In Vitro Kinetics of Coumarin 3,4-Epoxidation: Application to Species Differences in Toxicity and Carcinogenicity

Stephanie L. Born*,1, Douglas Caudill*, Bill J. Smith{dagger} and Lois D. Lehman-McKeeman*

* The Procter & Gamble Company, Human and Environmental Safety Division, Miami Valley Laboratories, P.O. Box 538707, Cincinnati, Ohio 45253-8707; and {dagger} CNS Discovery/Drug Metabolism, Central Research Division, Pfizer, Inc., Groton, Connecticut 06340

Received December 15, 1999; accepted July 25, 2000


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Coumarin, a natural product and fragrance ingredient, is a well recognized rat liver toxicant, and dietary administration at toxic dosages increased the incidence of rat cholangiocarcinomas and parenchymal liver-cell tumors in a chronic bioassay. Hepatotoxicity in rats is site- and species-specific, and is thought to result from the formation of coumarin 3,4-epoxide and its rearrangement product, o-hydroxyphenylacetaldehyde (o-HPA). The goals of the current study were to describe the in vitro kinetics of the metabolic activation of coumarin, and determine whether species differences in susceptibility to liver injury correlate with coumarin bioactivation determined in vitro. Coumarin 3,4-epoxidation was quantified via the formation of o-HPA in pooled hepatic microsomes from female B6C3F1 mice, male F344 rats, and individual humans (n = 12 subjects), and the apparent kinetic constants for o-HPA production were calculated using nonlinear regression and fitting to either a one-enzyme or two-enzyme model. Eadie-Hofstee analyses indicated that o-HPA formation was biphasic in both rat and mouse liver. Although the apparent high affinity Km in rat and mouse liver microsomes was 38.9 and 47.2 µM, respectively, the overall rate of o-HPA formation was far greater in mouse than in rat liver microsomes. Furthermore, the total clearance (CLint) of coumarin via o-HPA formation in mouse liver microsomes was 4-fold greater than in rat liver microsomes. Since mice are relatively resistant to hepatotoxicity, the data indicated that rates of o-HPA formation in rat and mouse liver microsomes were not directly predictive of liver toxicity in vivo, and further suggested that o-HPA detoxification played a role in modulating coumarin-mediated toxicity. The current studies also indicated that coumarin 3,4-epoxidation in human hepatic microsomes was minimal. In human liver microsomes (n = 12), the kinetics of o-HPA formation were best described by a single enzyme model, with the Km for o-HPA formation ranging from 1320–7420 µM. In the most active human sample, the intrinsic clearance of coumarin via the 3,4-epoxidation pathway was 1/9 and 1/38 that of the rat and mouse, respectively. The in vitro kinetics of o-HPA formation, and in particular, the large quantities of coumarin required for o-HPA production in human liver microsomes, strongly suggest that humans are unlikely to produce toxicologically relevant concentrations of this metabolite following low level coumarin exposures.

Key Words: coumarin; epoxidation; liver; kinetics; toxicity; in vitro; rat; mouse; human.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Coumarin (cis-o-coumarinic acid lactone, 1,2-benzopyrone) is a natural product used widely as a fragrance enhancer and stabilizer. Coumarin is also administered clinically as a treatment for high-protein lymphedemas (Jamal et al., 1989), and as an antineoplastic agent in the treatment of renal cell carcinoma (Marshall et al., 1994Go) and malignant melanoma (Marshall et al., 1989Go). Although reports of adverse effects in humans resulting from coumarin administration are rare (Cox et al., 1989Go; Egan et al., 1990Go), it is generally recognized that coumarin is a rat liver toxicant (Lake, 1984Go; Lake et al., 1989aGo, 1994Go). Furthermore, at levels that exceeded the maximum tolerated dose (5000 ppm), chronic dietary coumarin administration to SD rats resulted in an increased incidence of cholangiocarcinoma and parenchymal liver cell tumors (Carlton et al., 1996Go), an effect linked to the toxic effects of this chemical.

Species differences in coumarin toxicity are profound and are thought to be metabolism-mediated. Coumarin metabolism in humans occurs predominately via 7-hydroxylation (Fig. 1Go), a high affinity reaction catalyzed by CYP2A6 (Draper et al., 1997Go; Yamano et al., 1990Go). 7-Hydroxycoumarin and its glucuronide and sulfate conjugates are nontoxic, and constitute 40–97% of human urinary metabolites following an oral dose in most subjects (Egan et al., 1990Go; Shilling et al., 1969Go). In contrast, rats exhibit little hepatic 7-hydroxylase activity (Honkakoski et al., 1993Go) and the 3,4-epoxidation pathway is favored. Recently, coumarin 3,4-epoxide (CE) was synthesized, and in an aqueous environment it was found to rapidly rearrange to o-hydroxyphenylacetaldehyde (o-HPA), which has previously been shown to be the major in vitro product formed by rat hepatic microsomes (Fentem et al., 1991Go; Lake et al., 1992bGo). Furthermore, the major metabolite of coumarin detected in rats in vivo is o-hydroxyphenylacetic acid (o-HPAA) (Cohen, 1979Go; Kaighen and Williams, 1961Go), the oxidation product of o-HPA. This metabolite has also been detected in the urine of human subjects administered coumarin, indicating that low levels of coumarin 3,4-epoxidation may occur in humans in vivo (Hadidi et al., 1997Go, 1998Go; Meineke et al., 1998Go, Shilling et al., 1969Go).



View larger version (22K):
[in this window]
[in a new window]
 
FIG. 1. Proposed pathways of coumarin metabolism in rodents and man.

 
Coumarin-induced toxicity in the rat liver is dependent upon the formation of CE (Born et al., 1998aGo; Lake, 1984Go; Lake et al., 1989bGo; Lake et al., 1992aGo; Peters et al., 1991Go). Although the short half-life of coumarin epoxide precludes the direct examination of its effects (Born et al., 1997Go), the role of CE as a rat liver toxicant is supported by studies conducted with 3,4-dihydrocoumarin (DHC) (Lake et al., 1989aGo), a structural analog of coumarin that is saturated at the 3,4-position and will not form an epoxide. In contrast to coumarin, DHC was not toxic in the rat liver in vivo, nor did DHC-treated hepatocytes exhibit signs of toxicity (Lake et al., 1989aGo).

The toxic effects of coumarin are likely due not only to the formation of coumarin 3,4-epoxide, but also to the production of o-HPA (Born et al., 1997Go, 1998bGo). Aldehydes constitute a group of relatively reactive compounds, and the toxicity of these molecules is well-established (Feron et al., 1991Go). Although the molecular mechanisms of o-HPA-mediated toxicity have yet to be elucidated, this aldehyde is the only coumarin metabolite shown to be hepatotoxic in vitro, data that supports a role for its involvement in toxicity (Born et al. 2000Go; Lake, 1999Go).

Recognizing that CE rearranges quantitatively to o-HPA, it became possible to characterize the kinetics of hepatic coumarin 3,4-epoxidation across species. Therefore, the goal of the present work was to compare the 3,4-epoxidation of coumarin in hepatic microsomes from B6C3F1 mice, F344 rats, and humans. The results of this work provide the foundation for determining whether the selective toxic and carcinogenic effects of coumarin in the rat liver may be attributed to quantitative species differences in CE and o-HPA formation.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Materials.
Coumarin and 7-hydroxycoumarin were obtained from Aldrich Chemical Company (Milwaukee, WI). Benzene ring universally-labeled 14C-coumarin (2.15 GBq/mmol) was purchased from Amersham Life Science (Arlington Heights, IL).

Synthesis of authentic o-HPA standard.
Synthesis of o-HPA was carried out according to the method of Bruce and Creed (Bruce et al., 1970), and it was used as an authentic standard. The structure of o-HPA was verified using gas chromatography-mass spectrometry-light pipe Fourier-transformed infrared spectrometry (GC-MS-IR) (Born et al., 1997Go), and the purity of the compound exceeded 97%.

Animals.
Female B6C3F1 mice (20–25 g) and male F344 rats (210–220 g) were purchased from Charles River Laboratories, Portage, MI. The sex and strain of animals selected for study were based on the National Toxicology Program (1993) coumarin bioassay data indicating that rat liver necrosis was more severe in males, and that, in the absence of dose-related liver tumor effects in mice, female mice were more susceptible to coumarin-induced lung tumor formation. Animals were housed in humidity- and temperature-controlled rooms and allowed free access to food (Purina Laboratory Rodent chow, Ralston-Purina, St. Louis, MO.) and water.

Human liver microsomes.
Human hepatic microsomes (ID numbers 831921, 924941, 105952) were purchased from the International Institute for the Advancement of Medicine (IIAM; Exton, PA). Human liver microsomes (ID numbers H0017, H0018, H0019, H0026, H0036) were purchased from XenoTech LLC (Kansas City, KS). Table 1Go describes the donor characteristics for the human liver samples used in these studies. Four additional human liver microsomal samples (L2, L4, L10, L20) were a generous gift from Dr. Brian G. Lake (BIBRA International, Surrey, U.K.), but no details of the donor characteristics were provided to us.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Characteristics of Human Liver Donors
 
Preparation of rodent hepatic microsomes.
Hepatic microsomes were prepared from untreated female B6C3F1 mice (n = 10/pool, 5 pools) and untreated male F344 rats (n = 12/pool, 2 pools). Microsomes were prepared via differential centrifugation (Guengerich, 1989Go).

Metabolism in hepatic microsomes.
Microsomal protein was determined by the method of Bradford (Bradford, 1976Go) with bovine serum albumin as standard. Coumarin 7-hydroxylation was measured in mouse, rat, and human hepatic microsomes, according to the method of Greenlee and Poland (1978) as modified by Pearce et al. (1992). Briefly, 0.01–1 mg microsomal protein was incubated with 25 mM potassium phosphate buffer, pH 7.4, 1 mM EDTA, 3 mM MgCl2, and 50 µM coumarin in a final volume of 1 ml. This saturating substrate concentration was selected due to its common use in the literature to report coumarin 7-hydroxylase activities in vitro (Lovell et al., 1999Go; Steensma et al., 1994Go). After a 2-min pre-incubation at 37°C, 1 mM NADPH was added to the reaction mixture and the sample was incubated for 5–10 min at 37°C. Formation of 7-hydroxycoumarin was monitored using a Perkin-Elmer LS50-B luminescence spectrometer (Norwalk, CT). The limit of 7-hydroxycoumarin detection in rat liver microsomal incubations was 18-pmol/mg protein. Coumarin 7-hydroxylase (7-HC) activities determined for each pool of mouse liver microsomes (n = 5 pools) were highly similar, with the standard error of mean 7-HC values varying by <= 20%. Therefore, the 5 pools were treated essentially as a single pool. To determine the apparent Km and Vmax for coumarin 7-hydroxylation in mouse liver microsomes, 0.075 mg of protein was incubated with 0.25–10 µM coumarin as described above.

Production of o-HPA was measured in incubations containing 0.125–0.5 mg hepatic microsomal protein, 100 mM potassium phosphate, pH 7.4, 1 mM EDTA, 3 mM MgCl2, 5 mM glucose 6-phosphate, 20 IU/ml glucose 6-phosphate dehydrogenase, and coumarin (10–4000 µM in a final volume of 1 ml. Following a 2-min pre-incubation at 37°C, 1 mM NADP was added to the reaction mixture, and the sample was incubated for 5–30 min at 37°C. Reaction conditions were adjusted at low substrate conditions to prevent excess substrate depletion and optimize product formation. However, in human samples exhibiting high rates of 7-hydroxycoumarin formation (H0017, 831921, and 105952), reaction conditions could not be optimized sufficiently to reduce substrate depletion and maintain a signal-to-noise ratio of at least 2 for the detection of o-HPA at <= 50 µM coumarin. Reactions were terminated by the addition of 200 µl 30% perchloric acid. In the absence of strong acidification, no o-HPA was recovered from the microsomal incubations. After centrifugation at 6000 rpm for 10 min, the supernatant was extracted twice with 1 or 2 volumes of methylene chloride (Burdick and Jackson, GC2) containing 0.005 mg/ml n-tridecane (n-C13) as the internal standard. The extraction efficiency of 14C-coumarin was examined under these assay and extraction conditions and confirmed to be > 88%. Extracts were combined into a glass vial and gently evaporated under N2 to a volume of 200 µl. Sample extracts were not evaporated to dryness, as this process would result in a substantial loss of o-HPA. The final sample was transferred to a glass autosampler vial capped with a teflon-lined septum.

Gas chromatographic analysis of o-HPA formation.
Coumarin metabolites were separated and quantified using a Hewlett Packard 5890 gas chromatograph coupled with flame ionization detection. The gas chromatograph was equipped with a split/splitless injector and a Hewlett Packard HP-5 crosslinked 5% PH ME siloxane (30 m, 0.25 µm) column. The helium flow rate was 2 ml min–1. Splitless injection (1–2 µl) was performed with an autosampler and an injector temperature of 280°C. The oven temperature program was set initially at 40°C, held for 2 min, then increased at 14°C/min to 170°C. The rate was then increased to 32°C/min to 300°C, which was held for 5 min. The detector temperature was 250°C. The retention times for o-HPA, n-tridecane and coumarin were 10.17 min, 10.42 min, and 12.0 min, respectively (Fig. 2Go). No interfering peaks were detected in rodent or human liver microsomal samples in which NADP was omitted. The identity of o-HPA formed in microsomal incubations was confirmed via GC-MS-IR (results not shown). Quantification was accomplished by calculation of peak-area ratios relative to the internal standard, and use of standard curves obtained by analysis of dilutions of an authentic sample of o-HPA. The o-HPA standard curve was linear over the range of unknown samples, with a detection limit of 20 pmol on column. Using pooled rat, mouse, or individual human samples, formation of o-HPA was found to be linear with both time and the amount of microsomal protein added within the ranges tested (5–50 min and 0.125 to 0.6 mg/ml, respectively). The standard error of mean o-HPA values determined in pooled rat (n = 2 pools) or pooled mouse liver microsomes (n = 5 pools) varied by less than 20%. Therefore, pooled rat-liver and pooled mouse-liver microsomes were each treated as a single pooled sample.



View larger version (15K):
[in this window]
[in a new window]
 
FIG. 2. Chromatogram of o-HPA formation in B6C3F1 mouse liver microsomes. Microsomes were incubated with 1000 µM coumarin and the sample was extracted as described in Materials and Methods. o-HPA, n-tridecane (n-C13), and coumarin eluted at 10.17, 10.42, and 12.0 min, respectively. The peak eluting at 11.6 min was not identified.

 
Analysis of Data
The apparent kinetic constants Km and Vmax for mouse liver microsomal coumarin 7-hydroxylation were determined using nonlinear regression and fitting to a single-site model (Grafit, Erithacus Software Ltd, Middlesex, U.K.).

The results of the o-HPA kinetic experiments were initially evaluated using Michaelis-Menten and Eadie-Hofstee plots. The rat and mouse microsomal data appeared to be biphasic with a high-affinity site and a low-affinity site that was not saturated under the conditions of the experiment. Therefore, the rat and mouse data were fitted to a 2-site model using a procedure similar to that described previously for ethoxycoumarin by Carlile et al. (1999).

The high-affinity site is described by the parameters Vmax and Km that are the maximal rate of metabolism and the Michaelis constant, respectively. Vmax/Km is the clearance of the high-affinity site (CL1), and CL2 is the clearance term of the low-affinity site. Intrinsic clearance (CLint) is the sum of CL1 and CL2. Initial parameters used for the curve-fitting procedure were determined using an Eadie-Hofstee plot. The curve-fitting procedure was conducted using WinNonlin Professional version 2.1 (Pharsight Corporation, Palo Alto, CA) with the default Gauss-Newton (Levenberg and Hartley) non-linear model and a 1/Y(predicted) weighting. The human microsomal data was fitted to a 1-site model with parameters of Vmax and Km. Uniform weighting was used when fitting the human data.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Coumarin epoxidation and 7-hydroxylation were measured in pooled hepatic microsomes from rats and mice. In the rat, no coumarin 7-hydroxylase activity was detected (results not shown). In contrast, o-HPA formation was observed at low (10 µM) substrate concentrations, increasing to 3.4 ± 0.10 nmol/min/mg at 2000 µM coumarin (Fig. 3Go). In mouse, both coumarin 7-hydroxylation and epoxidation were observed. The Km and Vmax for coumarin 7-hydroxylation in pooled female B6C3F1 hepatic microsomes were 1.26 µM and 0.312 nmol/min/mg, respectively (results not shown). At 50 µM substrate, the rate of coumarin 7-hydroxylation was 0.282 ± 0.016 nmol/min/mg, whereas the rate of o-HPA formation was 2.04 ± 0.063 nmol/min/mg, exceeding 7-hydroxylation by 7-fold. Formation of o-HPA in mouse hepatic microsomes was greater than in rat at each coumarin concentration examined. In mice, a maximum rate of 9.2 ± 0.38 nmol/min/mg protein was achieved at 2000 µM coumarin, a rate 2.7 times higher than that observed in rat liver microsomes (Fig. 3Go).



View larger version (20K):
[in this window]
[in a new window]
 
FIG. 3. Mouse liver microsomal CE formation was greater than that in rat liver microsomes at every coumarin concentration examined. Data are the mean ± SE of triplicate or quadruplicate determinations. The best fit lines to the data were generated using the equation .

 
Eadie-Hofstee analysis suggested the involvement of at least 2 cytochrome P450 forms in the bioactivation of coumarin to o-HPA in rodents (data not shown). Recognizing the biphasic nature of o-HPA formation in mouse and rat liver microsomes, the apparent kinetic constants for the high-affinity, low-capacity enzymes were calculated using non-linear regression analysis (Table 2Go). In the absence of enzyme saturation at coumarin concentrations as high as 2000 µM, CL2 was calculated to reflect the clearance of coumarin attributable to a low-affinity enzyme. The high-affinity Km values for o-HPA formation in rats and mice were 38.9 and 47.2 µM, respectively, indicating that both species readily convert coumarin into toxic products. However, the Vmax of the high-affinity site in mice was 5 times that observed in rat liver microsomes. This apparent species difference in o-HPA formation is best reflected by total intrinsic clearance (CLint), with coumarin metabolism via the 3,4-epoxidation pathway being 4 times higher in mouse than in rat.


View this table:
[in this window]
[in a new window]
 
TABLE 2 Apparent Kinetic Constants For o-Hydroxyphenylacetaldehyde Formation In Rodent Hepatic Microsomes
 
Like the mouse liver, human liver microsomes catalyzed both the 7-hydroxylation and the 3,4-epoxidation of coumarin. The Km for coumarin 7-hydroxylation has been reported to be within the range of 0.2 to 3.6 µM (reviewed in Lake, 1999), and in the 12 human liver samples used in this study, coumarin 7-hydroxyation rates ranged from 0.104 to 4.11 nmol/min/mg at 50 µM (Table 3Go). o-HPA formation rates in human liver microsomes were consistently lower than in the rat, and varied up to 6-fold, with donors 924941 and 831921 being the lowest and highest producers of o-HPA, respectively (Fig. 4Go and Table 3Go). In general, human liver microsomal samples (n = 12) could be divided into groups with high (n = 3, mean activity = 1.50 nmol/min/mg), medium (n = 5, mean activity = 0.673 nmol/min/mg), or low (n = 4, mean activity = 0.351 nmol/min/mg) o-HPA activity at 2000 µM coumarin. Hepatic microsomes from donor 831921 exhibited a rate of 1.67 ± 0.23 nmol/min/mg (2000 µM), which increased to 2.23 ± 0.065 at 4000 µM, the highest activity observed among the 12 human samples. In subject 831921, CE production was observed at relatively low coumarin concentrations (33–50 µM). However, initial rates of o-HPA production could not be determined at these coumarin levels due to extensive substrate depletion via 7-hydroxycoumarin formation. In contrast, coumarin concentrations >= 100 µM were required before detectable quantities of CE were formed by 7 of the 12 human microsomal samples.


View this table:
[in this window]
[in a new window]
 
TABLE 3 7-Hydroxylation and 3,4-Epoxidation of Coumarin in Human Liver Microsomes
 


View larger version (25K):
[in this window]
[in a new window]
 
FIG. 4. Formation of o-HPA was observed in each of the 12 human liver samples examined. In general, the human liver samples could be divided into groups exhibiting high, medium, and low o-HPA formation rates. Human liver samples 831921 and H0017, H0019, and 924941 are representative of these 3 groups. Data are the mean ± SE of quadruplicate determinations performed with individual liver microsomal samples.

 
Eadie-Hofstee analysis suggested that coumarin 3,4-epoxidation in human liver microsomes was monophasic, with some samples showing limited evidence of biphasic enzyme kinetics (data not shown). Overall, the experimental data for each human sample was best fit, with the lowest statistical error, using a single enzyme model (Table 4Go). Studies using human (Zhuo, et al., 1999Go) and rat recombinant P450s (data not shown) indicate that CYP1A2 and CYP2E1 contribute to the epoxidation of coumarin in both rodent and human liver microsomes. The monophasic nature of o-HPA formation in human liver microsomes suggests that human CYP1A2 and CYP2E1 exhibit similar Km values for o-HPA formation. In contrast, the rodent forms may exhibit very different Michaelis constants for the same reaction. Additional studies will be needed to characterize the contributions of CYP1A and CYP2E1 to o-HPA formation.


View this table:
[in this window]
[in a new window]
 
TABLE 4 Apparent Kinetic Constants for o-Hydroxyphenylacetaldehyde Formation in Human Hepatic Microsomes
 
The apparent kinetic constants of CE formation in human liver microsomes are shown in Table 4Go. The high Km values (> 1000 µM) indicate that human liver microsomes do not readily form o-HPA at low coumarin concentrations. Comparison of the kinetic constants shown in Tables 2 and 4GoGo indicates that CE formation is greatest in mouse liver microsomes >> rat >> human.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The 3,4-epoxidation of coumarin is thought to be a requisite step in the cascade of metabolic events that leads to the hepatotoxic and carcinogenic effects of this chemical (Carlton et al., 1996Go; Cohen, 1979Go; Lake, 1999Go; National Toxicology Program, 1993Go). Indirect evidence of the involvement of CE in rat liver necrosis has been obtained via the use of coumarin structural analogs (Lake et al., 1994Go), and the discovery that mercapturic acid conjugates of coumarin are excreted in the urine of treated rats (Huwer et al., 1991Go). Further, at dosages that exceed the maximum tolerated dose, and under conditions known to result in the formation of CE (Huwer et al., 1991Go), coumarin has been associated with the formation of liver cholangiocarcinomas and parenchymal tumors in rats (Carlton et al., 1996Go). However, CE has not been isolated in vivo or from in vitro systems. Building on the foundation provided by the successful synthesis of coumarin 3,4-epoxide (Born et al., 1997Go), the demonstration that coumarin epoxide rearranges to o-HPA (Born et al., 1997Go), and the determination that o-HPA is hepatotoxic (Born et al., 1999), the present studies on the kinetics of o-HPA formation provide the first valid data on which to evaluate species differences in target organ sensitivity to coumarin.

Coumarin 7-hydroxylation and 3,4-epoxidation are generally recognized as the predominant coumarin metabolic pathways in humans and rodents. Coumarin 7-hydroxylation has been extensively studied and characterized (Cohen, 1979Go; Draper et al., 1997Go; Fernandez-Salguero et al., 1995bGo; Kaighen and Williams, 1961Go; Lindberg et al., 1992Go; Matsunaga et al., 1988Go; Oscarson et al., 1998Go; Yamano et al., 1990Go; Yun et al., 1991Go). Enzymes of the CYP2A subfamily have been of special interest, due to their polymorphic expression in the human population and the importance of these forms in catalyzing coumarin detoxification. Indeed, in humans and mice, where coumarin 7-hydroxylation is catalyzed by CYP2A6 and CYP2A5, respectively, the liver is relatively resistant to injury in vivo (Cox et al., 1989Go; Egan et al., 1990Go; Born et al., 1998; Carlton et al., 1996Go; Lake and Grasso, 1996). However, rat CYP2A1 and CYP2A2 do not catalyze the 7-hydroxylation of coumarin (Fernandez-Salguero et al., 1995a; Matsunaga et al., 1988Go), an observation that correlates with the susceptibility of this species to coumarin-mediated liver necrosis.

The metabolism of coumarin to other products, and the relationship between toxicity and metabolite formation, have been less well studied (Fentem et al., 1991Go, 1992; Lake et al., 1992bGo; Norman et al., 1984), being encumbered by the lack of appropriate standards and a lack of knowledge regarding coumarin 3,4-epoxidation. In vivo studies have clearly demonstrated a differential susceptibility to coumarin-mediated injury between F344 rats and B6C3F1 mice, with a single oral gavage dosage of coumarin (200 mg/kg) causing severe centrilobular necrosis and profound increases (150-fold over control) in clinical markers of hepatic injury (sorbital dehydrogenase [SDH] and alanine aminotransferase [ALT]). In contrast, an equimolar dosage administered orally to mice caused mild hepatic hypertrophy and small (2- to 6-fold) increases in SDH and ALT enzymes, respectively (Born et al., 1998aGo).

Although the relationship between toxicity and coumarin metabolite blood levels in rodents has not been characterized, it has been hypothesized that the species-specific hepatotoxic effects of coumarin in the rat result from a lack of coumarin detoxification via 7-hydroxylation, and the extensive formation of CE (Fentem and Fry, 1992Go; Lake et al., 1992aGo,bGo). However, the results of the present work suggest that species differences in coumarin-mediated hepatotoxicity cannot be simply explained by a balance between coumarin 7-hydroxylation and CE formation. At subsaturating concentrations (< 50 µM), coumarin is cleared primarily via the formation of 7-HC in mouse liver microsomes (CLint = 0.248 ml/min/mg), and competition between coumarin 7-hydroxylation and epoxidation would limit the percentage of a dose converted to toxic metabolites in vivo. In contrast, 7-hydroxylation was not observed in F344 rat liver microsomes. Due to 7-HC formation, a quantitatively important detoxification pathway in the mouse liver, coumarin clearance via o-HPA formation cannot alone be used to predict rodent liver toxicity at low substrate concentrations. However, as the substrate concentration increases into a range (>= 200 µM) where rat hepatocyte toxicity is observed (Born et al., 2000Go; Lake et al., 1989aGo) and injury in vivo is probable, coumarin epoxidation predominates, with o-HPA formation rates in mice being up to 4-fold greater than in rat liver microsomes. Thus, at dosages used in the National Toxicology Program (NTP) bioassay (200 mg/kg and 100 mg/kg in mice and rats, respectively), coumarin epoxidation is predicted to be the major route of coumarin clearance in both mouse and rat liver. These data demonstrate that no direct relationship exists between hepatic CE formation in rodents in vitro and toxic or carcinogenic outcomes in vivo.

It has also been presumed that humans polymorphic for CYP2A6 would respond to coumarin in a manner similar to that of the rat, exhibiting increased sensitivity to coumarin-mediated hepatotoxicity due to an increased clearance of coumarin via the epoxidation pathway (Hadidi et al., 1997Go, 1998Go). Although no polymorphic individuals were present in the pool of subjects (n = 12), 7-hydroxylation and 3,4-epoxidation appeared to parallel one other, with both activities being high or low in a particular subject (Table 3Go). Thus, the present data indicate that the degree to which coumarin is cleared via the 7-hydroxylation, or 3,4-epoxidation pathway, is a direct function of the kinetics of the respective reactions, and the amount of coumarin present. Furthermore, the results show that the epoxidation of coumarin is not a function of CYP2A activity in liver.

Whereas the results disprove the previously held notion that toxicity is inversely related to 7-hydroxylation, the fact that o-HPA production is greatest in the mouse, a species relatively refractory to the hepatic effects of coumarin, strongly suggests that susceptibility to injury is determined by the balance between CE formation and its detoxification. Previous studies in rats indicate that the 2 major excreted metabolites are o-HPAA (Kaighen and Williams, 1961Go; Huwer et al., 1991Go), and the glutathione conjugate of coumarin 3,4-epoxide (Huwer et al., 1991Go). Similar data are not available in mice, but the present results suggest that mice are either more efficient at conjugating CE, or are more likely to form o-HPAA, a non-toxic metabolite of coumarin (Lake et al., 1989aGo). These events are not without precedent, as the ability of mice to more effectively and efficiently form GSH conjugates of other potentially reactive intermediates, such as aflatoxin B1 8,9-epoxide, greatly exceeds that of the rat (Buetler et al., 1992Go; Ramsdell et al., 1990). o-HPAA is also detected in human urine samples, both in untreated individuals (Hadidi et al., 1997Go) and following coumarin administration (Hadidi et al., 1997Go; Meineke et al., 1998Go). Although o-HPAA excretion in normal humans is limited, and may not be dose-related (Meineke et al., 1998Go), its detection in human urine following coumarin administration may reflect a low level of CE formation in humans. Alternatively, it is possible that o-HPAA may form via pathways other than, or in addition to oxidation of o-HPA, as has been suggested by other groups (Lake et al., 1999). Additional studies will be needed to determine whether coumarin metabolites other than o-HPA are precursors of o-HPAA. Even in the case where o-HPAA derives exclusively from o-HPA, the kinetics of o-HPA formation and detoxification would dictate susceptibility to injury.

Coumarin is found naturally in a variety of foodstuffs, including cinnamon, peppermint, and green tea, and the average dietary consumption of coumarin is estimated to be 1.23 mg/day (Lake, 1999Go). Additional coumarin exposure attributed to perfumed products is estimated to be 2.0 mg/day (Lake, 1999Go). In clinical settings, coumarin dosages of up to 0.1–5 g daily are administered to treat kidney and skin cancers (Sharifi et al., 1993Go). In human subjects given 2 g of coumarin orally, the peak coumarin plasma level achieved was approximately 2 µg/ml, or 13.7 µM coumarin (Sharifi et al., 1993Go). Among the 12 liver microsomal samples examined in the current study, the lowest apparent Km for o-HPA formation was 1320 µM. Thus, even in clinical situations where large dosages of coumarin are administered to patients, the available data indicate that the epoxidation pathway will contribute very little to the elimination of coumarin. Moreover, at typical levels of exposure from dietary sources and consumer products, it is unlikely that CE formation would occur.

The present data also have major implications regarding the development of an accurate evaluation of the toxic and carcinogenic hazard associated with coumarin exposure in humans. A number of in vitro and in vivo studies have been conducted to determine the overall reactivity and potential genetic toxicity of coumarin (reviewed in Lake, 1999). In rats, an in vivo assay to assess unscheduled DNA synthesis was negative at dosages up to the maximum tolerated dose of 320 mg/kg (Edwards et al., 2000Go), a dose where peak plasma concentrations exceeding 900 µM (Ritschel and Hussain, 1988Go) were achieved. Similarly, an in vitro evaluation of unscheduled DNA synthesis in human liver slices was negative at concentrations up to 5 mM coumarin (Beamand et al., 1998Go). The current studies indicate that CE was formed at the high coumarin concentrations used in both of these genotoxicity studies, indicating that CE does not cause DNA damage in either test system.

In summary, the results of the present studies indicate that major quantitative species differences exist in the rate of coumarin 3,4-epoxidation, and most significantly, that low-level coumarin exposures resulting from dietary and consumer product sources are highly unlikely to result in CE formation or toxicity. Furthermore, in vitro rates of o-HPA formation are not predictive of hepatotoxicity in rats and mice, suggesting that the detoxification of coumarin epoxide and o-HPA play critical roles in determining toxic outcome. Studies are presently ongoing to determine the kinetics of GSH conjugation and oxidation of o-HPA to o-HPAA in order to determine whether these pathways differ across species, and the extent to which they contribute to species differences in toxicity.


    ACKNOWLEDGMENTS
 
This study was supported in part by the Research Institute for Fragrance Materials.


    NOTES
 
1 To whom correspondence should be addressed. Fax: (513) 627-0002. E-mail: born.sl{at}pg.com. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Beamand, J. A., Barton, P. T., Price, R. J., and Lake, B. G. (1998). Lack of effect of coumarin on unscheduled DNA synthesis in precision-cut human liver slices. Food Chem. Toxicol. 36, 647–653.[ISI][Medline]

Born, S. L., Fix, A. S., Caudill, D., and Lehman-McKeeman, L. D. (1998a). Selective Clara cell injury in mouse lung following acute administration of coumarin. Toxicol. Appl. Pharmacol. 151, 45–56.[ISI][Medline]

Born, S. L., Hu, J. K., and Lehman-McKeeman, L. D. (1998b). o-Hydroxyphenylacetaldehyde (o-HPA) is a hepatotoxic metabolite of coumarin. Toxicol. Sci. 42, 400.

Born, S. L., Hu, J. K., and Lehman-McKeeman, L. D. (2000). o-Hydroxyphenylacetaldehyde is a hepatotoxic metabolite of coumarin. Drug Metab. Dispos. 28, 218–223.[Abstract/Free Full Text]

Born, S. L., Rodriguez, P. A., Eddy, C. L., and Lehman-McKeeman, L. D. (1997). Synthesis and reactivity of coumarin 3,4-epoxide. Drug Metab. Dispos. 25, 1318–1324.[Abstract/Free Full Text]

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254.[ISI][Medline]

Bruce, J. M., and Creed, D. (1970). Light-induced and related reactions of quinones: VI. Reactions of some p-quinones carrying formyl groups. J. Chem. Soc. (C), 649–653.

Buetler, T. M., Slone, D., and Eaton, D. L. (1992). Comparison of the aflatoxin B1–8,9-epoxide conjugating activities of two bacterially expressed Alpha class glutathione S-transferase isozymes from mouse and rat. Biochem. Biophys. Res. Commun. 188, 597–603.[ISI][Medline]

Carlile, D. J., Hakooz, N., and Houston, J. B. (1999). Kinetics of drug metabolism in rat liver slices: IV. Comparison of ethoxycoumarin clearance by liver slices, isolated hepatocytes, and hepatic microsomes from rats pretreated with known modifiers of cytochrome P450 activity. Drug Metab. Dispos. 27, 526–532.[Abstract/Free Full Text]

Carlton, B. D., Aubrun, J.-C., and Simon, G. S. (1996). Effects of coumarin following perinatal and chronic exposure in Sprague-Dawley rats and CD-1 mice. Fund. Appl. Toxicol. 30, 145–151.[ISI][Medline]

Cohen, A. J. (1979). Critical review of the toxicology of coumarin with special reference to interspecies differences in metabolism and hepatotoxic response and their significance to man. Food Cosmet. Toxicol. 17, 277–289.[ISI][Medline]

Cox, D., O'Kennedy, R., and Thornes, R. D. (1989). The rarity of liver toxicity in patients treated with coumarin (1,2-benzopyrone). Hum. Toxicol. 8, 501–506.[ISI][Medline]

Draper, A. J., Madan, A., and Parkinson, A. (1997). Inhibition of coumarin 7-hydroxylase activity in human liver microsomes. Arch. Biochem. Biophys. 341, 47–61.[ISI][Medline]

Edwards, A. J., Price, R. J., Benwick, A. B., and Lake, B. G. (2000). Lack of effect of coumarin on unscheduled DNA synthesis in the in vivo rat hepatocyte DNA repair assay. Food Chem. Toxicol. 38, 403–409.[ISI][Medline]

Egan, D., O'Kennedy, R., Moran, E., Cox, D., Prosser, E. and Thornes, R. D. (1990). The pharmacology, metabolism, analysis, and applications of coumarin and coumarin-related compounds. Drug Metab. Rev. 22, 503–529.[ISI][Medline]

Fentem, J. H., and Fry, J. R. (1992). Metabolism of coumarin by rat, gerbil, and human liver microsomes. Xenobiotica 22, 357–367.[ISI][Medline]

Fentem, J. H., Fry, J. R., and Whiting, D. A. (1991). o-Hydroxyphenylacetaldehyde: A major novel metabolite of coumarin formed by rat, gerbil, and human liver microsomes. Biochem. Biophys. Res. Comm. 179, 197–203.[ISI][Medline]

Fernandez-Salguero, P., and Gonzales, F. J. (1995a). The CYP2A gene subfamily: species differences, regulation, catalytic activities, and role in chemical carcinogenesis. Pharmacogenetics 5, S123–S128.[ISI][Medline]

Fernandez-Salguero, P., Hoffman, S. M., Cholerton, S., Mohrenweiser, H., Raunio, H., Rautio, A., Pelkonen, O., Huang, J.-D., Evans, W. E., Idle, J. R., and Gonzalez, F. J. (1995b). A genetic polymorphism in coumarin 7-hydroxylation: Sequence of the human CYP2A genes and identification of variant CYP2A6 alleles. Am. J. Hum. Genet. 57, 651–660.[Medline]

Feron, V. J., Til, H. P., De Vrijer, F., Woutersen, R. A., Cassee, F. R., and Van Bladeren, P. J. (1991). Aldehydes: Occurrence, carcinogenic potential, mechanism of action, and risk assessment. Mutat. Res. 259, 363–385.[ISI][Medline]

Greenlee, W.F., and Poland, A. (1978). An improved assay of 7-ethoxycoumarin O-deethylase activity: Induction of hepatic enzyme activity in C57BL/6J and DBA/2J mice by phenobarbital, 3-methylcholanthrene and 2,3,7,8-tetrachlorodibenzo-p-dioxin. J. Pharmacol. Exp. Ther. 205, 596–605.[Abstract]

Guengerich, F. P. (1989). Analysis and characterization of enzymes. In Principles and Methods of Toxicology (A.W. Hayes, Ed.), pp. 777–814. Raven Press, New York.

Hadidi, H., Irshaid, Y., Vagbo, C. B., Brunsvik, A., Cholerton, S., Zahlsen, K., and Idle, J. R. (1998). Variability of coumarin 7-and 3-hydroxylation in a Jordanian population is suggestive of a functional polymorphism in cytochrome P450 CYP2A6. Eur J. Clin. Pharmacol. 54, 437–441.[ISI][Medline]

Hadidi, H., Zahlsen, K., Idle, J. R., and Cholerton, S. (1997). A single amino acid substitution (Leu160His) in cytochrome P450 CYP2A6 causes switching from 7-hydroxylation to 3-hydroxylation of coumarin. Food Chem. Toxicol. 35, 903–907.[ISI][Medline]

Honkakoski, P., Maenpaa, J., Leikola, J., Pasanen, M., Juvonen, R., Lang, M. A., Pelkonen, O., and Raunio, H. (1993). Cytochrome P4502A-mediated coumarin 7-hydroxylation and testosterone hydroxylation in mouse and rat lung. Pharmacol. Toxicol. 72, 107–112.[ISI][Medline]

Huwer, T., Altmann, H.-J., Grunow, W., Lenhardt, S., Przybylski, M., and Eisenbrand, G. (1991). Coumarin mercapturic acid isolated from rat urine indicates metabolic formation of coumarin 3,4-epoxide. Chem. Res. Toxicol. 4, 586–590.[ISI][Medline]

Jamal, S., and Casley-Smith, J. R. (1989). The effects of 5,6-benzo[a]pyrone (coumarin) and DEC on filaritic lymphoedema and elephantiasis in India, preliminary results. Ann. Trop. Med. Parasitol. 83, 287–290.[ISI][Medline]

Kaighen, M., and Williams, R. T. (1961). The metabolism of [3-14C]coumarin. J. Med. Pharm. Chem. 3, 25–43.[ISI]

Lake, B. G. (1984). Investigations into the mechanism of coumarin-induced hepatotoxicity in the rat. Arch. Toxicol. 7(Suppl.), 16–29.

Lake, B. G. (1999). Coumarin metabolism, toxicity and carcinogenicity: Relevance for human risk assessment. Food Chem. Toxicol. 37, 423–453.[ISI][Medline]

Lake, B. G., Evans, J. G., Lewis, D. F. V., and Price, R. J. (1994). Studies on the acute effects of coumarin and some coumarin derivatives in the rat. Food. Chem. Toxicol. 32, 357–363.[ISI][Medline]

Lake, B. G., Gaudin, H., Price, R. J., and Walters, D. G. (1992a). Metabolism of [3-14C]coumarin to polar and covalently bound products by hepatic microsomes from the rat, Syrian hamster, gerbil and humans. Food Chem. Toxicol. 30, 105–115.[ISI][Medline]

Lake, B. G., Gray, T. J. B., Evans, J. G., Lewis, D. F. V., Beamand, J. A., and Hue, K. L. (1989a). Studies on the mechanism of coumarin-induced toxicity in rat hepatocytes: Comparison with dihydrocoumarin and other coumarin metabolites. Toxicol. Appl. Pharmacol. 97, 311–323.[ISI][Medline]

Lake, B. G., Osborne, D. J., Walters, D. G., and Price, R. J. (1992b). Identification of o-hydroxyphenylacetaldehyde as a major metabolite of coumarin in rat hepatic microsomes. Food Chem. Toxicol. 30, 99–104.[ISI][Medline]

Lake, B. G., Walters, D. G., and Gangolli, S. D. (1989b). Comparison of the metabolism and disposition of [3-14C]coumarin in the rat and marmoset (Callithrix jacchus). Toxicol. Lett. 45, 299–306.[ISI][Medline]

Lindberg, R. L., Juvonen, R., and Negishi, M. (1992). Molecular characterization of the murine Coh locus: An amino acid difference at position 117 confers high and low coumarin 7-hydroxylase activity in P450coh. Pharmacogenetics 2, 32–37.[ISI][Medline]

Lovell, D. P., Van Iersel, M., Walters, D. G., Price, R. J., and Lake, B. G. (1999). Genetic variation in the metabolism of coumarin in mouse liver. Pharmacogenetics 9, 239–250.[ISI][Medline]

Marshall, M. E., Butler, K., Cantrell, J., Wiseman, C., and Mendelson, L. (1989). Treatment of advanced malignant melanoma with coumarin and cimetidine; A pilot study. J. Cancer Chemother. Pharmacol. 24, 65–66.

Marshall, M. E., Mohler, J. L., Edmonds, K., Williams, B., Butler, K., Ryles, M., Weiss, L., Urban, D., Bueschen, A., Markiewicz, M., and Cloud, G. (1994). An updated review of the clinical development of coumarin (1,2-benzopyrone) and 7-hydroxycoumarin. Cancer Res. Clin. Oncol. 120(Suppl.), 39–42.

Matsunaga, T., Nagata, K., Holsztynska, E. J., Lapenson, D. P., Smith, A., Kato, R., Gelboin, H. V., Waxman, D. J., and Gonzalez, F. J. (1988). Gene conversion and differential regulation in the rat P-450 IIA gene subfamily. Purification, catalytic activity, cDNA and deduced amino acid sequence, and regulation of an adult male-specific hepatic testosterone 15{alpha}-hydroxylase. J. Biol. Chem. 263, 17995–18002.[Abstract/Free Full Text]

Meineke, I., Desel, H., Kahl, R., Kahl, G.F., and Gundert-Remy, U. (1998). Determination of 2-hydroxyphenylacetic acid (2HPAA) in urine after oral and parenteral administration of coumarin by gas-liquid chromatography with flame-ionization detection. J. Pharm. Biomed. Anal. 17, 487–492.[ISI][Medline]

National Toxicology Program. (1993). Toxicology and carcinogenesis studies of coumarin in F344/N rats and B6C3F1 mice (gavage study). U.S. Department of Health and Human Services. NTP TR 422. National Institutes of Health, Public Health Service.

Norman, R. L., and Wood, A. W. (1984). o-Hydroxyphenylethanol, a novel lactone ring-opened metabolite of coumarin. Drug Metab. Dispos. 12, 543–549.[Abstract]

Oscarson, M., Gullsten, H., Rautio, A., Bernal, M. L., Sinues, B., Dahl, M-L., Sengard, J. H., Pelkonen, O., Raunio, H., and Ingelman-Sundberg, M. (1998). Genotyping of human cytochrome P450 2A6 (CYP2A6), a nicotine C-oxidase. FEBS Letters 438, 201–205.[ISI][Medline]

Pearce, R., Greenway, D., and Parkinson, A. (1992). Species differences and interindividual variation in liver microsomal cytochrome P450 2A enzymes: Effects on coumarin, dicoumarol, and testosterone oxidation. Arch. Biochem. Biophys. 298, 211–225.[ISI][Medline]

Peters, M. M. C. G., Walters, D. G., Van Ommen, B., Van Bladderen, P. J., and Lake, B. G. (1991). Effects of inducers of cytochrome P-450 on the metabolism of [3-14C]coumarin by rat hepatic microsomes. Xenobiotica 21, 499–514.[ISI][Medline]

Ramsdell, H. S., and Eaton, D. L. (1990). Species susceptibility to aflatoxin B1 carcinogenesis: Comparative kinetics of microsomal biotransformation. Cancer Res. 50, 615–620.[Abstract]

Ritschel, W. A., and Hussain, S. A. (1988). Transdermal absorption and topical bioavailability of coumarin. Meth. Find. Exptl. Clin. Pharmacol. 10, 165–169.[ISI][Medline]

Sharifi, S., Lotterer, E., Michaelis, H. C, and Bircher, J. (1993). Pharmacokinetics of coumarin and its metabolites. Preliminary results in three healthy volunteers. J. Ir. Coll. Physicians Surg. 22, 29–32.

Shilling, W. H., Crampton, R. F., and Longland, R. C. (1969). Metabolism of coumarin in man. Nature 221, 664–665.[ISI][Medline]

Steensma, A., Beamand, J. A., Walters, D. G., Price, R. J., and Lake, B., (1994). Metabolism of coumarin and 7-ethoxycoumarin by rat, mouse, guinea pig, Cynomolgus monkey, and human precision-cut liver slices. Xenobiotica 24, 893–907.[ISI][Medline]

Yamano, S., Tatsuno, J., and Gonzalez, F. J. (1990). The CYP2A3 gene product catalyzes coumarin 7-hydroxylation in human liver microsomes. Biochemistry 29, 1322–1329.[ISI][Medline]

Yun, C.-H., Shimada, T., and Guengerich, F. P. (1991). Purification and characterization of human liver microsomal cytochrome P-450 2A6. Molec. Pharmacol. 40, 679–685.[Abstract]

Zhuo, X., Gu, J., Zhang, Q.-Y., Spink, D. C., Kaminsky, L. S., and Ding, X. (1999). Biotransformation of coumarin by rodent and human cytochrome P-450: Metabolic basis of tissue-selective toxicity in olfactory mucosa of rats and mice. J. Pharmcol. Exp. Ther. 288, 463–471.[Abstract/Free Full Text]