Abstract
We assessed the impact of zonal factors on the hepatic reduced glutathione (GSH) conjugation of ethacrynic acid (EA). Uptake of EA by enriched periportal (PP) and perivenous (PV) rat hepatocytes was characterized by both saturable (Vmaxuptake = 3.4 ± 1.7 and 3.2 ± 0.8 nmol/min/mg protein andKmuptake = 51 ± 13 and 44 ± 15 μM) and nonsaturable (12 ± 5 and 12 ± 3 μl/min/mg protein) components. Values for the overall GSH conjugation rates of EA (200 μM) were similar among the zonal hepatocytes and resembled those for the influx transport rates. In the absence of the hepatocyte membrane, GSH conjugation in PV and PP hepatocyte cytosol was similar, but a higher perivenous GSH conjugation activity toward EA (PV/PP of 2.4) that mirrored the higher PV/PP ratios of immunodetectable GSTs Ya (1.7) and Yb2 (2.5) was found in cell lysates obtained by the dual-digitonin-pulse perfusion technique. The GSH conjugation rates in the subcellular fragments were, however, much greater than those observed for intact hepatocytes. Efflux rates of the glutathione conjugate EA-SG from zonal hepatocytes were similar, as were levels of the immunodetectable multidrug-resistance protein 2/canalicular multispecific organic anion transporter (Mrp2/cMoat) in the 100,000 g pellets. The composite results suggest that the GSTs responsible for EA metabolism are more abundant in the PV region, albeit that the gradient of enzymatic activities is shallow. Despite the existence of zonal metabolic activity, the overall GSH conjugation rate of EA is homogeneous among cells because the reaction is rate limited by uptake, which occurs evenly. Results on EA-SG efflux suggest the acinar homogeneity in Mrp2/cMoat function for canalicular transport.
Glutathione (GSH) conjugation is an important detoxification pathway for electrophiles, including some therapeutic agents, carcinogens, and reactive metabolites. Several factors have an impact on the rates of GSH conjugation in liver; these include the cellular uptake or in situ formation of acceptor substrates of GSH, the propensity for spontaneous GSH conjugation, the presence of the glutathione S -transferases (GSTs), and cosubstrate (GSH) availability. Once formed, the GSH-adduct is effluxed out of the hepatocyte by members of the multidrug resistance protein (MRP) family found at either the canalicular or basolateral membrane (Büchler et al., 1996; Hirohashi et al., 1998; Keppler et al., 1998; König et al., 1999). These processes are exemplified by the GSH conjugation of ethacrynic acid (EA) in vitro and in the perfused rat liver (Tirona and Pang, 1999). Data derived from in vitro spontaneous and enzymatic GSH conjugation and hepatocytic uptake of EA had been successfully scaled-up to describe observations from perfused rat liver with use of a physiologic, kinetic model that encompassed transport and bimolecular, spontaneous (nonenzymatic) and enzymatic metabolism. The data suggest that hepatic uptake rate limited the GSH conjugation of EA at low input concentrations (<50 μM), whereas cellular metabolism played an increasingly greater role at higher concentrations (∼200 μM). As intracellular levels of GSH became depleted as a result of consumption, cosubstrate availability invariably rate limited GSH conjugation.
There is, additionally, an increasing awareness that acinar heterogeneity in sinusoidal uptake and metabolism affects hepatic drug processing (Pang, 1995). Kwon and Morris (1997) demonstrated, in theory, that the total hepatic elimination of drugs would be influenced by zonal localization of transport and enzymatic activities. For GSH conjugation that is a bimolecular reaction, there is the extra consideration of the unevenness in cellular availability of the cosubstrate that could further affect the overall hepatic removal (Pang, 1995). There exists much evidence that the GSTs are more concentrated in the perivenous (PV) than the periportal (PP) regions (Redick et al., 1982). The rat liver GSTs that are constitutively expressed and capable of metabolizing EA (Ploemen et al., 1991, 1993), such as the α (subunits 1, 2, and 8) and μ (subunits 3 and 4) classes, are also more abundant in the PV zone (Sippel et al., 1996). Other GSTs (the microsomal GSTs and those belonging to the θ class) exhibit negligible activity toward EA (Ploemen et al., 1993), and these are also constitutive to the PV region (Mainwaring et al., 1996). The transport of EA into isolated rat hepatocytes was facilitated by a sodium-independent transporter whose identity and zonal distribution are currently unknown and by a nonsaturable (linear) system (Tirona and Pang, 1999). It is conceivable that zonal uptake exists for EA because this zonated event occurs within PV hepatocytes for substrates such as cysteine (Saiki et al., 1992), glutamate (Burger et al., 1989; Tan et al., 1999), and α-ketoglutarate (Moseley et al., 1992). Likewise, predominance of immunodetectable sinusoidal transport proteins was observed in PV hepatocytes for the rat glucose transporter 1 (GLUT1;Tal et al., 1990), the rat organic cation transporter 1 (rOCT1;Meyer-Wentrup et al., 1998), and the rat organic anion transporting polypeptide 2 (oatp2; Kakyo et al., 1999). By contrast, acinar homogeneity exists for the rat sodium-dependent taurocholate transporting polypeptide, ntcp (Stieger et al., 1994; Tan et al., 1999), and the organic anion transporting polypeptide from rat liver, oatp1 (T. N. Abu-Zahra, A. W. Wolkoff, R. B. Xim, and K. S. Pang, unpublished observations), in their uptake of substrates.
In this communication, we investigated the initial rates of uptake of EA and explored the possible roles of acinar metabolism and transport by isolated, enriched PP and PV rat hepatocytes. We further studied accumulation of the formed GSH adduct of EA (EA-SG) because it is recognized that GSH conjugates, including EA-SG, are product-inhibitors of GSTs (Ploemen et al., 1990), and the efficiency in efflux of GSH conjugates may play an important role in the cellular detoxification of electrophiles by the GSTs. The transporters responsible for this basolateral efflux of EA-SG in the rat have not been directly identified but are likely orthologs of human MRP family proteins (Zaman et al., 1997). However, considerably faster efflux of EA-SG occurs with the canalicular multispecific organic anion transporter, cMoat or Mrp2 (Büchler et al., 1996, Evers et al., 1998), and this accounted for the rapid appearance of EA-SG excretion in rat bile (Tirona and Pang, 1999). It is unknown whether acinar heterogeneity exists for the efflux processes in rat liver.
Experimental Procedures
Materials.
[14C]EA (specific activity, 15 mCi/mmol) was a kind gift from Dr. J. H. T. M. Ploemen (TNO, Ziest, the Netherlands) and was purified by high-performance liquid chromatography (HPLC) and solid-phase extraction (radiochemical purity >98%). [3H]Sucrose (specific activity, 10 Ci/mmol) was obtained from NEN Life Sciences (Boston, MA). EA, 1-chloro-2,4-dinitrobenzene (CDNB), GSH, and oxidized GSH (GSSG) were purchased from Sigma Chemical Co. (St. Louis, MO). EA-SG was synthesized as described previously (Tirona and Pang, 1999). Digitonin was obtained from Fluka Chemie (Buchs, Switzerland). Collagenase was purchased from Boehringer-Mannheim (Oakville, Ontario, Canada). Antisera raised against rat GST Ya (rGST 1-1) and Yb2 (rGST 4-4) were obtained from Biotrin International (Dublin, Ireland). Polyclonal antibodies against rat Mrp2/cMoat (EAG15; Büchler et al., 1996) and monoclonal antibodies against rat cytochrome P-450 1A (CYP1A) isozymes (MAb 1-7-1) were generously provided by Drs. D. Keppler (Deutsches Krebsforschungszentrum, Heidelberg, Germany) and H. V. Gelboin (National Institutes of Health, Bethesda, MD), respectively. All other reagents were of the highest available grade.
Isolation of PP and PV Rat Hepatocytes.
Enriched PP and PV hepatocytes from male Sprague-Dawley rats (275–325 g; Charles River Canada, St. Constant, Quebec, Canada) were harvested by the digitonin/collagenase perfusion method according to Lindros and Pentı̈lla (1985), with slight modifications as detailed by Tan et al. (1999). Hepatocyte viability (>90%) was assessed by Trypan blue exclusion. Zonal enrichment was routinely estimated by monitoring the activities of alanine aminotransferase (ALT) with a commercially available kit (Sigma) and of glutamine synthetase (GS) by a standard UV method (Tan et al., 1999). Protein was determined by the method ofLowry et al. (1951). Several low-speed centrifugations (50g) during the isolation procedure separated most of the nonparenchymal cells from hepatocytes; it was surmised that low levels of contamination persisted.
Uptake of EA by PP and PV Rat Hepatocytes.
The zonal hepatocytes were suspended in Krebs-Henseleit bicarbonate buffer (pH 7.4) supplemented with 5 mM glucose and 1 mM HEPES and were preconditioned for 10 min at 37°C. A mixture of [14C]EA, [3H]sucrose, and unlabeled EA was added to the cell suspension to result in final EA concentrations of 1 to 800 μM and ∼1.67 × 106 cells/ml. Samples were retrieved at 15- to 20-s intervals after admixture and were rapidly centrifuged through a layer of silicon oil as described previously (Tan et al., 1999). Cellular radioactivity was determined by liquid scintillation spectrometry (LSC, model 5801; Beckman Canada, Mississauga, Ontario, Canada), after correction of the adhered water layer defined by [3H]sucrose (Tan et al., 1999). Because the initial cellular accumulation of EA was linear over 80 s, the rate of uptake, vuptake, was estimated as the slope on regression of the accumulated amount-versus-time data. The kinetics of EA uptake were analyzed by fitting vuptake against the initial substrate concentration [EA] with use of the following equation (eq. 1, after elimination of other possibilities of single and multiple saturable systems) and an appropriate weighting scheme with a least-squares fitting routine with the software SCIENTIST (Micromath Scientific Software, Salt Lake City, UT):
Metabolism of EA by PP and PV Rat Hepatocytes.
We chose to study EA at the concentration of 200 μM because GSH conjugation of the drug would be influenced by both cellular uptake and enzymatic activity (Tirona and Pang, 1999). Hence, inherent acinar differences of these factors would be more apparent. EA, dissolved in physiological saline solution, was added to the hepatocyte suspension to achieve an initial concentration of 200 μM with ∼1.67 × 106cells/ml. At 1- to 5-min intervals throughout the 20-min incubation experiment conducted at 37°C, samples (500 μl) were retrieved and placed into 1.5-ml microcentrifuge tubes containing 70 μl of 70% perchloric acid. A separate sample (700 μl) was overlaid onto 250 μl of 1-bromododecane and centrifuged (Biofuge pico; Heraeus Instruments, Germany) for 10 s. An aliquot (500 μl) of the resulting supernatant was similarly placed into 70% perchloric acid to assay for contents in the extracellular space of the incubation mixture. The acidified samples were immediately mixed and stored at −70°C until analysis.
Analysis.
For quantitation of EA and EA-SG, each acidified sample (total incubation mixture and extracellular medium) was further combined with 200 μl of 1.2 mM 4-(2,4-dichlorophenoxy)-butyric acid (internal standard). After centrifugation, 100 μl of the supernatant was analyzed by HPLC according to a previously developed procedure (Tirona and Pang, 1999). Standard curves, prepared from solutions of known concentrations of EA (50–250 μM) and EA-SG (10–250 μM), were constructed in a similar fashion.
For the analysis of GSH and GSSG, 500 μl of the total suspension (cell and extracellular medium) and 500 μl of the extracellular medium were added to microcentrifuge tubes containing 100 μl of 70% perchloric acid and 50 μl of 15 mM bathophenanthroline-disulfonic acid. After centrifugation, the supernatant was immediately analyzed for GSH and GSSG with use of the derivatization and HPLC method ofFariss and Reed (1987). The intracellular concentrations of EA, EA-SG, and GSH/GSSG were estimated as the difference between the concentrations in the total suspension and the extracellular medium. Acid-precipitated hepatocytes were further analyzed for the GSH-protein mixed disulfide content with the same HPLC method (Fariss and Reed, 1987).
GST Activity of PP and PV Hepatocyte Cytosols.
Cytosol was obtained by homogenization of the zonal hepatocytes (Ultra-Turrax T25 homogenizer; Janke & Kunkel, Staufen im Briesgau, Germany), and the resultant supernatant fractions were sequentially centrifuged at 9,000g and 100,000g at 4°C. Cytosolic GST activity toward EA was determined within the linear protein concentration range by the spectrophotometric method of Satoh (1995) as previously described (Tirona and Pang, 1999) with 200 μM EA and 5 mM GSH, at 37°C and pH 7.2. Cytosolic GST activity toward CDNB was determined by standard UV methods with 1 mM CDNB and 1 mM GSH, at 25°C and pH 6.5 (Habig et al., 1974). The GST-catalyzed GSH conjugation activities toward EA and CDNB were obtained after correction of the (total) cytosolic GSH conjugation rates for the spontaneous reaction rates (in absence of cytosol) and normalization to the protein contents.
PP and PV Cell Lysate.
Preparation of the cell lysates from the most proximal and distal hepatocytes along the sinusoidal plate was performed according to the dual-digitonin-pulse perfusion method ofQuistorff and Grunnet (1987), with modifications. Paired zonal (PP and PV) lysates were prepared from the same liver. Rat livers were perfused with Hanks' buffer (pH 7.2) containing 10 mM HEPES, 0.5 mM EGTA, 4.2 mM NaHCO3, and 5 mM glucose (perfusion buffer) pregassed with 95% O2/5% CO2 at a flow rate of 30 ml/min into the portal vein in a prograde fashion. After 10 min, the flow rate was reduced to 12 ml/min, and the direction of flow was reversed (retrograde perfusion into the hepatic vein). After stabilization of the liver for 1 min, the perfusion medium was changed to the digitonin solution (3.25 mM digitonin, 150 mM NaCl, 6.7 mM KCl, and 50 mM HEPES) for perfusion at 6 ml/min until a spotted destruction pattern was observed on the liver surface (35 ± 11 s) for destruction of the PV region. The flow was then reverted to prograde flow, and perfusion buffer was used for perfusion at a rate of 20 ml/min. The eluate (PV lysate) was collected over 30 s. For continued preparation of the PP lysate, the direction of flow was reversed. Rat livers were perfused with perfusion buffer at the flow rate of 20 ml/min into the hepatic vein in a retrograde fashion for 2.5 min. Then, the flow rate was reduced to 12 ml/min for 1 min. Next, the digitonin solution was infused at 6 ml/min until a circular destruction pattern appeared (26 ± 8 s) for destruction of the PP region. Subsequently, the eluate (PP lysate) was collected for 30 s. PP and PV eluates were centrifuged at 100,000g for 60 min at 4°C, and the resultant supernatants were stored at −70°C.
Kinetic Modeling of EA Disposition by PP and PV Hepatocytes.
A kinetic model, whose scheme is shown in Fig.1, was used for analysis of the time-dependent disposition of EA in PP and PV hepatocytes. The saturable uptake (Kmuptake and Vmaxuptake) and the bidirectional (uptake and efflux) linear clearance (Pdiff) parameters, obtained from uptake experiments (Fig.2), were used to denote transmembrane transport. The spontaneous and GST-catalyzed GSH conjugations of EA were described by a second-order (spontaneous) reaction and a single enzyme-catalyzed, rapid-equilibrium, random sequential order scheme as reported previously (Tirona and Pang, 1999). The second-order constant (k2) for the spontaneous reaction and the apparent Michaelis-Menten constants for GSH (KmGSH) and EA (KmEA) for the enzymatic reactions were obtained from the previous in vitro studies (Tirona and Pang, 1999). The maximal enzymatic conjugation rate (Vmaxmetab) was estimated by the fitting procedure.
EA-SG formed within isolated hepatocytes was assumed to undergo net efflux into the extracellular medium and transport into intracellular sequestration vesicles (see Discussion). These linear clearance processes, denoted by the terms CLeffluxEA-SG and CLvesEA-SG, respectively, were estimated by fitting. Because control experiments indicated a lack of net change (over 20 min) in intracellular hepatocyte GSH (Figs.3A and 4A) and because GSH synthesis is expected to be insignificant in the absence of GSH precursors under the experimental conditions, intracellular GSH kinetics were simplified; the loss of GSH was due only to conjugation. However, there was a lack of mass balance because the amount of GSH consumed was less than the total amount of EA-SG formed. The intracellular GSH concentrations were scaled-up to allow for mass balance for the purposes of fitting; the GSH data were converted to concentration terms by assuming 10 μl of cell volume/ml of suspension. The mean data of the PP and PV hepatocyte experiments were simultaneously fitted with the model equations with use of a nonlinear least-squares procedure (SCIENTIST) and appropriate weighting schemes. The fitted parameter values, together with the assigned parameters, are shown (see Table 3).
The following mass balance rate equations are used in the model (Fig.1).
The equations describing the extracellular space (EC) for EA and EA-SG are
Immunoblot Analysis.
Cytosols, lysates (5 μg of protein containing the GSTs), and 100,000g pellets (10 μg of protein containing Mrp2 and the marker protein CYP1A2) derived from PP and PV hepatocytes were used for analysis. The immunoreactive proteins were resolved by SDS-polyacrylamide gel electrophoresis (PAGE) on 12, 9, and 7.5% gels for the GSTs, CYP1A2, and Mrp2, respectively, using the MiniProtean II system (Bio-Rad, Mississauga, Ontario, Canada). Protein was transferred onto nitrocellulose membranes (Hybond ECL; Amersham, Oakville, Ontario, Canada) with a semidry transfer unit (Bio-Rad). Subsequently, the membranes were blocked with 10% nonfat milk in Tris-buffered saline with 0.1% Tween 20 (TBST) for 2 h at room temperature. After washing with TBST, membranes were incubated with primary antibody (anti-GST Ya or Yb2 at 1:50,000 dilution, MAb 1-7-1 at 1:20,000 dilution, and EAG15 at 1:40,000 dilution) in TBST overnight at 4°C. After washing with TBST, the membranes were incubated with horseradish peroxidase-linked anti-rabbit (GST and Mrp2) or anti-mouse immunoglobulins (CYP1A2; Amersham or Bio-Rad) at 1:20,000 dilution in TBST for 2 h at room temperature. Detection was performed using enhanced chemiluminescence (ECL; Amersham), and membranes were visualized on Hyperfilm (Amersham). For the semiquantitation of the GSTs, CYP1A2, and Mrp2, the films were scanned (Umax Astra 1200S), and band intensities were integrated using the NIH Image software (developed at the National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/).
Statistical Analysis.
All data are presented as mean ± S.D. Unpaired and paired Student's t tests were used where appropriate; the level of significance was set at .05.
Results
Biochemical Characterization of PP and PV Hepatocytes and Lysates.
The activities of PP and PV marker enzymes (ALT and GS, respectively) are summarized in Table 1. Significant differences were observed in marker enzyme activities indicating the attainment of the zonal enrichment of PP and PV hepatocytes (p < .05) and PP and PV lysates (p < .05). The PP/PV activity ratios for ALT were higher for lysate (7.5) than for hepatocyte cytosol (1.8), confirming the steeper and decreasing (portal to venous) acinar gradient in enzyme content. The PP/PV activity ratios for GS were similar among lysates and hepatocyte cytosols (0.025 and 0.029, respectively), and these results are consistent with the confined localization of this enzyme to the terminal two or three hepatocytes in the PV region. These observations established the validity of the preparation on the enrichment of PP and PV hepatocytes or lysates.
Concentration-Dependent Uptake of EA by Zonal Hepatocytes.
EA uptake kinetics were similar among PP and PV hepatocytes (Fig. 2). The uptake parameter estimated on fitting of the data to eq. 1 furnished similar Kmuptake (51 ± 13 and 44 ± 15 μM) and Vmaxuptake (3.4 ± 1.7 and 3.2 ± 0.8 nmol/min/mg protein) values for the saturable uptake of EA for PP and PV hepatocytes (p > .05, n = 4); comparable values were also obtained for the nonsaturable component, Pdiff (12 ± 5 and 12 ± 3 μl/min/mg protein) for PP and PV hepatocytes (p > .05, n = 4).
Cellular GSH Concentration and GSH Conjugation Rates of EA by PP and PV Hepatocytes.
Control (saline-treated) PP and PV hepatocytes (n = 4) retained their initial intracellular GSH contents for at least 20 min, and the extracellular GSH levels remained constant throughout the incubation period (Figs. 3A and 4A, insets). GSSG, found mainly in the extracellular space, remained virtually constant during the incubation and accounted for approximately 10 to 15% of the total GSH equivalents in the system (data not shown).
The addition of EA (200 μM) to PP and PV hepatocyte suspensions (n = 12) greatly reduced GSH levels, which in turn affected the rate of GSH conjugation and disappearance of EA (Figs. 3A and 4A). Although the extracellular GSH levels remained constant throughout the incubation period in treated hepatocytes and were similar to those of the controls, the GSH within PP and PV hepatocytes was rapidly depleted within the first 5 min after EA treatment (Figs.3A and 4A). The pattern of depletion rate was, however, independent of the acinar origin of the hepatocytes.
EA disappeared at similar rates in total suspension and the extracellular medium of the incubation system with PP and PV hepatocytes (Figs. 3B and 4B). Cellular accumulation of EA was low in either PP or PV hepatocytes. The loss of EA was completely accounted for by the formation of EA-SG, and no other sequential metabolite (cysteinyl-glycine, cysteine and N -acetyl-cysteine adducts) was detected. Essentially, mass balance was conserved by EA and EA-SG throughout the 20-min incubation period; the dose recovery was 100 ± 5 and 100 ± 6% for the PP and PV hepatocyte incubation systems, respectively.
Both PP and PV hepatocytes produced EA-SG at equivalent rates (Figs. 3C and 4C). Because rapid efflux of EA-SG occurred on its formation within cells, the formation rate of EA-SG was more appropriately estimated by viewing the early-in-time data for total concentration in the incubation system before significant depletion of cellular GSH occurred. The initial formation rates (roughly estimated from upslope of the data between 0 to 3 min) of 6.2 ± 0.9 and 6.9 ± 1.1 nmol/min/mg protein, respectively, were comparable (p > .05) and were similar to those observed for uptake (5.4 ± 1.5 and 5.2 ± 0.9 nmol/min/mg protein, respectively; Table2) at 200 μM. This suggests that glutathione conjugation occurred very rapidly on cellular uptake of EA. The EA-SG formation rate rapidly approached its asymptotic values by 10 min (Figs. 3C and 4C) due to the depletion of GSH (Figs. 3A and 4A). The EA-SG formed within the zonally enriched hepatocytes rapidly escaped into the extracellular media in a monoexponential fashion. After 15 min, EA-SG remained sequestered within the hepatocyte with incubation time, and the net efflux of cellular EA-SG approached zero. Because the efflux rate equaled the rate of accumulation in extracellular medium, the initial EA-SG efflux rate was estimated from the early-in-time data for the extracellular medium (1.5–5 min, excluding zero time due to the short lag-time involved with the cellular formation/cellular distribution/appearance of EA-SG in medium). These were similar among the zonal hepatocytes (2.2 ± 0.2 and 2.3 ± 0.3 nmol/min/mg protein, respectively, p > .05), and the accumulation of EA-SG in extracellular medium approached their corresponding asymptotic values by 15 min. Although the rates of initial efflux of EA-SG were considerably slower than those of formation, the accumulation of intracellular EA-SG was short-lived because EA-SG formation ceased by 7.5 min due to the almost complete depletion of GSH (Figs. 3A and 4A).
It was noted that the total formation of the EA-SG (∼25 nmol/mg protein) at the end of the experiment exceeded the total loss of cellular GSH (from ∼15–17 nmol/mg protein to almost zero). Because the summed loss of EA equated with the production of EA-SG, it is unlikely that experimental errors occurred in the quantitation by HPLC. GSH synthesis within the hepatocytes could account for the difference because the rate of GSH synthesis by hepatocytes has been estimated to be ∼0.4 nmol/min/106 cells (Lu et al., 1991). This possibility, however, is remote due to the lack of amino acid precursors in the suspension medium and the lack of net change in GSH in control hepatocytes (Figs. 3A and 4A, insets). Furthermore, total EA-SG formation was discontinued by 10 min, an observation that contradicts the sustained GSH synthesis. The levels of GSH-protein mixed disulfides, measured in a separate group of hepatocytes, accounted for only ∼1% of the total GSH content. At this time, we have no explanation for this discrepancy.
Fitting of Data to Kinetic Model.
Reasonable fits to the data were obtained on scale-up of intracellular GSH to provide for mass-balance (Figs. 3 and 4). Values for the fitted Vmaxmetab, CLeffluxEA-SG, and CLvesEA-SG (Table 3) revealed that the processes governing the enzymatic formation of EA-SG and its transport were all similar among the zonal hepatocytes. We further tried out other kinetic models that do not include an intracellular sequestration space; all failed to provide adequate fits to the observations because these predicted complete intracellular depletion of EA-SG by the end of the incubation experiment (fits not shown).
The fits showed that the initial rates of EA-SG formation were somewhat underestimated by the model (Figs. 3C and 4C) when parameters observed for EA transport were assigned with values obtained from the uptake experiments. These values, based on [14C]EA influx in rapid uptake studies (Fig. 2, Table 2), predicted a slightly lower initial rate of EA uptake that resulted in a more gradual decline in EA disappearance and lowered formation of EA-SG during the first 3 min of incubation (Figs. 3 and 4). A systematic trend occurred with the fitting of the cellular and extracellular EA-SG data (Figs. 3C and 4C) and might have been the consequence of lack of modeling of the time-dependent decrease in CLeffluxEA-SG and concomitant increase in CLvesEA-SG with progressive internalization of Mrp2 during the time-span of the experiment. Moreover, transport of EA-SG from cell to extracellular medium was described only by a net efflux clearance parameter (CLeffluxEA-SG) and failed to incorporate bidirectional movement.
The present data for efflux of EA-SG from hepatocytes were further scaled-up with the scaling factor (α/β; where α is 1.25 × 108 cells/g liver and β is 1 × 106cells/mg protein) and compared with that obtained from the whole organ (Tirona and Pang, 1999). The calculation was based on the assumption that the total efflux clearance for EA-SG in hepatocytes equaled the sum of CLeffluxEA-SG and CLvesEA-SG (∼1.2 μl/min/mg), yielding an efflux clearance of 1.5 ml/min/10 g liver for whole liver. The estimated value was similar to the sum of sinusoidal and biliary clearance for EA-SG (1.1 ml/min/10 g liver) in perfused rat liver studies (Tirona and Pang, 1999).
The Vmaxmetab estimates (∼35 nmol/min/mg) obtained with model fitting (Table 3) with the hepatocyte experiments were much lower than that observed for the GST activity in vitro (in cytosol) toward EA-SG formation from 200 μM EA and 5 mM GSH at physiological conditions (∼200 nmol/min/mg, Table 2). This big discrepancy was the result of the very poor reliability of Vmaxmetab because only one concentration was studied and saturation of metabolism might not have been attained under the experimental condition. These led to the large standard deviation of the estimate (Table3).
GST Activities in Cytosolic Fractions of PP and PV Hepatocytes and Lysates.
The in vitro GST activities derived from cytosols of PP and PV hepatocytes and PP and PV lysates toward EA are summarized in Fig. 5A. No difference in GST activity was observed among zonally enriched hepatocytes, whereas a 2.4-fold greater activity (p < .05) was observed for the PV lysate compared with the PP lysate. The GST activities toward EA mirrored those toward CDNB (Fig. 5B). No difference in cytosolic GST activity was seen in zonal hepatocytes toward CDNB, whereas a 1.9-fold difference (p < .05) was observed between PV and PP lysates.
GSTs in PP and PV Hepatocytes and Lysates.
The levels of two constitutive rat liver GSTs (Ya and Yb2) were found to be similar for the cytosolic fractions of PP and PV hepatocytes (Fig.6) when these were assessed by SDS-PAGE and densitometry. However, the GST Ya and Yb2 proteins in PV lysate were 1.7 and 2.5 times those of the PP lysates (Fig.7).
Mrp2 in PP and PV Hepatocytes.
Levels of immunoreactive Mrp2 in crude membranes obtained from 100,000g pellets of homogenized PP and PV hepatocytes are shown in Fig.8. PP and PV hepatocytes contained similar amounts of Mrp2 protein. By contrast, the same membrane fractions contained a highly variable and 3- to 4-fold enrichment of constitutively expressed CYP1A2 in PV hepatocytes (p < .5), as anticipated for the marker protein.
Discussion
We had previously demonstrated utility of transport data from isolated rat hepatocytes and in vitro data on the spontaneous and enzymatic GSH conjugation in the prediction of GSH conjugation of EA in the whole liver (Tirona and Pang, 1999). It was concluded that various acinar factors on uptake, GSH availability, and distributions of the GSTs and GSH conjugate efflux systems among zonal cells could affect GSH conjugation. Although the identity of the EA transporter is uncertain and may be similar to that for bumetanide (Horz et al., 1996), a sodium-independent saturable system that is inhibitable by organic anions was found to exist for EA uptake with rat hepatocytes (Tirona and Pang, 1999). The EA transporter appears to be homogeneously distributed within the liver acinus because there is no difference for EA uptake among the enriched PP and PV hepatocytes (Fig. 2). The kinetic parameters obtained from uptake experiments were similar among the zonal regions and were not different than those obtained from hepatocytes prepared from all zonal regions (Tirona and Pang, 1999). Next to be considered is the aspect of zonal metabolism. In contrast to our anticipation of observing differences in metabolism at 200 μM EA (Tirona and Pang, 1999), initial rates of GSH conjugation within the PP and PV hepatocyte incubation were similar and closely resembled those for transport (Table 2). Moreover, GSH depletion rates were not different among PP and PV hepatocytes treated with EA (Figs. 3A and4A), suggesting the absence of zonal influence by GSH. The EA-SG formation rates were, however, much lower than those found in the corresponding cytosolic fractions of PP and PV hepatocytes (Table 2), inferring strongly that hepatocyte EA uptake rate-limits GSH conjugation regardless of the zonal position along the sinusoidal plate.
The difference in GST activities within the cytosolic fractions of the PP and PV hepatocytes could have been revealed in the in vitro studies, except for the shallow or modest gradient in PV enrichment of GST activities toward EA and CDNB (Fig. 5). The results on GST activities coincided with immunodetectable GST Ya and Yb2 levels measured from cytosols of enriched PP and PV hepatocytes (Fig. 6). As demonstrated by others for CDNB (Kera et al., 1987; Suolinna et al., 1989), the PV enrichment of cytosolic GST activities was modest (1.2- to 1.6-fold). The PP and PV lysates obtained by the dual-digitonin-pulse perfusion (Quistorff and Grunnet, 1987) offer an improved method for the study of metabolic heterogeneity in liver. With this technique, provision of the cellular contents of the most proximal or distal hepatocytes is accomplished by controlling the depth of digitonin penetration. These lysates proved to be more accurate in relating to PP and PV activities and showed 2- to 3-fold PV predominance in GST activity toward EA and CDNB (Fig. 5) and complement corresponding changes in immunoreactive GSTs Ya and Yb2 in the lysate (Fig. 7). These results, along with those from isolated PP and PV hepatocytes, confirm that acinar gradients in GST activity toward EA and CDNB exist, and the gradients are relatively shallow. Comparable observations were reported for CDNB GST activities (Kera et al., 1987) and GST protein contents in PP and PV lysates (Lindros et al., 1998). The slightly lower GST activity in PP lysate compared with PV lysate (155 ± 44 and 364 ± 42 nmol/min/mg cytosolic protein respectively, Fig. 6A) toward EA would not rate-limit EA elimination because the transport activity is much lower.
Our metabolic data demonstrate that limitations existed in the intact, isolated PP and PV hepatocyte system for the study of functional metabolic heterogeneity. In examination of the PP marker enzyme ALT, the PP/PV activity ratio (∼1.8) normally obtained was considerably lower than that (∼7.5) in lysates (Table 1). This difference underscores the fact that enriched zonal hepatocytes are harvested from approximately half of the sinusoidal length and are inevitably cross-contaminated to a higher degree (especially from midzonal regions) than for lysates, which are derived from only a small population of hepatocytes at the most distal and proximal acinar regions. Therefore, depending on the shape and steepness of the zonal distribution of enzymatic activities, difficulties persist in identifying shallow gradients on metabolic activities, as found in this study. However, the enrichment in the PV/PP ratio of GS was not much improved with the lysate over the hepatocytes because the distribution of GS is mostly confined to the last cells around the hepatic venules (Burger et al., 1989) and will not be perturbed much by the contamination of cells from other zonal regions in which GS was absent. Analogously, we cannot exclude the possibility that a shallow gradient exists for EA transporters, although we did not observe a difference in EA uptake among the zonal cells. There also is the possibility that the contents of nonparenchymal cells are sampled during dual-digitonin-pulse perfusion; this may also contribute to the differences seen in the metabolic activities observed in lysates and hepatocyte cytosols.
The aspects of accumulation and efflux of EA-SG were further addressed. Because product inhibition of the GSTs by EA-SG was known to exist (Ploemen et al., 1990, 1993), Mrp2 may play a role in GSH conjugation due to the removal of EA-SG at the canalicular membrane. We have shown that 90% of the formed EA-SG was rapidly excreted into bile of the rat liver preparation (Tirona and Pang, 1999), suggesting that canalicular transport predominates over cellular efflux mediated perhaps by the Mrp family proteins present on the lateral or basolateral membranes (Zaman et al., 1997). Although human MRP2 was found to transport EA-SG (Evers et al., 1998), rat Mrp isoforms have not been directly shown to transport EA-SG. However, there appears to be a close substrate specificity between the human and rat orthologs (Keppler et al., 1998), and ATP-dependent uptake of the typical Mrp2 substrate, dinitrophenyl-GSH conjugate, into rat canalicular membrane vesicles was almost completely inhibited by EA-SG (Ballatori and Truong, 1995). It is hence reasonable to assume that Mrp2 is primarily responsible for the efflux of EA-SG by hepatocytes. From our studies, it was noted that the initial EA-SG efflux rates among PP and PV hepatocytes were similar (Table 2), suggesting homogeneity in Mrp2 function within the liver acinus. The observation is consistent with the lack of zonal differences in the levels of immunoreactive Mrp2 protein as determined by immunoblotting (Fig. 8) and acinar homogeneity in MRP2 described by Kool et al. (1999) in human liver using an immunohistochemical technique. Recently, Morrow et al. (1998)demonstrated that the combined effect of enhanced expression of GST P1-1 and MRP1 in MCF7 breast carcinoma cells conferred resistance to the cytotoxic effects of EA. Elevated expression of either GST P1-1 or MRP1 alone did not cause significant resistance to toxicity. Given the dual presence of GSTs and Mrp2 in hepatocytes throughout the acinus, all hepatocytes are amply endowed with cytoprotective defenses against exposure to electrophiles such as EA.
The greater GSH conjugation over the efflux activity usually suggests an accumulation of EA-SG; however, this was not observed. EA-SG accumulated only transiently at the onset of incubation within isolated zonal hepatocytes due to the cessation of EA-SG formation as GSH became depleted and due to the rapid efflux into extracellular medium (Figs. 3 and 4). It remained plausible that product inhibition of the GSTs ensued during the early time period. The cessation of EA-SG formation and rapid efflux brought about the decline in intracellular concentrations, but EA-SG persisted within the cell after 20 min, albeit at a low and constant level. This observation is consistent with findings on the redistribution or internalization of Mrp2 from the cell surface to endosomes and lysosomes shortly after isolation of the hepatocytes (Oude Elferink et al., 1993; Roelofsen et al., 1995), although in the native intact liver, the majority of Mrp2 protein is localized on the apical membrane. Therefore, the intracellular EA-SG remaining beyond 20 min of incubation most likely signifies that the conjugate is sequestered within the intracellular vesicles (Fig. 1). Indeed, the fitting exercise supports the above conjecture. Absence of the sequestered pool or intracellular vesicles inevitably led to poorer fits to the data.
In conclusion, we demonstrated that the GST activities responsible for the GSH conjugation of EA were high, and that a relatively shallow and increasing acinar gradient existed in rat liver. Despite this zonal heterogeneity in metabolic activity, uptake was uniform throughout the acinus, and this, together with low GSH levels, rate-limited the GSH conjugation of EA. The major efflux system for EA-SG occurs most likely via Mrp2, and acinar homogeneity in Mrp2 function was observed. However, the present study dealt only with uptake and metabolism in hepatocytes. In extrapolating data to the whole liver, one must be cognizant of the impact of nonparenchymal cells (Kupffer, Ito, and endothelial cells), which also contain GSTs, albeit at reduced levels (Steinberg et al., 1989, Parola et al., 1993, Lee et al., 1994). The presence of other cell types and associated heterogeneities might further affect the hepatic GSH conjugation of EA.
Acknowledgments
We thank Dr. Dietrich Keppler (Deutsches Krebsforschungszentrum, Heidelberg, Germany) and Dr. Harry V. Gelboin (National Cancer Institute, National Institutes of Health, Bethesda, MD) for providing us with the antibodies toward Mrp2/cMoat and CYP1A, respectively, and Dr. J. H. T. M. Ploemen (TNO, Ziest, the Netherlands) for supplying the [14C]EA.
Footnotes
-
Send reprint requests to: Dr. K. Sandy Pang, Faculty of Pharmacy, University of Toronto, 19 Russell St., Toronto, Ontario, Canada M5S 2S2. E-mail: pang{at}phm.utoronto.ca
-
↵1 This work was supported by the Medical Research Council of Canada (Grants MA9104 and MT15657). R.G.T. was supported by fellowships from Merck-Frosst Canada, the Pharmaceutical Manufacturers Association of Canada–Health Research Foundation and MRC, and the University of Toronto Open Fellowship. This work was presented in part at the Annual Meeting of the American Association for the Study of Liver Diseases, 1998, Chicago, IL.
-
Received for publication June 11, 1999.
- Abbreviations:
- GSH
- reduced glutathione
- GSSG
- oxidized glutathione
- EA
- ethacrynic acid
- EA-SG
- ethacrynic acid-glutathione conjugate
- CDNB
- 1-chloro-2,4-dinitrobenzene
- PP
- periportal
- PV
- perivenous
- ALT
- alanine aminotransferase
- GS
- glutamine synthetase
- CYP
- cytochrome P-450
- Mrp2
- rat multidrug resistance protein 2
- MRP2
- human multidrug resistance protein 2
- cMoat
- rat canalicular multispecific organic anion transporter
- GST
- glutathioneS-transferase
- Accepted September 2, 1999.
- The American Society for Pharmacology and Experimental Therapeutics