Abstract
Long-chain unsaturated fatty acids inhibit several cytochrome P450 and UDP-glucuronosyltransferase (UGT) enzymes involved in drug metabolism, including CYP2C8, CYP2C9, UGT1A9, UGT2B4, and UGT2B7. Bovine serum albumin (BSA) enhances these cytochrome P450 and UGT activities by sequestering fatty acids that are released from membranes, especially with human liver microsomes (HLM) as the enzyme source. Here, we report the effects of BSA on CYP1A2-catalyzed phenacetin (PHEN) O-deethylation and lidocaine (LID) N-deethylation using HLM and Escherichia coli-expressed recombinant human CYP1A2 (rCYP1A2) as the enzyme sources. BSA (2% w/v) reduced (p < 0.05) the Km values of the high-affinity components of human liver microsomal PHEN and LID deethylation by approximately 70%, without affecting Vmax. The Km (or S50) values for PHEN and LID deethylation by rCYP1A2 were reduced to a similar extent. A fatty acid mixture, comprising 3 μM concentrations each of oleic acid and linoleic acid plus 1.5 μM arachidonic acid, doubled the Km value for PHEN O-deethylation by rCYP1A2. Inhibition was reversed by the addition of BSA. Ki values for the individual fatty acids ranged from 4.7 to 16.7 μM. Single-point in vitro-in vivo extrapolation (IV-IVE) based on the human liver microsomal kinetic parameters obtained in the presence, but not absence, of BSA predicted in vivo hepatic clearances of PHEN O-deethylation and LID N-deethylation that were comparable to values reported in humans, although in vivo intrinsic clearances were underpredicted. Prediction of the in vivo clearances of the CYP1A2 substrates observed here represents an improvement on other experimental systems used for IV-IVE.
Introduction
Data from in vitro kinetic studies are widely used to predict in vivo hepatic clearance (CLH), extraction ratio, and drug-drug interaction potential for compounds metabolized by cytochrome P450 (P450) and UGT enzymes using in vitro-in vivo extrapolation (IV-IVE) approaches (Houston 1994; Iwatsubo et al., 1997; von Moltke et al., 1998; Obach 1999; Proctor et al., 2004; Riley et al., 2005; Miners et al., 2006, 2010; Hosea et al., 2009). Typically, however, IV-IVE generally results in underprediction of CLH and the magnitude of inhibitory drug-drug interactions (Boase and Miners, 2002; Ito et al., 2004; Ito and Houston, 2005; Miners et al., 2010). Overestimation of in vitro Km and Ki values contribute to the underprediction of the in vivo kinetic parameters.
However, recent studies from this and other laboratories have demonstrated that the predictivity of IV-IVE for substrates (and inhibitors) of several P450 and UGT enzymes is improved significantly when in vitro kinetic parameters are generated from incubations performed in the presence of bovine serum albumin (BSA), fatty acid-free human serum albumin, or human intestinal fatty acid-binding protein. Addition of the individual proteins to incubations of human liver microsomes (HLM) or recombinant enzyme enhances activity, primarily via a decrease in Km and hence an increase in the in vitro intrinsic clearance (CLint). Increased activity in the presence of BSA (2% w/v), fatty acid-free human serum albumin (2% w/v), and/or intestinal fatty acid binding protein (1% w/v) has been demonstrated for CYP2C8 (Wattanachai et al., 2011), CYP2C9 (Tang et al., 2002; Rowland et al., 2008a; Kilford et al., 2009), UGT1A9 (Rowland et al., 2008b), UGT2B4 (Raungrut et al., 2010), and UGT2B7 (Rowland et al., 2007, 2009; Kilford et al., 2009). Furthermore, the Ki values of inhibitors of UGT2B4 and UGT2B7 are reduced when incubations are supplemented with BSA (Rowland et al., 2006; Uchaipichat et al., 2006; Raungrut et al., 2010). The “albumin effect” appears to arise from sequestration of long-chain polyunsaturated fatty acids released from membranes during the course of an incubation. In particular, it has been demonstrated that arachidonic acid, linoleic acid, and oleic acid are potent inhibitors of human liver (or kidney) microsomal and recombinant CYP2C8, CYP2C9, UGT1A9, and UGT2B7 (Yamazaki and Shimada, 1999; Tsoutsikos et al., 2004; Yao et al., 2006; Rowland et al., 2007, 2008a, 2009; Wattanachai et al., 2011).
CYP1A2 plays an important role in the metabolism of many drugs including caffeine N-demethylation (Tassaneeyakul et al., 1992, 1994), lidocaine N-demethylation (Wang et al., 2000; Orlando et al., 1994), phenacetin O-deethylation (Tassaneeyakul et al., 1993), and the N-demethylation of theophylline and other dimethylxanthines (Tassaneeyakul et al., 2004). Furthermore, CYP1A2 has the capacity to metabolize arachidonic acid (Choudhary et al., 2004), and arachidonic acid and linoleic acid are known competitive inhibitors of CYP1A2 (Yamazaki and Shimada, 1999; Yao et al., 2006). Given that the albumin effect is proposed to arise from sequestration of inhibitory membrane long-chain unsaturated fatty acids, particularly arachidonic acid, we hypothesized that supplementation of incubations of HLM and recombinant human CYP1A2 with BSA will decrease the Km values of CYP1A2 substrates and improve in vivo clearance prediction using IV-IVE.
As noted above, phenacetin (PHEN) and lidocaine (LID) are known substrates of CYP1A2. In particular, the high-affinity components of human liver microsomal PHEN and LID deethylation have been shown to be catalyzed exclusively by CYP1A2 (Tassaneeyakul et al., 1993; Venkatakrishnan et al., 1998; Wang et al., 2000). Thus, the present study investigated the effects of BSA (2% w/v) on the kinetics of PHEN O-deethylation and LID N-deethylation by HLM and recombinant human CYP1A2 (rCYP1A2), and IV-IVE was used to predict in vivo clearances from kinetic data generated with HLM as the enzyme source in the absence and presence of BSA. In addition, inhibition of rCYP1A2 by arachidonic acid, linoleic acid, and oleic acid and reversal of inhibition by BSA was demonstrated.
Materials and Methods
Chemicals and Reagents.
PHEN, acetaminophen (APAP), lidocaine hydrochloride monohydrate, monoethylglycinexylidide (MEGX), BSA (product number A7906), arachidonic acid (C20: 4n–6, sodium salt), linoleic acid (C18:2n–6, sodium salt), and oleic acid (C18:1n–9, sodium salt) were purchased from Sigma-Aldrich (Sydney, Australia). All other reagents and solvents were of analytical reagent grade.
Human Liver Microsomes and Recombinant CYP1A2.
Human livers (HL7, HL10, HL12, HL29, and HL40) were obtained from the human liver bank of the Department of Clinical Pharmacology, Flinders University. Approval was obtained from the Flinders Medical Centre Research Ethics Committee and Khon Kaen University for the use of human liver tissues in xenobiotic metabolism studies. Microsomes were prepared by differential centrifugation, as described previously (Bowalgaha et al., 2005). rCYP1A2 and rat NADPH cytochrome P450 oxidoreductase were coexpressed in Escherichia coli according to the general procedure of Boye et al. (2004). The CYP1A2/NADPH cytochrome P450 oxidoreductase ratio was unity.
Phenacetin O-Deethylation Assay.
PHEN O-deethylation (i.e., the conversion of PHEN to APAP) was measured according to Tassaneeyakul et al. (1993), with modifications. In brief, incubations (200 μl) contained phosphate buffer (0.1 M, pH 7.4), PHEN (2–2000 μM), and human liver microsomes (HLM) (0.2 mg/ml). For incubations conducted in the presence of BSA (2% w/v), HLM protein content was reduced to 0.15 mg/ml. After preincubation for 5 min, reactions were initiated by the addition of an NADPH-generating system (1 mM NADP, 10 mM glucose 6-phosphate, 2 IU/ml glucose-6-phosphate dehydrogenase, and 5 mM MgCl2). Reactions were performed for 25 min at 37°C in a shaking water bath and then were terminated by the addition of ice-cold acetonitrile (600 μl) and cooling on ice for 10 min. Samples were centrifuged at 18,000g for 5 min, and the supernatant layer (500 μl) was evaporated to dryness using a TurboVap LV concentration evaporator (Caliper Life Sciences, Hopkinton, MA) at 50°C. The residue was reconstituted with 120 μl of 20% methanol, and an aliquot (10 μl) was injected onto the HPLC column (Waters, Milford, MA). HPLC conditions were as described previously (Polasek et al., 2006a). The retention times for APAP and PHEN under these conditions were 3.3 and 13 min, respectively. APAP formation was quantified using a standard curve in the range 0.2 to 6 μM. The overall assay within-day reproducibility was determined by measuring APAP formation by the same batch of HLM on nine occasions; coefficients of variation were 5.2 and 2.0% at PHEN concentrations of 10 and 2000 μM, respectively.
Incubations with rCYP1A2 as the enzyme source were as described above for HLM, except for rCYP1A2 content (5 pmol of P450/ml), PHEN concentration (2–80 μM), incubation time (7.5 min), and standard curve range (0.2–4 μM). For incubations conducted in the presence of BSA (2% w/v), rCYP1A2 P450 content was reduced to 3.5 pmol of P450/ml, and the substrate concentration range was 2 to 30 μM.
Lidocaine N-Deethylation Assay.
LID O-deethylation activity was determined as MEGX formation. Incubations, in a total volume of 200 μl, contained phosphate buffer (0.1 M, pH 7.4), HLM (0.4 mg/ml), BSA (0 or 2% w/v), and LID (25–8000 μM). After a 5-min preincubation, reactions were initiated by the addition of a NADPH-regenerating system (1 mM NADP, 10 mM glucose 6-phosphate, 2 IU/ml glucose-6-phosphate dehydrogenase, and 5 mM MgCl2). Incubations were performed at 37°C in a shaking water bath for 15 min. Reactions were terminated by the addition of perchloric acid (2 μl, 70% v/v) and cooling on ice. Samples were subsequently centrifuged at 5000g for 10 min at 10°C, and a 7-μl aliquot of the supernatant fraction was injected directly onto the HPLC column. Similar conditions were used for experiments with rCYP1A2 as the enzyme source, except for CYP1A2 content (5 pmol of P450/ml).
MEGX formation was determined by HPLC. Analytes were separated on a Synergi Hydro RP analytical column (150 × 3.0 mm, 4-μm particle size; Phenomenex, Sydney, Australia) using a mobile phase comprising 20 mM phosphate buffer (pH 4.6) with 11% acetonitrile, at a flow rate of 1.0 ml/min. Column eluant was monitored by UV absorbance at 205 nm. Retention times for MEGX and LID were 4.4 and 7.0 min, respectively. MEGX formation was quantified by comparison of peak areas with those of an authentic MEGX standard curve prepared over the concentration range 1 to 50 μM. Overall assay reproducibility, assessed by measuring MEGX formation by the same batch of HLM on 10 occasions, was 4.3 and 2.0% for LID concentrations of 50 and 8000 μM, respectively.
Binding of Phenacetin and Lidocaine to HLM, E. coli, and BSA.
The binding of PHEN and LID to HLM and E. coli membranes (at the protein concentrations used in incubations) and to mixtures of BSA (2% w/v) with each enzyme source was measured by equilibrium dialysis using a Dianorm (Munich, Germany) equilibrium dialysis apparatus according to the general method of McLure et al. (2000). The concentrations of PHEN and LID used in dialysis experiments spanned the concentration ranges used in kinetic experiments. The dialysis cell assembly was immersed in a water bath maintained at 37°C and rotated at 12 rpm for 4 h. A 200-μl aliquot was collected from each compartment, protein was precipitated using 200 μl of acetic acid (4%) in methanol (PHEN) or ice-cold methanol alone (LID), and the sample was cooled on ice. Samples were subsequently centrifuged at 4000g for 10 min at 10°C, and an aliquot of the supernatant fraction (5 μl) was analyzed by HPLC.
HPLC conditions for the measurement of PHEN were as described above, except that the mobile phase comprised 5 mM sodium acetate buffer (pH 4.3) containing 30% acetonitrile (isocratic elution). The retention time for PHEN was 3.6 min. Separation of LID was achieved on a NovaPak C18 analytical column (3.9 × 150 mm, 4 μm particle size; Waters) using a mobile phase comprising 2 mM triethylamine (pH adjusted to 2.5 with perchloric acid) with 20% acetonitrile, delivered at a flow rate of 1 ml/min. Column eluant was monitored at 205 nm. The retention time for LID under these conditions was 2.8 min. PHEN and LID concentrations in dialysis samples were determined by comparison of peak areas with those of standard curves that spanned the concentration ranges observed in dialysis samples. Recovery of PHEN and LID in equilibrium dialysis experiments ranged from 94 to 102%.
Kinetic Analysis of Phenacetin O-Deethylation and Lidocaine N-Deethylation.
All kinetic experiments were performed in duplicate, and data points represent the mean of the duplicate measurements (<10% variance). The single- and two-enzyme Michaelis-Menten equations (see below) were fit to the kinetic data for APAP and MEGX formation using EnzFitter (version 2.0.18.0; Biosoft, Cambridge, UK) on the basis of the unbound PHEN and LID concentrations present in incubations. Goodness of fit to each equation was assessed from comparison of the parameter S.E. of fit, 95% confidence intervals, coefficient of determination (r2), and F-statistic. Statistical comparisons of kinetic parameters (t test or Mann-Whitney rank-sum test) were performed using SigmaStat version 3.11. Values of p < 0.05 were considered significant.
The Michaelis-Menten equation is as follows: where v is the rate of metabolite formation, Vmax is the maximum velocity, Km is the Michaelis constant (substrate concentration at 0.5 Vmax), and [S] is the substrate concentration. The two-enzyme Michaelis-Menten equation is as follows: where Vmax1 and Vmax2 are the maximum velocities of the high- and low-affinity component, and Km1 and Km2 are the Michaelis constants of the high- and low-affinity component.
Inhibition of Phenacetin O-Deethylation by Fatty Acids.
To investigate the effects of fatty acids on PHEN O-deethylase activity, a fatty acid mixture (FAM) comprising 3 μM concentrations each of oleic acid and linoleic acid and 1.5 μM arachidonic acid, which corresponds to 5% of their known content in HLM (Rowland et al., 2007), was added to an incubation mixture (200 μl) containing phosphate buffer (0.1 M, pH 7.4), rCYP1A2 (5 pmol P450/ml), and PHEN (2–80 μM). The sample preparation and HPLC conditions were as described for the PHEN O-deethylation assay. Subsequent experiments were performed to determine the type and potency of phenacetin O-deethylation inhibition by the individual fatty acids. PHEN and arachidonic acid, linoleic acid, or oleic acid were added to an incubation mixture containing rCYP1A2 (5 pmol of P450/ml) as the enzyme source. Inhibition of PHEN O-deethylation was assessed at each of three substrate concentrations (7.5, 15, and 45 μM) at four fatty acid concentrations in the following ranges: arachidonic acid, 3.75 to 30 μM; linoleic acid, 5 to 25 μM; and oleic acid, 4 to 20 μM. Sample preparation and HPLC conditions were as described for the PHEN O-deethylation assay. Inhibition data were modeled using the expression for competitive inhibition.
Prediction of Phenacetin and Lidocaine Intrinsic and Hepatic Clearances via O- and N-Deethylation.
Estimation of clearances from in vivo studies.
Pharmacokinetic parameters from clinical studies were used to calculate the in vivo intrinsic clearance (CLint, in vivo) and CLH of PHEN and LID via the O- and N-deethylation pathways. In vivo hepatic clearance was determined as CLH = fm × CL, where CL is systemic clearance and fm is the fraction of the intravenous dose of PHEN and LID metabolized via O- and N-deethylation, respectively. In vivo CLint from clinical studies was then calculated using the following equation: where QH is liver blood flow (taken as 90 l/h) and fub is the fraction of drug unbound in blood. Values of fm for PHEN and LID deethylation were taken as 0.95 and 0.70, respectively, based on metabolite recovery in urine (Nation et al., 1977; Veronese et al., 1985). The fraction of drug unbound in blood was calculated as fup/RB, where fup is the fraction unbound in plasma and RB is the blood/plasma concentration ratio. For PHEN, fup was taken as 0.60 (Vesell et al., 1975) and RB as 1.01 (Shibata et al., 2002), whereas for LID fup was 0.30 (Wing et al., 1984) and RB was 0.84 (Shibata et al., 2002). These values provided unbound fractions of PHEN and LID in blood of 0.59 and 0.36, respectively.
In vitro-in vivo extrapolation.
Single-point predictions of the intrinsic and hepatic clearances of PHEN and LID via O- and N-deethylation, respectively, were calculated by scaling in vitro kinetic parameters. With the use of data generated with HLM as the enzyme source, whole liver intrinsic clearance was calculated as follows: where MPPGL is the mass of microsomes per gram of human liver tissue, taken as 38 mg/g (Rowland et al., 2008a) and LW is the average weight of a human liver (1500 g). Microsomal intrinsic clearances (Vmax/Km) for the high- and low-affinity components of each deethylation reaction were summed. The in vivo hepatic clearances of PHEN and LID by deethylation were subsequently determined using the equation for the well stirred model, for which terms have been defined previously.
Results
Binding of Phenacetin and Lidocaine to HLM, E. coli Membranes, and BSA.
The binding of PHEN and LID to HLM and E. coli membranes in the presence and absence of 2% BSA (w/v) is reported as the unbound fraction present in the incubation mixture (fuinc). Binding of PHEN and LID to HLM and E. coli membranes was negligible (mean fuinc range 0.98–1.02) in the absence of BSA. However, PHEN and LID bound significantly to mixtures of HLM (or E. coli membranes) plus BSA (2% w/v). Mean fuinc values for PHEN binding to HLM plus BSA and E. coli membranes plus BSA were 0.78 ± 0.02 and 0.80 ± 0.02, respectively, whereas the mean fuinc value for the binding of LID to the HLM-BSA mixture was 0.88 ± 0.03. The binding of both PHEN and LID to the enzyme source/BSA mixture was concentration-independent over the substrate concentration ranges investigated. PHEN and LID concentrations were corrected for binding to incubation constituents, and kinetic parameters were calculated on the basis of the free drug concentration in vitro.
Kinetics of Phenacetin O-Deethylation.
The conversion of PHEN to APAP by HLM in both the absence and presence of BSA exhibited two-enzyme Michaelis-Menten kinetics (Fig. 1), whereas APAP formation by rCYP1A2 exhibited hyperbolic (Michaelis-Menten) kinetics (Fig. 1). Derived kinetic constants are shown in Table 1. Addition of 2% BSA to incubations resulted in a 76% reduction (p < 0.05) in the mean Km value of the high-affinity component (Km1) for PHEN O-deethylation, but BSA was without effect on Vmax1. Of interest, the Km for the high-affinity component of PHEN O-deethylation by HL29 was 3.8-fold higher than the mean Km (namely, 17.8 μM) for the other four livers (Fig. 1, top right; Table 1), although in the presence of BSA the Km was reduced to a value that was essentially identical to that in the other livers. Although BSA had no effect on Km2, there was a significant increase in Vmax for the low-affinity reaction. Like the high-affinity component of human liver microsomal PHEN O-deethylation, supplementation of incubations with BSA resulted in a 72% decrease in the Km with rCYP1A2 as the enzyme source without affecting Vmax (Table 1).
Kinetics of Lidocaine N-Deethylation.
Representative kinetic plots for MEGX formation by LID N-deethylation by HLM, with and without 2% BSA, are shown in Fig. 2, and derived kinetic parameters are summarized in Table 2. MEGX formation in the absence of BSA followed two-enzyme Michaelis-Menten kinetics with microsomes from four of the five livers, whereas LID N-deethylation by microsomes from HL29 exhibited hyperbolic (Michaelis-Menten) kinetics. Of note, the high-affinity component of microsomal MEGX formation was not observed with HL29. However, in the presence of 2% BSA, two-enzyme Michaelis-Menten kinetics was observed for all five livers. For microsomes from HL7, HL10, HL12, and HL40, addition of 2% BSA to incubations decreased the mean Km for the high-affinity reaction (Km1) by 68% (p < 0.05), without a significant effect on Vmax. BSA was without effect on the kinetic parameters for the low-affinity reaction (Table 2). In contrast to PHEN O-deethylation, LID N-deethylation by rCYP1A2 exhibited negative cooperative Hill kinetics (Fig. 2). The mean S50 was reduced by approximately 80% in the presence of BSA (Table 2). There was also a significant reduction in mean Vmax (by 29%), but the Hill coefficient (n) was unchanged.
Inhibition of Phenacetin O-Deethylation by Unsaturated Long-Chain Fatty Acids.
The effects of the FAM on the kinetics of PHEN O-deethylation by E. coli-expressed CYP1A2 in the absence and presence of BSA are shown in Fig. 3. The FAM comprised 3 μM oleic acid and linoleic acid and 1.5 μM arachidonic acid, which corresponds to approximately 1/20 of the content of these fatty acids present in HLM (Rowland et al., 2008b). In the absence of BSA, the FAM doubled Km (13.6–27.4 μM) without affecting Vmax (39.7 versus 39.9 pmol/min · pmol P450). In contrast, Km (3.8 versus 4.7 μM) and Vmax (46.5 versus 53.9 pmol/min · pmol P450) values determined in the presence of BSA (2% w/v) were similar between incubations conducted with and without the FAM.
Each of the three unsaturated long-chain fatty acids inhibited PHEN O-deethylation by rCYP1A2 in a competitive manner (Fig. 4). The Ki values for arachidonic acid, linoleic acid, and oleic acid inhibition of rCYP1A2-catalyzed PHEN O-deethylation were 4.7 ± 0.3, 13.6 ± 1.0, and 16.7 ± 2.0 μM (parameter ± S.E. of parameter fit), respectively.
In Vitro-In Vivo Extrapolation of Phenacetin O-Deethylation and Lidocaine N-Deethylation Intrinsic Clearance.
In vitro intrinsic clearances for PHEN and LID deethylation were scaled to CLint, liver and CLH following the procedure described under Materials and Methods. Predicted values of CLint, liver for PHEN O-deethylation calculated from kinetic data determined in the absence and presence of BSA were 102 and 403 l/h, respectively, whereas predicted values of CLH were 36 and 65 l/h, respectively. Likewise, predicted values of CLint, liver for LID N-deethylation calculated from kinetic data generated in the absence and presence of BSA were 28 and 75 l/h, respectively, whereas predicted values of CLH were 9.1 and 21 l/h, respectively.
Discussion
It has been reported that CYP1A2 is inhibited by long-chain unsaturated fatty acids, including arachidonic acid and linoleic acid (Yamazaki and Shimada, 1999; Yao et al., 2006). Because enhancement of the activities of several P450 and UGT enzymes by BSA apparently arises from sequestration of inhibitory fatty acids, including arachidonic, linoleic, oleic, or other long-chain unsaturated fatty acids depending on the enzyme source (Rowland et al., 2007, 2008a,b; Wattanachai et al., 2011), the present study investigated the effects of BSA on the high-affinity components of human liver microsomal PHEN and LID deethylation, and rCYP1A2 catalyzed PHEN O-deethylation and LID N-deethylation. Supplementation of incubations of HLM with BSA (2%, w/v) resulted in an approximate 70% reduction in the mean Km values for the high-affinity components of PHEN and LID deethylation without an effect on Vmax. BSA reduced the Km values for PHEN O-deethylation by rCYP1A2 to a similar extent. Of interest, LID N-deethylation by rCYP1A2 exhibited negative cooperative Hill kinetics in contrast to the Michaelis-Menten kinetics observed for PHEN O-deethylation. Nevertheless, a significant reduction in the S50 for LID N-deethylation by rCYP1A2 still occurred in the presence of BSA.
Similar to the demonstrated inhibitory effects of long-chain unsaturated fatty acids on CYP2C8, CYP2C9, UGT1A9, and UGT2B7 activities (Tsoutsikos et al., 2004; Yao et al., 2006; Rowland et al., 2007, 2008a,b; Wattanachai et al., 2011), arachidonic acid, linoleic acid, and oleic acid were shown here to competitively inhibit rCYP1A2-catalyzed PHEN O-deethylation with Ki values ranging from 4.7 to 16.7 μM. Furthermore, a combination of these fatty acids doubled the Km of PHEN O-deethylation. Inhibition of PHEN O-deethylation by the fatty acid mixture was reversed in the presence of BSA, consistent with the hypothesis that BSA sequesters inhibitory long-chain unsaturated fatty acids released from membranes during the course of an incubation. It should be noted that heptadecanoic acid is the predominant long-chain unsaturated fatty acid present in E. coli membranes, with lower proportions of C18 and C20 unsaturated acids (Rowland et al., 2008a). Although heptadecanoic acid appears to be a weaker inhibitor of P450 enzymes than C18 and C20 unsaturated acids (Rowland et al., 2008a), the Ki values obtained here for arachidonic, linoleic, and oleic acids are almost certainly overestimated because of the simultaneous release of inhibitory long-chain unsaturated fatty acids from the E. coli membranes.
The results of the present study are in agreement with the effects of BSA on the Km values of substrates of CYP2C8 (Wattanachai et al., 2011), CYP2C9 (Rowland et al., 2008a; Kilford et al., 2009), UGT1A9 (Rowland et al., 2008b), UGT2B4 (Raungrut et al., 2010), and UGT2B7 (Rowland et al., 2007, 2009). As observed here for LID N-deethylation by rCYP1A2, BSA has also been reported to marginally alter the Vmax of paclitaxel 6α-hydroxylation by recombinant CYP2C8 (Wattanachai et al., 2011), codeine 6-glucuronidation by recombinant UGT2B4 (Raungrut et al., 2010), and entacapone glucuronidation by UGT1A9 (Manevski et al., 2011). However, BSA does not reduce the Km values of substrates of UGT1A1, UGT1A4, and UGT1A6 (Rowland et al., 2006, 2008b), presumably because these enzymes are not inhibited by long-chain unsaturated fatty acids. Effects of BSA on the activities of hepatic drug-metabolizing P450s other than CYP1A2, CYP2C8, and CYP2C9 are yet to be investigated directly. It has been reported, however, that arachidonic acid is a potent inhibitor of CYP2C19 (Yamazaki and Shimada 1999; Yao et al., 2006), and hence it might be speculated that BSA will cause a reduction in the Km values for substrates of this enzyme. In contrast, arachidonic acid is only a weak inhibitor of CYP2A6, CYP2B6, CYP2D6, CYP2E1, and CYP3A4 (Yamazaki and Shimada 1999; Yao et al., 2006). Consistent with weak inhibition of CYP3A4 by arachidonic acid, kinetic constants for the low-affinity component of human liver microsomal LID N-deethylation (catalyzed predominantly by CYP3A4; see later discussion) were unaltered by BSA in the present study.
In the absence of BSA, the respective Km values of the high-affinity component of PHEN O-deethylation by HLM and rCYP1A2 (11.4–67.8 and 13.6 μM) were similar to previously published ranges (9–38.4 and 12.3–30.9 μM) (Gillam and Reilly, 1988; Tassaneeyakul et al., 1993; Venkatakrishnan et al., 1998; Kobayashi et al., 1999; Polasek et al., 2006b; Donato et al., 2010). Likewise, Km values of the high-affinity component of LID N-deethylation by HLM (135–246 μM) were similar to data reported by Wang et al. (2000). Kinetic data for PHEN O-deethylation and LID N-deethylation by microsomes from liver HL29 warrant comment. The Km value for the high-affinity component of PHEN O-deethylation by HL29 was 3- to 6-fold higher than those for other livers, although the Km determined in the presence of BSA was essentially identical to the Km values of the other four livers. Of interest, a high-affinity component of LID N-deethylation was not observed with HL29. However, the high-affinity reaction was observed in the presence of BSA and, as with PHEN O-deethylation, the Km was similar to those for the other livers. The reason for these anomalous results is not clear because the donor of HL29 was not receiving agents known to alter P450 activity and liver pathology was normal. Observation of the expected kinetic behavior in the presence of BSA could suggest greater release of inhibitory fatty acids from microsomes prepared from HL29.
There was no statistically significant effect of BSA on the Km values of the high-affinity components of PHEN and LID deethylation. BSA caused a 50% increase in the Vmax of the low-affinity component of PHEN O-deethylation, but there was no effect on the Vmax for the low-affinity component of LID N-deethylation. Multiple P450 enzymes contribute to the low-affinity component of PHEN O-deethylation (Venkatakrishnan et al., 1998; Kobayashi et al., 1999), whereas CYP3A4 appears to be the main enzyme responsible for the low-affinity component of LID N-deethylation (Wang et al., 2000).
Single-point IV-IVE was used to compare the prediction of in vitro kinetic parameters generated in the absence and presence of BSA. Mean values of CLint, liver for PHEN O-deethylation calculated from kinetic data determined in the absence and presence of BSA were 102 and 403 l/h, respectively, whereas predicted mean values of CLH were 36 and 65 l/h, respectively. Use of data reported by Raaflaub and Dubach (1975) after intravenous administration of PHEN (250 mg) to healthy volunteers provided mean values of 933 and 71 l/h for CLint, in vivo and CLH, respectively. Thus, kinetic data generated in the presence of BSA predicted the hepatic clearance of phenacetin via O-deethylation well, although in vivo intrinsic clearance was underpredicted. Mean values of CLint, liver for LID N-deethylation calculated from kinetic parameters generated in the absence and presence of BSA were 28 and 75 l/h, respectively, whereas predicted values of CLH were 9.1 and 21 l/h, respectively. CLint, in vivo and in vivo CLH were calculated from data reported for healthy volunteers administered intravenous LID (1–1.5 mg/kg) in four recent kinetic studies (Orlando et al., 2003, 2004; Olkkola et al., 2005; De Martin et al., 2006). Mean values of CLint, in vivo and in vivo CLH from the four studies ranged from 142 to 198 and from 32 to 40 l/h, respectively. Whereas in vitro kinetic data determined in the presence of BSA gave higher values of CLint, liver and predicted CLH, both parameters were lower than the reported corresponding in vivo clearances.
In summary, this study has demonstrated that supplementation of incubations with BSA (2% w/v) enhances human liver microsomal and rCYP1A2 activities because of an approximate 70% reduction in the Km (or S50) values of the probe substrates (PHEN and LID). This effect results from BSA sequestration of inhibitory long-chain unsaturated fatty acids that are apparently released from membranes of the enzyme source during the course of in vitro incubations. Single-point IV-IVE based on kinetic parameters generated from incubations of HLM supplemented with BSA predicted the in vivo CLH of the high-clearance drug PHEN well, but the in vivo CLH for LID N-deethylation was underpredicted by 50 to 100%. Nevertheless, prediction of the in vivo clearances of the CYP1A2 substrates observed here represents an improvement on other experimental systems used for IV-IVE.
Authorship Contributions
Participated in research design: Tassaneeyakul, Knights, and Miners.
Conducted experiments: Wattanachai, Rowland, Elliot, and Bowalgaha.
Performed data analysis: Wattanachai, Tassaneeyakul, Rowland, and Miners.
Wrote or contributed to the writing of the manuscript: Wattanachai, Tassaneeyakul, Rowland, Knights, and Miners.
Acknowledgments
We thank Dr. Benjamin Lewis for providing the recombinant CYP1A2 used in kinetic experiments.
Footnotes
This study was supported by the National Health and Medical Research Council of Australia [Grant 480417]; the Thailand Research Fund; the Higher Education Research Promotion and National Research University Project of Thailand; and the Office of the Higher Education Commission through the Health Cluster (SHeP-GMS), Khon Kaen University. N.W. and W.T. received support from the Thailand Research Fund through the Royal Golden Ph.D. Program [PHD/0167/2548].
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
ABBREVIATIONS:
- CLH
- hepatic clearance
- P450
- cytochrome P450
- UGT
- UDP-glucuronosyltransferase
- IV-IVE
- in vitro-in vivo extrapolation
- BSA
- bovine serum albumin
- HLM
- human liver microsomes
- PHEN
- phenacetin
- LID
- lidocaine
- rCYP1A2
- recombinant human CYP1A2
- APAP
- acetaminophen
- MEGX
- monoethylglycinexylidide
- HPLC
- high-performance liquid chromatography
- FAM
- fatty acid mixture.
- Received November 30, 2011.
- Accepted February 7, 2012.
- Copyright © 2012 by The American Society for Pharmacology and Experimental Therapeutics