Biphasic effects of the flavonoids quercetin and naringenin on the metabolic activation of 2-amino-3,5-dimethylimidazo[4,5-f]quinoline by Salmonella typhimurium TA1538 co-expressing human cytochrome P450 1A2, NADPH-cytochrome P450 reductase, and cytochrome b5

Mutation Research 545 (2004) 37–47

Biphasic effects of the flavonoids quercetin and naringenin on the metabolic activation of 2-amino-3,5-dimethylimidazo[4,5-f]quinoline by Salmonella typhimurium TA1538 co-expressing human cytochrome P450 1A2, NADPH-cytochrome P450 reductase, and cytochrome b5 Il-Hyun Kang a,b , Hyun-Jung Kim a , Hyeyoung Oh b , Young In Park a , Mi-Sook Dong a,∗ a

School of Life Sciences and Biotechnology, Korea University, Seoul 136-701, South Korea b Korea Food and Drug Administration, Seoul, South Korea Received 15 May 2003; received in revised form 25 July 2003; accepted 11 August 2003

Abstract Heterocyclic amines (HCAs) produced by cooking meat products at high temperatures are promutagens that are activated by cytochrome P450 (CYP) lA2. Using a newly developed Salmonella typhimurium TA1538/1A2bc-b5 strain, we tested the effect of quercetin and naringenin on the mutagenicity of 2-amino-3,4-dimethylimidazo[4,5-f]quinoline (MeIQ). TA1538/1A2bc-b5 bears two plasmids, one expressing human CYP1A2 and NADPH-P450 reductase (NPR), and the other plasmid which expresses human cytochrome b5 (cyp b5). TA1538/1A2bc-b5 cells showed high activities of 7-ethoxyresorufin O-deethylase (EROD) and methoxyresorufin O-demethylase (MROD) associated with CYP1A2 and are very sensitive to mutagenesis induced by several HCAs. MeIQ was found to be the strongest mutagen among the HCAs tested in this system. Mutagenicity of MeIQ was enhanced 50 and 42% by quercetin at 0.1 and 1 ␮M, respectively, but suppressed 82 and 96% at 50 and 100 ␮M. Naringenin also increased the MeIQ-induced mutation about 37 and 22% at 0.1 and 1 ␮M, but suppressed it 32 and 63% at 50 and 100 ␮M concentrations, respectively, in TA 1538/1A2bc-b5 cells. Thus, they stimulated the MeIQ induced mutation at low concentrations, but strongly suppressed it at high concentrations. This biphasic effect of flavonoids was due to the stimulation or the inhibition of CYP1A2 activity in a dose-dependent manner judging by the activities of EROD or MROD in the Salmonella cells. These results indicate that quercetin and naringenin can exhibit inhibitory or stimulating effects on CYP1A2 mediated mutagenesis by MeIQ, depending on their concentrations. © 2003 Elsevier B.V. All rights reserved. Keywords: Human cytochrome P450; NADPH-cytochrome P450 reductase; Cytochrome b5 ; Salmonella typhimurium TA1538; 2-Amino-3,4-dimethylimidazo[4,5-f]quinoline (MeIQ); Quercetin; Naringenin

∗ Corresponding author. Tel.: +82-2-3290-4146; fax: +82-3290-3951. E-mail address: [email protected] (M.-S. Dong).

0027-5107/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.mrfmmm.2003.08.002

38

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47

1. INTRODUCTION A substantial proportion of human cancer is now believed to be associated with diet [1]. A number of heterocyclic amines (HCAs) have been identified in cooked meat and fish as potent mutagens in the Ames/Salmonella assay [2,3]. Several of these compounds have been classified as possible or suspected human carcinogens by the IARC (1993) on the basis of the results of long-term animal feeding studies [4–6]. HCAs undergo bioactivation catalyzed mainly by the cytochrome P450 (CYP) [7], but also by prostaglandin synthase [8], and NADPH-cytochrome P450 reductase (NPR) [9]. However, the major route of bioactivation involves N-hydroxylation by the hepatic CYP1A isoforms, particularly CYP1A2. Thus, the modification of the liver CYP1A2 can alter the production in vivo of genotoxic or carcinogenic metabolites of HCAs [10,11]. Flavonoids are naturally occurring polyphenolic compounds that are widely distributed in fruits, vegetables, whole grains and beverages such as red wine and teas [12]. Moreover, extracts of many flavonoids are now available in health food stores. They attract special attention because they are consumed daily in considerable amounts and exhibit a wide variety of health-protective properties. Epidemiological studies and laboratory studies have indicated that the frequent intakes vegetables and fruits that contain sufficient amounts of flavonoids are associated with a lower incidence of carcinogen-induced tumors in a variety of organs [13,14]. The protective effects of flavonoids can be mediated though wide variety of mechanisms. One possible mechanism responsible for chemopreventive activities of flavonoids is their ability to modulate xenobiotic metabolism [10,11,15–17]. Modulation of CYP activities by flavonoids has drawn great concern due to inter-individual differences in susceptibility to certain chemical carcinogens [18]. The activity of the specific form(s) of CYP450 is likely to be a major determinant of susceptibility to chemically induced carcinogenesis between which varies among between individuals due to different dietary habits as well as genetic characteristics. Quercetin and naringenin (Fig. 1) are representative flavonoids that not only exert antiestrogenic, cholesterol-lowering and antioxidant activities but also can modulate the metabolism of many xenobiotics [19,20]. These flavonoids are in-

Fig. 1. Chemical structures of naringenin and quercetin.

deed potent inhibitors of cytochrome P450-dependent microsomal metabolism [15–17]. People consume cooked meat or fish together with various vegetables containing substantial amounts of quercetin and naringenin that can modify the enzyme activity of CYP1A2 to stimulate or to inhibit the mutagenic activities of HCAs. Most of the previous studies evaluating chemopreventive potential of flavonoids or vegetables were undertaken with relatively high doses sufficiently to inhibit the enzyme activity [15–17]. In this study, we established a bacterial mutagenicity tester strain, Salmonella strain, TA1538/1A2bc-b5 expressing human CYP1A2, NADPH-cytochrome P450 reductase, and cytochrome b5 (cyp b5). Using this strain, we examined the effect of varying doses of quercetin and naringenin on the mutagenicity of HCAs.

2. Materials and methods 2.1. Chemicals 2-Amino-3,4-dimethylimidazo[4,5-f]quinoline (MeIQ), 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx), 2-amino-1-methyl-6-phenylimidazo [4,5b]pyridine (PhIP) were purchased from Wako Pure Chemical Industries (Osaka, Japan), glucose 6-phosphate (G-6-P), G-6-P dehydrogenase, NADP+ , ␦-aminolevulinic acid, ␣-naphthoflavone, ampicillin, naringenin and quercetin were supplied by Sigma–Aldrich

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47

(St. Louis, MO, USA). Bactotryptone, bactopeptone, yeast extract, nutrient broth and bactoagar were obtained from Difco Lab. (Detroit, MI, USA). Other chemicals used were of the highest grade commercially available. 2.2. Salmonella typhimurium tester strains and plasmids The plasmid pCYP1A2bc carrying human CYP1A2 and hNPR cDNA was kindly provided by Dr. F.P. Guengerich of Vanderbilt University, USA (Fig. 2A) [21]. Construction of the Escherichia coli expression vector pCypb5 will be described elsewhere (Fig. 2B; [22], manuscript preparation). Expression plasmids pCYP1A2bc carrying human CYP1A2 and hNPR cDNA and pCypb5 was first modified by introduction in Salmonella typhimurium strain LB5000 (R− M+ ) (provided from the Salmonella Genetic Stock Center, University of Calgary, AB, Canada), and then the plasmids were introduced into TA1538 by the elec-

39

trophoretic method. Salmonella typhimurium tester strains and plasmids used in this study are summarized in Table 1. These strains were tested for histidine auxotrophy, uvrB deletion, rfa mutation, and ampicillin resistance, and they exhibited the same characteristics as S. typhimurium TA1538, except for the ampicillin and chloramphenicol resistance acquired as a result of the pCW and pACYC184 plasmids. 2.3. Expression of human CYP1A2, NPR and cytochrome b5 in S. typhimurium TA1538 A single colony of bacteria on a plate was picked up and inoculated into 10 ml of Luria–Bertani medium supplemented with ampicillin (100 mg/ml) and chloramphenicol (40 mg/ml). The bacteria were grown with shaking at 37 ◦ C for 12 h. One ml of culture was inoculated into 100 ml of modified Terrific Broth [23], and the bacteria were grown in a shaking chamber at 25 ◦ C for 6 h prior to induction with 1 mM IPTG. The expression of recombinant proteins was achieved by incubation at 25 ◦ C for additional 18 h with shaking. Table 1 Salmonella typhimurium strains and plasmids established in the present studya Strains/plasmids Plasmids pCW p1A2bc pACYC184 pCyp b5 Strains LB5000

Fig. 2. Plasmid Structure of (A) pCYP1A2bc and (B) pCyp b5 used to co-express CYP1A2, hNPR and cytochrome b5 in S. typhimurium TA1538. The 8.6 kb of bicistronic operons of pCYP1A2bc plasmid consists of cDNAs for CYP1A2 as the first cistron and NPR as the second. A 4.7 kb of expression plasmid pCypb5 contains cDNA of human cytochrome b5 in pACYC184 containing trc promoter.

TA1538 TA1538/pCWpACYC184 TA1538/pCW-b5 TA1538/1A2bc-b5

Characteristics CYP cDNA expression vector, pBR322 derivative, Ampr pCW derivative bearing human CYP1A2 and NPR, Ampr Use for human cyp b5 expression, Cmr , Tcr pACYC184 derivative bearing human cyp b5, Cmr metA22, metE551, trpC2, ilv-452 H1-b2-e,n,x, fla-66, rpsL120, xyl-404, leu, hsdL6 hsdSA29, hsdSB hisD3052, uvrB, rfa TA1538 harboring pCW and pACYC184; Ampr , Cmr , Tcr TA1538 harboring pCW and pCyp b5; Ampr , Cmr TA 1538 harboring pCYP1A2bc and pCyp b5; Ampr , Cmr

a Ampr : ampicillin resistance gene , Cmr : chloramphenicol resistance gene, bc: bicistronic system, Cyp b5: cytochrome b5 .

40

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47

2.4. Determination of catalytic activities Bacterial membrane and cytosol fractions were prepared as reported by Sandhu et al. [23]. The content of CYP in the membrane fractions was determined by the method of Omura and Sato [24] and verified by the immunoblot with rabbit antihuman CYP1A2 serum. The NADPH-driven cytochrome c reduction rate was measured according to the method described by Yasukochi and Masters [25] using an extinction coefficient of 21 mM−1 cm−1 . A specific activity of 3200 nmol of reduced cytochrome c per nmol reductase was used for calculating the concentration of hNPR. 7-Ethoxyresorufin O-deethylase (EROD) and methoxyresorufin O-demethylase (MROD) activities of whole cells and the membrane fraction were assayed essentially as described by Dong et al [26]. To assess the enzyme activity of whole cells, one ml of expressing culture was centrifuged, and the cell pellet was washed twice with cold potassium phosphate buffered saline (pH 7.4). The incubation mixture contained 100 mM potassium phosphate buffer (pH 7.4), cells from 1 ml of culture expressed enzymes, 5 ␮M 7-ethoxyresorufin or 7-methoxyresorufin in a final volume of 0.5 ml. To determine the enzyme activity of membrane fraction, the reaction mixture contained 100 mM potassium phosphate buffer (pH 7.4), 25 pmol CYP, 0.2 mM NADPH and 5 ␮M 7-ethoxyresorufin or 7-methoxyresorufin in a final volume of 0.5 ml. Incubation was carried out at 37 ◦ C for 30 min. 2.5. Mutagenicity assay The number of cells was adjusted to give 1 × 109 to 2 × 109 cells/ml by dilution with modified Terrific Broth. The assay was carried out as described by Maron and Ames [27] with minor modifications. Briefly, a mixture containing 0.5 ml of potassium phosphate buffer (0.1 M, pH 7.4) or 0.5 ml of rat S9 mix, 0.1 ml of diluted culture (0.1 ml) and a mutagen dissolved in DMSO (or DMSO alone for negative controls) were incubated at 37 ◦ C for 30 min with shaking. The volume of DMSO used was less than 10 ␮l per 0.6 ml. Top agar (2 ml) was added, and the mixtures were poured onto minimal lactose plates and then incubated at 37 ◦ C for 48 h. Experiments to test the effects of flavonoids were performed as above, except

that the buffer, cells and flavonoids were incubated at 37 ◦ C for 10 min before the addition of the mutagen. For the S9-dependent assays, rat liver S9 mixture (Aroclor1254-induced male Sprague–Dawley rat) was prepared as for the Ames test [27] at a concentration of 1 mg of S9 protein/plate (0.5 ml of S9 mix). Assays were carried out at least twice, with triplicate plates at each dose. 2.6. Statistical analysis Comparison of data was carried out according to Student’s t-test and the Wilcoxon test.

3. Results 3.1. Expression of CYP1A2 and NADPH-P450 reductase with cytochrome b5 in S. typhimurium TA1538 The expression of CYP1A2 protein in S. typhimurium TA1538 was determined in membranes by the CO difference spectra and immunoblotting (Fig. 3). The level of CYP1A2 and the NADPHcytochrome c reductase activity in membranes were about 0.15 nmol/mg protein and 6.82 nmol cytochrome c reduced/mg protein, respectively (Table 2). The membranes from TA1538/CYP1A2bc-b5 strain showed high EROD and MROD activities, whereas the EROD or MROD activity in membranes of TA1538/pCW-pACYC184 was not detectable. The catalytic activities of CYP1A2 expressed in TA1538 cells were 50.4 ± 7.8 and 68.6 ± 1.5 pmol/min/ml culture for EROD and MROD, respectively (Table 2). 3.2. Effect of expression time on the induction of revertants To obtain the maximal metabolic activity of CYP1A2, we examined the effect of expression time on the MeIQ-induced mutagenesis. TA1538/ CYP1A2bc-b5 cells were expressed at 25 ◦ C for 5, 10, 14, and 23 h. As shown in Fig. 4, the numbers of revertant formation by MeIQ increased, dependent on expression time, which reached the highest level at 23 h. Therefore, we selected this time point for subsequent experiments.

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47

41

Fig. 3. Expression of CYP1A2 (A) and NPR (B) in the S. typhimurium TA1538/CYP1A2bc-b5 (I) and TA1538/pCW-b5 (II). TA1538/CYP1A2bc-b5 was co-transfected pCYP1A2bc and pCyp b5 and TA1538/pCW-b5 pCW and pCyp b5. The enzymes were expressed for 23 h and the membranes of the Salmonella strains were prepared by the methods described in materials and methods. The activities of CYP (A) and NPR (B) were measured by Fe2+ –CO vs. Fe3+ difference spectra and the reduction rate of cytochrome c as described in materials and methods, respectively. The results are expressed mean ±S.D. (n = 3). (C) Western blot analysis of CYP1A2 in the membranes fraction prepared from salmonella strains. Lane 1, TA1538/CYP1A2bc-b5; Lane 2, TA1538/pCW-b5; and Lane 3, purified CYP1A2.

duction of His+ revertants in the TA1538/CYP1A2bcb5 strain by MeIQ was compared with that in the parental TA1538 strain in the presence S9 fraction, TA1538/CYP1A2bc-pACYC184, TA1538/pCW-b5,

3.3. Mutagenicity assay with TA1538/1A2bc-b5 For the evaluation of the sensitivity of the genetically engineered Salmonella strains to HCAs, the in-

Table 2 CYP1A2 contents and catalytic activities of CYP1A2 and hNPR in membrane fractions and whole cells of Salmonella strains

TA1538/pCWpACYC184 TA1538/1A2bc-b5

P450 (pmol/mg protein)

hNPR activity (nmol cyt c reduced/mg protein)

EROD activity (nmol product/min/ nmol of P450)

MROD activity (nmol product/min/ nmol of P450)

EROD activity (pmol product/ min/ml culture)

MROD activity (pmol product/ min/ml culture)

0

0.03 ± 0.01

<0.02

<0.02

<0.02

<0.02

151 ± 23

6.82 ± 1.31

0.99 ± 0.01

1.47 ± 0.11

50.4 ± 7.8

68.6 ± 1.5

25 ◦ C

CYP1A2 and hNPR in Salmonella strains were expressed at difference spectra. Values are presented as mean ± S.D. (n = 3).

for 24 h. CYP1A2 content was determined by Fe2+ –CO vs. Fe2+

42

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47

6000 5 hr 10 hr 14 hr 23 hr

4000

1A 2bc -b5 1A 2bc -pAC Y C184 pC W -b5 pC W -pA CY C184 TA 15 38+ rat S 9

5000 Revertants per plate

Revertants per plate

5000

3000

2000

4000

3000

2000

1000 1000

0 0

100

101

102

0

103

0

MeIQ (pmol)

and TA1538/pCW-pACYC184. The TA1538/CYP1A2bc-b5 strain showed high sensitivity to MeIQ comparable to TA1538 with the S9 fractions obtained from rats treated with Aroclor 1254 (Fig. 5). However, other Salmonella strains did not show the enhanced mutagenicity of MeIQ. The mutagenic potency of HCAs in the TA1538/CYP1A2bc-b5 was the highest with MeIQ and decreased with other HCAs in the order; MeIQ > MeIQx > PhIP (Fig. 6). At 10−3 pmol of MeIQ after 23 h incubation, a dramatic decreased in revertants seemed due to the cytotoxicity induced by reactive metabolites of MeIQ judging by bacterial lawns in plates (Figs. 4–6). 3.4. Inhibition of MeIQ activation in TA1538/1A2bc-b5 cells by α-naphthoflavone. When various concentrations of ␣-naphthoflavone, a known CYP1A inhibitor, were added to an incubation mixture, the compound showed dose-dependent inhibition of MeIQ-induced mutagenicity with IC50 about 8.3 nM. This result indicates that the CYP1A2 expressed in the TA1538/1A2bc-b5 strain is catalytically active (Fig. 7).

101

102 103 MeIQ (pmol)

Fig. 5. Sensitivity of genetically engineered salmonella strains to MeIQ. ( ) TA1538/1A2bc-b5; ( ) TA1538/1A2bc-pACYC184; ( ) TA1538/PCW-pCyp b5; ( ) TA1538/pCW-pACYC184; ( ) TA1538 + rat S9. Various Salmonella strains were expressed enzymes for 23 h. In TA1538 + rat S9, MeIQ were metabolically activated with rat liver S9. The results are expressed mean ± S.D. (n = 3).

3.5. Effect of quercetin and naringenin on the MeIQ induced mutagenesis We tested quercetin and naringenin in the varying concentration ranges with 1000-fold differences for 5000

Revertantsper Plate

Fig. 4. Effect of expression time on the induction of revertant by MeIQ in TA1538/CYP1A2bc-b5 cells: ( ) 5 h; ( ) 10 h; ( ) 14 h; ( ) 23 h. The results are expressed mean ± S.D. (n = 3). The optimal expression time in TA2538/CYP1A2bc-b5 cells was 23 h for the maximal revertant induction by MeIQ.

100

4000

MeIQ MeIQx PhIP

3000

2000

1000

0 10-1

101 103 105 Heterocyclic Amines (pmol)

Fig. 6. Sensitivity of TA1538/CYP1A2bc-b5 cells to HCAs. TA1538/CYP1A2bc-b5 cells were expressed enzymes for 23 h. Salmonella cells were treated with MeIQ ( ), MeIQx ( ), or PhIP ( ). The results are expressed mean ± S.D. (n = 3).

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47 120

120

Naringenin quercetin

100

80

80

% of inhibition

% of control (Revertantsper plate )

43

60 40 20

40

0

-40

0 0 10-3 10-2 10-1 100 101 102

-80

α-Naphthoflavone (µmol)

0.1

their effects on MeIQ-induced mutagenesis. Quercetin and naringenin were shown to have a biphasic effect on the mutation of TA1538/CYP1A2bc-b5 against MeIQ depending on the concentrations treated (Fig. 8). The

0.1 ␮M quercetin increased the mutagenicity induced by MeIQ about 38% compared to the DMSO controls. However, by further increasing the doses, the mutation rate was decreased and the mutagenicity was inhibited 100

EROD MROD

80

40 20 0 -20

EROD MROD

60

% of Inhibition

% of Inhibition

60

(A)

100

Fig. 8. Concentration-dependent biphasic effect of quercetin ( ) and naringenin ( ) on the activation of MeIQ in TA1538/CYP1A2bc-b5. Salmonella cells and flavonoids were incubated at 37 ◦ C for 10 min before the addition of MeIQ. Data points, the mean percentage of control (the final concentration 100 pM MeIQ, revertant 1769 ± 58) in the absence of inhibitor (n = 3); bars, S.D.

100

-40 0.01

10

Flavonoids (µM)

Fig. 7. Inhibition of the mutagenic activation of MeIQ by ␣-naphthoflavone in TA1538/CYP1A2bc-b5 cells. ␣-Naphthoflavone was added when the bacterial and MeIQ were mixed. Plates were incubated at 37 ◦ C for 2 days and the number of revertants counted. Data points, mean percentage of control (the final concentration 100 pM MeIQ, revertant 2385 ± 166) in the absence of inhibitor (n = 3); bars, S.D.

80

1

40 20 0 -20

0.1

1

10

Quercetin (µM)

100

-40 0.01

(B)

0.1

1

10

100

Naringenin (µM)

Fig. 9. Concentration-dependent biphasic effect of quercetin (A) and naringenin (B) on the activities of MROD and EROD in TA1538/CYP1A2bc-b5 cells. Bacterial cells and flavonoids were preincubated at 37 ◦ C for 10 min, ethoxyresorufin or methoxyresorufin were added to be a final concentration of 10 ␮M and incubated for 30 min. Enzyme activity without flavonoid was 50.4 ± 7.83 pmol per reaction per 30 min in MROD, 68.6 ± 1.51 pmol per reaction per 30 min in EROD. These results are means ± S.D. of triplicate experiments.

44

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47

by about 95% at 100 ␮M of quercetin. Naringenin also showed the biphasic effect results similar to quercetin. Thus, quercetin and naringenin increased the mutagenic rate of TA1538/1A2-b5 by MeIQ at low concentrations, but inhibited it at high concentrations. 3.6. Effect of quercetin and naringenin on the activity of CYP1A2 in S. typhimurium expressed CYP1A2 and reductase To determine the activity of CYP1A2 in TA1538/ 1A2bc-b5 cells treated with various concentrations of quercetin or naringenin, the activities of MRDO and EROD were measured. Quercetin and naringenin were also shown to have biphasic effects on the EROD and MROD activities depending on the doses treated to TA1538/CYP1A2bc-b5 cells. As illustrated in Fig. 9, both flavonoids stimulated the activities of EROD and MROD at low concentrations (0.05–1 ␮M), but inhibited them at high concentrations (10–100 ␮M). These results are consistent with the profile observed in the mutation assay. Thus, it is likely that the biphasic effects of quercetin and naringenin on the MeIQ-induced mutagenesis in S. typhimurium TA1538/CYP1A2bc-b5 were due to their differential modification of the CYP1A2 activity in these cells.

4. Discussion In the present study, quercetin and naringenin were examined using a new salmonella strain, TA1538/CYP1A2bc-b5, treated with HCA. We have developed a new bacterial tester strain (S. typhimurium TA1538/CYP1A2bc-b5) which co-expresses human CYP1A2, NPR and cyp b5. We used a bicistronic construct consisting of a human CYP1A2 encoded by the first cistron and the auxiliary protein NPR by the second, to express both CYP1A2 and NPR. The expression levels of these two proteins were very low compared to the E. coli system [21]. However, the sensitivity of TA1538/CYP1A2bc-b5 to HCAs was similar to that achieved in previously developed salmonella tester strain TA1538/ARO which expresses CYP1A2, NPR and an acetyl CoA-dependent N-hydroxyarylamine O-acetyltransferase (OAT) [27]. HCAs are considered to be N-hydroxylated by CYP1A2 at their amino groups as an initial step, and

the resulting hydroxyamino metabolites can be further activated to electrophilic O-acetyl or O-sulfonyl esters by an N-acetyltransferase or a sulfotransferase to exert their mutagenicity and carcinogenicity. Thus, the mutagenic assay systems for HCAs usually co-express CYP1A2, NPR and acetyltransferase [28,29]. However, the new tester strain (TA1538/CYP1A2bc-b5) employed in our present study co-expresses CYP1A2, NPR and cyp b5, but not OAT. Cyp b5 is known to transfer the 2nd electron from NADPH to CYP, particularly in CYP3A4 and CYP2C, but not in the CYP1 family isoforms [30]. Co-expression of cyp b5 increased the expression level of CYP1A2 in Salmonella (data not shown). When we co-expressed cyp b5 with CYP and NPR in E. coli, the expression levels of CYP and mRNA were increased by mRNA stabilization ([22] manuscript preparation). To increase the expression level of CYP1A2 in the Salmonella strain, we used the chloramphenicol as a selection marker for the pCyp b5 plasmid. Chloramphenicol is known to increase the mRNA levels of proteins by stabilization of mRNA. However, in our system, it does not appear to have the same effect. The expression level of CYP1A2 was not high in the Salmonella strain co-transfected pCYP1A2bc with pACYC184 compared with that in the Salmonella transfected with pCYP1A2bc only (data not shown). Thus, cyp b5 enhances the expression of CYP1A2 in Salmonella by increasing CYP1A2 expression through the stabilization of mRNA. There have been conflicting results in the literature concerning the mutagenicity of quercetin [31–33]. In the present study, neither quercetin nor naringenin alone showed the mutagenicity in the TA1538/CYP1A2bc-b5 tester strain (data not shown). This result is in agreement with that of Calomme et al. [29] which revealed that quercetin acted as an antimutagen against 2-aminofluorene in a Salmonella strain (TA1538) while the flavonoid also lacked any appreciable mutagenicity. Quercetin and naringenin showed the biphasic dose responses to the MeIQ-induced mutagenicity and the CYP1A2 activity in TA1538/CYP1A2bc-b5. At low concentrations of flavonoids, the CYP1A2 enzyme activity was stimulated with concomitant enhancement of MeIQ mutagenicity, possibly due to the elevated formation of the N–OH metabolite of MeIQ. However, at high concentrations, the enzyme activity was strongly inhibited,

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47

and MeIQ-induced mutagenicity was also suppressed. The modulation extent of the enzyme activity was not agreed with that of genotoxicity. EROD and MROD was stimulated about 15 and 20%, but the MeIQ induced revertants were increased about 40 and 50% at 0.1 pmol of quercetin and naringenin, respectively. The similar discrepancy in the extent of inhibition between the enzyme activity and the genotoxicity was also observed at high concentrations of them. Ueng et al. [34] was reported that 100 ␮M naringenin can suppress the genetoxicity of benzo(␣)pyrene activated by mouse liver microsome to the background value but only inhibit the activity of benzo(␣)pyrene hydroxylase by about ∼25%. These results may suggest that the genotoxicity induced by the metabolic activation of MeIQ is more sensitive than the modulation of enzyme activity in Ames test. However, it needs to be verify the mechanism by which they enhance the mutagenicity of MeIQ at low concentrations and inhibit at it high concentrations. A biphasic dose effect was previously observed when we examined the effect of resveratrol on the EROD activity in the human hepatic microsomes [35]. Liu and Gorrod [36] reported that N1-oxidation of 9-benzyladenine catalyzed by the CYP system in animal hepatic microsomes increased with lower ATP concentrations (0.5–2.5 mM), but decreased at higher ATP concentrations (>5 mM). Such differential effects of flavonoids on enzyme activity and mutagenicity can be explained in terms of the hormesis concept by Calabrese [37] who pointed out that the biphasic dose response appears to be quite common in the biomedical and toxicological literature. Calabrese and Baldwin showed that a large number of toxicology studies revealed a dose–response relationship of an hormetic-like biphasic nature depending on the chemical class, physical stressor, animal model, age/gender of subject and biological point [38]. They also described that the width of the low-dose stimulatory range was approximate 95% of several thousand examples within a 100-fold range. In our results, the stimulation of enzyme activity and genotoxicity by quercetin and naringenin was also within a 100-fold range at low dose. People usually took cooked meat or fish together with salads, vegetables and/or fruits which contained substantial amounts of quercetin and naringenin. Most of previous studies assessing chemopreventive potential of flavonoids, fruits or vegetables were carried

45

out with relatively high dose sufficiently to inhibit the enzyme activity [15–17,34]. When we used the 1000-fold range concentrations (0.1–100 pmol) of quercetin and naringenin, we can observe a biphasic dose effects on the enzyme activity and MeIQ induced genotoxicity. Our results indicate that if people consume meat with a small amount of vegetables enough to stimulate the enzymes to bioactivate the HCA to the mutagenic metabolites, the mutation of DNA might be increased. Thus, the consumption of sufficient amounts of vegetables to inhibit the enzyme activity should be emphasized in preventing malignancies, particularly those associated with HCA exposure. The chemopreventive agents which are known to function by modifying directly the metabolic activation of pro-carcinogen might need to be observed the low dose effect. In conclusion, the results of present work support the validation of the newly established strain Salmonella TA1538/CYP1A2bc-b5, expressing CYP1A2, NPR and cyp b5, for the induction of HCAinduced mutagenesis. Quercetin and naringenin showed the biphasic dose-response to MeIQ induced mutagenesis by affecting the activity of CYP1A2 in TA1535/CP1A2bc-b5.

Acknowledgements We thank Professor Young-Joon Surh for revise the manuscript. This study was supported by grant #HMP-98-D-5-0046 from the Ministry of Health and Welfare in Korea.

References [1] R. Doll, R. Peto, The Causes of Cancer, Oxford Medical Publications, Oxford University Press, Oxford, 1981, pp. 1266–1235. [2] M. Yamagishi, M. Natsume, A. Nagaki, T. Adachi, N. Osakabe, T. Takizawa, H. Kumon, T. Osawa, Antimutagenic activity of cacao: inhibitory effect of cacao liquor polyphenols on the mutagenic action of heterocyclic amines, J. Agric. Food Chem. 48 (2000) 5074–5078. [3] S. Pavanello, P. Simioli, G. Mastrangelo, S. Lupi, G. Gabbani, P. Gregorio, E. Clonfero, Role of metabolic polymorphisms NAT2 and CYP1A2 on urinary mutagenicity after a pan-fried hamburger meal, Food Chem. Toxicol. 40 (2002) 1139–1144.

46

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47

[4] IARC, some naturally occurring substances: food items and constituents. heterocyclic aromatic amines and mycotoxins. IARC Monogr. 56 (1993) 163–242. [5] H. Ohgaki, S. Takayama, T. Sugimura, Carcinogenicities of heterocyclic amines in cooked food, Mutat. Res. 259 (1991) 399–410. [6] R.H. Adamson, S. Takayama, T. Sugimura, U.P. Thorgeirsson, Induction of hepatocellular carcinoma in nonhuman primates by the food mutagen 2-amino-3-methylimidazo[4,5-f]quinoline, Environ. Health Perspect. 102 (1994) 190– 193. [7] F.P. Guengerich, W.G. Humphreys, C.-H. Yun, G.J. Hammons, F.F. Kadlubar, Y. Seto, O. Okazaki, M.V. Martin, Mechanisms of cytochrome P450 1A2-mediated formation of N-hydroxy arylamines and heterocyclic amines and their reaction with guanyl residues, in: R.H. Adams, J.-A. Gustafsson, N. Ito, M. Nagao, T. Sugimura, K. Wakabayashi and Y. Yamazoe (Eds.), Proceedings of the 23rd International Symposum of the Princess Takamatsu Cancer Research Fund, Princeton Scientific Publishng, Princeton, NJ, 1995, pp. 78–84. [8] Y. Liu, G.N. Levy, Activation of heterocyclic amines by combinations of prostaglandin H synthase-1 and -2 with N-acetyltransferase 1 and 2, Cancer Lett. 133 (1998) 115–123. [9] H. Maeda, K. Sato, T. Akaike, Superoxide radical generation from heterocyclic amines, In: R.H. Adams, J.-A. Gustafsson, N. Ito, M. Nagao, T. Sugimura, K. Wakabayashi and Y. Yamazoe (Eds), Proceedings of the 23rd International Symposum of the Princess Takamatsu Cancer Research Fund, Princeton Scientific Publishng, Princeton, NJ, 1995, pp. 102–112. [10] R. Edenharder, R. Rauscher, K.L. Platt, The inhibition by flavonoids of 2-amino-3-methylimidazo[4,5-f]quinoline metabolic activation to a mutagen: a structure-activity relationship study, Mutat. Res. 379 (1997) 21–32. [11] H. Lee, H. Yeom, Y.G. Kim, C.N. Yoon, C. Jin, J.S. Choi, B.R. Kim, D.H. Kim, Structure-related inhibition of human hepatic caffeine N3-demethylation by naturally occurring flavonoids, Biochem Pharmacol. 55 (1998) 1369–1375. [12] K.A. Steinmetz, J.D. Potter, Vegetables, fruits and cancer. Part I. Epidemiology, Cancer Causes Control 2 (1991) 325–357. [13] G. Block, B. Patterson, A. Subar, Fruit, vegetables, and cancer prevention: a review of the epidemiological evidence, Nutr. Cancer 18 (1992) 1–29. [14] J.A. Ross, C.M. Kasum, Dietary flavonoids: bioavailability, metabolic effects, and safety, Annu. Rev. Nutr. 22 (2002) 19–34. [15] K. Kanazawa, T. Yamashita, H. Ashida, G.I. Danno, Antimutagenicity of flavones and flavonols to heterocyclic amines by specific and strong inhibition of the cytochrome P450 1A family, Biosci. Biotechnol. Biochem. 62 (1998) 970–977. [16] J.Y. Moon, D.W. Lee, K.H. Park, Inhibition of 7-ethoxycoumarin O-deethylase activity in rat liver microsomes by naturally occurring flavonoids: structure-activity relationships, Xenobiotica 28 (1998) 117–126. [17] S. Lautraite, A.C. Musonda, J. Doehmer, G.O. Edwards, J.K. Chipman, Flavonoids inhibit genetic toxicity produced

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27] [28]

[29]

[30]

by carcinogens in cells expressing CYP1A2 and CYP1A1, Mutagenesis 17 (2002) 45–53. F.J. Gonzalez, H.V. Gelboin, Role of human cytochromes P450 in the metabolic activation of chemical carcinogens and toxins, Drug Metab. Rev. 26 (1994) 165–183. N.M. Borradaile, L.E. de Dreu, P.H. Barrett, C.D. Behrsin, M.W. Huff, Hepatocyte apoB-containing lipoprotein secretion is decreased by the grapefruit flavonoid, naringenin, via inhibition of MTP-mediated microsomal triglyceride accumulation, Biochemistry 42 (2003) 1283–1291. R. Edenharder, J.W. Sager, H. Glatt, E. Muckel, K.L. Platt, Protection by beverages, fruits, vegetables, herbs, and flavonoids against genotoxicity of 2-acetylaminofluorene and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) in metabolically competent V79 cells, Mutat. Res. 521 (2002) 57–72. A. Parikh, E.M.J. Gillam, F.P. Guengerich, Drug metabolism by Esherichia coli expressing human cytochromes P450, Nature Biotechnol. 15 (1997) 784–788. H.-J Kim, Y.I. Park, M.-S. Dong, Stabilization of CYP3A4 mRNA by co-expression of cytochrome b5 in E. coli, in preparation. P. Sandhu, Z. Guo, T. Baba, M.V. Martin, R.H. Tukey, F.P. Guengerich, Expression of modified human cytochrome P450 1A2 in Escherichia coli: stabilization, purification, spectral characterization, and catalytic activities of the enzyme, Arch. Biochem. Biophys. 309 (1994) 168–177. T. Omura, R. Sato, The carbon monoxide-binding pigment of liver microsomes. Part I. Evidence for its hemoprotein nature, J. Biol. Chem. 239 (1964) 2370–2378. Y. Yasukochi, B.S.S. Masters, Some properties of a detergentsolubilized NADPH-cytochrome c(cytochrome P-450): reductase purified by biospecific affinity chromatography, J. Biol. Chem. 251 (1976) 5337–5344. M.-S. Dong, H. Yamazaki, Z. Guo, F.P. Guengerich, Recombinant human cytochrome P450 1A2 and an N-terminal-truncanted form: Construction, purification, aggregation properties, and interactions with flavodoxin, ferredoxin, and NADPH-cytochrome P450 reductase, Arch. Biochem. Biophys. 327 (1996) 11–19. D.M. Maron, B.N. Ames, Revised methods for the Salmonella mutagenicity test, Mutat. Res. 113 (1983) 173–215. A. Suzuki, H. Kushida, H. Iwata, M. Watanabe, T. Nohmi, K. Fujita, F.J. Gonzalez, T. Kamataki, Establishment of a Salmonella tester strain highly sensitive to mutagenic heterocyclic amines, Cancer Res. 58 (1998) 1833–1838. P.D. Josephy, D.H. Evans, A. Parikh, F.P. Guengerich, Metabolic activation of aromatic amine mutagens by simultaneous expression of human cytochrome P450 1A2, NADPH-cytochrome P450 reductase, and N-acetyltransferase in Escherichia coli, Chem. Res. Toxicol. 11 (1998) 70–74. H. Yamazaki, M. Nakamura, T. Komatsu, K. Ohyama, N. Hatanaka, S. Asahi, N. Shimada, F.P. Guengerich, T. Shimada, M. Nakajima, T. Yokoi, Roles of NADPH-P450 reductase and apo- and holo-cytochrome b5 on xenobiotic oxidations catalyzed by 12 recombinant human cytochrome P450s expressed in membranes of Escherichia coli, Protein Exp. Purif. 24 (2002) 329–337.

I.-H. Kang et al. / Mutation Research 545 (2004) 37–47 [31] L.F. Bjeldanes, G.W. Chang, Mutagenic activity of quercetin and related compounds, Science 197 (1977) 577–578. [32] J. da Silva J, S.M. Herrmann, V. Heuser, W. Peres, N. Possa Marroni, J. Gonzalez-Gallego, B. Erdtmann, Evaluation of the genotoxic effect of rutin and quercetin by comet assay and micronucleus test, Food Chem Toxicol. 40 (2002) 941– 947. [33] M. Calomme, L. Pieters, A. Vlietinck, D. Vanden Berghe, Inhibition of bacterial mutagenesis by Citrus flavonoids, Planta Med. 62 (1996) 222–226. [34] Y.F. Ueng, Y.L. Chang, Y. Oda, S.S. Park, J.F. Liao, M.F. Lin, C.F. Chen, In vitro and in vivo effects of naringin on cytochrome P450-dependent monooxygenase in mouse liver, Life Sci. 65 (1999) 2591–2602.

47

[35] M.-S. Dong, S. K Chang, H.-J. Kim, E.M.J. Gillam, F.P. Guengerich, Y.I. Park, Inhibition of 7-alkoxyresorufin O-dealkylation activities of recombinant human CYP1A1 and CYP1B1 by resveratrol, Environ. Mut. Carcinogens 22 (2002) 169–174. [36] E.J. Calabrese, A.L.A. Baldwin, Hormesis: the dose-response revolution, Annu. Rev. Pharmacol. Toxicol. 43 (2003) 175– 197. [37] E.J. Calabrese, A.L.A. Baldwin, The hormetic dose–response model is more common than the threshold model in toxicology, Toxicol. Sci. 71 (2003) 246–250. [38] D.Y. Liu, J.W. Gorrod, Effects of cAMP-dependent protein kinase and ATP on N1-oxidation of 9-benzyladenine by animal hepatic microsomes, Life Sci. 66 (2000) 77–88.