Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Phenacetin [N-(4-ethoxyphenyl)acetamide] is widely used as an analgesic antipyretic and is converted to acetaminophen and p-phenetidine by oxidative O-deethylation and deacetylation, respectively, in humans.

Cytochrome P450 in human liver microsomes is responsible for the O-deethylation of phenacetin (Boobis et al., 1981). It is generally accepted that cytochrome P450 (CYP) 1A2 is specifically involved in the in vitro O-deethylation of phenacetin at low concentrations of the substrate (Butler et al., 1989; Sesardic et al., 1990; Bourrié et al., 1996; Rodrigues et al., 1997), although the in vitro kinetics of the reaction in human liver microsomes is biphasic with high- and low-affinity components, suggesting that the phenacetin O-deethylation is mediated by at least two different enzymes with distinct affinities (Boobis et al., 1981; Tassaneeyakul et al., 1993; von Moltke et al., 1996). Recently, Venkatakrishnan et al. (1998) reported that the phenacetin O-deethylation in human liver microsomes is catalyzed by six distinct cytochrome P450 isoforms: CYP1A2, -2A6, -2C9, -2C19, -2D6, and -2E1. CYP1A2 is the only high-affinity isoform and accounts for more than 80% of the reaction rate at low substrate concentrations. Furthermore, CYP2C9 contributes significantly to the reaction at high concentrations of substrate, with minor contributions made by CYP2A6, -2C19, -2D6, and -2E1.

In contrast to the many studies on the cytochrome P450-mediated reaction of the phenacetin O-deethylation, only limited information is available regarding the deacetylation of phenacetin, particularly in human tissue preparations. Hosokawa et al. (1995) elucidated the interindividual variation in phenacetin deacetylation in human liver microsomes and reported a 5.3-fold difference in catalytic activity among the samples from 12 different donors. Takai et al. (1997) isolated two carboxylesterases (i.e., carboxylesterase pI 4.5 and pI 5.3) from human liver and characterized their activity in the hydrolysis of a wide variety of ester- and amide-type drugs, including phenacetin. However, these two enzymes are not responsible for the phenacetin deacetylation, although they could catalyze the hydrolysis of various ester-type drugs and an amide-type drug, aniracetam, to both anisic acid and anisamidobutyric acid.

In general, it is considered that most carboxylic ester hydrolases (carboxylesterases) can also hydrolyze aromatic amides (Heymann, 1980; Satoh, 1987), based on findings using rat liver preparations (Mentlein et al., 1980; Heymann et al., 1981; Heymann and Mentlein, 1981; Hosokawa et al., 1987; Luan et al., 1997) and liver preparations from various animals (Hosokawa et al., 1990) as an enzyme source. Carboxylesterase is a member of the serine hydrolase superfamily (Satoh and Hosokawa, 1995) and has many distinct isoforms (Satoh and Hosokawa, 1998). Several carboxylesterase isoforms were isolated from human liver, and their hydrolytic activities were characterized (Ketterman et al., 1989; Probst et al., 1991; Brzezinski et al., 1994, 1997; Kamendulis et al., 1996; Pindel et al., 1997; Zhang et al., 1999). However, the catalytic properties of these carboxylesterases in phenacetin deacetylation are not known. In addition, no information is available, to our knowledge, regarding the intracellular distribution of phenacetin deacetylase activity in the liver and the extrahepatic distribution of the activity in human tissues, although it is generally believed that most carboxylesterase activity is found in the liver microsomal fraction, based on findings in rat liver (Mentlein et al., 1988).

In this study, we examined the microsomal and cytosolic distribution of phenacetin deacetylase activity in human liver and kidneys. The in vitro kinetic properties of the phenacetin deacetylation were also studied in human liver microsomes. In addition, chemical inhibition and stimulation studies in vitro were conducted to investigate the enzymatic and catalytic properties of phenacetin deacetylase in human liver microsomes.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals and Reagents. Phenacetin, p-phenetidine, p-toluidine hydrochloride, acetanilide, aniline sulfate, p-nitrophenyl acetate, p-nitrophenyl propionate, prazosin hydrochloride, acetylsalicylic acid, o-acetamidophenol, p-acetamidophenol, clofibrate, sodium phenobarbital, and procaine hydrochloride were purchased from Wako Pure Chemical Industries (Osaka, Japan). Flutamide, phenylmethylsulfonyl fluoride (PMSF), aniracetam, captopril, dilazep dihydrochloride, and oxybutynin chloride were purchased from Sigma Chemical Co. (St. Louis, MO). Procainamide hydrochloride was obtained from Aldrich Chemical Company, Inc. (Milwaukee, WI). Bis(4-nitrophenyl)phosphoric acid (BNPP) was obtained from Tokyo Kasei Kogyo Co., Ltd. (Tokyo, Japan). Sodium 1-hexanesulfonate was obtained from Nacalai Tesque, Inc. (Kyoto, Japan). All other reagents and solvents were of analytical grade.

Human Tissue Preparations. Pooled human liver microsomes and cytosol from 10 donors and individual human kidney microsomes and cytosol from 3 donors were supplied by Biomedical Research Institute, HAB Discussion Group (Chiba, Japan). Human livers and kidneys were homogenized with 0.25 M sucrose containing 3 mM Tris and 0.1 mM EDTA (pH 7.4), and then the microsomes and cytosol were isolated by differential centrifugation using an ordinary method. Washed microsomes were resuspended in 100 mM Tris-HCl buffer (pH 7.4) containing 1 mM EDTA and 20% glycerol at protein concentrations of 20 mg/ml for the livers and 9 to 11 mg/ml for the kidneys. Cytosolic protein concentrations in the liver and kidneys were 15 and 9 to 12 mg/ml, respectively. Kidney microsomes from individual donors were combined to prepare the pooled subcellular fraction. Cytosol was combined in the same way. Human tissue preparations were all stored at 80°C until use.

Phenacetin Deacetylation. The phenacetin deacetylase activity in human tissue preparations was assayed by the transformation of phenacetin to p-phenetidine. In the experiments on the subcellular distribution of the activity, a 0.5-ml reaction mixture containing 2 mg/ml microsomal or cytosolic protein and 5 mM phenacetin in 50 mM Tris-HCl buffer (pH 8.6) was incubated for 30 min at 37°C.

Acetanilide Deacetylation. The acetanilide deacetylase activity was assayed by determining aniline formation. Incubations contained acetanilide (10 mM) and 2 mg/ml microsomal or cytosolic protein in a total volume of 0.5 ml of 50 mM Tris-HCl buffer (pH 8.6) and were performed at 37°C for 30 min as described by Heymann et al. (1981). The reaction was initiated by the addition of the substrate and terminated by the addition of 0.5 ml of 1 N HCl containing 100 µg/ml p-toluidine hydrochloride, which was used as an internal standard. After centrifugation (10 min at 21,600g), aniline and the internal standard were extracted by solid phase as described in the section regarding phenacetin deacetylation. Then, 30 µl of the resultant extract was analyzed by an HPLC assay. The activity of acetanilide deacetylation was expressed as nanomoles per minute per milligram of protein.

p-Nitrophenylacetate and p-Nitrophenylpropionate Hydrolysis. The activities of p-nitrophenylacetate and p-nitrophenylpropionate hydrolase were colorimetrically determined as described by Heymann et al. (1981). The rate of p-nitrophenol formation was expressed as micromoles per minute per milligram of protein.

HPLC Analysis. The HPLC apparatus included a model 510 HPLC pump (Waters), a model 717 auto sample processor (Waters), a model 680 solvent programmer (Waters), a model 486 tunable absorbance detector (Waters), a model 503 degasser (M&S Instruments Trading Inc., Tokyo, Japan), and a Chromatopac C-R6A (Shimadzu Ltd., Kyoto, Japan). A TSK-gel ODS-80_TS column (5-µm particle size, 4.6 mm i.d. × 150 mm; Tosoh, Tokyo, Japan) equipped with a TSK-gel Guardgel ODS-80_TS guard column (3.2 mm i.d. × 15 mm; Tosoh) was used for the analysis. The mobile phases used for phenacetin and acetanilide deacetylations were a solution of 10 mM sodium 1-hexanesulfonate containing 18% acetonitrile and 1% acetic acid for p-phenetidine and 10 mM sodium 1-hexanesulfonate containing 15% acetonitrile and 1% acetic acid for aniline, respectively. The flow rate was 1.0 ml/min, and UV detection was performed at 240 nm. The retention times for the deacetylated metabolite and the internal standard were 11.5 and 8.5 min, respectively, for phenacetin deacetylation and 6.8 and 14.0 min, respectively, for acetanilide deacetylation.

Kinetic Analyses. The apparent K_m and V_max values were calculated from a nonlinear regression analysis with a computer program WinNonlin Standard (Version 1.5; Scientific Consulting, Inc., Apex, NC). A graphic analysis of Eadie-Hofstee plots was also conducted for liver microsomal phenacetin deacetylation. The intrinsic clearance (CL_intrinsic) was calculated using the following equation: CL_intrinsic = V_max/K_m. The K_i value for prazosin on phenacetin deacetylation was calculated from the linear regression line obtained by Dixon plots. In the study on the stimulation of the phenacetin deacetylase activity by flutamide, Hill coefficient values were obtained by the graphical analysis of Hill plots.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Distribution of Carboxylesterase Activities. Table 1 shows the distribution of various carboxylesterase activities in microsomal and cytosolic fractions of the human liver and kidneys. When compared with liver microsomes, the carboxylesterase activities tested were all lower in other tissue fractions but were still detectable. The activity of phenacetin deacetylation localized predominantly in liver microsomes, whereas the distribution of acetanilide deacetylase activity was relatively broad. In the kidneys, the specific activity of acetanilide deacetylation in the microsomal fraction was approximately one-half of that in the cytosolic fraction. The activities of p-nitrophenylacetate and p-nitrophenylpropionate deesterifications in kidney microsomes were both only approximately 10% of those in liver microsomes. Cytosolic deesterification accounted for approximately 40% of the activities in kidney microsomes, demonstrating that the level of the activity in the kidneys was relatively high in the cytosolic fraction compared with that in the liver.

Kinetics of Phenacetin Deacetylation in Liver Microsomes. The kinetic parameters for the deacetylation of phenacetin to p-phenetidine were assessed in human liver microsomes with substrate concentrations ranging from 0.1 to 4 mM. Eadie-Hofstee plots for p-phenetidine formation were monophasic, indicating a single-enzyme catalytic reaction (Fig. 1B). The data plots on substrate concentration versus the initial velocity are shown in Fig. 1A. The apparent K_m and V_max values estimated with a nonlinear regression analysis were 4.7 mM (coefficient of variation, 4.5%) and 5.54 nmol/min/mg of protein (coefficient of variation, 2.9%), respectively. The intrinsic clearance was 1.18 µl/min/mg of protein.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   A, two representative S-V plots for phenacetin deacetylation in human liver microsomes. B, Eadie-Hofstee plots of data presented in A. , experimental data points. The line in B is the fitted function.

Inhibition and Stimulation Studies. Table 2 shows data on the effects of coincubation of various chemicals on the formation of p-phenetidine from phenacetin at the concentration of 2.5 mM, which is approximately one-half the K_m value described earlier, with human liver microsomes. Numerous ester- and amide-type compounds have neither inhibitory nor stimulatory effects on phenacetin deacetylation, but both prazosin (Fig. 2) and BNPP (Fig. 3) showed a potent inhibition. In contrast, flutamide stimulated the activity in a concentration-dependent manner (Fig. 4).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Profile of the inhibition of phenacetin deacetylation by prazosin. Phenacetin (2.5 mM) and prazosin at various concentrations were coincubated with 4 mg/ml microsomal protein in the assay medium. The control activity (without prazosin) in the reaction was 2.01 nmol/min/mg of protein. Enzyme incubation and metabolite analysis were carried out in duplicate, and the IC50 value was estimated using the computer program PD model 107, inhibitory effect sigmoid Emax, in WinNonlin Standard (Version 1.5).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Effects of BNPP and PMSF on phenacetin deacetylase activity in human liver microsomes. Phenacetin deacetylase activities were determined in duplicate using standard assay conditions after 5-min preincubation with the inhibitors at 37°C at the concentrations indicated. Control activities in those reactions were 1.87 and 1.73 nmol/min/mg of protein for BNPP and PMSF studies, respectively.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Stimulatory effect of flutamide on phenacetin deacetylase activity in human liver microsomes. Phenacetin (2.5 mM) and flutamide (1-200 µM) were coincubated with 4 mg/ml microsomal protein at 37°C for 30 min. The control activity (without flutamide) in the reaction was 1.80 nmol/min/mg of protein. Enzyme incubation and metabolite analysis were carried out in duplicate. Maximum stimulation and ED50 (the stimulator concentration giving half-maximum stimulation) were estimated using computer program PD model 106, sigmoid Emax, in WinNonlin Standard (Version 1.5). The ED50 value was 31.4 µM (coefficient of variation, 4.7%), and the maximum stimulation rate was 349% (coefficient of variation, 1.6%) as noted in Table 2.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Dixon plots for the inhibition, by prazosin, of phenacetin deacetylase activity in human liver microsomes. Enzyme incubation and metabolite analysis were carried out in duplicate. The linear regression lines were y = 0.059x + 1.066 (2 = 0.999) and y = 0.030x + 0.597 (2 = 0.998) for the substrate concentrations of 1 and 2 mM, respectively, yielding an estimate for the Ki value of 19.0 µM.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Effect of flutamide on the kinetics of phenacetin deacetylation in human liver microsomes. Phenacetin (0.05-8 mM) was incubated in the assay buffer at 37°C with human liver microsomes (4 mg/ml) in the presence of flutamide (3-300 µM) for 30 min. Kinetic parameters Km and Vmax were calculated by nonlinear regression analysis and the estimated values are noted in Table 3.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Phenacetin deacetylation is a typical reaction probe for carboxylesterase-amidase activity (Heymann et al., 1981), but our preliminary studies showed that the ordinary assay method for the determination of phenacetin deacetylase activity was not applicable in human tissue preparations because of the low activity. Therefore, we developed an assay for the phenacetin deacetylation that measures p-phenetidine using the solid-phase extraction and ion-pair HPLC. The validation data showed that the assay reported here is accurate, precise, and sufficient to determine the activity.

In humans, carboxylesterase activity is distributed throughout a wide variety of organs and tissues, including the brain, liver, stomach, small intestine, colon, plasma, and blood cells (Satoh and Hosokawa, 1998). Although the highest activity level is found in the liver and moderate levels are found in the kidneys of animals (Satoh, 1987), the distribution of the carboxylesterase activity in the kidneys is poorly understood in humans. In addition, nonspecific carboxylesterases in the liver are generally considered to be located in both the microsomal and cytosolic fractions. However, the intracellular distribution, particularly in microsomes and cytosol, of the carboxylesterase activity involved in the hydrolysis of the amide-linkage is not yet fully characterized in animals and humans. With this background, we first assessed the distribution of phenacetin deacetylase activity in human liver and kidneys. The results showed that the activity of not only phenacetin deacetylase but also other carboxylesterases (i.e., acetanilide deacetylase, p-nitrophenylacetate hydrolase, and p-nitrophenylpropionate hydrolase) predominantly localized in the liver microsomal fraction. Considerable levels of activity of these carboxylesterases were also found in liver cytosol and kidney microsomes and cytosol. The distribution of acetanilide deacetylase activity in each fraction was relatively broad and differed from that of other carboxylesterase activities. This result may be explained by the involvement of multiple isoforms of carboxylesterase and/or other hydrolysis enzymes in the deacetylation of acetanilide, although this remains to be substantiated.

The localization of phenacetin deacetylase activity in liver microsomes was more remarkable than that of the other carboxylesterase activities. We thus focused on the deacetylation of phenacetin and studied the in vitro kinetic properties of the reaction in human liver microsomes. The Eadie-Hofstee plot for p-phenetidine formation was monophasic, indicating a single-enzyme catalytic reaction. The apparent K_m, V_max, and intrinsic clearance for the reaction with pooled liver microsomes were 4.7 mM, 5.54 nmol/min/mg of protein, and 1.18 µl/min/mg of protein, respectively. Two distinct metabolic pathways exist in phenacetin disposition: one is the deacetylation by carboxylesterase, and the other is the oxidative deethylation producing acetaminophen by cytochrome P450, particularly CYP1A2. Rodrigues et al. (1997) reported that Michaelis-Menten parameters, K_m, V_max, and intrinsic clearance of the phenacetin deethylation by human liver microsomes, were 54 µM, 0.23 nmol/min/mg of protein, and 5.00 µl/min/mg of protein, respectively. When compared with the intrinsic clearance in two different pathways, the value for the cytochrome P450-mediated deethylation was approximately 4 times greater than that for the carboxylesterase-mediated deacetylation. This suggests that the cytochrome P450-mediated metabolism of phenacetin contributes to the in vivo disposition of phenacetin in humans rather than the carboxylesterase-mediated metabolism (Fig. 7). In fact, the in vivo extent of biotransformation of phenacetin to acetaminophen appears to be abundant (Sloan et al., 1978).

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Two distinct metabolic pathways of phenacetin in human liver microsomes.

The K_m values of various compounds for carboxylesterase have been estimated using human liver microsomes and enzymes isolated from the liver as an enzyme source. The affinity of the following substrates for the carboxylesterase originating from the human liver is of the same order of magnitude as that of phenacetin deacetylation: hCE-1-catalyzed meperidine (K_m = 1.9 mM), heroin (6.3 mM), 6-acetylmorphine (8.3 mM) hydrolysis, and hCE-2-catalyzed heroin (6.8 mM) hydrolysis (Zhang et al., 1999); the mid pI carboxylesterase-catalyzed diethyl succinate (1.3 mM) hydrolysis (Ketterman et al., 1989); carboxylesterase pI 5.3-catalyzed irinotecan (1.5 mM), ONO-5046 (1.1 mM), cilazapril (1.3 mM), and delapril (1.5 mM) hydrolysis; and carboxylesterase pI 4.5-catalyzed camostat mesilate (2.7 mM), ONO-5046 (2.1 mM), cilazapril (1.3 mM), aspirin (2.3 mM), procaine (3.3 mM), and oxybutynin (1.1 mM) hydrolysis (Takai et al., 1997). In contrast, the K_m value for the following carboxylesterase-mediated reaction is lower than that of phenacetin deacetylation, meaning a relatively high-affinity reaction: liver microsomal irinotecan (0.02-0.05 mM) (Haaz et al., 1997) and 2-acetylaminofluorene (0.3-0.7 mM) hydrolysis (Probst et al., 1994); hCE-1-catalyzed cocaine (0.12 mM) and 4-methylumbelliferylacetate (0.8 mM) hydrolysis and hCE-2-catalyzed cocaine (0.39 mM), 6-acetylmorphine (0.13 mM), and 4-methylumbelliferylacetate (0.15 mM) hydrolysis (Zhang et al., 1999); carboxylesterase pI 5.3-catalyzed camostat mesilate (0.71 mM), dilazep (0.15 mM), benazepril (0.73 mM), quinapril (0.13 mM), temocapril (0.79 mM), imidapril (0.29 mM), and aniracetam (0.09-0.1 mM) hydrolysis; carboxylesterase pI 4.5-catalyzed dilazep (0.09 mM), irinotecan (0.24 mM), benazepril (0.79 mM), quinapril (0.12 mM), temocapril (0.33 mM), and aniracetam (0.3-0.41 mM) hydrolysis (Takai et al., 1997); and the mid pI carboxylesterase-catalyzed p-nitrophenylacetate (0.87 mM) and malathion (0.01 mM) hydrolysis (Ketterman et al., 1989). This would suggest that carboxylesterase in the human liver has a wide range of K_m values and that phenacetin deacetylation in the organ is a relatively low-affinity catalytic reaction.

The organophosphorus compound BNPP and the serine protease inhibitor PMSF are known inhibitors of carboxylesterases (Krisch, 1966; Mentlein et al., 1980, 1988; Tanaka et al., 1987; Luttrell and Castle, 1988; van Lith et al., 1989; Minagawa et al., 1995). We thus evaluated their capacity to inhibit the phenacetin deacetylation. The results showed that phenacetin deacetylase was sensitive to BNPP but not to PMSF. Similar inhibition profiles for these inhibitors are noted in the 2-acetylaminofluorene deacetylation by human liver microsomes (Probst et al., 1991, 1994). However, phenacetin deacetylase does not exhibit the characteristic inhibition pattern described for other carboxylesterases (Hosokawa et al., 1990; Luan et al., 1997; Takai et al., 1997; Zhang et al., 1999).

Although numerous ester- and amide-type compounds had no or a poor inhibitory effect on the phenacetin deacetylation, prazosin potentially inhibited the reaction. Inhibitory effects of prazosin have not been reported previously, to our knowledge, for any carboxylesterases. The K_i values, which was estimated by the Dixon plots, and the IC₅₀ value of prazosin were similar, being 26.6 and 19.0 µM, respectively, and thus consistent in two different studies. The Dixon plots also indicated that prazosin is a noncompetitive inhibitor of phenacetin deacetylase in human liver microsomes. Noncompetitive inhibition of phenacetin deacetylation by prazosin suggests that prazosin binds simultaneously to the enzyme-substrate complex, yielding an active enzyme-substrate-inhibitor complex. Namely, this supports the assumption that phenacetin deacetylase possesses at least two different ligand-binding sites: one is the active site, and the other is probably the modulation site.

In contrast, flutamide stimulated the phenacetin deacetylation by human liver microsomes. This is the first report of in vitro stimulation of the carboxylesterase activity. The substrate stimulation involving the positive cooperative binding of more than a single substrate molecule was not observed in the phenacetin deacetylation because there was no sigmoidicity in the Michaelis-Menten plots of the reaction. That Hill coefficients are approximately 1.0 in the flutamide stimulation supports no involvement of substrate stimulation. To explain the stimulatory effect of flutamide on the phenacetin deacetylation, an allosteric mechanism with two distinct binding sites may be assumed, because the decrease in the K_m value most likely indicated an increase in the affinity of phenacetin deacetylase in the presence of flutamide. This can probably be explained by an effector binding site that, on binding of an effector molecule, alters the character of the substrate binding site. This hypothetical model for the mechanism of the flutamide stimulation appears to be applicable, in turn, to the mechanism of the prazosin inhibition. Thus, phenacetin deacetylase in human liver microsomes appears to possess a positive and/or negative modulation site or sites in its molecule. However, further studies are needed to clarify whether the binding sites of the positive effector flutamide and the negative effector prazosin are the same.

As previously mentioned, the carboxylesterases represent a multigene family. Recently, Satoh and Hosokawa (1998) proposed to classify carboxylesterase isoforms into four families: CES 1, CES 2, CES 3, and CES 4. CES1A1 includes the major isoforms of carboxylesterase in humans. The CES 4 family includes only one isoform, aryacetamide deacetylase, which could catalyze the deacetylation of 2-acetylaminofluorene (Probst et al., 1991, 1994) in human liver. Both ES46.5K (Watanabe et al., 1993) and E6123 hydrolase (Kusano et al., 1996), originating from mouse and monkey, respectively, are considered to be part of the CES 4 family. It is still unclear which isoform or isoforms are involved in the phenacetin deacetylation in human liver microsomes, but the nature of the inhibition of phenacetin deacetylase by BNPP or PMSF would appear to be similar to that of CES 4. Studies to identify the CES isoform involved in the deacetylation of phenacetin in human liver microsomes are in progress.

In conclusion, the present in vitro study demonstrates that phenacetin deacetylase activity predominantly localizes in liver microsomes in humans and is a single-enzyme catalytic reaction. Prazosin noncompetitively inhibits the phenacetin deacetylation. In contrast, flutamide stimulates the catalytic reaction, indicating that phenacetin deacetylase has a positive and/or negative modulation site or sites.