TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

108,436-447

(199

I)

Bioactivation of Aflatoxin B, by Human Liver Microsomes: Role of Cytochrome P450 WA Enzymes HOWARD A. CRAIG

S. RAMSDELL,* EDDY,$

ANDREW AND

DAVID

PARKINSON,?

L.

EATON*,’

*Department of Environmental Health and Institufe.for Environmental Studies: $Department ofSurgery. University of Washington, Seattle, Washington 98195; and -fDepartmenl of Pharmacology, To,x-icologv and Therapeutics, Center,for Environmentaf and Occupational Heahh. University oyKansas Medical Center, Kansas City, Kansas 66103

Received September 24. 1990; accepted January 13. 1991 Bioactivation of Aflatoxin B, by Human Liver Microsomes: Role of Cytochrome P450 IIIA Enzymes. RAMSDELL, H. S., PARKINSON, A., EDDY. A. C.. AND EATON, D. L. (1991). Toxical. Appl. Pharmacol. 108, 436-447. Based on our previous observations (H. S. Ramsdell and D. L. Eaton, 1990, Cancer Res. 50,6 15-620) that the proportion of aflatoxin B, (AFB,) converted to the highly reactive AI%,-8.9-epoxide in microsomal incubations varies with substrate concentration. we have examined the hypothesis of T. Shimada and F. P. Guengerich (1989, Proc. Natl. Acad. Sci. USA 86, 462-465) that cytochrome P450 IIIA4 is principally responsible for the activation (epoxidation) of AFB, by human liver microsomes. The initial rates of formation of AFBi-8,9-epoxide and hydroxylated AFB, metabolites were determined in microsomes prepared from livers of organ donors (n = 14) at AFB, concentrations of 124 and 16 PM. Microsomal oxidation of nifedipine. catalyzed primarily by P450 IIIA enzymes, was also determined by HPLC. Rates of formation of AFB, metabolites and nifedipine oxidation were poorly correlated at either AFB, concentration (r2 = 0.13-0.41). A somewhat better correlation between AFB, epoxidation and nifedipine oxidation was observed at 124 JIM AFB, (r* = 0.41) than at 16 pM AFB, (rZ = 0.26). Treatment of pooled microsomes with troleandomycin. an apparently specific inhibitor of P450 IIIA enzymes, resulted in 35% inhibition of AFB,-8,9-epoxide formation at the high AFB, level but had little effect at 16 pM AFB, An antibody against rat cytochrome P450 IRA1 significantly inhibited AFB, epoxidation at high, but not low, AFB, concentrations, whereas AFQ, formation was strongly inhibited at all substrate levels examined. These results are consistent with the hypothesis that cytochrome P450 IIIA enzyme(s) can form AFB,-8,9-epoxide, but are effective at only relatively high substrate concentrations. Another P450 enzyme(s) appears to be principally responsible for AFB,-8,9-epoxide formation at the low AFB, levels that would be typical for dietary exposures. ICI 1991 Academic PBS. IN.

Aflatoxin B1 (AFB,) is a potent hepatocarcinogen in animals (Busby and Wogan, 1984)and there is epidemiological evidence that humans may be subject to its effects (Groopman et al., 1988). Carcinogenesis by AFB, requires activation by microsomal cytochrome P450 (Essigmann et al., 1982), which is apparently me’ To whom correspondence should be addressed at Department of Environmental Health. SC-34. University of Washington, Seattle, WA 98 195. 004 1-008X/9

I $3.00

Copyright 0 1991 by Acadcmc Press. Inc. All nghis of reproduction m any form reserved.

436

diated by the formation of the highly reactive AFB,-8,9-epoxide (Swenson et al., 1977). Microsomal oxidation of AFBr also results in the formation of the hydroxylated metabolites aflatoxin Q, (AFQ,) and aflatoxin MI (AFM,), both of which are lesstoxic than the parent compound (Campbell and Hayes, 1976; Coulombe et al., 1982). In previous work, we found that the pattern of AFB, biotransformation by human liver microsomesisdependent on the concentration

MICROSOMAL

ACTIVATION

of AFB, present in the incubation mixture (Ramsdell and Eaton, 1990). The proportion of AFBi converted to the epoxide by human liver microsomes is inversely related to substrate concentration, suggesting that at the low tissue concentrations of AFB, that would result from dietary exposures, human liver would be expected to activate a large proportion of the dose. Thus, it is critical to consider the concentration dependence of microsomal biotransformation reactions involving AFB, . A recent report presented evidence that a particular cytochrome P450 enzyme, P450 IIIA42, is principally responsible for the bioactivation of AFB, in human liver (Shimada and Guengerich, 1989). In that study, the rate of formation of AFBr-8,9-epoxide was determined indirectly with a bacterial genotoxicity assay (Oda et al., 1985; Shimada and Nakamura, 1987) conducted in the presence of human liver microsomes or purified human cytochromes P450. Activation of AFB, was correlated with nifedipine oxidase activity and the effects of known activators and inhibitors of P450 IIIA4 on AFB, activation were also examined. Most of the nifedipine oxidase activity in human liver has been attributed to cytochrome P450 IIIA4 (Gonzalez et al., 1988; Guengerich et al., 1986a). Cytochrome P450 IIIAS, detected in lo-20% of the adult human livers tested, had similar nifedipine oxidation activity as P450 IIIA4 when each was expressed in a vaccinia virus system (Aoyama et al., 1989). The nifedipine oxidase activity of other P450 IIIA enzymes has not been characterized, so it is unclear what contribution they make to ’ The nomenclature of cytochrome I’450 isoenzymes used in this paper adheres, where possible, to that recommended by Nebert et al. ( 1989). Trivial names are used for HLp3, HLp2. and HFLa; sequence data for the first has not been published and that for the latter two (Komori et al.. 1989: Schuetz ef al.. 1989) has appeared too recently to be included in the systematic nomenclature (Nebert e! ~1.. 1989). The general designation “P450 IRA” is used to refer collectively to the members of the subfamily which may be present in adult human liver: P450 IRA3 (HLp), P450 IIIA4 (P-450,r). P450 IIIA5 (hPCN3). HLp2. and HLp3.

OF

437

AF’B,

nifedipine oxidation. At least four other enzymes in the P450 IIIA family have been purified from human liver. Cytochrome P450 IIIA3 (HLp) is closely related to P450 IIIA4 (Molowa et al., 1986; Watkins et al., 1985). Both HLp2, a major cytochrome P450 in fetal human liver (Schuetz et al., 1989; Wrighton and Vandenbranden, 1989) and HLp3 (Wrighton et al., 1989) have been shown to be expressed polymorphically in adult human liver. Another enzyme closely related to HLp2, P450 HFLa, has been characterized in human fetal liver (Kitada et al., 1985; Komori et al., 1989). We have investigated the hypothesis that the cytochrome P450 IIIA enzyme(s) is involved in the bioactivation of AFB, by human liver microsomes by utilizing an HPLC-based determination of initial rates of AFB, oxidation in vitro. This assay allows sensitive and specific determination of the rates of formation of the major microsomal metabolites of AFB,. AFB,-8,9-epoxide, AFQ,, and AFM, (Monroe and Eaton, 1987; Ramsdell and Eaton, 1990). In the present study, two AFB, concentrations were utilized: the high concentration (124 PM) represents a saturating level under the in vitro assay conditions used, whereas the low concentration ( 16 PM) reflects the shift in metabolite patterns with decreasing AFB, concentrations described previously (Ramsdell and Eaton, 1990). On the basis of correlations with nifedipine oxidation assays and inhibition and activation experiments, we conclude that P450 IIIA enzymes are capable of AFB, epoxidation at relatively high substrate concentrations, but that at lower substrate concentrations, another enzyme(s) is involved. Thus, at tissue levels expected from dietary AFB, exposure, P450 IIIA enzymes. including P450 IIIA4. would not be expected to substantially contribute to the activation of AFB, to AFB,-8,9-epoxide. MATERIALS

AND

METHODS

Chemicals. Aflatoxins B,, Q,, M,, P,, and G,. NADP. glucose 6-phosphate, glucose-6-phosphate dehydrogenase. glutathione (GSH), nifedipine, nitrendipine, troleando-

438

RAMSDELL

mycin. and Tris base were obtained from Sigma Chemical Co. (St. Louis. MO). Inorganic salts were analytical reagent grade from commercial sources. HPLC-grade solvents from J. T. Baker, Inc. (Phillipsburg. NJ), were used. The pyridine derivative of nifedipine (dehydronifedipine) was synthesized according to a published method (Loev and Snader, 1965). The product gave a single peak upon HPLC and CC analysis and had a melting point (IO&107°C) UV, and mass spectra consistent with literature data (Ebel ef ~11.. 1978; Testa et al.. 1979). A4icrosomal preparations. Human liver samples were obtained from nine organ donors through the University of Washington hospitals (seven male. two female, ages 1446 years), frozen immediately in liquid NZ, and stored at -80°C until microsomes were prepared as previously described (Ramsdell and Eaton, 1988). Equal volumes of these microsomes were combined for use in experiments using pooled samples. Human microsomes were also obtained from the livers of five additional organ donors obtained through the University of Arizona Medical Center; (two male. two female. one not identified, ages IS-49 years). For these five livers subcellular fractions were prepared immediately. Microsomes were isolated and stored (-70°C) according to the procedures of Van der Hoeven and Coon (1974) and Halpert et al., ( 1983). The drug histories of the donors are as follows: no known drug treatment, 6: labetolol. 1; heavy alcohol use, 1; propanolol and heavy alcohol use. 1: no drug history was available for 5 of the donors. Protein concentrations were determined using the bicinchoninic acid method with bovine serum albumin as the standard (Smith rt a/.. 1985). Cytochrome P450 content was measured using carbon monoxide difference spectra (Omura and Sato. 1964). .-1FB ouidulion assaj~. Microsomal biotransformation of AFB, was determined by a modification of a method described previously (Monroe and Eaton. 1987). Incubations of AFB, with microsomes (in duplicate or triplicate as noted in the legends to the figures) were carried out in the presence of hepatic cytosol from BHA-treated mice and GSH to quantitatively trap AFB,-8,9-epoxide as the GSH conjugate (Monroe and Eaton. 1987). The incubation mixtures included 2-3 mg BHA-treated mouse liver cytosolic protein/ml. 5 mM GSH. 5 mM glucose 6-phosphate, 0.5 units/ml glucose 6-phosphate dehydrogenase, and 2 mM NADP in a buffer containing 190 mM sucrose. 60 mM potassium phosphate. 80 mM Tris, 15 mM NaCl. 5 rnM KCI. and 4 mM MgClz (pH 7.6 at room temperature). AFB, in dimethyl sulfoxide (DMSO) was added to give a final concentration of 124 or 16 FM as determined by uv absorbance spectrophotometry of the stock solutions (Busby and Wogan. 1984). Equal volumes of the solvent (4%~ X,/V) were added to all incubation mixtures. The GSH conjugate. AFQ,, and AFM, were quantified by reversedphase HPLC with aflatoxin G, as an internal standard. The formation of metabolites was linear with time for at least 10 min and with protein from 50 to 375 pg per incubation at both AFB, concentrations,

ET

AL.

Nijhdipine oxidation assal). The oxidation of nifedipine to its pyridine derivative by human liver microsomes was measured by a substantial modification of the procedure of Guengerich et al. (I 986a). Due to the light sensitivity of nifedipine and nitrendipine solutions (Ebel ef al.. 1978) all handling was done with indirect, subdued lighting; solutions were stored in amber containers protected from room light whenever possible: incubations were done in amber microcentrifuge tubes: and amber autoinjector vials were used for HPLC analysis. Incubation mixtures (250 +I total volume) contained 2 mM NADP. 5 mM glucose h-phosphate. 0.5 units/ml glucose-6-phosphate dehydrogenase. and 1 mg/ml microsomal protein in 80 mM potassium phosphate buffer (pH 7.6). Following preincubation for 5 min at 37°C. nifedipine (in 2 ~1 acetone) was added to give a final concentration of 200 PM. After incubation for 10 min. the reaction was stopped by the addition of an equal volume of ice-cold methanol containing nitrendipine as an internal standard. The tubes were immediately vortexed and placed on ice. The samples were then kept at -20°C for at least 2 hr to allow precipitation of protein. No change in the amount of dehydronifedipine was observed in samples stored at -20°C for up to 5 days. Following centrifugation. an aliquot of the supernatant was transferred to an autoinjector vial for analysis by HPLC. At room temperature. the dehydronifedipine content in such a supernatant decreased by about 0.07% per hour (a negligible loss), so no correction was applied for the different periods oftime that samples remained at room temperature in the autoinjector before analysis. The samples were analyzed by HPLC with an Econosphere Cl8 cartridge column (Alltech Assoc.) (4.6 mm X 15 cm) eluted with water and a linear gradient from 45 to 70% methanol in 12.5 mitt at room temperature. This was followed by a 3-min flush with 90% methanol before reequilibration at the starting conditions. Peaks were detected by absorbance at 254 nm and areas were determined with a computing integrator. The retention times of dehydronifedipine, nifedipine, and nitrendipine were 6.9, 8.4, and 12.0 min. respectively. No interfering peaks were observed in samples in which NADP. nifedipine, or nitrendipine was omitted. Quantification was accomplished by calculation of peak areas relative to the internal standard. and use of standard curves obtained by analysis of dilutions of an authentic sample ofdehydronifedipine. The standard curve was Iinear over the range of the unknown samples with an approximate detection limit of 5 ng dehydronifedipine injected. Using a pooled sample of human liver microsomes. oxidation of nifedipine was found to be linear with both time and the amount of microsomal protein added within the ranges tested (2-20 min and 0.2-2.0 mg/ml, respectively). Inhibition and activafion experimenls. Inhibition of microsomal AFB, oxidation by nifedipine was determined by addition of nifedipine (various concentrations in acetone) to incubation mixtures after preincubation and just prior to addition of AFB, An equal amount of acetone

MICROSOMAL

ACTIVATION

(0.8% of the total incubation volume) was added to control incubations. Pretreatment of microsomes (0.25 mg protein) with 20 fiM troleandomycin was done in the presence of 2 mM NADP, 5 mM glucose 6-phosphate, and 0.5 units/ml of glucose 6-phosphate dehydrogenase. Following incubation ofthis mixture for 30 min at 37°C. the samples were placed on ice and additional components were added to make up the reaction mixtures as described above for both nifedipine and AFB, oxidation assays. Microsomes were pretreated with an equal volume of vehicle (DMSO, 1.3% v/v) for control incubations. The effects of 7,8-benzoflavone on AFB, and nifedipine oxidation were determined by addition of the compound at various concentrations in DMSO (representing 0.8% of the total incubation volume) before preincubation. Control incubations received the same volume of DMSO alone. Antibody inhibition experiments were carried out with rabbit IgG raised against cytochrome P450 IIIAl (Halvorson et al., 1990). Control incubation mixtures contained an equivalent amount of preimmune rabbit IgG.

RESULTS An initial experiment was conducted to assess the ability of nifedipine to inhibit the oxidation of AFB, by a pooled sample (n = 9) of human liver microsomes. Figure 1 shows the effects of nifedipine (2 to 200 PM) on the formation of AFB,-8,9-epoxide, AFQl , and AFM, at 124 and 16 PM AFB, . At both substrate concentrations, the rate of epoxidation of AFB, was the least sensitive to inhibition in the presence of nifedipine, whereas AFQ, formation was inhibited to the greatest extent. The data at 200 FM nifedipine (20-fold higher than the apparent K, measured for nifedipine oxidation by human liver microsomes; Guengetich et al., 1986a) show 69, 52, and 9% inhibition of AFQ, , AFM,, and AFB,-8,9-epoxide formation, respectively, at 124 PM AFB, , and 9 l%, 35%, and no inhibition, respectively, at 16 yM AFB I . To address more specifically the potential involvement of cytochrome P450 IIIA enzymes in the activation of AFB, , the inhibitory effects of troleandomycin (TAO) pretreatment of microsomes on AFB, oxidation were investigated. Figure 2 shows the results obtained with a pooled sample of human liver microsomes. At the higher concentration of AFBl

-

OF

439

AFB, [AFB]

I

p”o1

= 124

10 Nifedipine

100 Concentration

[AFE]

1

1000 (PM)

= 16 PM

10 Nifedipine

PM

100 Concentration

1000 (PM)

FIG. 1. Inhibition of AFB, oxidation by nifedipine. Human liver microsomes (pooled sample, n = 9) were incubated in duplicate with AFB, and nifedipine at the concentrations indicated. Error bars indicate the range of the data. Control values for AFB,-8.9-epoxide, AFQ,, and AFM, were 196, 978, and 49 pmol/mg protein/min, respectively, at 124 PM AFB, . and 3 I .4,42.3, and 15.0 pmol/ mg protein/min, respectively, at 16 PM AFB,

tested, some inhibition of AFB,-8,9-epoxide and AFM, formation was observed (35 and 26%, respectively). Striking inhibition of AFQ, formation (90%) was observed under the same conditions. With a substrate concentration of 16 PM, TAO pretreatment of microsomes caused a 49% inhibition of AFQl formation but no change in AFB,-8,9-epoxide formation. Formation of AFM, was below the limit of detection (approximately 5 pmol/min/mg protein) in both control and TAO-treated samples at the lower AFB, concentration, apparently due to the partial loss of cytochrome P450 activity during the pretreatment period (30 min at 37°C). Pretreatment of the pooled human liver microsome sample with TAO resulted in 74% inhibition of nifedipine oxidation activity (data not shown). The use of a pooled microsome sample may have obscured effects of TAO on cytochrome P450 IIIA enzymes which may be present in

440

RAMSDELL [AFB] 500

= 124

PM

1

0

Control

0

TAO-treated

a9 2

PI,,. AFB

epoxide

AFQ

[AFB] c

2E

40

6= k

30

I

AFM

= 16 PM

q Control 0

TAO-treated

20 a 2 f

ET

AL

inhibition was observed at the higher AFB, level (4 l-70%; mean 56%). There was no segregation of the extent of inhibition of AFBi8,9-epoxide formation at either substrate level with the nifedipine oxidase activity level. AFQ , formation (Fig. 3B) was strongly inhibited by TAO pretreatment at both 16 PM AFB, (7787%; mean 81%) and 124 PM AFB, (87-95% inhibition; mean 9 1%). Inhibition of AFQ, formation by TAO appeared to be somewhat stronger in samples with high nifedipine oxidase activity than those with lower activity; the small sample size precluded statistical analysis of the significance of such differences.

10 0

AFS-Epoxide

A

AFB

epoxide

AFQ

AFY

FIG. 2. Inhibition of AFB, oxidation by troleandomycin pretreatment of human liver microsomes. Pooled human liver microsomes (n = 9) were pretreated with troleandomycin (TAO) as described under Materials and Methods and then incubated with AFB,. AFB, metabolites were quantified by HPLC analysis. Error bars indicate the SE of triplicate incubations except for the TAO-treated group at 124 JLM AFB,, for which the error bars indicate the range of data from duplicate incubations. The values shown for the TAO-treated samples are the percentage of the respective control values. Under the same conditions, nifedipine oxidation was inhibited by 74%.

5L

100

E 8

75

5 E 8 ; 0

50 25 0 1

2 Low NFO Microsome

B

AFQ

3

4

High NFO Sample

Formation

60

0E

individuals in varying amounts. The inhibitory effects of TAO were also examined in a subset of four individual microsomal samples, chosen on the basis of their level of nifedipine oxidation activity (see below); two had high nifedipine oxidation activity and two had low activity. Figure 3 shows the results for the effects of pretreating the microsomes with TAO on AFB , epoxidation (A) and AFQ i formation (B). The inhibitory effects of TAO on nifedipine oxidation are also shown in Fig. 3. The results obtained with the individual samples were generally the same as those shown in Fig. 2 for pooled microsomes. Inhibition of nifedipine oxidation ranged from 46 to 83% (mean 64%) in the individual samples. Again, little inhibition of AFB, epoxidation (Fig. 3A) was observed at the low AFBr concentration (412%: mean 8%). whereas a greater degree of

Formation

q [AF61=16~M

40

[AFB] = 124 ,,M

6 E 8 tii P

20 0 1

Low NF02 Microsome

3

HiQh NF04 Sample

FIG. 3. Effects of troleandomycin pretreatment on individual human liver microsome samples. Samples of microsomes were pretreated with troleandomycin as described under Materials and Methods and then incubated with either AFB, or nifedipine. Oxidative metabolites were quantified by HPLC analysis of duplicate incubations. Control values for nifedipine oxidation (NFO) were I .70, 1.48.3.09, and 3.74 nmol/mg protein/min for samples I4, respectively. Control values for AFB, epoxidation were 50, 21, 27, and 96 pmol/mg protein/min at 16 MM AFB,, and 207, 172.205, and 43 1 pmol/mg protein/min at 124 fiM AFB, for samples 1-4, respectively. Control values for AFQ, formation were 37,40,48, and 79 pmol/mg protein/ min at 16 @M AFB,. and 906. 1021. 1210, and 21771 pmol/mg protein/min at 124 PM AFB, for samples 1-4, respectively.

MICROSOMAL AFB: 400

RA2

15.6

ACTIVATION

OF

441

AFB,

pM

AFB: 1200

a

= 0.262

1 R”2

124

pM

= 0.412

1 3OOi

200-

100 -

0

0-l

0

4 Nifedlplne

400

1 R”2

300

B (nmollminlnmol

P450)

Nifedipine

B (nmol

12

/min/nmol

P450)

A

4 Nlfediplne

B

i

01 0

12

(nmollmlnlnmol

n

4

Nlfedlplne

P450)

150

RA2

4

= 0.156

01 0

I 0

12

300-

RA2

= 0.246

B

12

(nmol/min/nmol

P450)

= 0.129 n

1 100 -

200-

100-

“I

0

4 Nlfedlplne

B (nmol/mln/nmol

12 P450)

0 Nlfediplne

8

I

I

4

B

12

(nmol/mln/nmol

P450)

FIG. 4. Correlation of AFB, and nifedipine oxidation by human liver microsomes. Biotransformation of AFB, and nifedipine by individual human liver microsome samples was quantified by HPLC analysis. Assays were conducted in triplicate for each sample. Coefficients of determination (r’) were determined by leastsquares regression analysis.

Oxidation of nifedipine to its pyridine metabolite by individual human liver microsome samples was measured. Specific activity values ranged from 0.85 to 5.23 nmol product/mg

protein/min, comparable with values reported in the literature (Gonzalez et al., 1988; Shimada and Guengerich, 1989; Shimada et al., 1989). Plots comparing these data with the

442

RAMSDELL 600,

500 a. .Z .? 5 5L E s 84

-7

AFB

epoxide

400 -

300

200 :/

NM

ET AL.

The effects of 7,8-benzohavone (50, 100, or 200 PM) on AFB, and nifedipine oxidation by a pooled sample of human liver microsomes are shown in Fig. 5.7,8-Benzoflavone at these levels had no effect on nifedipine oxidation, whereas AFB , epoxidation was increased fourto five-fold at both substrate concentrations. In contrast, AFQ, formation was inhibited by 40-X0% under the same conditions. Formation of AFM, was detected only at 124 PM AFB, and was completely abolished at all 7,8benzoflavone concentrations tested.

7,8-benzoflavone

FIG. 5. Effect of 7.8benzoflavone on oxidation of AFB, and nifedipine by human liver microsomes. Pooled human liver microsomes (n = 9) were incubated with either AFB, or nifedipine in the presence of 7.8benzoflavone. Assays were carried out in duplicate and error bars indicate the range of the data. For AFB,-8,9-epoxide and AFQ,, the open symbols indicate data obtained at an AFB, concentration of 16 pM and the solid symbols denote results observed at 124 JAM AFB,. Control values were 1.6 nmol/ mg protein/min for nifedipine oxidation. 29 and 35 pmol/ mg protein/min for AFB,-8.9-epoxide and AFQ,, respectively, at 16 pM AFB, and 149 and 892 pmol/mg protein/ min for AFB,-8,9-epoxide and AFQ, , respectively, at 124 PM AFB,.

specific activities observed for the formation of each of the AFB, metabolites at both substrate concentrations are shown in Fig. 4. Coefficients of determination (r2) obtained by linear regression analysis of each data set ranged from 0.16 to 0.4 1. The correlation between AFB,-8,9-epoxide formation and nifedipine oxidation was somewhat better at the higher AFB, concentration (r2 = 0.41) than at the lower level (r2 = 0.26) suggesting that an enzyme with nifedipine oxidase activity may contribute to AFB, epoxidation at higher concentrations of AFB, . When specific activities were calculated on the basis of microsomal protein content, similar correlation plots were obtained. Coefficients of determination were 0.2 1, 0.29, and 0.15 for AFB,8,9-epoxide, AFQ, , and AFM, , respectively, at 16 PM AFBr , and 0.54, 0.43, and 0.19 for AFB,-8,9-epoxide, AFQ, and AFM, , respectively, at 124 PM AFB, .

AFB-Epoxide

AFB Epoxide at 124 pM AFB 25-

1

0-l 0

pg

1

anti-IIIAl/pg

2

microsomal

protein

B 100

75

50 ?

0

AFB

100

200

300

400

500

PM AFB FIG. 6. Antibody inhibition of human liver microsomal oxidation of AFB, . Pooled human liver microsomes (n = 9) were incubated with different amounts of anticytochrome P450 IIIAl (A). or varying concentrations of AFB, in the presence of 250 #g of preimmune rabbit I& (control) or 250 pg IgG raised against cytochrome P450 IllAl, to yield a final IgG concentration of I pg IgG protein/pg of microsomal protein (B). The values represent the mean of duplicate determinations and the error bars indicate the range of the data.

MICROSOMAL

ACTIVATION

Oxidation of AFB, was inhibited by antibodies against P450 IIIA 1. The degree of inhibition was proportional to the amount of IgG added to incubations (Fig. 6A). The inhibition of AFB, epoxidation activity in human liver microsomes by anticytochrome P450 IIIAl IgG was strongly dependent on the concentration of substrate (Fig. 6B). No inhibition of epoxidation was observed at AFB, concentrations of 8 or 16 PM, whereas approximately 45% of the epoxidation activity was inhibited at 124 and 478 pM. In contrast, the rate of formation of AFQi was inhibited by 65-80% by the antibody over the entire range of substrate concentrations tested. Formation of AFMl was not significantly inhibited by anti-P450 IIIA 1 IgG (data not shown). DISCUSSION We have measured the initial rates of biotransformation of AFB, by human liver microsomes to investigate the hypothesis of Shimada and Guengerich ( 1989) that P450 IIIA4 is the principal cytochrome P450 enzyme responsible for the formation of AFB,-8,9-epoxide in human liver. The use of two different ‘substrate concentrations has provided an indication of the involvement of different cytochrome P450 enzymes with varying affinities for AFB I . The higher AFB , concentration ( 124 PM) used in most of the experiments here is representative of a saturating level under the conditions used in our assay (Ramsdell and Eaton, 1990). The lower concentration (16 PM) was selected to be more reflective of the nonsaturating conditions that would be expected in human liver following dietary AFB, ingestion, yet provides sufficient formation of product to be above the assay detection limits (by at least a factor of 2). The ability of nifedipine to act as an inhibitor of microsomal AFB, oxidation was the first approach used to examine the role of P450 IIIA enzymes. Very little inhibition of AFB,8,9-epoxide formation occurred at either AFB, concentration, suggesting that the enzyme(s)

OF

AFB,

443

responsible for epoxidation has a higher affinity for AFBi than for nifedipine. Nifedipine inhibition of AFM, formation was somewhat greater than, but parallel to, that for AFB,8,9-epoxide (Fig. 1). In contrast, AFQ, formation was strongly inhibited by nifedipine. The results for AFQ, are not conclusive in proving a role of P450 IIIA4 in its formation, because a compound may show high affinity for a P450 enzyme for which it is not a substrate (Guengerich et al., 1986b). However, the failure of nifedipine to act as an inhibitor of AFBl epoxidation argues against a major role of cytochrome P450 IIIA enzymes with nifedipine oxidation activity, including P450 IIIA4. The macrolide antibiotic troleandomycin forms an inactive metabolite complex with cytochrome(s) P450 in both rats and humans (Pessayre et al., 198 1; Pessayre et al., 1982). In the case of rat liver cytochrome P450, this interaction has been shown to be characteristic of P450 IIIAl (Wrighton et al., 1985). Thus, with the assumption that TAO also interacts specifically with human P450 IIIA enzymes, the inhibition of AFB, oxidation by TAO pretreatment was tested (Figs. 2 and 3). The extent of inhibition was dependent on the concentration of AFB, in the incubations, with some effect on AFB,-8,9-epoxide at 124 pM AFB, but none at 16 pM AFB, . These observations are consistent with the hypothesis that P450 IIIA enzymes contribute to AFB, epoxidation only at relatively high substrate concentrations. The extent of correlation between AFB, and nifedipine oxidation rates among different human liver microsome samples was determined. Poor correlations (Fig. 4) resulted when AFB, oxidation data obtained at 16 pM AFB, were compared with nifedipine oxidation rates (r2 = 0.16-0.26). The better correlation between nifedipine oxidation and AFB, epoxidation rates at the higher AFB, concentration is again consistent with a greater contribution of P450 IIIA enzymes with nifedipine oxidation activity to AFB, epoxidation with increasing substrate concentration. The level of nifedipine oxidase activity in human liver mi-

444

RAMSDELL

crosomes has been correlated with the amount of P450 IIIA4 determined by immunoblotting (Gonzalez et al., 1988; Guengerich et al., 1986a), but other human P450 IIIA isoenzymes (P450 IIIA3, P450 IIIAS, HLp2, HLp3, HFLa) may contribute to nifedipine oxidation and the amount of cross-reacting P450 immunoquantifiable by antibody against P450 IIIA4 (Bork et al., 1989; Aoyama et al., 1989). 7,8-Benzoflavone has been reported to enhance activation of AFB, by human liver microsomes (Buening et al., 1981; Shimada and Okuda, 1988; Shimada and Guengerich, 1989; Shimada et al., 1989) and to stimulate P450mediated oxidation of several other substrates, including some associated with P450 IIIA6 (rabbit 3c) (Kapitulnik et al., 1977; Schwab et al., 1988). The results shown in Fig. 5 demonstrate a striking increase in the rate of AFB, epoxidation by human liver microsomes in the presence of 7,8-benzoflavone, whereas no effect was seen on nifedipine oxidation in the same range of 7,8-benzoflavone concentrations. This differential effect is inconsistent with a major role of the same enzyme in the oxidation of both substrates. It appears that 7,8-benzoflavone may exert its activating effects by altering the substrate affinity of cytochrome P450 (Johnson et al., 1988; Schwab et al., 1988). A differential alteration of the affinity for AFB, and nifedipine of the same cytochrome P450 enzyme(s) is an alternative hypothesis that would be consistent with the results shown in Fig. 5. Stimulation of cytochrome P450 activity by 7,8-benzoflavone is not limited to P450 IIIA (Huang et al., 1981) so other isoenzymes may be involved in the increase in AFB, epoxidation. The stimulation of AI%,-8,9-epoxide formation in vitro by the flavonoid 7,8-benzoflavone raises the possibility that natural flavonoids, to which humans are routinely exposed in their diet (Kuhnau, 1976), may result in enhanced bioactivation of ingested AFBr . Antibody inhibition experiments with P450 IIIAl antibody (Fig. 6) provide additional evidence against the involvement of P450 IIIA enzymes in the epoxidation of AFBl at low

ET

AL.

substrate concentrations. The partial inhibition of AFB,-8,9-epoxide formation at the higher substrate concentrations indicates that P450 IIIA isoenzymes, including P450 IIIA4, are catalyzing AFB, epoxidation, but with relatively low efficiency (e.g., low affinity for AFB,). The results we obtained with human liver microsomes and 124 pM AFB, are comparable to those reported previously from experiments with 10 PM AFB, (Shimada and Guengerich, 1989), whereas our experiments with 16 j&M AFB, yielded quite different results. This discrepancy may be due to the different assay systems used: the previous work measured the production of genotoxic metabolites from AFB, with a bacterial mutation system and incubation periods of 2 hr (Shimada and Guengerich, 1989), whereas we used lo-min incubations and measured the appearance of oxidative metabolites of AFBi by HPLC. We also used approximately 10 times as much P450 (in nmol/ml) in our incubations, although metabolite production was still directly proportional to protein concentration and incubation time. The results of this study suggest that P450 IIIA enzymes are not the only P450 enzymes capable of activating AFB, to the AFB-8,9-epoxide, and that these other P450 enzymes by virtue of their higher affinity are primarily responsible for epoxidating AFB, at a low substrate concentration. Our conclusions are supported by the recent demonstration that five different human cytochrome P450 enzymes can activate AFB, to genotoxic metabolites (Aoyama et al., 1990). Isoenzymes P450 IA2, IIA3, IIB7, IIIA3, and IIIA4, expressed using cDNAs and a vaccinia virus system, were capable of mediating Al%,induced reversion in Salmonella typhimurium TA98 and binding of [3H]AF’BI to endogenous DNA in cells expressing the human P45Os. Crespi et al. ( 1990) have also demonstrated that human cytochrome P450 IIA3, expressed in a human B lymphoblastoid cell line, is capable of activating APB, to cytotoxic and mutagenic metabolites, presumably AFB-8,9-epoxide. In addition to a loo-fold increase in

MICROSOMAL

ACTIVATION

mutagenicity of AFB in cells expressing P450 IIA3, DNA binding of [3H]AFB was greatly elevated above that found in cells not expressing P450 IIA3. Our results are consistent with the hypothesis that P450 IIIA4 in human liver microsomes can convert AFB, to the 8,9-epoxide (Shimada and Guengerich, 1989), but we have demonstrated that cytochrome P450 IIIA isoenzymes are quantitatively important only at AFB 1 concentrations approaching saturating conditions. Another cytochrome P450 enzyme(s) appears to be involved at lower substrate concentrations. Recent experiments utilizing cell lines transfected with genes for specific human cytochromes P450 have also demonstrated that enzymes in addition to the cytochromes P450 IIIA subfamily are capable of activating AFBl , including P450 IA2, IIB7, and IIA3 (Aoyama et al., 1990; Crespi et al., 1990). The determination as to which of these forms is primarily responsible for the activation of aflatoxin B1 at the low concentrations encountered in the human diet will require careful kinetic evaluation of each of these forms. ACKNOWLEDGMENTS This work was supported by NIH Grants T32 ES-07032, ES-03933, ES-04696. and CA-4756 1. A.P. is the recipient of a Research Career Development award from NIH (ES 00 166) and is supported by grants from the Wesley Foundation and NIH (GM37044). We thank Dr. Mont Juchau for his helpful comments regarding this manuscript. The authors are grateful for the assistance of Drs. William Trager and Alan Rettie, Department of Medicinal Chemistry, University of Washington, for access to human liver tissue samples. Additional human liver microsomes were generously provided by Drs. Glenn Sipes and John Barr, Department of Pharmacology and Toxicology, University of Arizona, with support from NIH (NO I -ES-55 112). Note added in proqf Since the revision of this manuscript, two papers have been published which support the hypothesis that human hepatic microsomal P450 IIIA4 is not the principle P450 which activates AFB, to AFBO at low substrate concentrations. Forrester et al. (Proc. Nat/. Acad. Sci. USA 87,8306-83 10. 1990) demonstrated that P45Os other than P450 IIIA4, notably P450 IA2, P450 IIA I. and P450 IIC. were effective at activating AFB, .

OF

AFB,

445

Crespi et al. (Carcinogenesis 12, 355-359. 1991) provide evidence that human P450 IA2 is lo-fold more effective than P450 IIIA4 at activating AFB, to mutagenic metabolites at low substrate concentrations, suggesting that P450 IA2 may be the low K, form.

REFERENCES AOYAMA. T., YAMANO. S.. GUZELIAN, P. S., GELBOIN, H. V., AND GONZALEZ. F. J. (1990). Five of 12 forms of vaccinia virus-expressed human hepatic cytochrome P450 metabolically activate aflatoxin B1. Proc. Natl. Acad. Sci. US.4 87,4790-4793. AOYAMA, T., YAMANO, S., WAXMAN, D. J.. LAPENSON. D. P., MEYER, U. A., FISCHER, V., TYNDALE, R., INABA. T.. KALOW, W., GELBOIN, H. V., AND GONZALEZ, F. J. (1989). Cytochrome P-450 hPCN3, a novel cytochrome P-450 IIIA gene product that is differentially expressed in adult human liver. J. Biol. Chem. 264, 10,388-10.395. BORK. R. W., Muro, T., BEAUNE, P. H.. SRIVASTAVE. P. K.. LLOYD, R. S., AND GUENGERICH. F. P. (1989). Characterization of mRNA species related to human liver cytochrome P-450 nifedipine oxidase and the regulation of catalytic activity. J. Biol. Chem. 264, 910919. BUENING, M. K.. CHANG, R. L., HUANG. M.-T., FORTNER. J.G.. WOOD. A. W.,AND~ONNEY. A. H. (1981). Activation and inhibition of benzo[a]pyrene and aflatoxin B, metabolism in human liver microsomes by naturally occurring flavonoids. Cancer Res. 41, 67-72. BUSBY. W. F.. AND WOGAN. G. N. (1984). Aflatoxins. Chemical Carcinogens (C. E. Searle. Ed.), pp. 945- 1136. American Chemical Society, Washington DC. CAMPBELL, T. C.. AND HAYES, J. R. (1976). The role of aflatoxin metabolism in its toxic lesion. Toxicol. .4ppl. Pharmacol. 35, 199-222. COULOMBE, R. A., SHELTON, D. W., SINNHUBER, R. 0.. AND NIXON, J. E. (1982). Comparative mutagenicity of aflatoxins using a Salmonella/trout hepatic enzyme activation system. Carcinogenesis 3, 126 1-1264. CRESPI. C. L.. PENMAN, B. W.. LEAKEY. J. A. E., ARLOTTO. M. P.. STARK. A.. PARKINSON. A., TURNER. T.. STEIMEL. D. T., Ruw, K., DAVIES, R. L.. AND LANGENBACH, R. (1990). Human cytochrome P450IIIA3: Sequence, role of the enzyme in the metabolic activation of promutagens, comparison to nitrosamine activation by human cytochrome P450IIE 1. Carcinogenesis 11, 1293- 1301. CRESPI. C. L., PENMAN, B. W., STEIMEL, D. T.. GELBOIN. H. V., AND GONZALEZ, F. J. (199 1). The development of a human cell line stably expressing human CYP3A4: role in the metabolic activation of aflatoxin Bl and comparison to CYPlA2 and CYP2A3. Carcinogenesis 12,355-359. DUIGNAN, D. B., SIPES, I. G., LEONARD, T. B., AND HALPERT, J. R. (1987). Purification and characterization of

446

RAMSDELL

the dog hepatic cytochrome P-450 isozyme responsible for the metabolism of 2.2’,4.4’,5.5’-hexachlorobiphenyl. Arch. Biochem. Biophys. 255,290-303. EBEL, S., SCHUTZ, H., AND HORNITSCHEK, A. (1978). Untersuchungen zur Analytik von Nifedipine unter besonderer Berucksichtigung der bei Lichtexposition entstehenden Umwandlungsprodukte. .4tzneim.-Forsch. 28,2188-2193.

ESSIGMANN. J. M.. CROY, R. G., BENNETT. R. A., AND WOGAN, G. N. (1982). Metabolic activation of aflatoxin B, : Patterns of DNA adduct formation, removal, and excretion in relation to carcinogenesis. Drug MetaD. Rev. 13,581-602.

FORRESTER.L. M., NEAL, G. E.. GLANCEY, M. J., AND WOLF, R. C. (1990). Evidence for involvement of multiple forms of cytochrome P-450 in aflatoxin Bl metabolism in human liver. Proc. Nutl. ,lcad. Sci. US.4 87,8306-8310.

GONZALEZ. F. J., SCHMID, B. J.. UMENO. M., MCBRIDE, 0. W., HARDWICK, J. P.. MEYER. U. A.. GELBOIN, H. V., AND IDLE, J. R. (1988). Human P450 PCNl: Sequence. chromosome localization and direct evidence through cDNA expression that P450 PCNl is nifedipine oxidase. DN.4 I, 79-86. GROOPMAN, J. D., CAIN, L. G.. AND KENSLER, T. W. ( 1988). Aflatoxin exposure in human populations: Measurements and relationship to cancer. CRC Crit. Rev. Toxicol. 19, I 13- 145. GUENGERICH, F. P.. MARTIN. M. V.. BEAUNE, P. H.. KREMERS. P.. Wolff, T.. AND WAXMAN. D. J. (1986a). Characterization of rat and human liver microsomal cytochrome P-450 forms involved in nifcdipine oxidation. a prototype for genetic polymorphism in oxidative drug metabolism. J. Biol. Chem 261, 505 I-5060. GUENGERICH. F. P., MULLER-ENOCH. D., AND BLAIR. I. A. (1986b). Oxidation of quinidine by human liver cytochrome P-450. Mol. Pharmucol. 30, 287-295. HALPERT, J., NAUSLAND. B.. AND BETMAN, I. (1983). Suicide inactivation of rat liver cytochrome P-450 by chloramphenicol it1 viva and in vitro. Mol. Phartnacol. 23.445-452.

HALVORSON. M.. GREENWAY, D.. EBERHART. D., FITZGERALD, K., AND PARKINSON, A. (1990). Reconstitution of testosterone oxidation by purified rat cytochrome P450 p (IIIAI). Arch. Biochem. Biophys. 277, 166-180. HUANG, M.-T., JOHNSON,E. F.. MULLER-EBERHARD. U., KOOP. D. R., COON, M. J.. AND CONNEY. A. H. ( 198 1). Specificity in the activation and inhibition by flavonoids of benzo[a]pyrene hydroxylation by cytochrome P-450 isoenzymes from rabbit liver microsomes. J. Biol. Chem. 256, 10,897-10.901. JOHNSON. E. F., SCHWAB, G. E., AND VICKERY, L. E. ( 1988). Positive effectersof the binding of an active sitedirected amino steroid to rabbit cytochrome P-450 3c. J. Biol. Chetn. 263. 17,672-17.677. KAPITULNIK, J., POPPERS,P. J.. BUENING, M. K., FORTNER. J. G.. AND CONNEY. A. H. (1977). Activation of

ET AL. monooxygenase in human liver by 7.8-benzoflavone. Clin. Pharmacol. Ther. 22, 475-484. KITADA, M., KAMATAKI, T.. ITAHASHI, K., RIKIHISA. T., KATO, R., AND KANAKUBO, Y. (1985). Purification and properties of cytochrome P-450 from homogenates of human fetal livers. Arch. Biochem. Biophys. 241, 275280.

KOMORI. M.. NISHIO, K.. OHI. H., KITADA. M., AND KAMATAKI, T. (1989). Molecular cloning and sequence analysis of cDNA containing the entire coding region for human fetal liver cytochrome P-450. J. Biochem. 105, 161-163. KUHNAU, J. (1976). The flavanoids. A class of semi-essential food components: Their role in human nutrition. World Rev. Nutr. Diet. 24, I 17-191. LOEV, B.. AND SNADER, K. M. (1965). The Hantzch reaction. 1. Oxidative dealkylation of certain dihydropyridines. J. Org. Chetn. 30, 1914-1916. MOLOWA. D. T., SCHEUTZ. E. G., WRIGHTON, S. A.. WATKINS. P. B., KREMERS, P., MENDEZ-PICON. G.. PARKER.G. A., AND GUZELIAN, P. S. (1986). Complete cDNA sequence of a cytochrome P-450 inducible by glucocorticoids in human liver. Proc. Natl. Acad. .Qi. us483,5311-5315.

MONROE, D. H.. AND EATON, D. L. (1987). Comparative effects of butylated hydroxyanisole on hepatic in vivo DNA binding and in vitro biotransformation of aflatoxin B, in the rat and mouse. Toxicol. .4ppl. Pharmacol. 90, 401-409. NEBERT, D. W., NELSON, D. R.. ADESNIK. M., COON, M. J., ESTABROOK. R. W., GONZALES, F. J., GUENGERICH,F. P.. GUNSALUS, I. C., JOHNSON,E. F., KEMPER,B.. LEVIN. W., PHILLIPS,I. R., SATO, R., AND WATERMAN. M. R. (1989). The P450 superfamily: Updated listing of all genes and recommended nomenclature for the chromosomal loci. DNA 8, l-13. ODA, Y., NAKAMURA. S.-I., OKI. I., KATO, T., AND SHINAGAWA, H. (1985). Evaluation ofthe new system (umutest) for the detection of environmental mutagens and carcinogens. Murat. Res. 147, 219-229. OMURA, T.. AND SATO, R. (1964). The carbon monoxide binding pigment of liver microsomes. I. Evidence for its hemoprotein nature. J. Biol. Chem. 239, 2370-2378. PESSAYRE, D., DESCATOIRE. V., KONSTANTINOVAMITCHEVA, M., WANDSHEER.J.-C..COBERT,B.. LEVEL, R.. BENHAMOU, J. P.. JAQUEN, M., AND MANSUY, D. (198 1). Self-induction by triacetyloleandomycin of its own transformation into a metabolite forming a stable 456 nm-absorbing complex with cytochrome P-450. Biochetn. Pharmucol. 30, 553-558. PESSAYRE. D., LARREY. D.. VITAUX. J., BREIL. P.. BELGHITI, J., AND BENHAMOU, J. P. (1982). Formation of an inactive cytochrome P-450 Fe(II)-metabolite complex after administration of troleandomycin in humans. Biochem. Pharmacol. 31, 1699. RAMSDELL. H. S., AND EATON, D. L. (1988). Modification of allatoxin B, biotransformation in vitro and DNA

MICROSOMAL

ACTIVATION

binding in vivo by dietary broccoli in rats. J. To.xicol. Environ. Health 25, 269-284. RAMSDELL, H. S., AND EATON, D. L. (1990). Species susceptibility to aflatoxin B, carcinogenesis: Comparative kinetics of microsomal biotransformation. Cancer Res. 50,6 15-620. SCHUETZ, J. D., MOLOWA, D. T.. AND GUZELIAN. P. S. (1989). Characterization of a cDNA encoding a new member of the glucocorticoid-responsive cytochromes P450 in human liver. Arch. Biochem. Biophys. 214,355365.

G. E., RAUCY, J. L., AND JOHNSON,E. F. (1988). Modulation of rabbit and human cytochrome P-450catalyzed steroid hydroxylations by a-naphthoflavone. Mol. Pharmacol. 33,493-499. SHIMADA. T., AND GIJENGERICH,F. P. (1989). Evidence for cytochrome P-450NF. the nifedipine oxidase. being the principal enzyme involved in the bioactivation of aflatoxins in human liver. Proc. Natl. Acad. Sci. USA SCHWAB,

86,462-465.

SHIMADA. T., AND NAKAMURA, S.-I. (1987). Cytochrome P-450-mediated activation of procarcinogens and promutagens to DNA-damaging products by measuring expression of umu gene in Salmonella typhimurium TA1535/pSK1002. Biochem. Pharmacol. 36, 19791987. SHIMADA. T., AND OKUDA, Y. (1988). Metabolic activation of environmental carcinogens and mutagens by human liver microsomes. Biochem. Pharmacol. 31,459465.

SHIMADA, T.. MARTIN, M. V., PRUESS-SCHWARTZ,D., MARNETT. L. J.. AND GUENGERICH,F. P. (1989). Roles of individual human cytochrome P-450 enzymes in the bioactivation of benzo(a)pyrene, 7,8-dihydroxy-7,8dihydrobenzo(a)pyrene, and other dihydrodiol derivatives of polycyclic aromatic hydrocarbons. Cancer Rex 49,6304-6312.

OF AFB,

447

SMITH. P. K., KROHN. R. I.. HERMANSON.G. T., MALLIA, A. K.. GARTNER, F. H., PROVENZANO, M. D.. FUJIMOTO, E. K.. GOEKE, N. M., OLSON, B. J., AND KLENK, D. C. (1985). Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76-85. SWENSON,D. H., LIN, J.-K., MILLER, E. C., AND MILLER, J. A. (1977). Aflatoxin B,-2,3-oxide as a probably intermediate in the covalent binding of aflatoxins B, and BZto rat liver DNA and ribosomal RNA in vivo. Cancer Res. 37, 172-181. TESTA. R.. DOLFINO, E., RESCHIOTI’O, C., SECCHI, C.. AND BIONDI, P. A. (1979). GLC determination of nifedipine, a light sensitive drug, in plasma. II Farmaco Ed. Pr. 34,463-473. VANDER HOEVEN. T. A., AND COON, M. J. (I 974). Preparation and properties of partially purified cytochrome P-450 and reduced NADPH reductase from rabbit liver microsomes. J. Biol. Chem. 249. 6302-63 10. WATKINS, P. B., WRIGHTON. S. A.. MAUREL. P.. SCHUETZ, E. G., MENDEZ-PICON, G., PARKER, G. A., AND GUZELIAN, P. S. (1985). Identification of an inducible form of cytochrome P-450 in human liver. Proc. Nat. Acad. Sci. USA 82, 6310-63 14. WRIGHTON. S. A., MAUREL, P.. SCHUETZ, E. G., WATKINS, P. B., YOUNG, B., AND GUZELIAN, P. S. (1985). Identification of the cytochrome P-450 induced by macrolide antibiotics in rat liver as the glucocorticoid responsive cytochrome P-450. Biochemistry 24, 2 I7 l2178. WRIGHTON. S. A., RING. B. J.. WAXINS, P. B.. AND VANDENBRANDEN, M. (1989). Identification of a polymorphically expressed member of the human cytochrome P-450111 family. Mol. Pharmacol. 36, 97-105. WRIGHTON, S. A.. AND VANDENBRANDEN, M. (1989). Isolation and characterization of human fetal liver cytochrome P450 HLp2: A third member of the P450 III gene family. Arch. Biochem. Biophys. 268, 144- 15 1.