Dominant Contribution of P450 3A4 to the Hepatic ... - ACS Publications

Department of Pharmacology, Johannes Gutenberg UniVersity, Mainz, Germany. ReceiVed December 23, 2005. The hepatic carcinogen aflatoxin B1 (AFB1) is ...
0 downloads 0 Views 182KB Size
Download PDF
Chem. Res. Toxicol. 2006, 19, 577-586

577

Dominant Contribution of P450 3A4 to the Hepatic Carcinogenic Activation of Aflatoxin B1 Landry K. Kamdem,† Ingolf Meineke,† Ute Go¨dtel-Armbrust,‡ Ju¨rgen Brockmo¨ller,† and Leszek Wojnowski*,‡ Department of Clinical Pharmacology, Georg-August UniVersity, Goettingen, Germany, and Department of Pharmacology, Johannes Gutenberg UniVersity, Mainz, Germany ReceiVed December 23, 2005

The hepatic carcinogen aflatoxin B1 (AFB1) is metabolized in the liver by at least four different P450s, all of which exhibit large interindividual differences in the expression levels. These differences could affect the individual risk of hepatocellular carcinoma (HCC). We investigated the metabolism of AFB1 in a panel of 13 human liver microsomal preparations using a hepatic abundance model, which takes into account the specific kinetic parameters and the expression levels of these P450s. We found a 12fold variability in the production rate of the carcinogenic metabolite AFB1-8,9-epoxide (AFBO) and a 22-fold variability in the production of the detoxification product AFQ1. The ratio between the AFBO and the AFQ1 production rates varied between 1:19 and 1:1.7. P450 3A4 contributed a majority of AFBO and AFQ1, and its expression level was the most important determinant of the AFB1 disposition toward these primary metabolites. P450 3A5, which exclusively produced AFBO, was the second-most important enzyme activating AFB1 to AFBO, followed by P450 3A7 and P450 1A2. The relative contribution of AFBO by P450 3A5 strongly depended on the concomitant expression of P450 3A4, and it was as high as 15% in a P450 3A5 high expressor with the lowest P450 3A4 expression of all livers. The P450 1A2-specific AFB1 detoxification product AFM1 was not detected. In conclusion, the variable expression of P450s has a major effect on the carcinogenic activation of AFB1, which may affect the individual predisposition to HCC. P450 3A4 expression is the most important determinant of AFB1 activation to AFBO. The contribution of P450 1A2 to AFB1 metabolism appears to be negligible and may have been overestimated. Targeted chemoprevention of AFB1-associated HCC should consider P450 3A4 inhibitors and avoidance of P450 3A4 inducers. Introduction Hepatocellular carcinoma (HCC) is the most common type of liver cancer and the fifth most common cancer in the world. Annually, HCC affects more than 500000 people worldwide, and five year mortality exceeds 95% (1). The main risk factors for HCC are hepatitis B and C viruses and exposure to aflatoxin (AF) B1. AFB1 is produced by certain Aspergillus species, which contaminate human and animal foods, such as peanuts and corn, during plant growth and after harvest. The highest exposure to AFB1 has been observed in parts of Africa, China, and Southeast Asia, which are also characterized by a high incidence of HCC (2). Biotransformation is necessary for the carcinogenic effect of AFB1 in the liver. The main direct products of AFB1 metabolism are AFM1, AFQ1, and AFB1-8,9-epoxide (AFBO).1 The last one, an epoxidation product, is a mix of the stereoisomers AFB1-endo-8,9-epoxide and AFB1-8,9-exo-epoxide. Only the last compound is genotoxic. It binds to DNA to form predominantly the 8,9-dihydro-8-(N7-guanyl)-9-hydroxy-AFB1 (AFB1-N7-Gua) adduct. The metabolites are poorer substrates for epoxidation than AFB1 and, consequently, less mutagenic, carcinogenic, and toxic than AFB1 (3). * To whom correspondence should be addressed. Tel: +49-61313933460. Fax: +49-941 5992 36727. E-mail: [email protected]. † Georg-August University. ‡ Johannes Gutenberg University. 1 Abbreviations: AFBO, AFB -8,9-epoxide; AFBO-GSH, glutathione 1 conjugate of AFB1-8,9-epoxide; HLM, human liver microsomes; Km, Michaelis-Menten constant; n, Hill coefficient; P450, cytochrome P450; Vmax, maximum reaction velocity.

Previous studies implicated P450 3A4 and P450 1A2 in the in vivo AFB1 metabolism to AFBO (4-9). P450 1A2 has been predicted to predominate AFBO production due to the higher affinity for AFB1 (5, 6). Accordingly, cells expressing P450 1A2 were more sensitive to AFB1 than those expressing P450 3A4 (10). On the other hand, a correlation approaching statistical significance has been reported between P450 3A4 (but not P450 1A2) expression and AFB1-DNA adduct levels in a panel of adult human livers (11). The assessment of the relative importance of these P450s for AFBO production has also been complicated by their other, exclusive AFB1 metabolites. Thus, in addition to the production of the genotoxic AFBO, P450 3A4 detoxifies AFB1 via 3-R-hydroxylation to AFQ1. In contrast, P450 1A2 converts AFB1 to another hydroxylated metabolite, AFM1 (3). Even less clear is the in vivo role of P450 3A5, which was investigated less extensively following reports that it was relatively inefficient in AFB1 bioactivation (7, 12, 13) and expressed in the liver at much lower levels than P450 3A4 (14, 15). More recent research indicates that P450 A5 may account for 30-40% of the hepatic P450 3A pool, at least in some individuals. Indeed, P450 3A5 exhibits a classic polymorphism, with expression reduced 10-20-fold in homozygous carriers of the point mutation known as P450 3A5*3, which leads to aberrant splicing (16). In Africans, the expression of the isozyme may be abolished by at least two further common gene variants, P450 3A5*6 and P450 3A5*7 (17), which should be considered in the prediction of P450 3A5 expression in this ethnic group

10.1021/tx050358e CCC: $33.50 © 2006 American Chemical Society Published on Web 03/24/2006

578 Chem. Res. Toxicol., Vol. 19, No. 4, 2006

(18). About 90% Caucasians, 67% Japanese, and 30-45% African-Americans are P450 3A5 low expressors (16, 17, 19). Most importantly, unlike P450 3A4 and P450 1A2, P450 3A5 preferentially transforms AFB1 into the genotoxic AFBO (12, 20, 21). For these reasons, P450 3A5 may play a more important role in the AFB1 activation than originally thought. In agreement with this hypothesis, we recently reported increased blood plasma levels of AFB1-albumin adducts in P450 3A5 high expressors from Gambia, particularly in those with low concomitant P450 3A4 expression (18). Finally, AFBO production could be contributed to by P450 3A7, which is expressed in some individuals at levels comparable to P450 3A5 levels in P450 3A5 high expressors (22). The unequivocal identification of the enzyme(s) responsible for the formation of the carcinogenic AFBO and their relative contributions is important, since P450 3A4, P450 3A5, P450 3A7, and P450 1A2 are expressed in livers at inter- and intraindividually highly variable levels (16, 23, 24). This variability reflects the effect of the individual genetic makeup (e.g., the P450 3A5 and P450 3A7 polymorphisms) and of environmental factors on P450 transcription (induction). It is conceivable that the individual expression levels of P450s producing AFBO could affect the individual risk to develop HCC. If this was confirmed, preventive measures could be developed and applied to individuals at particularly high risk. These considerations are of practical importance, since genetic and phenotypic markers for the individual P450s are becoming available. In a shorter term, it could become easier to evaluate how other, nongenetic factors, such as hepatitis status and nutrition, influence AFB1 metabolism. For all of these reasons, we investigated the metabolism of AFB1 in a bank of liver samples phenotyped for the expression of P450 3A4, P450 3A5, and P450 1A2. This was followed by a first calculation of the relative contributions of these P450s to AFBO formation in human livers at AFB1 concentrations encountered in exposed individuals.

Experimental Procedures Caution: The following chemicals are hazardous and should be handled carefully: all AFs, particularly AFB1. ProtectiVe clothing should be worn. Crystalline material presents an inhalation hazard because the crystals deVelop an electrostatic charge and cling to dust particles. After use, treatment of AF solutions and of contaminated Vessels with commercial bleach solutions is recommended for decontamination. Chemicals and Reagents. AFs B1, Q1, M1, and G1, reduced glutathione (GSH), and purified rat liver glutathione S-transferase (GST) (a mixture of R- and µ-class enzymes) were purchased from Sigma-Aldrich (Germany). NADPH was purchased from Roche (Mannheim, Germany). The bank of 13 phenotyped liver samples [BD Gentest single donor human liver microsomes (HLM)] with catalog nos. HG64, HG74 (P450 3A5 low expressors), HH54, HH47, HH91, HG95, HH108, HH86, HH3, HH1, HH89, HH48, and HH9 (P450 3A5 high expressors), as well as BD Gentest’s baculovirus-derived microsomes expressing P450 1A2/OR (catalog no. P203, lot 28), P450 3A4/OR (catalog no. P207, lot 19), P450 3A7/OR/b5 (P235, lot 7) and P450 3A5/OR (P235, lot 18), P450 2A6/OR (catalog no. P204, lot 14), P450 2B6/OR (catalog no. P210, lot 9), P450 2C8/OR/b5 (catalog no. P252, lot 10), and P450 2C9*1/ OR (catalog no. P218, lot 15) were obtained from Natutec (Frankfurt/Main, Germany). Purified cytochrome b5 (catalog no. P2252, lot 24548B) and anti-P450 1A2 monoclonal antibody (catalog no. P2733) were purchased from Panvera (Invitrogen Corp., Karlsruhe, Germany). The glutathione conjugate of AFBO was a mix of endo and exo stereoisomers (ratio, 1:10). and it was was synthesized by Dr. A. Seidel (Biochemical Institute of Environmental Carcinogens, Grosshansdorf, Germany).

Kamdem et al. Phenotyping of HLMs. The quantifications of the HLM total P450 content, of P450 3A4, P450 3A5 expression, and of the P450 1A2 activity marker phenacetin O-deethylase were carried out by the manufacturer, as described in product information sheets. The quantification of the P450 1A2 protein expression was carried out by immunoblotting using a monoclonal antibody against human P450 1A2 (Panvera, Invitrogen). Briefly, microsomal protein (0.005-0.05 pmol of recombinant P450 1A2 standard and 5 µg of liver microsomal protein) was denatured for 3 min at 100 °C in 125 mM Tris buffer (pH 6.8) containing 2 mM EDTA, 4% SDS, 20% glycerol, 2% β-mercaptoethanol, and 0.02% bromphenol blue. Recombinant P450 1A2 (Natutec) was used to generate calibration curves in concentrations ranging from 0.005 to 0.05 pmol/lane. Protein was separated on a 7.5% SDS-polyacrylamide gel and transferred to polyvinylidene fluoride membranes (Roche Diagnostics) by electroblotting at 70 mA for 1 h in 25 mM Tris/40 mM DL-norleucin/20% methanol transfer buffer. The blots were blocked for 1 h with 3% BSA in TBS-T (0.1% Tween-20, 15 mM sodium chloride, and 1 mM Tris; pH 7.4). The blots were then incubated for 1 h with the primary antibody (1:2500 in 1% BSA/TBS-T buffer), washed, incubated with anti-mouse IgG peroxidase conjugate (1:20 000 in 1% BSA/TBS-T buffer) for 1 h, and washed again. P450 1A2 protein was detected by means of a chemoluminescence reaction (SuperSignal WestDura; Pierce Biotechnology, Rockford, IL). Blots were exposed to film, developed, and quantified by use of Quantity One 4.21 software (Biorad). The calibration curve was linear within the range of 0.005-0.05 pmol. AFB1 Metabolism by HLM and Baculovirus-Expressed Cytochrome P450 (P450). Stock solutions of AFB1 were prepared in methanol and were allowed to evaporate under N2. Components were added as follows: 100 mmol/L potassium phosphate buffer (pH 7.4), 5 mmol/L reduced glutathione, purified rat liver glutathione S-transferase (0.2 mg/mL), and HLM or baculovirusexpressed P450. The final incubation volume was 100 µL. The end concentration of AFB1 ranged from 25 to 500 µM. The reaction mixture with P450 3A4/OR or P450 3A5/OR was additionally supplemented with purified cytochrome b5 (P450:b5 ratio of 1:1) following a previously described method (25). The reaction was initiated through addition of NADPH (final concentration, 1 mM), allowed to proceed for 30 min, and stopped by addition of 100 µL of ice-cold methanol containing 10 µM AFG1 as an internal standard. The incubation mixtures were then centrifuged at 10000g for 5 min. Supernatants were transferred into new tubes, and 100 µL was used for the HPLC analysis. The formation of the main AFB1 metabolites was linear with time between 10 and 30 min and with protein ranging from 100 to 500 µg for HLMs (AFQ1 and AFBO) and from 20 to 100 pmol for baculovirus-expressed cytochromes (AFQ1, AFM1, and AFBO). The substrate consumption was 0.05). The major AFB1 metabolite detected in these HLMs was AFQ1 followed by AFBO (last panel of Figure 1, Figure 2E,H and Table 2). AFM1 was detected in none of the liver samples. There was a 12-fold variation in the Vmax of the AFBO formation. The variation in Vmax of AFQ1 formation was 22fold. There was a statistically significant correlation between Vmax values of AFBO and AFQ1 formation (Figure 3). The ratio of AFQ1 and AFBO Vmax median values in HLMs (7.2:1) was almost identical to the corresponding ratio obtained for the baculovirus-expressed P450 3A4 (8:1) given above. We com-

pared the kinetic parameters for the formation of these metabolites between the recombinant P450 3A4 and HLM. This was done with the consideration of the P450 3A4 expression levels in HLMs (Table 2). The median Vmax (1.91 pmol/pmol P450/ min), Km (225 µM), and Vmax/Km (0.01 µL/pmol P450/min) values for AFBO formation, as well as median Vmax (13.74 pmol/ pmol P450/min), Km (415 µM), and Vmax/Km (0.033 µL/pmol P450/min) values for AFQ1 formation in this liver bank were very similar to the corresponding values measured with recombinant P450 3A4 (Table 1). Contributions of the Individual P450s to the Hepatic AFB1 Metabolism. These results suggested that most of AFQ1 and AFBO production was catalyzed by P450 3A4. To verify this, we compared Vmax values of AFQ1 and AFBO productions measured in the individual HLMs with the expression levels of P450 3A4, P450 1A2, and P450 3A5 and with the P450 1A2 activity marker phenacetin O-deethylase. In agreement with the dominant role of P450 3A4 in AFQ1 production, Vmax of AFQ1 formation correlated statistically significantly with the expression of the P450 3A4 protein (Figure 4A). The corresponding values for P450 3A5 (Figure 4B) and P450 1A2 (both protein and phenacetin O-deethylase, Figure 4C, D) were low and statistically insignificant, in agreement with the incapability of these P450s to catalyze AFQ1 (Figure 1 and Table 1). The correlation coefficient r was statistically significant between the P450 3A4 protein expression and the Vmax of AFBO formation (Figure

Table 2. Kinetic Constants for the Conversion of AFB1 to AFBO and AFQ1 in a Bank of Human Liver Samplesa AFBO materials

Vmax

HG64 477 HG74 613 HH54 365 HH47 840 HH91 332 HG95 236 HH86 541 HH3 967 HH108 384 HH89 261 HH48 261 HH9 782 HH1 4633 mean median range

Km 307 186 102 414 106 350 163 267 90 125 260 249 230

770 374 573 225 281-3366 104-1553

AFQ1 n

Vmax

1.1 3560 NA 4308 NA 2865 1 11281 NA 6924 1.3 657 2 4046 NA 1649 1.6 1587 NA 2261 NA 1335 1.3 8669 NA 16871 5552 4124 753-16871

Km 312 450 283 1762 724 645 310 109 112 370 256 1098 1207

n NA NA NA 1.2 1.1 1.7 2 NA NA 1.2 1.5 1 NA

total P450 content 340 250 240 260 340 230 330 370 310 300 210 210 490

protein 3A4 NA NA 87 62 46 29 83 76 27 38 30 53 110

3A5 1.2 1.1 8.6 9.3 5.7 6.9 7.6 11 11 11 9.8 12 12

1A2 0.27 4.32 3.72 2.34 3.84 3.43 2.4 1.91 3.76 2.74 1.8 4.27 2.15

summary statistics for all livers (n ) 13) 685 299 58 8.2 2.84 415 300 53 9.3 2.74 120-1937 210-490 27-110 1.1-12 0.27-4.32

phenacetin O-deethylase activity (pmol/mg/min) 145 580 230 360 390 290 300 210 330 240 130 660 400 328 300 130-660

race Caucasian Caucasian Asian African-American Caucasian Hispanic Hispanic Caucasian Caucasian Caucasian Caucasian Caucasian Caucasian

age gender (years) M M F F F F M M F M M M F

63 32 62 53 55 47 57 38 27 33 62 51 31 47 42 27-63

a The units corresponding to the kinetic constants and protein levels described in this table are pmol/mg/min (V max), µM (Km), and pmol/mg (total P450, P450 3A4, P450 3A5, and P450 1A2).

582 Chem. Res. Toxicol., Vol. 19, No. 4, 2006

Kamdem et al.

Figure 4. Correlation between Vmax of AFQ1 formation and (A) P450 3A4 protein, (B) P450 3A5 protein, (C) P450 1A2 protein, and (D) phenacetin O-deethylase activity in a panel of HLMs.

Figure 5. Correlation between Vmax of AFBO formation (measured as AFBO-GSH) and (A) P450 3A4 protein, (B) P450 3A5 protein, (C) P450 1A2 protein, and (D) phenacetin O-deethylase activity in a panel of HLMs.

5A). The corresponding values for P450 3A5 (Figure 5B) and P450 1A2 (both protein and phenacetin O-deethylase, Figure 5C,D) were lower and statistically insignificant, even though both P450s are capable of AFBO production (Figures 1 and 2B,D and Table 1). The absence of a correlation between the hepatic P450 3A5 expression level and AFBO production was in apparent dis-

agreement with the considerable clearance of AFB1 to AFBO by the recombinant P450 3A5 (Table 1). To resolve this discrepancy, we calculated the contribution of P450 3A5 to AFBO formation at an epidemiologically relevant AFB1 concentration of 0.1 µM using a hepatic abundance model. The model takes into account both the kinetic parameters (Table 1) and the hepatic expression levels of P450 3A5 (Table 2). Under

Cytochrome P450 and AFB1 Metabolism

Chem. Res. Toxicol., Vol. 19, No. 4, 2006 583

Discussion

Figure 6. Correlations between the relative contribution of P450 3A5 (A) and P450 1A2 (B) to AFBO formation (measured as AFBO-GSH) and the expression of the P450 3A4 protein in a panel of HLMs from P450 3A5 high expressors. Data calculated using a hepatic abundance model for the in vivo relevant AFB1 concentration of 0.1 µM using kinetics and protein expression data from Tables 1 and 2.

these conditions, the contribution of P450 3A5 showed an inverse, statistically significant correlation with the amount of P450 3A4 protein (Figure 6A; r ) 0.82; p < 0.001). The highest relative contribution of P450 3A5 to AFBO formation (15.3%) was observed in the liver HH108, which is a P450 3A5 high expresser with the lowest P450 3A4 expression of all livers investigated. A similar relationship was observed for P450 1A2. The highest relative contribution of P450 1A2 to AFBO formation (5.35%) was also observed in the liver HH108, which expressed the lowest P450 3A4 protein content and the highest P450 1A2 protein amount among the liver sample set. A similar statistically significant inverse correlation was observed between the relative contribution of P450 1A2 to AFBO formation and the P450 3A4 protein (Figure 6B; r ) 0.82; p < 0.001). Using the same model, we calculated the relative contributions of P450 3A4, P450 3A5, P450 3A7, and P450 1A2 to the primary conversion of AFB1 (0.1 µM). On average, 89% of AFB1 are converted to AFQ1 and 11% to AFBO (Figure 7), but depending on the liver, the ratio of the two metabolites may vary between 19:1 (95% AFQ1 and 5% AFBO) and 1.7:1 (63% AFQ1 and 37% AFBO). P450 3A4 is the exclusive source of AFQ1, except for individuals with increased P450 3A7 expression. In the latter group, P450 3A7 contributes up to 22% of AFQ1. This value results from the specific clearance of AFB1 to AFQ1 by P450 3A7 (Table 1) and the hepatic P450 3A7 expression levels of 24-90 pmol/mg protein (22). AFBO is catalyzed by P450 3A4, 3A5, 1A2, and 3A7. A majority of AFBO is contributed by P450 3A4, but its share varies, depending on the concomitant expression of P450 3A5 and P450 3A7. P450 3A5 may contribute between 4 and 15% of AFBO in P450 3A5 high expressers. The respective value for P450 3A7 high expressors is 5-7%, whereas P450 1A2 contributes less than 5% of AFBO.

The primary steps of detoxification and carcinogenic activation of AFB1 are catalyzed by P450 3A4, P450 3A5, P450 3A7, and P450 1A2. All of these enzymes exhibit substantial interindividual expression variability, which could affect the ratio between the AFB1 detoxification and the activation. This ratio is likely an important determinant of the individual risk to AFB1induced HCC. Our study constitutes a first investigation of the metabolism of AFB1 by these P450s at concentrations encountered in the liver tissue of exposed individuals, i.e., at around 0.1 µM (5, 11, 28). The metabolism of AFB1 was assessed at concentrations measurable with HPLC and adjusted for concentrations measured in exposed individuals using a hepatic abundance model. This model also takes into consideration the specific activities of the individual P450s and their hepatic expression levels. The AFB1 metabolite spectra and, where available, the kinetic parameters were in agreement with previous reports (5, 12, 20, 21). AFBO was the exclusive AFB1 product catalyzed by the recombinant P450 3A5, but it was also catalyzed by P450 3A4, P450 3A7, and P450 1A2. Although our HPLC assay does not separate AFB1-8,9-exo- and -endoepoxide stereoisomers, it is a fair measure of the AFB1 carcinogenic activation, since the less toxic stereoisomer AFB1endo-8,9-epoxide is only produced by P450 1A2, in marked contrast to P450 3A isoforms. These produce predominantly or exclusively AFB1-exo-8,9-epoxide (3). AFQ1, the major detoxification product, was catalyzed by P450 3A4 and P450 3A7, whereas AFM1 was specific for P450 1A2. The extent of P450 3A4 variability in HLMs was smaller than in some previous studies, but it was in the same range as recently reported by Sim et al. (22). It has been suggested that the variability in the expression of the P450 3A4 may have been overestimated in earlier studies, due to tissue collection artifacts (29). The expression of P450 3A5 was bimodal, as reported previously (16, 17). To allow for a first estimate of the contribution of P450 3A7 to AFB1 metabolism, we used P450 3A7 expression levels measured in human livers with a recently developed, specific P450 3A7 antibody (22). For the epidemiologically relevant AFB1 concentration of 0.1 µM, our data predict a substantial interindividual variability of the toxin’s disposition. The share of AFB1 converted to the carcinogenic AFBO varies between 5 and 37% even in this relatively small set of HLMs (Figure 7). Correspondingly, the rate between AFQ1, the primary detoxification product of AFB1, and AFBO widely varies (from 19:1 to 1.7:1). The hepatic abundance model predicts that a majority of both AFBO and AFQ1 is catalyzed by P450 3A4. This is supported by the high correlation coefficients between the HLM expression levels of P450 3A4 and Vmax values of, respectively, AFBO (Figure 5A) and AFQ1 (Figure 4A) production. It is also in agreement with the high correlation between Vmax values of AFBO and AFQ1 formations measured in HLMs (Figure 3), when taking into account the similarity of AFBO and AFQ1 kinetics in recombinant P450 3A4 and HLMs. Although recombinant P450 3A7 also generates AFQ1, its extent is negligible in most individuals in comparison to P450 3A4, since they only express traces of P450 3A7 (22). In P450 3A7 high expressors, which account for approximately 15% of Caucasians (22), this P450 may contribute on average almost a quarter of AFQ1 production. This estimate is based on P450 3A7 accounting for almost a quarter of the total P450 3A in P450 3A7 high expressors (22), combined with similar specific AFB1 clearances of P450 3A4 and P450 3A7 toward AFQ1. The frequency of the P450 3A7 polymorphism in populations

584 Chem. Res. Toxicol., Vol. 19, No. 4, 2006

Kamdem et al.

Figure 7. Hepatic metabolism of AFB1 in P450 3A5 or P450 3A7 high expressors at a concentration of 0.1 µM. Data calculated using a hepatic abundance model for the in vivo relevant AFB1 concentration of 0.1 µM using kinetics and protein expression data from Tables 1 and 2. Both the detoxification to AFQ1 and the oxidation to the genotoxic AFBO are catalyzed predominantly by P450 3A4. P450 3A5 contributes between 4 and 15% of AFBO formation, depending on the concomitant expression of P450 3A4. The contribution of P450 3A7 to AFQ1 and AFBO formation in P450 3A7 high expressers (in italics) was calculated using hepatic P450 3A7 expression levels reported by Sim et al. (22).

exposed to high AFB1 concentrations, such as Asians and Africans, is currently unknown. A dominant role of P450 3A4 in AFBO production is in agreement with the findings of Doi and colleagues, who detected a correlation approaching statistical significance between P450 3A4 expression and AFB1-DNA adduct levels in a panel of adult human livers (11). In addition to P450 3A4, AFBO production is contributed to by P450 3A5, P450 3A7, and P450 1A2, which may affect the share of AFB1 converted to AFBO. The contribution of P450 3A7 in P450 3A7 high expressers (22) is estimated to be between 5 and 7%, whereas that of P450 1A2 is between 1 and 5%. P450 3A5 may contribute up to 15.3% of AFBO in P450 3A5 high expressors. This value is still relatively modest in comparison with P450 3A4 (79-95%, Figure 7), and it is consistent with the low value of the correlation coefficient between the P450 3A5 protein expression level and Vmax of AFBO production in the HLMs investigated (Figure 5B). On the other hand, we found a significant, inverse correlation between the contribution of P450 3A5 to the hepatic AFBO production and the expression of the P450 3A4 protein (Figure 6). This means that the relative contribution of P450 3A5 to AFBO production is highly dependent on the concomitant expression of P450 3A4. This is in agreement with the recent observation of AFB1-albumin adducts blood plasma levels

increased especially strongly in AFB1-exposed P450 3A5 high expressors with low concomitant P450 3A4 expression (18). Similar relationships have been reported for several other substrates common to P450 3A4 and P450 3A5, with the effect of the P450 3A5 polymorphism especially pronounced against a background of low P450 3A4 expression (30-32). In conclusion, these findings indicate that following P450 3A4, P450 3A5 expression status is the second-most important determinant of the carcinogenic AFB1 activation to AFBO. The clinical importance of the P450 3A5 effect on AFB1 metabolism will be ultimately determined by genotyping HCC patients from areas of high AFB1 exposure. In our microsomal incubations, we detected no AFM1, which is an AFB1 metabolite specific for P450 1A2. Usually considered a detoxification product, AFM1 can be converted, e.g., by cultured liver cells, to the genotoxic AFM-8,9-epoxide (3335). Such a conversion could be responsible for the absence of AFM1 from our incubations of HLMs with AFB1. Alternatively, the production of AFM1 may have been simply too low to be detected. In agreement with this latter interpretation, the production of AFM1 by HLMs was a magnitude smaller than that of AFQ1 in the available reports on hepatic AFM1 production (4, 36). Likewise, median AFM1 concentrations in urine and feces of AFB1-exposed Chinese individuals accounted,

Cytochrome P450 and AFB1 Metabolism

respectively, for 1/60 and 1/260 of the corresponding AFQ1 values (37). In the same study, urinary excretion of AFQ1 (but not that of AFM1) correlated significantly with the excretion of the AFB1-N7-guanine adduct, which is considered to be the best predictor of HCC risk (38-40). Furthermore, Doi and colleagues (11) detected no correlation between P450 1A2 expression level and AFB1-DNA adducts in the same panel of microsomes from 10 human livers in which they reported a correlation approaching statistical significance between these adducts and P450 3A4 expression. In summary, all of these data suggest that the disposition of AFB1 to AFM1 in HLMs may be negligible. Similarly negligible may be the contribution of P450 1A2 to the formation of AFBO. Our conclusion that P450 3A4 is responsible for the majority of AFBO formation in human livers is in agreement with expectations from several studies with purified or recombinant P450 (8, 9, 41-45). On the other hand, it is in opposition to the only work that we are aware of, which assessed AFB1 metabolism in microsomes from human livers (5). These authors predicted that at AFB1 concentrations encountered in vivo, over 99% of the hepatic AFBO production would be contributed by P450 1A2. However, the critically important hepatic expression levels of P450 1A2 and P450 3A4 were not considered in the calculation. The postulated dominant role of P450 1A2 in AFBO production was also not supported by the Vmax of AFBO production, which was highest in the liver with the highest P450 3A4 expression and not in the liver with the highest P450 1A2 expression, as would be expected (Table 2 in ref 5). Furthermore, only 20-46% of AFBO production was inhibited by the specific P450 1A2 inhibitor furafylline (compare Tables 1 and 2 in ref 5). Could our findings have any immediate implications for public health decisions in areas with a high prevalence of AFB1associated HCC? Preventive measures, such as vaccination or chemoprotection, are expected to reduce the burden of HCC, especially effectively if they are applied to individuals at particularly high risk. Because all AFB1 metabolizing enzymes are expressed in the liver at individually variable levels, targeted chemoprotection by P450 inhibition represents a valid hypothesis (46). Our results argue against targeting any preventive measures specifically to P450 3A5, P450 3A7, or P450 1A2 high expressors. Considering the prominent contribution of P450 3A4 to AFBO formation, the inhibition of P450 3A4 in individuals with particularly high hepatic levels of this P450 should be much more effective. However, this would require genetic or phenotypic P450 3A4 activity markers, the development of which has proven difficult. More practical would be the reduction of P450 3A4 expression by avoiding pharmacologic and dietary P450 3A4 inducers. These substances can sometimes enhance the turnover of P450 3A4 drug substrates to an extent where their effectiveness is abolished (47) and they likely increase AFBO levels in individuals exposed to AFB1. Preventing P450 3A4 induction would have the added benefit of a similar effect on P450 3A5 and P450 3A7 in the respective high expressors of these P450. Indeed, all three P450 3A genes exhibit overlaps in the spectra of inducers, due to similarities in their transcription regulation (48, 49). Acknowledgment. This work was supported by the Deutsche Forschungsgemeinschaft (DFG) Grant WO505/2-1. We thank Professor Christopher P. Wild (Leeds University, United Kingdom) for comments on the manuscript.

References (1) Hall, A. J., and Wild, C. P. (2003) Liver cancer in low and middle income countries. BMJ 326, 994-995.

Chem. Res. Toxicol., Vol. 19, No. 4, 2006 585 (2) Wild, C. P., and Hall, A. J. (2000) Primary prevention of hepatocellular carcinoma in developing countries. Mutat. Res. 462, 381-393. (3) Wild, C. P., and Turner, P. C. (2002) The toxicology of aflatoxins as a basis for public health decisions. Mutagenesis 17, 471-481. (4) Ramsdell, H. S., Parkinson, A., Eddy, A. C., and Eaton, D. L. (1991) Bioactivation of aflatoxin B1 by human liver microsomes: Role of cytochrome P450 IIIA enzymes. Toxicol. Appl. Pharmacol. 108, 436447. (5) Gallagher, E. P., Kunze, K. L., Stapleton, P. L., and Eaton, D. L. (1996) The kinetics of aflatoxin B1 oxidation by human cDNAexpressed and human liver microsomal cytochromes P450 1A2 and 3A4. Toxicol. Appl. Pharmacol. 141, 595-606. (6) Gallagher, E. P., Wienkers, L. C., Stapleton, P. L., Kunze, K. L., and Eaton, D. L. (1994) Role of human microsomal and human complementary DNA-expressed cytochromes P4501A2 and P4503A4 in the bioactivation of aflatoxin B1. Cancer. Res. 54, 101-108. (7) 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. U.S.A. 87, 4790-4793. (8) Ueng, Y. F., Shimada, T., Yamazaki, H., and Guengerich, F. P. (1995) Oxidation of aflatoxin B1 by bacterial recombinant human cytochrome P450 enzymes. Chem. Res. Toxicol. 8, 218-225. (9) Forrester, L. M., Neal, G. E., Judah, D. J., Glancey, M. J., and Wolf, C. R. (1990) Evidence for involvement of multiple forms of cytochrome P-450 in aflatoxin B1 metabolism in human liver. Proc. Natl. Acad. Sci. U.S.A. 87, 8306-8310. (10) Van Vleet, T. R., Klein, P. J., and Coulombe, R. A., Jr. (2002) Metabolism and cytotoxicity of aflatoxin b1 in cytochrome p-450expressing human lung cells. J. Toxicol. EnViron. Health A 65, 853867. (11) Doi, A. M., Patterson, P. E., and Gallagher, E. P. (2002) Variability in aflatoxin B(1)-macromolecular binding and relationship to biotransformation enzyme expression in human prenatal and adult liver. Toxicol. Appl. Pharmacol. 181, 48-59. (12) Wang, H., Dick, R., Yin, H., Licad-Coles, E., Kroetz, D. L., Szklarz, G., Harlow, G., Halpert, J. R., and Correia, M. A. (1998) Structurefunction relationships of human liver cytochromes P450 3A: Aflatoxin B1 metabolism as a probe. Biochemistry 37, 12536-12545. (13) Xue, L., Wang, H. F., Wang, Q., Szklarz, G. D., Domanski, T. L., Halpert, J. R., and Correia, M. A. (2001) Influence of P450 3A4 SRS-2 residues on cooperativity and/or regioselectivity of aflatoxin B(1) oxidation. Chem. Res. Toxicol. 14, 483-491. (14) Wrighton, S. A., Brian, W. R., Sari, M. A., Iwasaki, M., Guengerich, F. P., Raucy, J. L., Molowa, D. T., and Vandenbranden, M. (1990) Studies on the expression and metabolic capabilities of human liver cytochrome P450IIIA5 (HLp3). Mol. Pharmacol. 38, 207-213. (15) Wrighton, S. A., Ring, B. J., Watkins, P. B., and VandenBranden, M. (1989) Identification of a polymorphically expressed member of the human cytochrome P-450III family. Mol. Pharmacol. 36, 97-105. (16) Kuehl, P., Zhang, J., Lin, Y., Lamba, J., Assem, M., Schuetz, J., Watkins, P. B., Daly, A., Wrighton, S. A., Hall, S. D., Maurel, P., Relling, M., Brimer, C., Yasuda, K., Venkataramanan, R., Strom, S., Thummel, K., Boguski, M. S., and Schuetz, E. (2001) Sequence diversity in CYP3A promoters and characterization of the genetic basis of polymorphic CYP3A5 expression. Nat. Genet. 27, 383-391. (17) Hustert, E., Haberl, M., Burk, O., Wolbold, R., He, Y. Q., Klein, K., Nuessler, A. C., Neuhaus, P., Klattig, J., Eiselt, R., Koch, I., Zibat, A., Brockmoller, J., Halpert, J. R., Zanger, U. M., and Wojnowski, L. (2001) The genetic determinants of the CYP3A5 polymorphism. Pharmacogenetics 11, 773-779. (18) Wojnowski, L., Turner, P. C., Pedersen, B., Hustert, E., Brockmoller, J., Mendy, M., Whittle, H. C., Kirk, G., and Wild, C. P. (2004) Increased levels of aflatoxin-albumin adducts are associated with CYP3A5 polymorphisms in The Gambia, West Africa. Pharmacogenetics 14, 691-700. (19) Tateishi, T., Watanabe, M., Moriya, H., Yamaguchi, S., Sato, T., and Kobayashi, S. (1999) No ethnic difference between Caucasian and Japanese hepatic samples in the expression frequency of CYP3A5 and CYP3A7 proteins. Biochem. Pharmacol. 57, 935-939. (20) Gillam, E. M., Guo, Z., Ueng, Y. F., Yamazaki, H., Cock, I., Reilly, P. E., Hooper, W. D., and Guengerich, F. P. (1995) Expression of cytochrome P450 3A5 in Escherichia coli: Effects of 5′ modification, purification, spectral characterization, reconstitution conditions, and catalytic activities. Arch. Biochem. Biophys. 317, 374-384 [Gillam, E. M., Guo, Z., Ueng, Y. F., Yamazaki, H., Cock, I., Reilly, P. E., Hooper, W. D., and Guengerich, F. P. (1995) Published erratum. Arch. Biochem. Biophys. 318 (2), 498]. (21) Ueng, Y. F., Kuwabara, T., Chun, Y. J., and Guengerich, F. P. (1997) Cooperativity in oxidations catalyzed by cytochrome P450 3A4. Biochemistry 36, 370-381. (22) Sim, S. C., Edwards, R. J., Boobis, A. R., and Ingelman-Sundberg, M. (2005) CYP3A7 protein expression is high in a fraction of adult

586 Chem. Res. Toxicol., Vol. 19, No. 4, 2006

(23) (24) (25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

(33) (34) (35)

(36)

human livers and partially associated with the CYP3A7*1C allele. Pharmacogenet. Genomics 15, 625-631. Schweikl, H., Taylor, J. A., Kitareewan, S., Linko, P., Nagorney, D., and Goldstein, J. A. (1993) Expression of CYP1A1 and CYP1A2 genes in human liver. Pharmacogenetics 3, 239-249. Lamba, J. K., Lin, Y. S., Schuetz, E. G., and Thummel, K. E. (2002) Genetic contribution to variable human CYP3A-mediated metabolism. AdV. Drug DeliVery ReV. 54, 1271-1294. Evert, B., Eichelbaum, M., Haubruck, H., and Zanger, U. M. (1997) Functional properties of CYP2D6 1 (wild-type) and CYP2D6 7 (His324Pro) expressed by recombinant baculovirus in insect cells. Naunyn-Schmiedebergs Arch. Pharmacol. 355, 309-318. Venkatakrishnan, K., von Moltke, L. L., Court, M. H., Harmatz, J. S., Crespi, C. L., and Greenblatt, D. J. (2000) Comparison between cytochrome P450 (CYP) content and relative activity approaches to scaling from cDNA-expressed CYPs to human liver microsomes: Ratios of accessory proteins as sources of discrepancies between the approaches. Drug Metab. Dispos. 28, 1493-1504. Yuan, R., Madani, S., Wei, X. X., Reynolds, K., and Huang, S. M. (2002) Evaluation of cytochrome P450 probe substrates commonly used by the pharmaceutical industry to study in vitro drug interactions. Drug Metab. Dispos. 30, 1311-1319. Groopman, J. D., Hall, A. J., Whittle, H., Hudson, G. J., Wogan, G. N., Montesano, R., and Wild, C. P. (1992) Molecular dosimetry of aflatoxin-N7-guanine in human urine obtained in The Gambia, West Africa. Cancer Epidemiol. Biomarkers PreV. 1, 221-227. Floyd, M. D., Gervasini, G., Masica, A. L., Mayo, G., George, A. L., Jr., Bhat, K., Kim, R. B., and Wilkinson, G. R. (2003) Genotypephenotype associations for common CYP3A4 and CYP3A5 variants in the basal and induced metabolism of midazolam in European- and African-American men and women. Pharmacogenetics 13, 595-606. Kamdem, L. K., Streit, F., Zanger, U. M., Brockmoller, J., Oellerich, M., Armstrong, V. W., and Wojnowski, L. (2005) Contribution of CYP3A5 to the in vitro hepatic clearance of tacrolimus. Clin. Chem. 51, 1374-1381. Huang, W., Lin, Y. S., McConn, D. J., 2nd, Calamia, J. C., Totah, R. A., Isoherranen, N., Glodowski, M., and Thummel, K. E. (2004) Evidence of significant contribution from CYP3A5 to hepatic drug metabolism. Drug Metab. Dispos. 32, 1434-1445. Kamdem, L. K., Meineke, I., Koch, I., Zanger, U. M., Brockmoller, J., and Wojnowski, L. (2004) Limited contribution of CYP3A5 to the hepatic 6beta-hydroxylation of testosterone. Naunyn-Schmiedebergs Arch. Pharmacol. 370, 71-77. Neal, G. E., Eaton, D. L., Judah, D. J., and Verma, A. (1998) Metabolism and toxicity of aflatoxins M1 and B1 in human-derived in vitro systems. Toxicol. Appl. Pharmacol. 151, 152-158. Bujons, J., Hsieh, D. P., Kado, N. Y., and Messeguer, A. (1995) Aflatoxin M1 8,9-epoxide: Preparation and mutagenic activity. Chem. Res. Toxicol. 8, 328-332. Loveland, P. M., Wilcox, J. S., Hendricks, J. D., and Bailey, G. S. (1988) Comparative metabolism and DNA binding of aflatoxin B1, aflatoxin M1, aflatoxicol and aflatoxicol-M1 in hepatocytes from rainbow trout (Salmo gairdneri). Carcinogenesis 9, 441-446. Ramsdell, H. S., and Eaton, D. L. (1990) Species susceptibility to aflatoxin B1 carcinogenesis: Comparative kinetics of microsomal biotransformation. Cancer Res. 50, 615-620.

Kamdem et al. (37) Mykkanen, H., Zhu, H., Salminen, E., Juvonen, R. O., Ling, W., Ma, J., Polychronaki, N., Kemilainen, H., Mykkanen, O., Salminen, S., and El-Nezami, H. (2005) Fecal and urinary excretion of aflatoxin B1 metabolites (AFQ1, AFM1 and AFB-N7-guanine) in young Chinese males. Int. J. Cancer 115, 879-884. (38) Groopman, J. D., Wang, J. S., and Scholl, P. (1996) Molecular biomarkers for aflatoxins: From adducts to gene mutations to human liver cancer. Can. J. Physiol. Pharmacol 74, 203-209. (39) Groopman, J. D., Wild, C. P., Hasler, J., Junshi, C., Wogan, G. N., and Kensler, T. W. (1993) Molecular epidemiology of aflatoxin exposures: Validation of aflatoxin-N7-guanine levels in urine as a biomarker in experimental rat models and humans. EnViron. Health Perspect. 99, 107-113. (40) Qian, G. S., Ross, R. K., Yu, M. C., Yuan, J. M., Gao, Y. T., Henderson, B. E., Wogan, G. N., and Groopman, J. D. (1994) A follow-up study of urinary markers of aflatoxin exposure and liver cancer risk in Shanghai, People’s Republic of China. Cancer Epidemiol. Biomarkers PreV. 3, 3-10. (41) Guengerich, F. P., and Kim, D. H. (1990) In vitro inhibition of dihydropyridine oxidation and aflatoxin B1 activation in human liver microsomes by naringenin and other flavonoids. Carcinogenesis 11, 2275-2279. (42) Raney, K. D., Coles, B., Guengerich, F. P., and Harris, T. M. (1992) The endo-8,9-epoxide of aflatoxin B1: A new metabolite. Chem. Res. Toxicol. 5, 333-335. (43) Raney, K. D., Meyer, D. J., Ketterer, B., Harris, T. M., and Guengerich, F. P. (1992) Glutathione conjugation of aflatoxin B1 exo- and endoepoxides by rat and human glutathione S-transferases. Chem. Res. Toxicol. 5, 470-478. (44) Raney, K. D., Shimada, T., Kim, D. H., Groopman, J. D., Harris, T. M., and Guengerich, F. P. (1992) Oxidation of aflatoxins and sterigmatocystin by human liver microsomes: significance of aflatoxin Q1 as a detoxication product of aflatoxin B1. Chem. Res. Toxicol. 5, 202-210. (45) Shimada, T., and Guengerich, 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. U.S.A. 86, 462-465. (46) Kensler, T. W., Qian, G. S., Chen, J. G., and Groopman, J. D. (2003) Translational strategies for cancer prevention in liver. Nat. ReV. Cancer 3, 321-329. (47) Lucey, M. R., Kolars, J. C., Merion, R. M., Campbell, D. A., Aldrich, M., and Watkins, P. B. (1990) Cyclosporin toxicity at therapeutic blood levels and cytochrome P-450 IIIA. Lancet 335, 11-15. (48) Burk, O., Koch, I., Raucy, J., Hustert, E., Eichelbaum, M., Brockmoller, J., Zanger, U. M., and Wojnowski, L. (2004) The induction of CYP3A5 in human liver and intestine is mediated by the xenobiotic sensors PXR and CAR. J. Biol. Chem. 279, 38379-38385. (49) Burk, O., Tegude, H., Koch, I., Hustert, E., Wolbold, R., Glaeser, H., Klein, K., Fromm, M. F., Nuessler, A. K., Neuhaus, P., Zanger, U. M., Eichelbaum, M., and Wojnowski, L. (2002) Molecular mechanisms of polymorphic CYP3A7 expression in adult human liver and intestine. J. Biol. Chem. 277, 24280-24288.

TX050358E