Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Chlorogenic acid (CHL) derivatives were recently identified as novel, potent, and specific inhibitors of the glucose 6-phosphate (Gl-6-P) translocase (Arion et al., 1997, 1998; Hemmerle et al., 1997; Schindler et al., 1998). Gl-6-P translocase is an essential component of the hepatic glucose-6-phosphatase (Gl-6-Pase) system (Arion et al., 1975) mediating access of Gl-6-P to the lumen of the endoplasmic reticulum. Pharmacodynamic studies in isolated perfused rat liver and in vivo demonstrated that the CHL derivative S 3483 caused inhibition of both glucose-producing pathways, gluconeogenesis and glycogenolysis, by interference at the level of the Gl-6-Pase system (Herling et al., 1998). Pharmacological modulation of hepatic glucose production by inhibition of the Gl-6-Pase system is a new experimental approach for the treatment of type 2 diabetes, where inappropriately increased rates of hepatic glucose production are present (DeFronzo, 1988; Reaven, 1997). The increased hepatic glucose production contributes to the elevated blood glucose concentrations, a well known surrogate parameter in diabetes.

Because the liver is the major target organ for the pharmacological action of CHL derivatives, knowledge about the hepatic uptake mechanism may significantly contribute to the design of new drugs and might further facilitate the understanding of the pharmacokinetic-pharmacodynamic relationship.

The recent molecular identification and subsequent cloning of hepatic transport proteins resulted in a more detailed understanding of the basic mechanisms responsible for the hepatic clearance of xenobiotics and endogenous compounds. The sodium-dependent bile salt cotransporting polypeptide (Ntcp) mainly transports bile salts into hepatocytes (Hagenbuch et al., 1991; Boyer et al., 1994; Hagenbuch and Meier, 1994; Schwab et al., 1997; Baringhaus et al., 1999; Kramer et al., 1999) and is a rather specific transport system for this endogenous class of compounds. The family of the organic anion transporting polypeptides (Oatps) seems to be more important for drug transport into the liver (Meier et al., 1997). Oatp1, cloned from rat liver (Jacquemin et al., 1994) exhibits a broad substrate specificity, e.g., bile salts, glucuronidated and sulfated steroids, but also neutral compounds such as ouabain, the peptidomimetic compound CRC200, and even the cationic ajmalinium (Kullak-Ublick et al., 1994; Bossuyt et al., 1996; Eckhardt et al., 1999) were shown to be substrates. Oatp2, originally cloned from rat brain (Noe et al., 1997), but highly expressed in the liver (Reichel et al., 1999), has a similar substrate pattern as Oatp1, but does not transport bromosulfophthalein (BSP) and sulfolithocholyltaurine (SLCT) (Reichel et al., 1999). In addition, Oatp2 exclusively mediates high-affinity uptake of digoxin (Noe et al., 1997), whereas Oatp1 hardly transports this compound. A recently published additional family member, rat liver-specific organic anion transporter, rlst-1, expressed exclusively in rat liver, only mediates transport of cholyltaurine (Kakyo et al., 1999).

In the present study, we examined the mechanism of hepatocellular uptake of the anionic CHL derivatives. By comparison of apparent uptake inhibition constants obtained in freshly isolated rat hepatocytes with those obtained in heterologous expression systems we provided evidence that hepatic uptake of CHL derivatives is facilitated by Oatp1.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. CHL derivatives were synthesized by the Chemistry Department at Aventis Pharma Deutschland GmbH, Frankfurt am Main, Germany. Estimates of the log octanol/water partition coefficients (log P) for the CHL derivatives used in this study were calculated with the software Kowwin version 1.65 (SRC, Syracuse, NY) and were in the range between 2.2 and 3.1. CHL itself had a calculated log P value of 1. The measured pK_a values for the mono- and diacidic compounds were in the range between 3.3 and 4.7 as measured by titration. The radiolabeled analog of S 1743, [³H]S 1743, was synthesized by acetylation of the amine precursor in the conventional way, using [³H]acetic anhydride (9.7 µmol, 10.3 Ci/mmol; Amersham Pharmacia Biotech, Uppsala, Sweden) and pyridine (Hemmerle et al., 1997). Tritium-labeled cholyltaurine (3.4 Ci/mmol) was obtained from DuPont (Dreieich, Germany). All other reagents were commercially available products of highest analytical grade.

Animals. Adult male Sprague-Dawley rats (Moellegaard, Lille Skensved, Denmark), weighing 180 to 230 g, were used for hepatocyte isolation. They were housed in groups of up to five per cage in a temperature-controlled room with a 12/12-h light/dark cycle. All animals had free access to water and to a standard pellet rat chow (Altromine 1320) unless otherwise indicated.

Isolation of Hepatocyte Suspensions. Hepatocytes were isolated by a standard two step perfusion protocol with Ca²⁺-free medium and collagenase as described in Seglen (1976). Nonviable cells were removed by iso-density Percoll centrifugation (Kreamer et al., 1986). Quality of the final cell preparations was assessed by trypan blue exclusion. Viability of hepatocytes was usually above 95%.

Uptake Studies in Hepatocytes. Uptake of [³H]S 1743 into freshly isolated rat hepatocytes was determined using the centrifugal filtration technique recently described (Schwab et al., 1997). Briefly, uptake was initiated by adding 400 µl of cells to 400 µl of the appropriate substrate, dissolved in standard buffer at 37°C. Every 15 s (up to 90 s), aliquots of 100 µl were withdrawn and subsequently centrifuged. Separated cell pellets were dissolved in 400 µl of Biolute (Zinsser Analytic, Frankfurt, Germany). After addition of 5 ml of Quickszint 501 (Zinsser Analytic) radioactivity was determined by liquid scintillation counting (Beckman LB 2800; Beckman Instruments GmbH, Munich, Germany).

Uptake Studies in Xenopus laevis Oocytes. pSPORT plasmids containing the cDNAs for rat Oatp1 (Jacquemin et al., 1994) and Oatp2 (Noe et al., 1997) were linearized with NotI and capped cRNA was synthesized using T7 RNA polymerase as described in Hagenbuch et al. (1991). X. laevis oocytes were prepared as described previously (Hagenbuch et al., 1996). After an overnight incubation at 18°C, healthy oocytes were injected with 5 ng of Oatp1 or Oatp2 cRNA or water. After 3 days in culture, uptake of radiolabeled S 1743 was measured at 25°C in a medium containing 100 mM choline chloride, 2 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, and 10 mM HEPES/Tris, pH 7.5, as described previously (Hagenbuch et al., 1990). Subsequently, oocytes were washed with ice-cold PBS. Single oocytes were lysed with 0.1 N SDS and radioactivity was determined by liquid scintillation counting.

Cell Culture. CHO cells were grown in Dulbecco's modified Eagle's medium, supplemented with 10% fetal calf serum, 2 mM L-glutamine, and 50 µg/ml fungizone (amphotericin B) at 37°C with 5% CO₂ and 95% humidity. Selective medium contained additionally 400 µg/ml geneticin sulfate. For uptake studies, cells were seeded on six-well plates and were grown to confluence in the same medium as described above, except for omission of phenol red and geneticin sulfate. To increase uptake rates, cells were incubated for 24 h before the experiment in the presence of 5 mM sodium butyrate (Palermo et al., 1991; Eckhardt et al., 1999).

Uptake Studies in CHO Cells. Oatp1-expressing CHO cells as described by Schroeder et al. (1998), and wild-type CHO cells were used for transport studies. One-minute uptake of 1 µM [³H]cholyltaurine into CHO cells was performed as recently described (Schroeder et al., 1998) using six-well plates (4.9-cm² area) in the presence or absence of the indicated inhibitors. Cell layers were subsequently washed five times with 3 ml of ice-cold PBS each. Cells were lysed in 0.1 N NaOH and 1% SDS, and cell-associated radioactivity was determined by liquid scintillation counting. The uptake of cholyltaurine assessed in parallel in each experiment was linear for at least 1 min.

Protein Content. The BCA test (Pierce, Rockford, IL) in the presence of 1% SDS was used to determine protein content by applying the microtiter plate protocol as recommended by the supplier. Bovine serum albumin (Pierce) was used as standard.

Preparation of Microsomes and Determination of Gl-6-P Translocase Activity. Gl-6-P translocase activity was determined by assessing Gl-6-Pase activity in intact and disrupted microsomes. Microsomes were prepared from 10% (w/v) liver homogenates obtained from rats fasted for 20 h as has been reported in detail previously (Nordlie and Arion, 1966). Intactness of the preparations was assessed by determination of the latency of the "low K_M" mannose-6-phosphatase activity (Arion, 1989), which was usually above 97%. Fully disrupted microsomes were prepared by exposing thawed microsomes for 30 min at 0°C to optimal concentrations of the detergent Triton X-100.

Data Analyses. All studies with isolated hepatocytes were performed with at least two different cell preparations. Initial uptake rates were calculated by linear regression analysis from the slope of the linear portion of the time-dependent uptake curves, measured in 15-s intervals from 15 up to 90 s.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Structures of chlorogenic acid derivatives used in this study are shown in Table 1. The core structure was substituted by substituents R1 to R4. Substituent R1 represents a hydroxy group, except for S 1743, where it represents a sulfonamine group. Due to the substituents at R1, all presented CHL derivatives were regarded at least as monoanionic compounds at physiological pH. For compound S 3025, an additional negative charge at substituent R4 is present, resulting in the only dianionic CHL derivative in this study.

To investigate the mechanism of uptake of CHL derivatives into freshly isolated rat hepatocytes, the tritium-labeled CHL derivative [³H]S 1743 was used. Uptake of [³H]S 1743 into isolated hepatocytes was characterized by measuring time-dependent uptake of [³H]S 1743 in the presence (Fig. 1A) or absence of sodium. Analysis of initial uptake rates revealed a sodium-independent saturable uptake process (Fig. 1, B and C). Up to 200 µM, no diffusion component could be detected, suggesting the involvement of solely facilitated transport processes. The initial uptake kinetics of S 1743 could be described by an apparent half-saturation concentration (K_M) of 0.74 ± 0.26 µM and a maximal flux rate (J_max) of 0.15 ± 0.02 nmol/(min · mg of protein).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   A, time dependence of S 1743 uptake into isolated rat hepatocytes. Uptake of [3H]S 1743 was determined in the presence of 0.5 µM (), 5 µM (), and 15 µM () unlabeled S 1743. Initial uptake rates were determined by linear regression analysis of the linear part of the respective curves. Results from a representative experiment are shown. For the sake of clarity, only three substrate concentrations are shown. B, comparison of initial uptake rates of S 1743 into isolated rat hepatocytes in presence of Na+ (closed bars) and in Na+-depleted (open bars) medium. Results from a representative experiment are shown. Columns represent mean values of two incubations performed under identical conditions. C, concentration dependence of initial flux rates of S 1743 into isolated rat hepatocytes in the presence of Na+. Initial uptake rates were determined from the linear range between the interval of 15 to 90 s, in the presence of different S 1743 concentrations and constant concentrations of [3H]S 1743. The initial uptake rates were plotted using the Michaelis-Menten equation assuming one simple transport system.

Given the anionic nature of S 1743 and the broad substrate specificity of the Oatps (Meier et al., 1997), we tested the hypothesis that S 1743 is a substrate of the liver Oatps Oatp1 and Oatp2. As demonstrated in Fig. 2A, both members of the Oatp family mediated uptake of [³H]S 1743 into cRNA-injected oocytes well above the values obtained with water injected control oocytes with Oatp1-expressing cells showing uptake values of about twice of those observed with Oatp2-expressing cells. In addition, also Oatp1-expressing CHO cells demonstrated a time-dependent uptake of S 1743, whereas in control cells, uptake could not be demonstrated (Fig. 2B). These results clearly show that Oatp1 and Oatp2 were capable to transport S 1743. 

View larger version (17K): [in this window] [in a new window]   Fig. 2.   A, uptake of [3H]S 1743 (0.5 µM) into oocytes injected with water (n = 12), Oatp1-cRNA (n = 12), and Oatp2-cRNA (n = 12). ***p < 0.01 versus water-injected oocytes. Oocytes were incubated for 45 min in the presence of 0.5 µM [3H]S 1743. Subsequently, oocytes were washed with ice-cold PBS. Single oocytes were lysed with 0.1 N SDS and radioactivity was determined by liquid scintillation counting. B, time-dependent uptake of S 1743 (0.5 µM) (, ) into Oatp1-expressing CHO cells () and wild-type CHO cells () and of cholyltaurine (0.7 µM) (, ) into Oatp1-expressing CHO cells () and wild-type CHO cells () (n = 3). Cells were incubated with the respective radiolabeled substrate for the indicated time period, and were subsequently washed five times with ice-cold PBS. After lysing the cells by 0.1 N NaOH and 1% SDS, radioactivity was determined by liquid scintillation counting.

To get more insights into the characteristics of the involved transport process, we studied uptake of S 1743 into freshly isolated rat hepatocytes in the presence of different unlabeled CHL derivatives. Potency for inhibition of uptake of these CHL derivatives was also determined using Oatp1-expressing CHO cells (Table 2). A similar approach to elucidate functional involvement of cloned Oatp1 or Ntcp in the hepatic uptake of different substrates was published recently (Kouzuki et al., 2000). For practical reasons, cholyltaurine instead of [³H]S 1743 was used as a standard substrate for Oatp1-expressing CHO cells. The use of cholyltaurine instead of S 1743 as a standard substrate assumes that the inhibition constants are independent of the substrate used. Furthermore, in Oatp1-expressing CHO cells, interference with other transport systems is not expected, whereas hepatocytes exhibit different multiple transport systems, which may act in parallel for different substrates. Therefore, the use of cholyltaurine instead of S 1743 as a model substrate for Oatp1 using Oatp1-expressing CHO cells did not compromise the results of the study.

Most of the CHL derivatives studied inhibited uptake of [³H]S 1743 into hepatocytes with K_{i app} values in a range of 1.1 to 11 µM, irrespective of their potency as inhibitors of Gl-6-P translocase (Table 2). This holds true also for the dianionic CHL derivative S 3025. However, the naturally occurring compound chlorogenic acid, reported as a poor inhibitor of Gl-6-P translocase, did not inhibit uptake of the CHL derivative [³H]S 1743 at a concentration 100 µM. Unlabeled S 1743 revealed under these experimental conditions an K_{i app} value of 1.1 ± 0.2 µM, which is close to the obtained K_M value of 0.74 ± 0.26 µM. The determined K_{i app} values for S 1743 uptake into hepatocytes were in good agreement with the K_{i app} values determined for cholyltaurine transport in Oatp1-expressing CHO cells. This is the first evidence that Oatp1 could be involved in the uptake of S 1743 into rat hepatocytes.

To further characterize these uptake processes, uptake inhibition constants for additional compounds were determined for both S 1743 transport into rat hepatocytes and cholyltaurine transport into Oatp1-expressing CHO cells. The results are summarized in Table 3 and demonstrate again that similar K_{i app} values were obtained for both transport processes. In agreement with the substrate specificity of Oatp1, which does not mediate transport of the organic anions -ketoglutarate and p-aminohippurate, no inhibition was obtained with these substrates of the organic anion transporter family (Sekine et al., 1997) as well as with the anions geneticin sulfate and 5-sulfosalicylate, and with the organic cation N-methylnicotinamide. On the other hand, strong inhibition of uptake was demonstrated in the presence of SLCT, BSP, cholyltaurine, estrone-3-sulfate, digoxin, ouabain, bumetanide, phenol red, probenecide, and verapamil. Thus, also this inhibition pattern clearly suggested that Oatp1 is the major transport system for S 1743 uptake in rat hepatocytes. This hypothesis was corroborated by the similarity of the determined K_{i app} values and the published K_M values of Oatp1 (Reichel et al., 1999) for substrates such as cholyltaurine, BSP, estrone-3-sulfate, and ouabain (Table 3), and is visualized in Fig. 3, where all values obtained in the presented study are plotted.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Double logarithmic plot of Ki app (apparent inhibition constant) values for Oatp1 from experiments with Oatp1-expressing CHO cells using cholyltaurine as substrate and for uptake of S 1743 into hepatocytes. Ki app values were determined as described in Table 2.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The present study characterized the hepatic uptake mechanism of CHL derivatives. Uptake of CHL derivatives into rat hepatocytes per se was already strongly suggested by the fact that CHL derivatives showed inhibition of hepatic glucose production in isolated perfused rat liver and in vivo, as recently published (Herling et al., 1998). Kinetic analysis of the underlying uptake mechanism showed that hepatic uptake of S 1743 is solely facilitated by sodium-independent transport mechanism(s). The absence of a diffusion component in the transport of CHL derivatives is most likely explained by the complex structures of the derivatives that might reduce their potential for interactions with biological membranes. The range of the calculated log P values provided an estimate of the hydrophilic/lipophilic properties of the derivatives and does not support the hypothesis that uptake by simple diffusion through the cytoplasmic membrane is prevented by a purely hydrophilic character of the compounds.

The organic anion S 1743 proved to be a substrate for Oatp1 and Oatp2, as demonstrated by mediated transport of S 1743 in Oatp1- and Oatp2-expressing oocytes, as well as in Oatp1-expressing CHO cells. This supported the hypothesis that the structurally complex anionic compounds are substrates for the Oatps. A more detailed investigation of the underlying transport mechanism(s) was performed by inhibition studies. For this, a variety of CHL derivatives, organic anions, and organic cations were chosen as inhibitors of cholyltaurine uptake into Oatp1-expressing CHO cells, exhibiting solely Oatp1 transport function, and S 1743 uptake into rat hepatocytes, as a suitable model representing rat liver transport function. Initial uptake rates of S 1743 into rat hepatocytes were not only inhibited by CHL derivatives (Table 2), but also by other compounds known to interact with Oatps (Table 3). Known substrates of Oatp1, such as cholyltaurine, BSP, SLCT, estrone-3-sulfate, and ouabain (Bossuyt et al., 1996; Kanai et al., 1996; Meier et al., 1997; Eckardt et al., 1999) inhibited uptake of S 1743 into isolated hepatocytes (Table 3). Comparison of the K_{i app} values obtained from transport inhibition of Oatp1, using cholyltaurine as model substrate, with those obtained by transport inhibition of S 1743 into isolated hepatocytes revealed the same ranking in both systems (Table 3). A double logarithmic plot was used to visualize the correlation between K_{i app} obtained in hepatocytes and Oatp1-expressing CHO-cells over 4 orders of magnitude (Fig. 3). Furthermore, all other compounds known for lack of interaction with Oatps also did not show any inhibition of S 1743 uptake into hepatocytes, strongly suggesting that Oatp(s) plays a major role in hepatic uptake of S 1743.

Different types of inhibition change the meaning of the K_{i app} value, which was used to quantify potency of inhibition of uptake for different compounds in this study. Depending on the transport mechanisms of two competing substrates, mutual inhibition of both substrates does not necessarily indicate a competitive type of inhibition, also noncompetitive and uncompetitive types of inhibition could be possible (Krupka and Deves, 1983). Accordingly, a positive correlation of K_{i app} values of different compounds does not necessarily imply a competitive inhibition type for these compounds.

Since the first publication of the successful cloning of Oatp1 (Jacquemin et al., 1994), Oatp2 was cloned and found to be expressed in the liver (Noe et al., 1997; Reichel et al., 1999). Functionally, high similarities in the transport of Oatp1 and Oatp2 were found, except for the striking difference in digoxin, BSP, and SLCT transport (Noe et al., 1997; Reichel et al., 1999). Although Oatp1 failed to transport digoxin, Oatp2 exhibited a K_M in the submicromolar range (0.24 µM) for digoxin (Reichel et al., 1999). In the present study, digoxin inhibited uptake of S 1743 into hepatocytes and cholyltaurine into Oatp1-expressing CHO cells, characterized by K_{i app} values of 93 and 103 µM, respectively. These inhibition constants were much higher than the K_{i app} values determined for digoxin for Oatp2. A pronounced inhibition of S 1743 uptake by digoxin is expected if Oatp2 would exhibit a major contribution in the uptake process of S 1743 into rat hepatocytes. Because no relevant difference between the K_{i app} values of digoxin for hepatocytes and Oatp1 were present, Oatp2 can be clearly excluded from being a major transport system for S 1743 uptake into rat hepatocytes.

Uptake of S 1743 into Oatp1- and Oatp2-expressing oocytes was significantly higher than uptake into water-injected oocytes. Oatp1 exhibited an about 2-fold higher transport rate for S 1743 compared with Oatp2, further supporting Oatp1 being the major transport system for S 1743 uptake. Assuming similar expression levels and kinetics for both Oatps expressed in oocytes as well as in hepatocytes, digoxin should have shown a much lower mixed K_{i app} value as found in the present study. Under these assumptions, it is suggested that Oatp2 exhibits a very low expression in freshly isolated rat hepatocytes compared with Oatp1.

In addition to compounds well known to interact with Oatps, other compounds were also included in the present study. Phenol red, bumetanide, probenecid, and verapamil were competent inhibitors of Oatp1, in hepatocytes, as well as in Oatp1-expressing oocytes.

Although probenecid was already reported to inhibit BSP transport through Oatp (Kanai et al., 1996), and p-aminohippurate and -ketoglutarate failed to inhibit, phenol red and bumetanide were reported to have no effect on Oatp1. However, concentrations of phenol red and bumetanide were 10 µM (Kanai et al., 1996), which is significantly lower than the apparent inhibition constants found herein (Table 3).

Verapamil inhibited uptake of S 1743 into hepatocytes and cholyltaurine uptake into Oatp1-expressing CHO cells with K_{i app} values of 21.3 and 40.5 µM, respectively (Table 3), demonstrating interference of the organic cation with the uptake of S 1743. Direct transport of an organic cation by Oatp1 could be demonstrated by Bossuyt et al. (1996), who showed that the amphipathic organic cation N-propylajmaline was transported by Oatp1-expressing oocytes. Bumetanide has been shown to inhibit bile salt carriers but not to be a substrate (Horz et al., 1996). Due to the interaction potential of phenol red with Oatp1, phenol red-free media were used for culturing of the Oatp1-expressing CHO cells, whereas geneticin sulfate showed no potential for interaction.

CHL derivatives also acted as potent inhibitors of S 1743 uptake. Only chlorogenic acid failed to inhibit uptake of S 1743 at concentrations of 100 µM. All CHL derivatives included in the present study exhibited a narrow range of K_{i app}, from 1.1 to 11 µM. Furthermore, K_{i app} values determined in Oatp1-expressing oocytes and hepatocytes were similar. This correlation suggests that the major uptake for the different CHL derivatives into hepatocytes was facilitated by Oatp1. The narrow range of K_{i app} values for various CHL derivatives for the uptake of S 1743 in rat hepatocytes is in contrast to the much wider range of potency found for the pharmacological activity of CHL derivatives in rat liver microsomes (Table 2). This could be explained by a rather unspecific interaction with Oatp1, which has been demonstrated to have a very broad substrate specificity compared with a more specific interaction with the target system Gl-6-P translocase. It should be noted, however, that K_{i app} values do not necessarily resemble the K_M values of the respective compounds for the uptake of the inhibitors into cells. Furthermore, no estimation of corresponding maximal flux rates of the inhibitors is possible based on the determined K_{i app} values. Only direct transport measurements can prove real substrates.

Combining the available evidence from pharmacokinetic and pharmacodynamic experiments, CHL derivatives exhibited a specific Oatp1-mediated uptake mechanism into rat liver. Whereas CHL derivatives exhibited a potential for interaction with Oatp1 only in a narrow range of concentrations of about 1 order of magnitude, pharmacological potency of the compounds was exhibited over a concentration range of at least 5 orders of magnitude. Liver-specific targeting is regarded as a prerequisite to achieve the necessary concentrations of the drug at the intracellular target in the liver. Furthermore, liver-specific uptake may also decrease the exposure of other tissues to the drug, thereby decreasing possible side effects in tissues not involved in the desired pharmacological action. The statin pravastatin was recently described as another example for a drug exhibiting a low penetration into nonhepatic tissues due to a liver-specific targeting to the target organ (Koga et al., 1992). Therefore, all available data are in agreement with the assumption that Oatp1 is at least a major component in transport of CHL derivatives.