1994). The criterion for statistical signiﬁcance was p < 0.05. A two-tailed Student’s t-test was used to evaluate statistical differences between BAAT mRNA levels in day-1, 3 and -4 SCRH (GraphPad Prism 5 for Windows; GraphPad Software Inc., San Diego, CA). 3. Results
Total RNA was isolated from SCRH using TRIzol reagent (Ambion/Life Technologies). RNA was reverse transcribed using Superscript II (Invitrogen) according to manufacturer’s instructions. qPCR was performed on MyIQ™ single-color real-time PCR Detection System (Biorad), using a reaction mix containing Fast start Universal SYBR Green Master (Roche) forward (50 GGAAGGCCGAATCCGGGG30 ) and reverse primer (50 GCAAAGCCATGACTGGCCAGG30 ), and 12.5 ng cDNA. Target gene mRNA expression was normalized to b-actin. Data are expressed as fold-increase in mRNA levels detected in day-1 SCRH, arbitrarily set at 100%. 2.10. Data analysis The Emax model was used to describe the concentration-dependency of the inhibitory effect of various BAs on the capacity of hepatocytes to produce urea:
E ¼ Emax ðEmax E0 Þ
Cn n C þ ICn50
where E is the urea production by the hepatocytes, Emax is the urea production under control condition (without inhibitor), E0 is the urea production at the maximum inhibitory effect of inhibitor, (Emax E0) is the maximum inhibitory effect. The IC50 is the BA concentration causing 50% inhibition of urea formation. The parameter ‘‘n’’ denotes the Hill factor. The best ﬁts of the above equation to the individual urea formation data sets were obtained by non-linear regression analysis with the NLS2 package in R version 2.15.1. The inverse of the experimentally obtained standard deviations were used for weighing. In the absence of Ca++/Mg++, the tight junctions between hepatocytes were disrupted, resulting in release of biliary content from the canalicular space (Liu et al., 1999). The accumulation values obtained from wells treated with Ca++/Mg++-free buffer were considered to reﬂect intracellular BA accumulation only. The amounts excreted in bile canaliculi were determined by subtracting the amounts of BAs in cell lysates prepared from cultures treated with Ca++/Mg++-free buffer, from amounts measured in cultures treated with standard buffer. In an adapted incubation protocol with day-4 SCRH, cultured in 6-well plates, net canalicular efﬂux of CDCA and GCDCA was calculated by subtracting the BA levels in the standard efﬂux buffer from the BA levels in the Ca++/Mg++-free efﬂux buffer. Note that this canalicular efﬂux reﬂects BA amounts that have been excreted and retained in the bile networks during the 4 h
3.1. Comparison of BA toxicity between day-1 and day-3 SCRH The effect of culture time (day-1 versus day-3) on BA toxicity was evaluated for CDCA, DCA, UDCA, and GCDCA in two batches of rat hepatocytes by comparing urea formation capacity (Fig. 1). Higher BA concentrations were required in day-3 cultures, as compared to day-1 cultures, to decrease the urea formation to the same extent. This implies that BAs are signiﬁcantly more toxic to day-1 SCRH than day-3 SCRH. The cytotoxicity differences are reﬂected by the signiﬁcantly higher IC50 values obtained in day-3 cultures for both hepatocyte batches (Table 2). 3.2. Intracellular accumulation of CDCA, DCA and their conjugates upon extracellular exposure of day-1 and day-3 SCRH to CDCA and DCA As differences in BA toxicity between day-1 and day-3 hepatocyte cultures are possibly related to different intracellular exposure to BAs, concentration-dependent accumulation of CDCA and DCA, as well as their formed conjugates was determined. For this purpose, hepatocytes were incubated with CDCA and DCA at 4 different concentrations, covering the no observed adverse effect level (NOAEL) range obtained from the toxicity data presented in Fig. 1. Fig. 2 depicts BA accumulation after 4 h exposure to 5, 25, 50 or 100 lM of CDCA or DCA, in three different batches of SCRH. Note that the 4 h incubation with BAs was followed by rapid rinsing of the cells and subsequent 10 min incubation with Ca++/Mg++free buffer. By day-3, SCRH have developed bile canaliculi that release their contents upon treatment with Ca++/Mg++-free buffer. Therefore, this particular incubation design (4 h medium + 10 min buffer) was necessary to determine intracellular accumulation without canalicular content in day-3 SCRH. For consistency, day1 hepatocytes were treated identically, even though no canaliculi are present at this time in culture. Fig. 2 illustrates that the hepatocytes converted the administered unconjugated BAs CDCA and DCA to corresponding glycine and taurine conjugates. The combined amounts of CDCA and its conjugates GCDCA and TCDCA that were retained in cells after 4 h exposure to 5, 25, 50 or 100 lM of CDCA (and followed by 10 min efﬂux in Ca++/Mg++-free buffer) are shown in Fig. 2A. The corresponding data following exposure to DCA are shown in Fig. 2E. The combined BA levels did not differ signiﬁcantly between day-1 and day-3 SCRH. Also when the intracellular levels of the unconjugated BAs only were considered, as illustrated in Fig. 2B
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
Fig. 1. Culture time-dependent toxicity of bile acids (BAs): day-1 (s) versus day-3 (D) toxicity proﬁles of CDCA, DCA, UDCA and GCDCA in two SCRH batches (column A and B for batch 1 and 2, respectively) cultured in 24-well plates, following 24 h incubation with BAs. The solid and the dotted lines represent the best ﬁt of the sigmoid Emax model to the day-1 and day-3 toxicity proﬁles in SCRH. Urea formation is expressed as % of control (hepatocytes treated with 0.3% DMSO). The toxicity proﬁles of the four BAs are statistically signiﬁcantly different (p < 0.05) between day-1 and day-3 SCRH.
for CDCA and Fig. 2F for DCA, no signiﬁcant differences between day-3 and day-1 cultures could be observed. However, the respective glycine and taurine conjugates reached signiﬁcantly higher intracellular levels in day-1 SCRH compared to day-3 SCRH
(Fig. 2C and D for CDCA conjugates and Fig. 2G and H for DCA conjugates). Fig. 2 further illustrates that 4 h exposure to (C)DCA results in 10–20 times lower taurine conjugate levels as compared to glycine conjugate levels.
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
Fig. 2. Culture time-dependent accumulation of CDCA, DCA and their formed conjugates: (A–D) Intracellular accumulation of: CDCA and its conjugates (A), CDCA (B), GCDCA (C), and TCDCA (D), in day-1 (s) or day-3 (D) SCRH, after incubation with 5, 25, 50 and 100 lM of CDCA for 4 h, followed by 10 min incubation with Ca++/Mg++-free buffer. (E– H) Intracellular accumulation of: DCA and its conjugates (E), DCA (F), GDCA (G), and TDCA (H), in day-1 (s) or day-3 (D) SCRH, after incubation with 5, 25, 50 and 100 lM of DCA for 4 h, followed by 10 min incubation with Ca++/Mg++-free buffer. Data represent the mean (± SEM) accumulation in three different batches of SCRH (n = 3; for TDCA n = 2) cultured in 24-well plates, each measured at least in triplicate. Indicates that the accumulation proﬁles are statistically signiﬁcantly different between day-1 and day-3 SCRH (p < 0.05).
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
depicts corresponding medium concentrations (lM) of GCDCA and TCDCA as measured immediately after 4 h CDCA exposure. Consistent with their total formation proﬁles, GCDCA and TCDCA concentrations were about 2 times higher in the medium of day1 SCRH compared to day-3 SCRH, at intracellular CDCA concentrations of 20 lM onwards.
CDCA DCA UDCA GCDCA
123 ± 16.3 141 ± 20.7 341 ± 35.3 91.9 ± 22.5
205 ± 17.0 396 ± 18.4 836 ± 152 398 ± 45.2
130 ± 10.9 74.3 ± 9.50 317 ± 29.7 62.7 ± 8.30
202 ± 12.7 331 ± 20.6 840 ± 98.4 434 ± 142
3.4. In vitro biliary excretion of CDCA, DCA and their conjugates by day-3 SCRH
Table 2 Mean (±SD) IC50 values (lM) for inhibition of urea formation following 24 h exposure of two batches of day-1 and day-3 SCRH to CDCA, DCA, UDCA or GCDCA. IC50 (lM)
Sandwich-cultured rat hepatocytes Batch 1
3.3. Effect of culture time on formation and sinusoidal efﬂux of GCDCA and TCDCA in day-1 and day-3 SCRH Subsequent incubations were conducted to delineate the exact role of changing BA conjugation activity as underlying mechanism in culture time-dependent BA disposition. For this purpose, day-1 and day-3 hepatocytes were exposed to different concentrations of CDCA for 4 h, before measuring CDCA (and conjugate) levels both in hepatocyte cultures and in the medium. Importantly, the 10 min incubation with Ca++/Mg++-free buffer was omitted in this incubation design. Thus, total conjugating capacity towards BA could be assessed for the various CDCA concentrations in both day-1 and day-3 cultures (Fig. 3A and C for GCDCA and TCDCA). Note that the total conjugate levels were plotted against intracellular CDCA concentrations (X-axis) as these constitute the driving force for BA conjugation. Intracellular concentrations were calculated based on measured intracellular CDCA amounts and estimated total intracellular hepatocyte volume (Uhal and Roehrig, 1982; Brouwer, 2010). Fig. 3A and C conﬁrms that after 4 h exposure to CDCA, total glycine and taurine conjugate levels produced by day-1 hepatocytes clearly exceeded those in day-3 hepatocytes for intracellular CDCA concentrations above 20 lM. Fig. 3B and D
Apart from elimination by conjugation and subsequent sinusoidal efﬂux, BAs added to day-3 SCRH may also undergo biliary excretion, because of formation of bile canaliculi in day-3 SCRH. To further explore the contribution of this elimination mechanism, SCRH were ﬁrst exposed to CDCA or DCA for 4 h; subsequently, the medium was rinsed off and hepatocytes were incubated for 10 min with either standard or Ca++/Mg++-free buffer. For a selected batch of day-3 SCRH, mean (± SD) amounts of individual BAs accumulating in cells and bile networks (standard buffer) were found to be higher than BAs accumulating in cells only (Ca++/Mg++-free buffer) (see Supplementary Fig. 1). The differences between accumulation following standard versus Ca++/Mg++-free buffers represent biliary excretion of (C)DCA and corresponding conjugates. Fig. 4A depicts the biliary excretion of (C)DCA in function of (calculated) intracellular (C)DCA concentration. In contrast to the clear saturation of conjugate formation (Fig. 3), biliary excretion of unconjugated (C)DCA did not reveal saturation. Fig. 4B and C illustrates that after formation, the glycine conjugates of CDCA and DCA were excreted into the bile networks. Upon 4 h exposure to CDCA followed by 10 min efﬂux phase, the intracellular levels of GCDCA increase with increasing intracellular CDCA concentration. In contrast, the biliary excretion of GCDCA did not increase proportionally beyond 25 lM of extracellular CDCA (Fig. 4B). The maximum amount of GCDCA in
Fig. 3. Total formation and culture medium concentration of GCDCA and TCDCA in day-1 and day-3 SCRH: Total amounts (in hepatocytes + medium) of GCDCA (3A) and TCDCA (3C) and culture medium concentration of GCDCA (3B) and TCDCA (3D) after incubating day-1 (s) or day-3 (D) SCRH with 5, 25, 50 and 100 lM CDCA for 4 h. Incubations were followed by measurement of BA levels in the culture medium and in the hepatocytes. Data represents mean (± SD) of triplicate measurements in a single batch of SCRH, cultured in 6-well plates. The X-axis represents intracellular CDCA concentrations after 4 h of incubation.
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
Ca++/Mg++-free buffer were much higher compared to that in the standard buffer throughout the concentration range 5–100 lM during the 10 min incubation (Fig. 5). In line with accumulation data, efﬂux of TCDCA was about 10 times less than the efﬂux of GCDCA (data not shown). Moreover, corresponding BA accumulation data revealed biliary excretion, mainly of GCDCA when higher CDCA concentrations were used (Supplementary Fig. 2). 3.6. Culture time-dependent expression of BAAT mRNA in day-4 SCRH Further to the observed decrease in BA conjugate formation in day-3 SCRH (Figs. 2 and 3), possible culture time-dependent changes in mRNA expression of the BA conjugating enzyme, BAAT were investigated. Unexpectedly, a fourfold increase in BAAT mRNA levels was observed in day-3 and day-4 SCRH compared to day-1 SCRH. The mRNA levels of BAAT in day-4 SCRH were not found to be signiﬁcantly different from day-3 SCRH (Fig. 6). 3.7. Effect of incubation time on disposition of CDCA in day-1 and day3 SCRH
Fig. 4. Biliary excretion of (C)DCA and G(C)DCA in day-3 SCRH: (A) Amount of unconjugated CDCA (h) and DCA (j) directly excreted in bile as a function of CDCA and DCA intracellular concentrations, after exposure to extracellular concentrations of 5, 25, 50 and 100 lM CDCA or DCA. (B and C) Cellular accumulation (o) and biliary excretion (h) of GCDCA (B) and GDCA (C) after incubating day-3 SCRH for 4 h with 5, 25, 50 and 100 lM CDCA or DCA. Next, cellular and canalicular BA levels were determined following incubations with standard or Ca++/Mg++-free buffer for 10 min, as described in Section 2. Data represent the mean (± SD) accumulation of triplicate measurements in a selected batch of SCRH, cultured in 24-well plates.
bile was detected at 25 lM of extracellular CDCA exposure. For GDCA, the amount excreted into the bile was lower than intracellular GDCA amount at each extracellular DCA concentration (Fig. 4C), however, no saturation was observed. BA accumulation in standard versus Ca++/Mg++-free buffer treated cells did not differ signiﬁcantly in two other batches of SCRH (data not shown). 3.5. Disposition of CDCA and conjugates in day-4 SCRH As mentioned in Section 3.4, the extent of biliary excretion of (C)DCA and their conjugates in day-3 SCRH exhibited considerable interbatch variability. Therefore, an adapted incubation design was used to further explore the relevance of the biliary excretory pathway for CDCA and its conjugates. For this purpose, sinusoidal and canalicular efﬂux of CDCA and GCDCA were determined in day-4 SCRH that had been exposed to various concentrations of CDCA for 4 h. Amounts of CDCA and GCDCA that were efﬂuxed in the
As preliminary experiments with distinct SCRH batches suggested unexpected incubation time-dependent BA accumulation, the incubation time-dependency of CDCA accumulation and conjugation in day-1 versus day-3 SCRH was determined after incubating the same batch of SCRH for 30 min and 4 h with CDCA. These incubations were immediately followed by intracellular measurement of CDCA and conjugates. No subsequent efﬂux phase was applied for this purpose. This implies that for day-3 SCRH the combined accumulation in both cells and bile networks was determined (as opposed to data in Fig. 2). Interestingly, these accumulation data revealed that total BA accumulation hardly increased or even signiﬁcantly decreased between 30 min and 4 h (240 min) incubation of day-1 and day-3 SCRH. In addition to the unexpected decrease (or only slight increase) in total BA accumulation between 30 min and 4 h, relative (and absolute) amounts of TCDCA and GCDCA that accumulated (and remained) in the hepatocytes (and also bile networks for day-3 SCRH) were found to be highly incubation time-dependent. After 30 min incubation with CDCA, higher levels of TCDCA were seen as compared to GCDCA, whereas after 4 h incubation, the opposite was observed (see also Figs. 2 and 3). This phenomenon was observed at each concentration of extracellular exposure of CDCA and in both day-1 (Fig. 7A) and day-3 (Fig. 7B) SCRH. Decreasing BA accumulation with culture time and higher TCDCA levels after 30 min were conﬁrmed when the 4 h incubation with CDCA was followed by a 10 min efﬂux with standard buffer as well (Supplementary Fig. 3). 4. Discussion The cellular and molecular mechanisms involved in the development of cholestatic diseases remains elusive. However, excessive intrahepatic accumulation of BAs and/or their metabolites is thought to play a pivotal role in mediating the hepatic injury observed in cholestatic diseases (Schmucker et al., 1990). Intracellular BA levels in the liver are the result of intricate inﬂuences of hepatic BA metabolism along with hepatic uptake and efﬂux processes. Using sandwich-cultured rat hepatocytes, we presently substantiate the link between (excessive) intracellular BA accumulation and compromised hepatocyte viability and function. The goal of this study was to speciﬁcally address the culture timedependent changes in toxicity of BAs, in light of anticipated culture time-dependent changes in exposure to different BA species in SCRH. In function of culture time, dynamic changes in enzymes and transporters activities involved in BA disposition have been
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
Fig. 5. Efﬂux of CDCA and GCDCA by day-4 SCRH: Sinusoidal [in standard buffer (s)], combined ‘sinusoidal + canalicular’ [in Ca++/Mg++-free buffer (h)] and canalicular efﬂux (D) of CDCA and GCDCA by day-4 SCRH cultured in 6-well plates. Following a 4 h exposure of day-4 SCRH to 5–100 lM CDCA, hepatocytes were rinsed (to remove residual extracellular BAs) and immediately incubated for 10 min with standard or Ca++/Mg++-free buffer. BA levels were measured afterwards in standard or Ca++/Mg++-free buffer. Canalicular efﬂux was obtained by subtracting the levels of sinusoidal efﬂux (standard buffer) from combined ‘sinusoidal + canalicular’ efﬂux (Ca++/Mg++-free buffer). The levels of CDCA in lower concentrations (5 and 25 lM) are shown in the inset. Data represents mean (± SD) amounts of CDCA or GCDCA efﬂux measured in triplicate.
Fig. 6. Time-dependent changes in BAAT mRNA expression in sandwich-cultured rat hepatocytes from day-1 to day-4: total RNA was isolated from 1, 3, and 4-dayold cultured primary rat hepatocytes as described in Section 2. Quantitative realtime PCR was performed and mean (± SD) of quadruplicate measurements are indicated for each culture time. Data represents mean of fold change with respect to the BAAT mRNA content of day-1 SCRH. Indicates statistically signiﬁcant difference compared to day-1 mRNA expression (p 6 0.0007).
reported. These changes include increased expression (Annaert et al., 2001; Li et al., 2009; Draheim et al., 2010; Tchaparian et al., 2011) and activity (Turncliff et al., 2006; Li et al., 2009) of basolateral and canalicular BA efﬂux transporters as well as decreased hepatic BA uptake transporter expression and activity (Jigorel et al., 2005; Tchaparian et al., 2011; Noel et al., 2013). As a consequence of these previously documented changes in BA disposition-related enzyme/transporters, we initially hypothesized that exogenously added BAs would exert their cytotoxic effect in SCH in a culture time-dependent fashion (Fig. 8). The literature cited above strongly suggests that BA handling will differ substantially between day-1 and day-3 SCRH. Our present observation (Fig. 1) that BA-induced cytotoxicity in SCRH was signiﬁcantly lower in biliary excretion-competent (i.e., day-3) hepatocytes corroborates this concept of culture time-dependent capacity in BA handling. To further explore the culture time-dependent BA exposure-toxicity association, intracellular BA levels were quantiﬁed following exposure of day-1 versus day-3 SCRH to exogenously added BAs. A particular advantage of this approach is that the use of transporter and/or enzyme inhibitors (with often questionable speciﬁcity) of hepatic BA disposition processes can be circumvented.
In the present work, compromised urea synthesis capacity was chosen as an indicator for concentration-dependent in vitro cytotoxicity of BAs in SCRH (Supplementary Fig. 4, Fig. 1). Determination of urea formation by BA-exposed hepatocytes enables detecting (sub-lethal) reduction in hepatocyte functionality (Dabos et al., 2004; Poll et al., 2006). To efﬁciently measure urea formation in SCRH for our purpose, the well-known urea assay (Goeyens et al., 1998) was ﬁrst miniaturized to a higher throughput 96-well format. We presently show that among all BAs evaluated, glycineconjugated BAs were the most potent in reducing the urea formation rate in day-1 SCRH upon 24 h exposure (Supplementary Fig. 4). Consistently, several earlier in vitro observations in rat hepatocytes (Spivey et al., 1993; Patel et al., 1994) illustrated that glycine-conjugated BAs were more toxic than their unconjugated or taurine-conjugated counterparts. This result supports the validity of the urea assay for the purpose of the present study. Concentration-dependent intracellular accumulation (Fig. 2) of CDCA and DCA (including their glycine/taurine conjugates) was studied, both after 1 day and 3 days in culture, in three batches of SCRH, at concentrations around the in vitro NOAEL for these BAs (5–100 lM). For this part of the study, a 4 h exposure period (rather than 24 h used for toxicity assessments) was selected to achieve low/moderate toxicity but avoid excessive cell death as this could mask the hepatic BA disposition processes. When ‘‘total’’ accumulation of BAs (unconjugated + glycine and taurine conjugated BAs) was compared between day-1 and day-3 SCRH (Fig. 2A and E), no statistically signiﬁcant differences were found. This may appear contrary to the previously reported culture time-dependent decrease in uptake transporter activity and simultaneous increase in sinusoidal efﬂux transporter activity (Liu et al., 1998; Tchaparian et al., 2011; De Bruyn et al., 2013). Moreover, accumulation values of the unconjugated BA species CDCA and DCA were actually similar or even higher in day-3 versus day-1 SCRH (Fig. 2B and F). This indicates that CDCA and DCA do not play a direct role in the culture time-dependent decrease in toxicity (Fig. 1), which was observed when SCRH were exposed to these BAs. While (C)DCA accumulation did not differ signiﬁcantly with culture time, the intracellular accumulation of the conjugates was signiﬁcantly higher in day-1 compared to day-3 SCRH. In addition, the intracellular levels of the taurine conjugates were about 10-times lower than the glycine conjugates. Taking into account that the glycine conjugates are the most toxic BA species, the substantially higher intracellular levels of glycine conjugates of CDCA
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
Fig. 7. Time-dependent accumulation of CDCA and its conjugates: Compositional bar charts for absolute amounts (nmol/cm2) of CDCA and its conjugates GCDCA and TCDCA in cells and bile networks in case of day-3 SCRH, after incubating the same batch of day-1 (column A) or day-3 (column B) SCRH (in 6-well plates) with 5, 25, 50 and 100 lM of CDCA for 30 min or 4 h, followed by measurement of intracellular accumulation. Data represents mean values of triplicate measurements for each time point.
and DCA in day-1 compared to day-3 cultures (Fig. 2C and G) suggest, that these glycine conjugates play a key role in the observed culture time-dependent toxicity when SCRH were exposed to (C)DCA. The reduced intracellular glycine conjugate levels in day-3 SCRH (Fig. 2C and G) may result from: (i) culture time-dependent decrease in overall formation of glycine conjugates, (ii) increased basolateral efﬂux of these conjugated BAs in day-3 cultures and/ or (iii) increased canalicular excretion of the BAs in day-3 SCRH as compared to day-1 SCRH (Fig. 8). To delineate between these three mechanisms, we ﬁrst addressed the difference in glycine conjugation activity across culture days. Total amounts of glycine
conjugates formed in SCRH were assessed by measuring combined BA levels in both culture media and hepatocytes (and including bile networks, in case of day-3 SCRH) (Fig. 3A and C). These data illustrate that the total amounts of formed GCDCA and TCDCA were higher in day-1 compared to day-3 SCRH when extracellular CDCA concentrations above 25 lM were applied. A very similar pattern was observed when only medium conjugate concentrations (instead of total amounts) were plotted against intracellular CDCA concentration (Fig. 3B and D). It can be safely inferred that the medium concentrations of the conjugates primarily result from the sinusoidal efﬂux that occurred during the 4 h incubation period. The presence of conjugates in both day-1 and day-3
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
Fig. 8. Schematic representation of culture time-dependent BA handling by SCRH: day-1 and day-3 SCRH were incubated with 5–100 lM (C)DCA, following measurements of the unconjugated and glycine and taurine conjugate levels. Toxicity of (C)DCA was assessed in day-1 and day-3 SCRH. In day-1 SCRH, (C)DCA is taken up inside the hepatocytes by uptake transporters (Oatp/Ntcp) present in the sinusoidal membrane. Once inside the hepatocytes, BAs can undergo conjugation with glycine and taurine (1) to form the respective glycine and taurine conjugates. The conjugates can be transported back into the medium via transporters Mrp3/4, Osta/b present in the sinusoidal membrane (2). Intracellular accumulation of glycine conjugates (4) results in toxicity (5). In day-3 SCRH, development of bile canaliculi facilitates the biliary excretion (3) of accumulated BAs. Decreased conjugation in day-3 SCRH leads to decreased intracellular exposure of SCRH to G(C)DCA and a consequential decrease in toxicity of unconjugated BAs ((C)DCA) in day-3 SCRH.
culture medium is consistent with the established role of sinusoidal efﬂux transporters (such as Mrp 3/4, Ost a/b) in transporting the BAs across basolateral membrane (Donner and Keppler, 2001; Soroka et al., 2001; Boyer et al., 2006). For both culture days, sinusoidal efﬂux of the conjugates was high, reaching up to 90% of the total amounts formed after 4 h. As a consequence, the higher glycine conjugate formation in day-1 versus day-3 SCRH was also closely reﬂected in the medium concentration proﬁles (Fig. 3B and D). This implies that lower intracellular glycine conjugate levels in day-3 SCRH cannot be explained in terms of a culture time-dependent increase in sinusoidal efﬂux. It is noteworthy that the culture time-dependent decrease in glycine conjugate levels was observed for CDCA incubation concentrations above 25 lM. This
corresponds to the concentration range around which CDCA becomes toxic in day-1 (but not in day-3) SCRH (Fig. 1). Therefore, decreased formation of glycine conjugates, but not increased sinusoidal efﬂux, in day-3 cultures appears to be the major mechanism explaining reduced glycine conjugate levels in day-3 cultures as shown in Fig. 2C and G. Further to BA metabolism by conjugation and sinusoidal efﬂux, biliary excretion of BAs will also contribute to hepatic BA elimination. Intracellular and intracanalicular BA accumulation were determined via differentially treating the hepatocytes with standard and Ca++/Mg++-free buffer, as explained in Sections 2 and 3 (Liu et al., 1999; Annaert et al., 2001; Annaert and Brouwer, 2005). As illustrated in Fig. 4, CDCA and DCA showed extensive biliary excretion in a selected batch of day-3 SCRH. GDCA showed non-saturable biliary excretion (Fig. 4C). In contrast, clear saturation of biliary GCDCA excretion was observed for intracellular CDCA concentrations between 25 and 320 lM (Fig. 4B), which is also consistent with the intracellular and total GCDCA exposure proﬁles in Figs. 2 and 3. Although accumulated BA levels as well as conjugate formation rates were found to be very similar between three batches of SCRH (Fig. 2), considerable interbatch variability was observed with respect to biliary excretion of CDCA, DCA and their conjugates. The inter-batch variability in biliary excretion may result from different factors. First, high extents of sinusoidal BA efﬂux by SCRH, which indirectly affects biliary excretion rates via changes in intracellular concentrations as driving force, is likely to be one of these factors. In day-1 SCRH, 78 (±10)% to 87 (±9)% of the total GCDCA formed is found in the medium, while in day-3 SCRH, GCDCA medium levels represent 89 (±6)% to 95 (±19)% of total GCDCA formation. Another factor potentially contributing to interbatch variability in biliary excretion might involve pulsatile and/or cyclic release of canalicular contents by sandwich-cultured hepatocytes (Brouwer et al., 2007). However, the present data provide little evidence for the contribution of this canalicular leakage pathway to sinusoidal BA levels. Indeed, day-1 absolute medium levels of glycine and taurine conjugates, which can only result from sinusoidal efﬂux (bile canaliculi not developed), signiﬁcantly exceed the observed day-3 medium levels. It follows that contamination of the already lower day-3 medium levels by biliary contents is highly unlikely, unless accompanied by a very pronounced decrease in sinusoidal BA efﬂux in day-3 SCRH as compared to day-1 SCRH. The latter is in sharp contrast with literature data on expression and activity of sinusoidal efﬂux transporters, which is increasing with culture time (Chandra et al., 2005; Zhang et al., 2005; Tchaparian et al., 2011). We therefore conclude that the sinusoidal efﬂux pathway rather than the pulsatile canalicular release provides SCRH with a protective mechanism to cope with rising intracellular and intracanalicular BA levels. Nonetheless, the biliary excretion data obtained for both CDCA and DCA in selected batches of day-3 (Fig. 4) and day-4 (Supplementary Fig. 2) SCRH illustrate that direct biliary excretion of these BAs remains an important elimination mechanism even when intracellular concentrations are rising. In addition, to further conﬁrm the role of the biliary excretion pathway, day-4 SCRH were subjected to a modiﬁed incubation protocol for biliary excretion assessment. This implied determination of the difference between sinusoidal and canalicular efﬂux (rather than accumulation) of BAs. The BA efﬂux data shown in Fig. 5 further support the importance of the biliary excretion pathway for disposition of (C)DCA and their conjugates in SCRH. It should however be noted that quantitative interpretation of the relative importance of sinusoidal versus canalicular BA efﬂux may not be reliable under the present conditions. This is at least partly related to the fact that efﬂux protocol favors canalicular accumulation for 4 h (and subsequent release in 10 min), whereas sinusoidal efﬂux only represents a 10 min incu-
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
bation period. Nevertheless, direct measurement of differential BA efﬂux in addition to differential BA accumulation clearly provides a valid and complementary approach for determining in vitro biliary excretion of these BAs in SCRH. The data obtained in the present study favor sinusoidal BA efﬂux as a dominant elimination pathway in SCRH, which is also consistent with the high basolateral efﬂux of BAs in SCRH as demonstrated previously (Jemnitz et al., 2010). However, it remains to be determined whether these in vitro ﬁndings represent a scenario that can also take place in vivo under conditions of rising BA levels and/or cholestatic conditions. The upregulation of basolateral transporters are reported in case of obstructive cholestasis (Donner and Keppler, 2001). Thus the high basolateral efﬂux-mediated variability in canalicular excretion can be a reﬂection of the in vivo phenomenon. To explain decreased formation of BA conjugates by SCRH with culture time, the mRNA levels of the BA conjugating enzyme BAAT were measured (Falany et al., 1994). The BAAT mRNA levels were expected to decrease with culture time to explain the culture time-dependent conjugate formation. However, the mRNA levels of BAAT in day-3 and day-4 SCRH were about 4 times higher than the levels in day-1 SCRH (Fig. 6). Thus, changes in mRNA levels of BAAT could not explain the culture time-dependent decrease in the formation of the conjugates form the unconjugated BAs. These results illustrated the utility of the present study concept, i.e., measuring enzyme/transporter function in addition to expression levels. As discussed above, disposition of (C)DCA was initially determined after exposure to CDCA or DCA for 4 h (Fig. 3) or 4 h + 10 min (Fig. 2). Interestingly, when incubations with BA lasted for only 30 min, the absolute and relative levels of the taurine conjugates of CDCA were much higher than after 4 h exposure (Fig. 7, Supplementary Figs. 3 and 5). For instance, after exposure of day-1 SCRH to 5 and 25 lM CDCA, TCDCA accounted for about 50% of total intracellular BA exposure (TCDCA > CDCA > GCDCA), whereas TCDCA levels were negligible following the 4 h exposure time. As the incubation medium is composed primarily of Williams’ E Medium, containing 50 mg/L of glycine (but no taurine), it is possible that the intracellular source of taurine was depleted by 4 h, thus leading to increased formation of glycine conjugates. This has been suggested for cholic acid (Rembacz et al., 2010). However, this does not refute our explanation for the observed culture time-dependent toxicity, as the same composition of culture medium was used for day-1 and day-3 SCRH. In addition, it can be inferred that in a (patho)physiological condition, when the plasma glycine/taurine ratio increases, the formation of glycine conjugated BAs (from unconjugated BA) will increase. Glycine conjugates by virtue of being the most toxic BA species, will make the liver susceptible to BA-mediated injury. Interestingly, the decrease in formation of conjugates in day-3 SCRH compared to day-1 SCRH was visible only after 4 h incubation with CDCA (Supplementary Fig. 5). This indicates the importance of adequate incubation times in addressing the toxicity ﬁndings. To our knowledge, this is the ﬁrst time that a direct link is revealed between BA toxicity and intracellularly formed glycine conjugate exposure in sandwich-cultured rat hepatocytes. Toxicity of several BAs in SCRH is culture time-dependent, with toxicity in day-3 being less pronounced than at day-1. The lower intracellular exposure to the glycine conjugates (formed from CDCA and DCA), appears to be a key factor in the culture time-dependent decline in sensitivity of SCRH towards BA-mediated toxicity. The decreased intracellular exposure to glycine conjugates in older cultures can primarily be attributed to a decreased formation of glycine conjugates. The decreased formation of glycine conjugates could not be explained by the culture time-dependent changes in mRNA levels of the conjugating enzyme, BAAT. Assuming that glycine
conjugates play a similar role in BA-mediated toxicity in human liver, our work has diagnostic and therapeutic implications for conditions of intrahepatic cholestasis in human. Conﬂict of Interest There are no conﬂicts of interest. Acknowledgments The present research was made possible by outstanding contributions from awesome friends and colleagues Sarinj Fattah, and Tom De Bruyn. This work was supported by DBOF grant (ID 57610).This study was carried out in part by private funds of Pieter Annaert and Sagnik Chatterjee. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.tiv.2013.10.020. References Alrefai, W.A., Gill, R.K., 2007. Bile acid transporters: structure, function, regulation and pathophysiological implications. Pharm. Res. 24, 1803–1823. Annaert, P.P., Brouwer, K.L.R., 2005. Assessment of drug interactions in hepatobiliary transport using rhodamine 123 in sandwich-cultured rat hepatocytes. Drug Metab. Dispos. 33, 388–394. Annaert, P.P., Turncliff, R.Z., Booth, C.L., Thakker, D.R., Brouwer, K.L.R., 2001. Pglycoprotein-mediated in vitro biliary excretion in sandwich-cultured rat hepatocytes. Drug Metab. Dispos. 29, 1277. Boyer, J.L., Trauner, M., Mennone, A., Soroka, C.J., Cai, S.Y., Moustafa, T., Zollner, G., Lee, J.Y., Ballatori, N., 2006. Upregulation of a basolateral FXR-dependent bile acid efﬂux transporter OSTa-OSTb in cholestasis in humans and rodents. Am. J. Physiol. Gastr. Liver Physiol. 290, G1124–G1130. Brouwer, K., 2010. In Vitro to In Vivo Translation Hepatic Intracellular Volume. Brouwer, K., Tian, X., Zhang, P., Hoffmaster, K., 2007. Pulsing of Bile Compartments in Sandwich-Cultured Hepatocytes. WO Patent WO/2007/146, 203. Chandra, P., Zhang, P., Brouwer, K.L.R., 2005. Short-term regulation of multidrug resistance-associated protein 3 in rat and human hepatocytes. Am. J. Physiol. Gastr. Liver Physiol. 288, G1252–G1258. Dabos, K.J., Nelson, L.J., Hewage, C.H., Parkinson, J.A., Howie, A.F., Sadler, I.H., Hayes, P.C., Plevris, J.N., 2004. Comparison of bioenergetic activity of primary porcine hepatocytes cultured in four different media. Cell Transplant. 13, 213–229. Dawson, P.A., Lan, T., Rao, A., 2009. Bile acid transporters. J. Lipid Res. 50, 2340. De Bruyn, T., Chatterjee, S., Fattah, S., Keemink, J., Nicolaï, J., Augustijns, P., Annaert, P., 2013. Sandwich-cultured hepatocytes: utility for in vitro exploration of hepatobiliary drug disposition and drug-induced hepatotoxicity. Expert Opin. Drug Metab. Toxicol. 9, 589–616. Donner, M.G., Keppler, D., 2001. Up-regulation of basolateral multidrug resistance protein 3 (Mrp3) in cholestatic rat liver. Hepatology 34, 351–359. Draheim, V., Reichel, A., Weitschies, W., Moenning, U., 2010. N-glycosylation of ABC transporters is associated with functional activity in sandwich-cultured rat hepatocytes. Eur. J. Pharm. Sci. 41, 201–209. Falany, C.N., Johnson, M.R., Barnes, S., Diasio, R.B., 1994. Glycine and taurine conjugation of bile acids by a single enzyme. Molecular cloning and expression of human liver bile acid CoA: amino acid N-acyltransferase. J. Biol. Chem. 269, 19375. Fischer, S., Beuers, U., Spengler, U., Zwiebel, F.M., Koebe, H.G., 1996. Hepatic levels of bile acids in end-stage chronic cholestatic liver disease. Clin. Chim. Acta 251, 173–186. Gibson, W.J., Roques, T.W., Young, J.M., 1994. Modulation of antagonist binding to histamine H1-receptors by sodium ions and by 2-amino-2-hydroxymethylpropan-1, 3-diol HCl. Br. J. Pharmacol. 111, 1262. Goeyens, L., Kindermans, N., Yusuf, M.A., Elskens, M., 1998. A room temperature procedure for the manual determination of urea in seawater. Estuar Coast Shelf Sci. 47. Hillaire, S., Ballet, F., Franco, D., Setchell, K.D.R., Poupon, R., 1995. Effects of ursodeoxycholic acid and chenodeoxycholic acid on human hepatocytes in primary culture. Hepatology 22, 82–87. Jemnitz, K., Veres, Z., Vereczkey, L., 2010. Contribution of high basolateral bile salt efﬂux to the lack of hepatotoxicity in rat in response to drugs inducing cholestasis in human. Toxicol. Sci. 115, 80–88. Jigorel, E., Le Vee, M., Boursier-Neyret, C., Bertrand, M., Fardel, O., 2005. Functional expression of sinusoidal drug transporters in primary human and rat hepatocytes. Drug Metab. Dispos. 33, 1418–1422.
S. Chatterjee et al. / Toxicology in Vitro 28 (2014) 218–230
Kullak-ublick, G.A., Stieger, B., Meier, P.J., 2004. Enterohepatic bile salt transporters in normal physiology and liver disease. Gastroenterology 126, 322–342. Li, N., Bi, Y.-A., Duignan, D.B., Lai, Y., 2009. Quantitative expression proﬁle of hepatobiliary transporters in sandwich cultured rat and human hepatocytes. Mol. Pharm. 6, 1180–1189. Liu, X., Brouwer, K.L.R., Gan, L.S.L., Brouwer, K.R., Stieger, B., Meier, P.J., Audus, K.L., LeCluyse, E.L., 1998. Partial maintenance of taurocholate uptake by adult rat hepatocytes cultured in a collagen sandwich conﬁguration. Pharm. Res. 15, 1533–1539. Liu, X., LeCluyse, E.L., Brouwer, K.R., Gan, L.S.L., Lemasters, J.J., Stieger, B., Meier, P.J., Brouwer, K.L.R., 1999. Biliary excretion in primary rat hepatocytes cultured in a collagen-sandwich conﬁguration. Am. J. Physiol. Gastr. Liver Physiol. 277, G12– G21. Martinez-Diez, M.C., Serrano, M.A., Monte, M.J., Marin, J.J.G., 2000. Comparison of the effects of bile acids on cell viability and DNA synthesis by rat hepatocytes in primary culture. Biochim. Biophys. Acta 1500, 153–160. Noel, G., Vee, M.L., Moreau, A., Stieger, B., Parmentier, Y., Fardel, O., 2013. Functional expression and regulation of drug transporters in monolayer-and sandwichcultured mouse hepatocytes. Eur. J. Pharm. Sci.. Patel, T., Bronk, S.F., Gores, G.J., 1994. Increases of intracellular magnesium promote glycodeoxycholate-induced apoptosis in rat hepatocytes. J. Clin. Invest. 94, 2183. Pauli-Magnus, C., Meier, P.J., Stieger, B., 2010. Genetic determinants of drug-induced cholestasis and intrahepatic cholestasis of pregnancy. Semin. Liver Dis. 30, 147– 159. Perez, M.J., Briz, O., 2009. Bile-acid-induced cell injury and protection. World J. Gastroenterol. 15, 1677. Poll, D.V., Sokmensuer, C., Ahmad, N., Tilles, A.W., Berthiaume, F., Toner, M., Yarmush, M.L., 2006. Elevated hepatocyte-speciﬁc functions in fetal rat hepatocytes co-cultured with adult rat hepatocytes. Tissue Eng. 12, 2965–2973. Rembacz, K.P., Woudenberg, J., Hoekstra, M., Jonkers, E.Z., Van Den Heuvel, F.A.J., Buist-Homan, M., Woudenberg-Vrenken, T.E., Rohacova, J., Marin, M.L., Miranda, M.A., Moshage, H., Stellaard, F., Faber, K.N., 2010. Unconjugated bile salts shuttle through hepatocyte peroxisomes for taurine conjugation. Hepatology 52, 2167–2176. Rodríguez-Garay, E.A., 2003. Cholestasis: human disease and experimental animal models. Ann. Hepatol. 2, 150–158. Rolo, A.P., Palmeira, C.M., Wallace, K.B., 2002. Interactions of combined bile acids on hepatocyte viability: cytoprotection or synergism. Toxicol. Lett. 126, 197–203.
Rolo, A.P., Palmeira, C.M., Wallace, K.B., 2003. Mitochondrially mediated synergistic cell killing by bile acids. Biochim. Biophys. Acta. 1637, 127–132. Rolo, A.P., Palmeira, C.M., Holy, J.M., Wallace, K.B., 2004. Role of mitochondrial dysfunction in combined bile acid-induced cytotoxicity: the switch between apoptosis and necrosis. Toxicol. Sci. 79, 196. Scherer, M., Gnewuch, C., Schmitz, G., Liebisch, G., 2009. Rapid quantiﬁcation of bile acids and their conjugates in serum by liquid chromatography–tandem mass spectrometry. J. Chromatogr. B 877, 3920–3925. Schmucker, D.L., Ohta, M., Kanai, S., Sato, Y., Kitani, K., 1990. Hepatic injury induced by bile salts: correlation between biochemical and morphological events. Hepatology 12, 1216–1221. Soroka, C.J., Lee, J.M., Azzaroli, F., Boyer, J.L., 2001. Cellular localization and upregulation of multidrug resistance-associated protein 3 in hepatocytes and cholangiocytes during obstructive cholestasis in rat liver. Hepatology 33, 783– 791. Spivey, J.R., Bronk, S.F., Gores, G.J., 1993. Glycochenodeoxycholate-induced lethal hepatocellular injury in rat hepatocytes. Role of ATP depletion and cytosolic free calcium. J. Clin. Invest. 92, 17. Stieger, B., Fattinger, K., Madon, J., Kullak-Ublick, G.A., Meier, P.J., 2000. Drug-and estrogen-induced cholestasis through inhibition of the hepatocellular bile salt export pump (Bsep) of rat liver. Gastroenterology 118, 422–430. Tchaparian, E.H., Houghton, J.S., Uyeda, C., Grillo, M.P., Jin, L., 2011. Effect of culture time on the basal expression levels of drug transporters in sandwich-cultured primary rat hepatocytes. Drug Metab. Dispos. 39, 2387–2394. Tribe, R.M., Dann, A.T., Kenyon, A.P., Seed, P., Shennan, A.H., Mallet, A., 2009. Longitudinal proﬁles of 15 serum bile acids in patients with intrahepatic cholestasis of pregnancy. Am. J. Gastroenterol. 105, 585–595. Turncliff, R.Z., Tian, X., Brouwer, K.L.R., 2006. Effect of culture conditions on the expression and function of Bsep, Mrp2, and Mdr1a/b in sandwich-cultured rat hepatocytes. Biochem. Pharmacol. 71, 1520–1529. Uhal, B.D., Roehrig, K.L., 1982. Effect of dietary state on hepatocyte size. Biosci. Rep. 2, 1003–1007. van Mil, S.W.C., van der Woerd, W.L., van der Brugge, G., Sturm, E., Jansen, P.L.M., Bull, L.N., van den Berg, I.E.T., Berger, R., Houwen, R.H.J., Klomp, L.W.J., 2004. Benign recurrent intrahepatic cholestasis type 2 is caused by mutations in ABCB11. Gastroenterology 127, 379–384. Zhang, P., Tian, X., Chandra, P., Brouwer, K.L.R., 2005. Role of glycosylation in trafﬁcking of Mrp2 in sandwich-cultured rat hepatocytes. Mol. Pharmacol. 67, 1334–1341.