Abstract
Contrast-enhanced magnetic resonance imaging (CE-MRI) is a valuable technique for the diagnosis of liver diseases. As gadocoletic acid trisodium salt (B22956/1), a new contrast agent showing high biliary excretion, may be potentially advantageous in hepatobiliary imaging, the aim of the study was to investigate the molecular mechanisms of hepatic transport of the B22956 ion in a cellular model of hepatic tumor. B22956 ion uptake was measured in tumoral (HepG2) and nontumoral (Chang liver) hepatic cell lines. Absolute quantitative real-time reverse transcriptase (RT)-polymerase chain reaction (PCR) analyses, using cloned PCR products as standards, were performed on total RNA of both cell lines and normal liver to evaluate the transcription of 12 transport genes: SLCO1A2, SLCO2B1, SLCO1B1, SLCO3A1, SLCO4A1, SLCO1B3, SLC22A7, SLC22A8, SLC22A1, SLC10A1, SLC15A1, and SLC15A2. B22956 transport was more efficient in Chang liver than in HepG2 cells and was inhibited by cholecystokinin-8, a specific substrate of OATP1B3. Real-time RT-PCR analyses revealed different transcription profiles in the tumoral and nontumoral cell lines. Compared with normal liver, the expression of SLCO1B1, SLCO3A1, and SLCO1B3 was greatly repressed in HepG2 cells, whereas SLCO2B1, SLC22A7, and SLC22A8 expression was either maintained or increased. On the contrary, in Chang liver cells, SLC22A7 and SLC22A8 genes were undetectable, whereas the expression of SLCO3A1, SLCO4A1, and SLCO1B3 was similar to normal liver. Transport studies and gene expression analyses indicated that B22956 ion is a good substrate to the liver-specific OATP1B3, reported to be poorly expressed or absent in human liver tumors. Therefore, B22956 may be helpful in detecting hepatic neoplastic lesions by CE-MRI.
Magnetic resonance imaging (MRI) produces high-resolution three-dimensional maps delineating morphological features of the tissues analyzed. The specificity of MRI can be further increased by using exogenous contrast agents such as the gadolinium chelates (Artemov, 2003). The development of molecular targeted MRI contrast agents directed to specific tissue entities could dramatically expand the range of MRI applications (Lorusso et al., 2005). Because of the intrinsically low sensitivity of MRI, high local concentrations of the contrast agents at the target site are required to evoke effective magnetic resonance contrast. This means that the contrast agent should recognize targeted cells with high affinity and sensitivity (Weinmann et al., 2003). Several compounds have been studied as potential hepatospecific contrast agents, and only a few of them are in the late-phase clinical trials or have received marketing authorization. These agents include hydrophilic paramagnetic chelates with specific uptake in the hepatocytes, such as Teslascan (mangafodipir trisodium; GE Healthcare, Chalfont St. Giles, UK), MultiHance [gadobenate dimeglumine (Gd-BOPTA); Bracco Imaging S.p.A., Milan, Italy], and [gadoxetic acid disodium [Gd-EOB-DTPA, Primovist); Schering AG, Berlin, Germany] (Weinmann et al., 2003). These three agents are characterized by a rather high hepatic uptake and a rapid biliary excretion, although the primary route of elimination is renal.
Recently, gadocoletic acid trisodium salt (B22956/1; Bracco Imaging S.p.A.), a new gadolinium-based magnetic resonance contrast agent, has been formulated to be used as a blood pool agent (Cavagna et al., 2002). The compound is a derivative of gadopentetate to which a bile acid-like lateral chain has been linked to increase the binding to the plasma proteins and the permanence in the circulation. Pharmacological studies in animals and in healthy human volunteers have shown that, unlike Gd-BOPTA and Gd-EOB-DTPA, the biliary excretion is the primary route of elimination of the contrastive moiety B22956 ion (La Noce et al., 2002). Recent studies also indicated that the ATP-dependent transporter MRP2 (ABCC2), localized to the hepatocyte bile canaliculus, is responsible for the biliary excretion of B22956, whereas the ubiquitous MRP1 (ABCC1) could be involved in contrasting urinary elimination (Lorusso et al., 2002).
No data have been produced so far on the mechanism for the basolateral hepatic uptake of this compound. Nevertheless its bile acid-like molecular structure leads us to hypothesize that B22956 is a potential substrate for the basolateral organic anion transporting polypeptides (OATPs) of the liver (Meier and Stieger, 2002).
OATPs are a family of transport proteins that are expressed in multiple tissues and involved in the intake or elimination from the body of a wide range of structurally unrelated compounds. Several members of this family have been studied, largely for their role in the hepatic clearance of albumin-bound substrates (Hagenbuch and Meier, 2003). It has been demonstrated that Gd-EOB-DTPA is substrate to rat Oatp1a1 but not to human OATP1A2 (van Montfoort et al., 1999), whereas contrasting data have been presented for Gd-BOPTA. The latter is not a substrate for human OATP1A2, and linear transport kinetics have been proposed in rat liver plasma membrane vesicles (Pascolo et al., 1999), whereas more recent studies provided indirect evidence on a role of OATPs in its hepatic transport (Pastor et al., 2003; Planchamp et al., 2004, 2005).
The most important OATPs in human liver are OATP1B1 (or human liver specific transporter) and OATP1B3, whereas OATP1A2, the first OATP cloned from human liver, seems to be expressed at much lower extent (Jung et al., 2001). OATP1B1 has been reported to be down-regulated during neoplastic transformation in the liver (Kinoshita and Miyata, 2002). It was also shown that OATP1B1 and OATP1B3 are not detectable in HepG2 cells, widely used as a model of human hepatic carcinoma (Cui et al., 2003).
In the present study we report on the assessment of transport studies for B22956 ion in two human liver-derived cell lines: HepG2, used as a model of hepatocellular carcinoma, and Chang liver, used as model of nontumor liver cells (Wirth et al., 1995; Seow et al., 2001a,b). Absolute quantitative RT-PCR analyses were also performed to evaluate the effective expression of 12 drug transport genes (SLCO1A2, SLCO2B1, SLCO1B1, SLCO3A1, SLCO4A1, SLCO1B3, SLC22A7, SLC22A8, SLC22A1, SLC15A1, SLC15A2, and SLC10A1), mainly organic anion, cation, and peptide transporters (van Montfoort et al., 2001, 2003). The gene expression was assessed in the two cell lines and compared with normal liver to understand the transporters potentially involved in the “in vivo” targeting of the contrast agent B22956 and to confirm the applicability of the cell models.
Materials and Methods
Chemicals. B22956/1 is the trisodium salt of a derivative of gadopentetate and its chemical name according to the Chemical Abstracts Service is trisodium [(3β,5β,12α)-3-[[(4S)-4-[bis[2-[bis[(carboxy-κO)methyl]amino-κN]ethyl]amino-κN]-4-(carboxy-κO)-1-oxybutyl]amino]-12-hydroxycholan-24-oato(6-)] gadolinate (3-) (de Haen et al., 2006). In aqueous solution B22956/1 dissociates into sodium ions and the contrastive component B22956 ion that will be indicated as B22956 throughout the text. B22956 (or Gd-B22950) is the code name for the gadolinium complex with the ligand B22950.
[147Pm]Cl3 in HCl was purchased from MP Biomedicals (Irvine, CA) and was used to prepare stock solutions of [147Pm]B22950 with a specific activity of 89.6 mCi/mmol as described previously (Lorusso et al., 2002). Labeled solutions of Gd-B22950 were obtained by adding the appropriate amount of the solution of the 147Pm complex to the solution of the corresponding Gd-B22950. 147Pm and Gd complexes were prepared as sodium salts. The purity of each compound was assessed by high-performance liquid chromatography and was >99%. The 147Pm-labeled compound used for in vitro studies will be referred as labeled B22956 throughout the text.
[3H]Taurocholate (specific activity 3.47 Ci/mmol) was obtained from PerkinElmer Life and Analytical Sciences (Boston, MA). CCK-8 was prepared by peptide synthesis as reported previously (Accardo et al., 2004) and kindly supplied by Department of Chemistry IFM, University of Torino (Torino, Italy). All other chemicals were of the highest commercially available purity.
Uptake in Cell Cultures from Human Liver. HepG2 and Chang liver cell lines were obtained from Istituto Zooprofilattico Sperimentale di Brescia (Italy). The two cell lines were cultured in Dulbecco's modified Eagle's medium with 10% (v/v) fetal bovine serum and 1% antibiotics (10,000 U/ml penicillin and 10 mg/ml streptomycin), under standard conditions. The cells were routinely maintained in 75 cm2 Falcon flasks for 3 days and then were harvested by exposure to a solution of 0.25% trypsin and 0.02% EDTA and transferred onto 35-mm diameter Petri dishes at a density of 5 × 105 cells/cm2. On the next day, the cells were used for the uptake experiments. Media were from Celbio (Milan, Italy) and plastic from Falcon (BD Biosciences, Franklin Lakes, NJ).
All uptake studies were carried out at 37°C. Cells grown on dishes were washed once with 1.5 ml of fresh culture medium, and uptake was initiated by the addition of 1.0 ml of uptake medium consisting of Dulbecco's modified Eagle's medium with 10 to 200 μM concentrations of labeled compounds (labeled B22956 or [3H]taurocholate). After incubation at different time intervals (from 0 to 180 min), uptake was stopped by removing the uptake medium and then by immediately washing the cells three times with 1.5 ml of ice-cold Hanks' solution (Sigma-Aldrich, Milan, Italy). Nonspecific transport (binding) was measured by incubation at 4°C or by short incubations (20 s) at room temperature.
The cells were then solubilized into 1 ml of 2% SDS-0.2 N NaOH, and 500 μl of the cell lysate were added to 10 ml of scintillation liquid (Filtercount; Packard, Groningen, The Netherlands), and radioactivity was counted. Protein content was measured on an aliquot of the cell lysate, using the bicinchoninic acid protein assay (Sigma-Aldrich), and the overall transport was expressed as picomoles of substrate per milligram of protein. The nonspecific binding (mainly 4°C incubations) was proportional to time of incubation and increased linearly at increasing concentration. The linear binding component was subtracted both in time course and Michaelis-Menten graphs where only net transport is shown. Uptake studies were also performed by the addition to the transport media of OATP substrates, such as bromosulfophthalein (BSP), taurocholate (all from Sigma-Aldrich), and CCK-8.
RNA Extraction and Reverse Transcription. Total RNA from cell cultures was extracted using TRI-Reagent (Sigma-Aldrich) following the manufacturer's instructions. The concentration of RNA and its purity were measured spectrophotometrically, and its integrity was checked by assessing the sharpness of ribosomal RNA bands on a 1% agarose gel.
The total RNA from normal human liver tissue was purchased from Ambion (Cambridge, UK), and 1 μg of RNA was transcribed using an iScript kit (Bio-Rad, Hercules, CA) following the instruction of the manufacturer. The cDNA samples were stored at –20°C.
Primers and Standards for Real-Time PCR Analysis. Primers for real-time RT-PCR were designed using Vector NTI Suite software (InforMax, Frederick, MD). One of the primers of each gene spans an exon-exon boundary to avoid amplification of genome DNA. The primer sequences were submitted to a BLAST search to avoid amplification of other than desired sequences. The sequences, access codes to reference sequences, and other specifications for primers are given in Table 1. The specificity of primers was checked by electrophoresis of PCR products on 3% agarose gel. All PCR products migrated as a single band at the expected size. In each run the melting curves of analyzed samples were compared with the melting curves of plasmid standards.
The pCR plasmids (pCR XL-TOPO or pCR 4Blunt-TOPO; Invitrogen, Carlsbad, CA) with inserted PCR products were prepared by using appropriate cloning kits (Invitrogen) as reported previously (Ceckova et al., 2006) and used as standards for real-time PCR analysis. The cDNA from the HepG2 cell line served as the source for PCR amplification of SLCO2B1, SLCO4A1, SLC22A7, SLC22A8, SLC22A1, SLC15A1, and SLC10A1 transcripts. The cDNA from Chang liver was used to generate amplicons of SLCO1B3, SLCO1B1, SLCO3A1, and SLC15A2 transcripts and PCR products; the SLCO1A2 amplicon was generated from normal human liver cDNA.
Successful cloning was tested by PCR by using specific primers for each insert and following electrophoresis of PCR products on 3% agarose gel. The plasmids were isolated from colonies carrying PCR-positive plasmid for each gene using a QIAGEN Plasmid Midi Kit (QIAGEN GmbH, Hilden, Germany). Their concentrations were determined by UV absorbance A260 measurement, and their quality was assessed from the UV absorbance A260/A280 ratio. The plasmids were then sequenced on an ABI PRISM 310 DNA sequencer (Applied Biosystems, Foster City, CA), and the sequences of inserts was compared with the reference sequences from the National Center for Biotechnology Information (NCBI) database. All of the plasmids contained the inserts of desired sequence.
Real-Time RT-PCR Analysis. Real-time PCR was performed using an iCycler (Bio-Rad). For each gene analyzed, 50 ng of reverse-transcribed RNA was amplified using SybrGreen PCR amplification master mix (Bio-Rad), following the manufacturer's instructions. The temperature profile was 95°C for 3 min, 50 times; 95°C for 15 s, 60°C for 15 s, and 72°C for 20 s; melting curve program 72 to 95°C. The calibration curves for each gene amplification were obtained from a series of decimally diluted plasmid standards. All of the calibration curves had the correlation coefficient ≥0.99%. The amplification efficiencies varied between 0.98 and 1.02. The triplicates of cDNA samples were amplified along with the decimally diluted plasmids in quantities 5 × 106 for 50 copies in one run of cycler for each gene. A nontemplate control for all genes was included in each run. The quantities of transcripts in the samples were extrapolated from calibration curves.
Western Blot Analysis. Plasma membrane vesicles of Chang liver, HepG2, and MDCKII cells were obtained, as described previously (Fernetti et al., 2001), by separation of a crude membrane fraction on a discontinuous sucrose gradient (38% w/v and 19% w/v) followed by centrifugation at 200,000g at 4°C for 120 min. The membrane fraction at the 19 to 38% sucrose interface was collected. Membrane preparations were stored in liquid nitrogen in 250 mM sucrose, 20 mM HEPES/Tris, pH 7.4 until use.
Membrane proteins (30 μg) were solubilized in Laemmli buffer, separated on 7.0% SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes by electroblotting, using 25 mM Tris base, 192 mM glycine, and 20% methanol as the transfer solution. The specific monoclonal antibody mMDQ (Cui et al., 2003) against OATP1B3 (and OATP1B1) was purchased from Progen Biotechink (Heidelberg, Germany). Blots were blocked in 4% (w/v) skim milk powder in Tris-buffered saline/Tween 20 and incubated for 1 h at room temperature with the primary antibody diluted in blocking solution. After washing in blocking solution, horseradish peroxidase-conjugates (goat anti-mouse IgG, dilution 1:5,000; Sigma-Aldrich) were added, and blots were incubated for 60 min. After washing five times for 5 min in Tris-buffered saline/Tween 20, immunoreactivity was visualized by the ECL-Plus Western Blotting detection system (Amersham Biosciences, Milan, Italy).
Statistical Analysis. Results are expressed as means ± S.D. of three assays per condition. Statistical analysis was performed using analysis of variance combined with Student's two-tailed t test. Differences among the conditions were considered significant at p < 0.05.
Results
Characterization of Taurocholate Uptake. The transport kinetics of [3H]taurocholate in the two cell lines is shown in Fig. 1. When measured at 50 μM [3H]taurocholate, both Chang liver and HepG2 cells showed a rapid initial uptake of the bile acid (Fig. 1, A and C), with similar steady-state levels after 10 min of incubation. The two cell lines showed similar [3H]taurocholate transport in terms of time course and kinetic constants. The Michaelis-Menten constants for [3H]taurocholate transport were obtained at 1 min of incubation and calculated using a mathematical fit according to the equation: V = [(Vmax × [S])/(Km + [S])] + Dc × [S], in which Dc is the diffusive component of transport. In Chang liver cells (Fig. 1B), the apparent Km was 149.4 ± 57.9 μM, the maximal uptake (Vmax) was 97.3 ± 32.9 pmol/mg protein/min, and the diffusive component was negligible (Dc was 0.0 ± 0.05). Likewise, in HepG2 cells (Fig. 1D), the Km was 174.6 ± 48.1 μM, and the Vmax was 194.5 ± 55.0 pmol/mg protein/min; the diffusive component was also negligible (Dc was 0.0 ± 0.05). These data indicate the presence in both cell lines of one or more carrier-mediated mechanisms for [3H]taurocholate with comparable affinity and capacity. It should be noted, however, that, despite a similar affinity, a significantly greater accumulation of the bile acids was found in HepG2 cells (Vmax HepG2 cells > Vmax Chang liver cells, p = 0.02). The transport of [3H]taurocholate measured in HepG2 cells is in good agreement with that reported previously (Kullak-Ublick et al., 1996; Lee et al., 2001).
Characterization of B22956 Uptake. The transport of B22956 in Chang liver and HepG2 cells was much slower than that of taurocholate. As shown in Fig. 2, B and D, the accumulation of the contrast agent increases linearly over time in the first 60 min of incubation in both cell lines. Because the uptake was linear in this time frame, incubation times of 15 and 20 min were chosen to measure initial uptake in HepG2 and Chang liver cells, respectively, and various concentrations were tested to evaluate kinetic constants. In Chang liver cells, B22956 uptake was saturative with a Km of 369 ± 138 μM, whereas the Vmax was 3270 ± 440 pmol/mg protein/20 min (Fig. 2A). The uptake of the contrast agent was also saturative in HepG2 cells (Fig. 2C) with an apparent affinity (Km value of 377 ± 47 μM) similar to that found in Chang liver cells but with a significantly lower Vmax (704 ± 41 pmol/mg protein/15 min). The overall efficiency transport, expressed as the ratio of Vmax/Km was ∼3.5 times higher in Chang liver than in HepG2 cells.
As shown in Table 2, B22956 transport experiments were performed in the presence of organic anions known to be substrates for OATP transporters (Hagenbuch and Meier, 2003). When transport was evaluated at a low B22956 concentration (50 μM), the presence of equimolar concentrations of either taurocholate or BSP reduced the uptake by 50% in Chang liver cells but not in HepG2 cells. In contrast, the addition of CCK-8 (20 μM), the OATP1B3 substrate (Ismair et al., 2001), reduced by >80% B22956 uptake both in Chang liver and HepG2 cells. To better define the inhibitory effect of taurocholate and CCK-8, experiments were repeated at a higher B22956 concentration (200 μM). In Chang liver cells, low taurocholate concentrations (10 and 50 μM) reduced B22956 transport by 30 to 40%, whereas CCK-8 (20 and 50 μM) confirmed a strong inhibitory effect, reducing transport by ≥60%. CCK-8 but not taurocholate was still effective in reducing the contrast agent transport in HepG2 cells, although the effect was rather weak with a decrease of 20 to 25% over the control.
The inhibition studies indicate that OATP proteins are involved in B22956 transport in both cell lines, with a crucial role of OATP1B3 suggested by the consistent reduction observed in the presence of low CCK-8 concentration, particularly in Chang liver cells. Of notice was the observation that transport of 50 μM taurocholate was not inhibited by the presence of B22956 (100 μM) and CCK-8 (20 μM), either in Chang liver or in HepG2 cells (data not shown), suggesting that several transporters contribute to the cell accumulation of taurocholate or B22956 and points to partial sharing of different mechanisms.
Quantitative RT-PCR Analyses. To determine which transporters may mediate B22956 transport in Chang liver and HepG2 cells and which may account for the different accumulation, we tested for the presence of various organic anion transporter transcripts from the OATP family (namely SLCO1A2, SLCO2B1, SLCO1B1, SLCO3A1, SLCO1B3, and SLCO4A1) and the OAT family (SLC22A7 and SLC22A8), which are reported to be expressed in liver cells (Hagenbuch and Meier, 2003). To better understand the pharmacological meaning of the data, transporters other than organic anion transporting polypeptides were also analyzed. In particular, we assessed the expression of the sodium dependent taurocholate cotransporter SLC10A1, the organic cation transporter SLC22A1 and the peptide transporters SLC15A1 and SLC15A2 (van Montfoort et al., 2003). The expressions of the selected transporters were analyzed in Chang liver and HepG2 cells and compared with the levels found in a human liver tissue from a healthy donor.
The results shown in Fig. 3 represent a comparison of the gene expression of the different genes in both cell lines, expressed as log units of transcripts per microgram of total RNA. The level of transcripts of any gene did not exceed 107 copies/μg of RNA, and the reported relative expression corresponds with localization on the scale 0 to 7 logs. The data indicate that HepG2 and Chang liver cells express almost the same amount of SLCO4A1 and SLC15A2 and the liver-specific SLCO1B1 genes. Significantly higher expression of SLCO2B1 was observed in HepG2 cells compared with Chang liver cells (Table 3). Conversely, the transcripts of SLCO3A1 and SLCO1B3 were more abundant in Chang liver cells, whereas the expression of SLCO1A2 has not been detected in any cell line. The transcripts for SLC22A7, SLC15A1, SLC10A1, and SLC22A8 were detected only in HepG2 cells (Table 3).
Real-time PCR analysis confirmed the expression of all of the observed genes in normal human liver, with the exception of SLC22A8 (Fig. 3). Differently from what was reported in cell lines (Lee et al., 2001) and similarly to what was reported in tissues (Briz et al., 2003), SLCO1A2 was significantly expressed in normal liver. This finding seems to indicate that OATP1A2 could be not specific for the hepatocyte but most probably for other less abundant cell types in liver. The other investigated OATPs were expressed at high and quite comparable levels. This result is particularly true for the so-called liver-specific SLCO1B1 that is found at the greatest expression level, second only to that registered for SLC22A1.
As shown in Fig. 3, HepG2 cells mimic the expression profile of the normal liver in a larger number of transporters than Chang liver cells. In line with previous reports (Lee et al., 2001), HepG2 cells do not express significant levels of SLCO1B1 and SLCO1B3, two hepatospecific transporters that seem to be greatly repressed in liver tumors (Cui et al., 2003).
The expression of SLCO2B1, SLCO4A1, SLC22A7, SLC22A1, SLC10A1, SLC15A1, and SLC15A2 is maintained in HepG2 cells with detectable SLC22A8. Chang liver cells resemble normal liver in the expression of a lower number of transporters and of particular interest is the maintained expression of SLCO1B3, SLCO4A1, and SLCO3A1. This kind of cells misses completely the expression of SLC22A7, SLC22A8, and SLC10A1 genes.
Western Blot Analyses for OATP1B3 Protein.Figure 4 shows the results of immunoblot experiments to test the expression of OATP1B3 on plasma membrane vesicles of Chang liver and HepG2 cells. Canine kidney-derived MDCKII cells were used as nonhuman controls. The commercial monoclonal antibody recognized OATP1B3 at the specific band of 120 kDa only in Chang liver cells. Additional signals at lower molecular mass are also present in these cells. These bands seem related to the OATP1B3 core-glycosylated (90 kDa) (Cui et al., 2003) or degraded forms more than to unspecific binding. The possible cross-reactivity with OATP1B1 (apparent molecular mass of 90 kDa), which is reported for the antibody (Cui et al., 2003), should be also considered, despite the very low mRNA expression level in both hepatic cell lines. Although no signal was detected in MDCK II cells, in HepG2 cells only faint bands at low molecular weight have been revealed and are consistent with a very marginal presence of differently glycosylated forms of OATP1B3 or unspecific detection.
Discussion
Pharmacological studies of new and old drugs take great advantage of the actual genetic knowledge on the cell membrane transport mechanisms, particularly in liver, kidney, and physiological barriers such as the blood-brain barrier and placenta. This is the case for the contrast media for MRI, which is a class of diagnostics formulated to remain in the bloodstream for the time interval needed to acquire images, to not accumulate in organs, and to be rapidly cleared by kidney or, in specific cases, by liver (Lorusso et al., 2005). When liver-specific MRI contrast agents were studied, biochemical and pharmacological studies were useful to demonstrate that two media (Gd-BOPTA and Gd-EOB-DTPA) currently used in clinical practice are substrates to the known canalicular transport system MRP2 (ABCC2) (de Haen et al., 1996). Although the biliary excretion of Gd-BOPTA and Gd-EOB-DTPA in rats ranges from 50 to 70% of the administered dose (Lorusso et al., 2005), in humans the biliary excretion of the two contrast agents is much lower (Lorusso et al., 2005). The reasons for such great differences may be explained by considering that the hepatic clearance of a drug is conditioned by the uptake mechanisms driving the passage of the compounds from blood into the hepatocytes. This reasoning is supported by the evidence that Gd-EOB.DTPA is a substrate to the rat Oatp1a1 but not to the nearest human ortholog OATP1A2 (Pascolo et al., 1999).
Previous studies indicated that B22956/1 is also a potential hepatospecific contrast agent, because its biliary excretion is much higher than that of Gd-EOB.DTPA and Gd-BOPTA, although mainly mediated by the MRP2 transporter (Lorusso et al., 2002). No data have been produced until now on the basolateral uptake mechanism in liver for this compound.
In this study we demonstrate that B22956 is a substrate to OATP proteins and, in particular, OATP1B3 is involved in the hepatic accumulation of the compound. Because the lateral chain of this new Gd complex derived from the DTPA backbone is a bile acid, it was important to compare transport kinetics of B22956 with those of taurocholate. The present results demonstrate that these two anions share only some transport mechanisms, possibly those mediated by the OATPs.
The transport kinetics of taurocholate were similar in HepG2 and Chang liver cells and comparable with those described previously (Lee et al., 2001). The taurocholate uptake in HepG2 cells seems to be mainly related to the OATP4A1 as suggested by the high transcript levels of its gene SLCO4A1, compared with other transporters, and high affinity to taurocholate (Hagenbuch and Meier, 2003), although the additional role of the very low expressed SLCO1B3 and of the relevant SLC22A7 (Ugele et al., 2003) must be considered. The OATP2B1 involvement is excluded because taurocholate is not substrate to this transporter (Meier and Stieger, 2002). The role of NTCP should also be marginal considering the low transcript levels of its gene SLC10A1 and, more important, previous studies demonstrating the absence of NTCP protein expression in HepG2 cells (Lee et al., 1996; Glasova et al., 2002). In Chang liver cells, taurocholate uptake is similar to that found in HepG2 cells. From the quantitative analyses of the transcripts of the different transporters in Chang liver cells, it can be concluded that OATs (SLC22A7 and SLC22A8) and NTCP (SLC10A1) proteins do not contribute to taurocholate accumulation as their transcripts are absent. On the contrary, the mRNA of SLCO3A1, SLCO4A1, and SLCO1B3 is present, and all of the corresponding transporters may potentially contribute to the uptake of the bile acid.
The characteristics of the B22956 uptake are different in the two cell lines. Although the apparent Km values of the kinetics were comparable, the Vmax found in Chang liver cells was significantly higher than that in HepG2 cells (3270 ± 440 pmol/mg protein/20 min versus 704 ± 41 pmol/mg protein/15 min, respectively). The inhibition studies (Table 2) clearly indicate that the molecular mechanisms involved are different with only a partial overlap with those accounting for the transport of taurocholate. Taurocholate was effective in inhibiting B22956 uptake in Chang liver cells, whereas it was ineffective in HepG2 cells. In addition, the transport of B22956 in HepG2 cells was not inhibited by BSP, whereas the cholephilic anion showed a clear inhibition in Chang liver cells. The uptake at low concentrations of B22956 in HepG2 was affected by CCK-8, a high-affinity substrate for OATP1B3. In Chang liver cells, SLCO1B3 expression was 40 times higher than that in HepG2, and in Western blot analyses, the OATP1B3 protein was clearly detected only in the nontumoral hepatic cells (Fig. 4). This difference may explain why CCK-8 inhibits B22956 uptake by 60% at a CCK-8/B22956 ratio of 1:10 (Table 3). In contrast with HepG2 cells, taurocholate inhibited B22956 uptake in Chang liver cells, at low and high concentrations of the contrast agent, with the maximal effect at equimolar concentration. Because taurocholate is transported with different affinities by various OATPs (Kullak-Ublick et al., 2000), it is not surprising that, on the other side, B22956 was unable to inhibit taurocholate uptake (data not shown).
Taken together, these data show that B22956 is transported by OATP1B3 (SLCO1B3) with a relatively high affinity, as demonstrated by the inhibition studies. Other transporters such as OATP3A1 (SLCO3A1) and OATP4A1 (SLCO4A1) seem to be involved in Chang liver cells, although with lower affinity. In HepG2 cells, the low gene expression of SLCO1B3 and the nearly null presence of the protein OATP1B3 demonstrated in Western blot experiments could seem anyway sufficient to account for some B22956 transport, if considering the results of the inhibition studies. To explain B22956 transport, it should be considered the additional role of unknown transport mechanisms also shared by CCK-8 and that of OATP2B1 (SLCO2B1) and OAT2 (SLC22A7).
These results may have a great practical relevance considering that the compound B22956 is an MRI contrast agent already in the process of being used in human diagnostics. The demonstration that B22956 is taken up by hepatic cells by carrier-mediated systems and that differently expressed transporters correspond to a different accumulation points to a possible use of the contrast agent to discriminate diseased from healthy cells. In particular, it has been reported that OATP1B1 and OATP1B3 expression is greatly repressed in hepatic tumors. Because OATP1B3, whose gene SLCO1B3 is almost exclusively expressed in the liver (Ismair et al., 2001), is involved in the hepatic accumulation of B22956, it may be expected that hepatic tumors have a lower accumulation of this contrast agent.
HepG2 and Chang liver cells are the models of hepatic cells used most often. Whereas Chang liver cells are considered an in vitro model of nonmalignant liver, HepG2 cells are a widely used cellular system for well-differentiated hepatocarcinoma. No comparison has been reported so far in terms of drug accumulation and expression of basolateral transporters on different cellular models for liver disease. In the present study, we report the comparison of the uptake and the expressions at the RNA level of different transport proteins to better characterize the two widely used cell lines.
As shown in Fig. 3, Chang liver cells retain high levels of expression of only some OATPs but lose or greatly reduce the expression of SLC22A7, SLC10A1, and SLC15A1, which are reported to be abundant in normal liver (Alcorn et al., 2002). The analysis of HepG2 cells suggests that during tumorigenesis not only are SLCO1B3 and SLCO1B1 repressed, as already reported (Cui et al., 2003; Le Vee et al., 2006), but also SLCO3A1 and SLC10A1 expression may be affected. Quite surprisingly, OATs (SLC22A7 and SLC22A8) expression is maintained and possibly increased compared with that in normal liver. All of these results are likely to be useful for therapeutic considerations for various families of drugs.
Footnotes
-
This work was supported by grants from Ministero dell'Istruzione, Università e Ricerca (FIST 2002/359Ric to C.T.), an Erasmus fellowship (to A.L.), Fondazione Cassa di Risparmio di Trieste (FCRT-01); Fondo Studio Fegato-ONLUS, and career development grants from Bracco Imaging S.p.A. (to C.F. and L.P.
-
Part of this work was presented at the Federation of European Biochemical Societies Congress and International Union of Biochemistry and Molecular Biology; 2005 July 2–7; Budapest, Hungary. Faculty of Pharmacy, Charles University in Prague, Heyrovského, Hradec Králové, Czech Republic.
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.106.106591.
-
ABBREVIATIONS: MRI, magnetic resonance imaging; GD, BOPTA, gadobenate dimeglumine; Gd-EOB-DTPA, gadoxetic acid disodium; B22956/1, gadocoletic acid trisodium salt, trisodium [(3β,5β,12α)-3-[[(4S)-4-[bis[2-[bis[(carboxy-κO)methyl]amino-κN]ethyl]amino-κN]-4-(carboxy-κO)-1-oxybutyl]amino]-12-hydroxycholan-24-oato(6-)] gadolinate (3-); MRP, multidrug resistance related protein; OATP, organic anion-transporting polypeptide; RT, reverse transcriptase; PCR, polymerase chain reaction; CCK-8, cholecystokinin octapeptide; BSP, bromosulfophthalein; OAT, organic anion transporter; NTCP, sodium-dependent taurocholate cotransporting polypeptide.
- Received April 20, 2006.
- Accepted August 7, 2006.
- The American Society for Pharmacology and Experimental Therapeutics