Escherichia coli MTC, a human NADPH P450 reductase competent mutagenicity tester strain for the expression of human cytochrome P450 isoforms 1A1, 1A2, 2A6, 3A4, or 3A5: catalytic activities and mutagenicity studies

Mutation Research 441 Ž1999. 73–83

Escherichia coli MTC, a human NADPH P450 reductase competent mutagenicity tester strain for the expression of human cytochrome P450 isoforms 1A1, 1A2, 2A6, 3A4, or 3A5: catalytic activities and mutagenicity studies Michel Kranendonk a,b,) , Filipa Carreira b, Patricia Theisen b, Antonio Laires b,c , Charles W. Fisher d , Jose´ Rueff b, Ronald W. Estabrook e, Nico P.E. Vermeulen a a

b

Leidenr Amsterdam Center for Drug Research (LACDR), DiÕision of Molecular Toxicology, Vrije UniÕersiteit Õan Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam, Netherlands Department of Genetics, Faculty of Medical Sciences, UniÕersidade NoÕa de Lisboa, R. de Junqueira 96, P-1349-008 Lisbon, Portugal c Faculty of Sciences and Technology, UniÕersidade NoÕa de Lisboa, P-2825 Monte de Caparica, Portugal d Texas Tech Health Sciences Center, School of Pharmacy, Department of Pharmaceutical Sciences, 1300 Coulter AÕenue, Amarillo, TX 79106-1920, USA e Department of Biochemistry, The UniÕersity of Texas, Southwestern Medical Center, 5323 Harry Hines BouleÕard, Dallas, TX 75235, USA Received 22 December 1998; received in revised form 12 February 1999; accepted 16 February 1999

Abstract We report here on the genetic engineering of four new Escherichia coli tester bacteria, coexpressing human CYP1A1, CYP2A6, CYP3A4 or CYP3A5 with human NADPH cytochrome P450 reductase ŽRED. by a biplasmid coexpression system, recently developed to express human CYP1A2 in the tester strain MTC. The four new strains were compared for CYP- and RED-expression levels and CYP activities with the formerly developed CYP1A2 expressing strain. CYP1A2 and CYP2A6 were expressed at the highest, CYP1A1 at the lowest and CYP3A4 and CYP3A5 at intermediate expression levels. Membranes of all five tester bacteria demonstrated similar RED-expression levels, except for the two CYP3A-containing bacteria which demonstrated slightly increased RED-levels. CYP-activities were determined as ethoxyresorufin deethylase ŽCYP1A1 and CYP1A2., coumarin 7-hydroxylase ŽCYP2A6. and erythromycin N-demethylase ŽCYP3A4 and CYP3A5. activities. Reaction rates were comparable with those obtained previously for these CYP–enzymes, except for CYP3A5 which demonstrated a lower activity. Benzow axpyrene and 7,12-dimethylbenzw axanthracene demonstrated mutagenicity in the CYP1A1 expressing strain with mutagenic activities, respectively, approximately 10-fold and 100-fold higher as compared with those obtained with the use of rat liver S9 fraction. Aflatoxin B1 demonstrated a significant mutagenicity with all CYP expressing strains, albeit lower as compared to those obtained with the use of rat liver S9. CYP1A2 was approximately 3-fold more effective in generating a mutagenic response of AFB1 as compared to CYP3A4. CYP3A5 and CYP3A4

AbbreÕiations: a-Naphthoflavone Ž a NF.; 2-Aminoanthracene Ž2AA.; 7,12-Dimethylbenzw axanthracene ŽDMBA.; Aflatoxin B1 ŽAFB1.; Benzow axpyrene ŽBw axP.; Cytochrome P450 ŽCYP.; Diethyldithiocarbamate ŽDTC.; Isopropyl b-D-thiogalactoside ŽIPTG.; Ketoconazole ŽKC.; NADPH cytochrome P450 reductase ŽRED. ) Corresponding author. Fax: q351-1-3622018; E-mail: [email protected] 1383-5718r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 7 1 8 Ž 9 9 . 0 0 0 3 2 - 7

74

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

demonstrated comparable capacities in AFB1 bioactivation which was equal as found for CYP1A1. It is concluded that these four new strains contain stable CYP- and RED-expression, significant CYP-activities and demonstrated significant bioactivation activities with several diagnostic carcinogens. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Human cytochrome P450; Bioactivation, NADPH cytochrome P450 reductase; Mutagenicity; Chemical carcinogen; Escherichia coli

1. Introduction Metabolism of chemical carcinogens plays a major role in the aetiology of cancer w1,2x. Chemical carcinogens are usually chemically inert and require metabolic activation to exert their detrimental effect w1,3,4x. The success of detecting potential carcinogens as mutagens by short-term mutagenicity assays is largely due to the incorporation of a rodent derived metabolic system, to mimic mammalian xenobiotic metabolism w5x. These systems contain high concentrations of cytochrome P450 Žiso.enzyme ŽCYPs.. CYPs play a central role in biotransformation w4,6x. Different members of the superfamily of CYPs catalyse the metabolism of a wide variety of endogenous and exogenous compounds, including steroids, therapeutic drugs, environmental pollutants and Žpro.carcinogens w7–9x. It is generally accepted that CYPs are of importance in the mechanisms leading to cancer w2,10x and that inter-individual differences may be of crucial importance in this regard w11x. Although the use of exogenous metabolic systems has greatly improved the utility of mutagenicity assays, these systems still suffer from a number of deficiencies w12,13x. In order to overcome these deficiencies, novel metabolically competent cell systems for use in mutagenicity assays are being developed w2,14x. One of the more effective approaches is the heterologous expression of mammalian enzymes in the target cell w13,15x. We have reported previously on the development and validation of the Escherichia coli mutagenicity tester strain BMX100 containing active human CYP1A2 w16x. This strain was derived from strain MX100 which was designed for the detection and mechanistic studies of mutagens w17,18x. Mutagenicity detection is based on monitoring the reversion to L-arg prototrophy which can occur via base-substitution mutations w18x. This strain has been comprehensively characterised for more than 20 endogenous

Žbacterial. activities and cofactors that might determine the balance of metabolic activation of a chemical to a mutagen w19x. A major concern in the use of bacteria transformed to express CYPs is the need to overcome insufficient electron transport from NADPH to the CYPs. To solve this insufficiency, we have recently designed a biplasmid coexpression system with the tester strain MTC, combining the expression of human NADPH CYP reductase ŽRED. with a human CYP w20x. This approach has the advantage over other systems, to coexpress any CYP isoform with RED without the need of cloning the RED de novo with each CYP. This is required in designing plasmid constructs for CYP–RED fusion constructs w21,22x, bicistronic constructs w23x, or ‘one plasmid– two promotors’ systems w24x. We have shown that heterologous expression of human CYP1A2 in tester strain MTC1A2 generated a CYP1A2 strain which was more competent in bioactivation of several carcinogens when compared to the corresponding MTC-strains containing a CYP1A2rRED fusion- or bicistronic-coexpression system w20x. Until now, CYP1A2 is the only CYP isoform expressed in mutagenicity tester bacteria w16,20,25– 27x. In this study we report on the development of MTC tester strains which coexpress human RED with human CYP1A1, CYP2A6, CYP3A4 or CYP3A5. These novel strains were characterised for CYP- and RED-expression, CYP activities and bioactivation and compared with the formerly developed strain MTC1A2, expressing CYP1A2.

2. Materials and methods 2.1. Reagents and enzymes L-Arginine, d-aminolevulinic acid, aflatoxin B1 ŽAFB1., 2-aminoanthracene Ž2AA., ampicillin, kanamycin sulphate, a-naphthoflavone Ž a NF., benzo-

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

w axpyrene ŽBw axP., coumarin, 7-hydroxycoumarin Žumbelliferone., cytochrome C, erythromycin, isopropyl b-D-thiogalactoside Ždioxane free. ŽIPTG., diethyldithiocarbamate ŽDTC., 7,12-dimethylbenz w axanthracene ŽDMBA. and thiamine were obtained from Sigma ŽSt. Louis, MO, USA.. Bacto agar, bacto tryptone, bacto yeast extract and bacto peptone were from Difco ŽDetroit, MI, USA.. Ketoconazole ŽKC. was obtained from ICN ŽCosta Mesa, CA, USA.. All other chemicals were of the highest quality. 2.2. Strains, plasmids and cultures Strains and plasmids used in this study for the biplasmid coexpression of human RED and CYP, are presented in Table 1. Development of strain MTC1A2 was previously described w20x. Construction of the E. coli expression vectors pCWh1A1, pCWh2A6,

75

pCWh3A4 and pCWh3A5, containing the modified cDNAs of human CYP1A1, CYP2A6, CYP3A4 or CYP3A5 respectively, will be described elsewhere. These constructions followed the basic design of previous constructs in which the 5X coding sequence is modified to both truncate and alter the amino acid sequence in this region of the protein. They contain optimised coding sequences w22,28x. Modifications to the 5X coding sequence of CYP3A4 have been described previously w22x. The 3X coding sequence was not modified other than to introduce a restriction site immediately following the stop-codon. All constructs were performed by PCR mutagenesis to modify the 5X- and 3X-ends of the sequence. Constructs were sequenced to ensure that modifications to the coding sequence did not occur during mutagenesis. Strain FP401 was transfected with plasmid pLCMhOR. The new strain ŽFP401rpLCMhOR. was subsequently transfected with the different CYP ex-

Table 1 E. coli strains and plasmids Strains

Genotype

Origin

AB1157

B. Bachman

AB1886 MR2100 MR2101 FP401 MX100 BMX100 MTC MTCq MTC1A1 MTC1A2 MTC2A6 MTC3A4 MTC3A5

thr-1, ara-14, leuB6, DŽ gpt–proA.62, lacY 1, tsx-33, qsry, supE44, galK 2, ly, racy hisG4, rfbD1, mgl-51, rpsL31, kdgK51, xyl-5, mtl-1, argE3, thi-1 AB1157, uÕrA6 AB1886, galE MR2100, LPS defective MR2101, hisq, proq, leuq, thrq FP401rpKR11 FP401rpLCM FP401rpLCMhORrpCWq FP401rpLCMrpCWq FP401rpLCMhORrpCWh1A1 FP401rpLCMhORrpCWh1A2 FP401rpLCMhORrpCWh2A6 FP401rpLCMhORrpCWh3A4 FP401rpLCMhORrpCWh3A5

B. Bachman w17x w17x w18x w18x w16x w20x this study this study w20x this study this study this study

Plasmids

Relevant genetic markers

Origin

pKM101 pKR11 pLCM pLCMhOR pCWq pCWh1A1 pCWh1A2 pCWh2A6 pCWh3A4 pCWh3A5

q

r

mucAB , Amp mucABq, Kanr, deletion derivative of pKM101 mucABq, Kanr, derivative of pACYC177 pLCM, containing cDNA of human P450 oxidoreductase under tac promotor pCWoriq ; E.coli exp. vector, Amp r, ptac.ptacrLacI g , without cDNA pCWoriq , containing modified cDNA of human CYP1A1 pCWoriq , containing modified cDNA of human CYP1A2 pCWoriq , containing modified cDNA of human CYP2A6 pCWoriq , containing modified cDNA of human CYP3A4 pCWoriq , containing modified cDNA of human CYP3A5

B.N. Ames w17x w16x w20x w28x C.W. Fisher w28x C.W. Fisher C.W. Fisher C.W. Fisher

76

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

pression vectors, resulting in strain MTC1A1, MTC2A6 MTC3A4 or MTC3A5. All strains were verified for the correct phenotype for mutagenicity testing, as previously described w18x. All transfections to tester strain FP401 and derivatives were carried out by the CaCl 2 method. Bacteria for heterologous expression were grown at 288C for 15 h in TB medium supplemented with peptone Ž2 g ly1 ., ampicillin Ž50 mg mly1 ., kanamycin Ž15 mg mly1 ., d-aminolevulinic acid Ž0.1 mM, except for strain MTC1A2 for which no daminolevulinic acid was added., thiamine Ž1 mg mly1 ., a mixture of trace elements Ž0.4 ml ly1 . w20x and IPTG Ž0.2 mM.. The cultures were analysed for their RED- and CYP-expression, CYP activities and applied in the mutagenicity assays. 2.3. Analysis of RED and CYP expression Preparation of bacterial membranes, determination of CYP- and RED-expression, ethoxyresorufinand methoxyresorufin-dealkylation activities, and protein contents were performed as previously described w20x. A specific activity of 3200 nmol of reduced cytochrome c per nmol reductase was used for calculating the concentration of RED w23x. Coumarin 7-hydroxylase and erythromycin N-demethylase activities for CYP2A6 and CYP3A4r5 respectively, were determined spectrophotometrically, as previously described w29,30x.

amount of CYP Ždetermined in whole cells. present in plated bacteria. CYP-inhibition studies were performed with a 15 min. preincubation of bacteria in incubation buffer with different concentrations of inhibitor, before addition of carcinogen. IC 50 values were determined from the slope as the least squareline of the linear portion of the logarithmic inhibition plots and expressed in mM, as present in preincubations.

3. Results 3.1. MTC-tester strains expressing different CYPs The biplasmid approach recently developed for the coexpression of human RED and CYP1A2 in tester strain MTC1A2 might also allow the generation of derivatives which coexpress RED with other CYPs. In the present study, strains MTC1A1, MTC2A6, MTC3A4 and MTC3A5 were obtained for the first time by introduction of the expression vec-

2.4. Mutagenicity assays The mutagenicity assays were performed using the liquid-preincubation assay technique, as previously described w20x. Stock solutions of carcinogens and inhibitors were freshly made in DMSO and working solutions were obtained by dilution in water. DMSO concentration in preincubations were F 1.3%. Experiments were performed at least in triplicate. Revertant colonies were counted after the usual 48 h of incubation at 378C. Mutagenic activities Žin revertants per nanomole of carcinogen. were determined from the slope as the least square-line of the linear portion of the dose–response curve. Specific mutagenic activities Žin revertants per nanomole of carcinogen and per picomole of CYP. were obtained by normalising the mutagenic activities with the

Fig. 1. Reduced-CO CYP difference spectra of membranes derived from MTC CYP-expressing bacteria. Strains were grown as for mutagenicity assays Žfor methods, see Table 2..

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

tors of human CYP1A1, CYP2A6, CYP3A4 or CYP3A5 in strain FP401rpLCMhOR Žsee Table 1 for the different strains and plasmids used.. In a former study we reported on the need to adapt growth conditions of strain MTC1A2 to prevent loss in sensitivity of the bacteria to detect mutagenicity, in particular when high levels of heterologously expressed CYP and RED were induced w20x. These adapted growth conditions, notably the lower medium concentration of IPTG Ž0.2 mM. used in the present study, were verified for all four newly developed strains and were found to be effective to prevent a low cell-viability, at high expression levels of CYPs and RED Ždata not shown.. The addition of the heme-precursor d-aminolevulinic acid to culture media in general leads to increased CYP-expression levels in bacteria w31x. It was verified if this addition was beneficial for culturing the present tester bacteria as well. The addition of the heme-precursor respectively resulted in an increase of approximately 20% ŽCYP3A4., 50% ŽCYP2A6. or 100% ŽCYP1A1 and CYP1A2. CYP-expression. Strains were subsequently cultured in the presence of d-aminolevulinic acid, except for the CYP1A2-expressing strain, which demonstrated an increase in the spontaneous rever-

77

sion, leading to a decrease of the sensitivity of strain MTC1A2. 3.2. Characterisation of RED and CYP expression and CYP actiÕities In order to characterise the newly engineered bacteria, membranes were derived from cultures as used for mutagenicity assays and analysed for CYPand RED-content as well as for CYP activities Žsee Fig. 1 and Table 2.. The highest CYP-contents were found for strains MTC1A2 Ž180 pmol CYP Žmg protein.y1 . and MTC2A6 Ž143 pmol CYP Žmg protein.y1 . and the lowest for MTC1A1 Ž14 pmol CYP Žmg protein.y1 .. The two CYP3A-expressing strains demonstrated similar levels, namely 43 ŽCYP3A4. and 40 ŽCYP3A5. pmol CYP Žmg protein.y1 . Membranes derived from the strains expressing CYP1A1, CYP3A4 and CYP3A5 demonstrated considerable levels of Žpartly. denatured CYP as indicated by the absorbance-maximum at 420 nm in their CO-reduced vs. reduced difference spectra ŽFig. 1.. RED was expressed similarly in all strains Žapproximately 12 pmol Žmg protein.y1 . except for the two CYP3A-strains which presented slightly higher

Table 2 Characterisation of membranes derived from strain MTCrCYP expressing five different human CYP isoforms and from the control strain MTCq Strain

MTC1A1 MTC1A2 MTC2A6 MTC3A4 MTC3A5 MTCq

CYP Reductase EROD MROD Coumarin Žpmol mgy1 . Žpmol mgy1 . pmol miny1 miny1 pmol miny1 miny1 7-hydroxylation mgy1 mgy1 pmol miny1 miny1 mgy1

pmol miny1 miny1 mgy1

14 " 6 180 " 23 143 " 10 43 " 8 40 " 6 n.d.

– – – 620 " 80 500 " 80 380 " 100

12 " 1 13 " 1 12 " 1 18 " 2 21 " 1 1"1

111 " 2 86 " 12 – – – 1"1

7.9 0.5

55 " 1 282 " 8 – – – 3"1

3.9 1.6

– – 118 " 2 – – 1"1

0.8

Erythromycin N-demethylation

5.6 c 3.0 c

Reference rates a Žminy1 .

8 b,d 0.7 b,d 1.4 e 5.6 f 23 f

Bacterial membranes were isolated from cells as grown for mutagenicity assays; enzymatic activities were determined for each isoform using specific substrates and the results are expressed in pmol miny1 Žmg membrane protein.y1 and in miny1 ; contents and activities" standard deviation Ž N G 3.; EROD: ethoxyresorufin deethylation; MROD: methoxyresorufin demethylation. a Enzymatic rates of purified CYP proteins. b Rates obtained for EROD. c Rates obtained after subtracting activity found for membranes of strain MTCq. d Ref. w52x. e Ref. w45x. f Ref. w36x. n.d., not detected; –, not determined.

78

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

levels. CYP activities were determined as ethoxyresorufin deethylase ŽEROD. and methoxyresorufin demethylase ŽMROD. for both CYP1A-strains, as coumarin 7-hydroxylase for the CYP2A6- and as erythromycin N-demethylase for the CYP3A4- and CYP3A5- expressing strains, respectively Žsee Table 2.. Membranes of strain MTC1A1 and MTC1A2 demonstrated an EROD activity which was approximately 100-fold higher as determined in the control strain MTCq. Strain MTC1A1 demonstrated a higher EROD activity when compared to the MROD activity, the reverse was found for strain MTC1A2. Membranes derived from strain MTC2A6 demonstrated also an approximately 100-fold increase in coumarin 7-hydroxylase activity as compared to the background activity found in the CYPrRED null strain, MTCq. The two MTC3A-strains contained high levels of erythromycin N-demethylase activities Ž620 and 500 pmol miny1 Žmg protein.y1 , respectively. but a considerable bacterial activity was found in the CYPrRED null-strain, as well Ž380 pmol miny1 mgy1 .. 3.3. Mutagenicity assays

Table 3 Mutagenic activities of chemical carcinogens in strain MTCrCYP expressing five different human CYP isoforms Carcinogen

Strain

2AA

MTC1A2 MTCqqS9

3433"447 108"4a

AFB1

MTC1A1 MTC1A2 MTC2A6 MTC3A4 MTC3A5 MTCqqS9

77"12 5437"555 421"10 2349"330 978"59 13247"732 a

Bw axP

MTC1A1 MTCqqS9

798"60 84"8 a

DMBA

MTC1A1 MTCqqS9

682"77 5"1b

a b

Mutagenic activity Žrevertants per nanomole "standard deviation.

Determined with non-induced strain MTC1A2. Determined with strain MX100 w18x.

dent procarcinogens, namely 2-aminoanthracene Ž2AA., benzow axpyrene ŽBw axP., 7,12-dimethylbenzw axanthracene ŽDMBA. and aflatoxin B1 ŽAFB1..

The different MTCrCYP-tester strains were used in mutagenicity assays, testing different CYP-depen-

Fig. 2. Mutagenicity of Bw axP using strains MTC1A1 ŽB. and MTCq Žv .. Points: mean of at least triplicate determinations; bars: standard deviation.

Fig. 3. Histogram of specific mutagenic activities of AFB1 Žrevertants per nanomole of AFB1 and picomole of CYP. using strains MTC1A1, MTC1A2, MTC2A6, MTC3A4 and MTC3A5; bars: standard deviation.

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

Bw axP and DMBA demonstrated mutagenicity with strain MTC1A1, as shown in Fig. 2 and Table 3. The mutagenic activities Žrevertants per nanomole of carcinogen. could be determined from dose–response curves and were 786 and 682 revertants Žnmol carcinogen.y1 , respectively. These activities were substantially higher when compared to those obtained with the CYPrRED void strain, MTCq plus added rat liver S9 fraction. The carcinogen AFB1 was tested with all strains. CYP-normalised mutagenic activities for AFB1 Žin revertants per nanomole of AFB1 and picomole of CYP. could be detected and are presented in Fig. 3. Strain MTC2A6 demonstrated the lowest specific mutagenic activity for AFB1 Ž30 revertants per nanomole of carcinogen and picomole of CYP. and strain MTC1A2 the highest Ž430 revertants Žnmol carcinogen.y1 Žpmol CYP.y1 .. The latter activity was approximately three times higher as found for CYP1A1, CYP3A4 or CYP3A5 Žapproximately 150 revertants Žnmol carcinogen.y1 Žpmol CYP.y1 .. All carcinogens were tested as well with the CYPrRED-null strain, MTCq, up to the highest concentration-levels as tested with the CYPrRED-containing strains, as shown for Bw axP in Fig. 2. No mutagenicity could be detected with any of the carcinogens in this CYPrRED-void strain. The mutagenicity of the carcinogens tested with the CYPrRED-containing strains could be reduced to background levels by chemical CYP-inhibitors Žsee Table 4... The potency of inhibition ŽIC 50 . was determined to be 1.7 mM a NF for the Bw axP muta-

Table 4 Inhibition of mutagenicity of carcinogens with chemical CYP-inhibitors with strain MTCrCYP, expressing five different human CYP isoforms Strain

Carcinogen Name Dose Žnmol.

Inhibitor a

b IC 50 ŽmM"S.D..

MTC1A1 MTC1A2 MTC2A6 MTC3A4 MTC3A5

Bw axP 2AA AFB1 AFB1 AFB1

a NF a NF DTC KC KC

1.7"0.2 0.04"0.001 82"3 2.7"0.3 3.2"0.1

a

0.63 0.31 2.88 0.32 0.51

a NF, a-naphtoflavone; DTC, diethyldithiocarbamate; KC, ketoconazole. b IC 50 was calculated as described in Section 2; S.D., standard deviation; N G 3.

79

genicity in strain MTC1A1. The mutagenicity of 2AA in the strain expressing CYP1A2 and of AFB1 in the strains expressing CYP2A6, CYP3A4 or CYP3A5 could be inhibited with a NF, diethyldithiocarbamate ŽDTC. and ketoconazole ŽKC., respectively. IC 50 values could be determined for these CYP inhibitions as well and were found to be 0.04, 82, 2.7 and 3.2 mM, respectively ŽTable 4.. 4. Discussion The application of E. coli genetically engineered to express various CYPs for bacterial mutagenesis assays provides a practical approach for the evaluation of chemicals potentially harmful to humans. Human CYPs can be easily expressed in E. coli w32x, so that questions about differences in patterns of metabolism between species may be avoided. Furthermore, the generation of reactive metabolites intracellularly permits the evaluation of both shortlived and long-lived intermediates without transmembrane migration, necessitated when using extracellularly added metabolising systems Ži.e., the rodent liver S9 fraction. w13,15x. Engineered E. coli provide a reproducible source of CYPs for use in automated systems required for the large scale analysis of chemicals generated by combinatorial chemistry methods w33x. We describe here the engineering and characterisation of four new mutagenicity tester bacteria coexpressing human RED and human CYP1A1, CYP2A6, CYP3A4 or CYP3A5. High levels of CYP and RED could be obtained in the different MTC tester strains, but with a subsequent loss in cell-viability. Expression of both proteins but foremost RED had to be reduced, notably by lowering the IPTG medium-concentration to 0.2 mM, to obtain cells functional for use in mutagenicity assays. The final CYP- and RED-contents were therefore lower when compared to those obtained in normal E. coli strains Žsuch as DH5a or JM109. w6,23x, where such restrictions were not present. All strains contained significant levels of CYP as demonstrated in the CO-reduced vs. reduced difference spectra ŽFig. 1.. Membranes of strain MTC1A1, MTC3A4 and MTC3A5 demonstrated beside the characteristic 450 nm absorbance-maximum also a considerable absorbance at 420 nm. This type of spectra has been shown previously for CYP1A1,

80

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

CYP2D6 and CYP3A4, heterologously expressed in the E. coli Žcloning. strain DH5a w23x which might indicate a higher sensitivity of these isoforms for mechanical disruption w34x. CYP enzyme activities were determined using specific substrates. The activities were low Žor nonexisting. in the CYPrRED void strain, MTCq, except for erythromycin N-demethylation. As was shown by us previously w20x, human CYP1A2 is more efficient in catalysing methoxyresorufin demethylation ŽMROD. as compared to ethoxyresorufin deethylation ŽEROD.. We demonstrate here that this preference is reversed for human CYP1A1, which is corroborated by results previously obtained by others w35x. The reaction rates of EROD, coumarin 7-hydroxylation and erythromycin N-demethylation in membrane preparations found in this study were comparable with those obtained previously with purified Žrecombinant. enzymes, except for CYP3A5 ŽTable 2.. CYP3A5 demonstrated a lower erythromycin N-demethylation turnover-rate Ž3.0 miny1 . as compared to purified CYP3A5 enzyme Ž23 miny1 . w36x and also lower than CYP3A4 Ž5.6 miny1 .. CYP3A5 generally demonstrates lower drug oxidation rates as compared to CYP3A4 w37x. With all five CYPrRED-expressing tester bacteria we could detect mutagenicity of known CYP-dependent procarcinogens ŽTable 3.. Polycyclic hydrocarbons of which Bw axP and DMBA are prototypes, are bioactivated by CYP1A1 w2x. Both carcinogens were mutagenic with the CYP1A1-expressing tester strain. The mutagenicity of 2AA with the CYP1A2expressing strain has been demonstrated previously w20x and showed a similar mutagenic activity. The potent hepatocarcinogen AFB1 was tested with all strains, as well ŽFig. 3.. Previously, multiple human CYP isoforms were shown to activate AFB1 to mutagenic intermediates, including CYP1A1, CYP1A2, CYP2A1, CYP2A3, CYP2A6, CYP3A3, CYP3A4, CYP3A5, CYP2B7 and CYP2C8 w38–43x. Several authors reported that CYP3A4 is the dominant isoform responsible for AFB1-bioactivation w44–46x. However, in two other studies it was shown that this is also the case for CYP1A2, when determined in direct comparison with CYP3A4 w41,47x. In the present study, CYP1A2 was most efficiently bioactivating AFB1, with an approximately 3-fold higher specific mutagenic activity Žin revertants

Žnmol AFB1.y1 Žpmol CYP.y1 . as compared to CYP3A4 ŽFig. 3.. These results confirm a principal role of CYP1A2 in the bioactivation of AFB1 in addition to CYP3A4 w47x. We demonstrated that CYP1A1 is also capable of bioactivating AFB1, with a specific activity similar to that of CYP3A4. Furthermore, our results demonstrated that human CYP3A5 is as effective in generating mutagenic intermediates as CYP3A4. Previously it has been reported that CYP3A4 is approximately 2-fold more effective in this type of bioactivation w43x, although in this study an extracellularly reconstituted CYPcontaining system was used for metabolic activation. Bw axP, DMBA and 2AA demonstrated higher mutagenic activities with the CYPrRED-containing strains as compared to CYPrRED null strain plus added rat liver S9 fraction ŽTable 3.. AFB1 however, demonstrated a higher mutagenic activity with the addition of the rat liver S9. The S9 mediated mutagenic activities of the first three carcinogens were one ŽBw axP and 2AA. or two ŽDMBA. orders of magnitude lower as compared to the CYPrREDcontaining strains. The differences in mutagenic activity between CYP competent tester bacteria and the CYP-void bacteria plus the addition of rat liver S9 demonstrate the problem of transmembrane migration of bioactivated, reactive intermediates when generated exogenously of the target-cell. Species-differences in AFB1 biotransformation is exemplified with the differences in the mutagenic activities observed for AFB1. A marked species difference in AFB1 bioactivation exists between rat and man w48x. Rat liver microsomes are far more efficient in generating reactive AFB1 intermediates, than human liver microsomes. The pertinent role of the various CYP-isoforms in the mutagenicity of several carcinogens tested with the MTCrCYP strains is demonstrated by the fact that no mutagenicity could be detected with the CYPrRED null tester strain, MTCq. Moreover, the CYP inhibitors a-naphtoflavone Žspecific for CYP1A., diethyldithiocarbamate Žfor CYP2A6. and ketoconazole Žfor CYP3A. w49,50x reduced the mutagenicity of Bw axP, 2AA and AFB1 respectively, to background levels observed with the corresponding CYP expressing strains. The IC 50 values described here are similar to those reported previously w49,50x. a-Naphtoflavone inhibited mutagenicity of 2AA

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

more effectively in the CYP1A2-containing strain as compared to the inhibition of Bw axP mutagenicity in the CYP1A1-containing strain. It has been shown previously that a-naphtoflavone preferentially inhibits CYP1A2 as compared to CYP1A1 w51x. Ketoconazole inhibited equally the mutagenicity of AFB1 in strain MTC3A4 and MTC3A5 demonstrated by the two similar IC 50 values ŽTable 4.. In conclusion, we described here the engineering of four new human CYP competent mutagenicity tester strains. These strains express functional CYPs, other than CYP1A2, the only CYP enzyme which has been expressed in tester bacteria until now. We showed here that the biplasmid coexpression of the different CYPs with human RED is an efficient mean to generate different CYP mutagenicity tester bacteria which can efficiently detect the mutagenicity of several known CYP-dependent procarcinogens. Since most research in this area is ultimately directed towards hazard assessment of human exposure to chemicals, there is an inherent advantage of using the human enzymes CYP1A1, CYP1A2, CYP2A6, CYP3A4 and CYP3A5, the isoforms now expressed in strain MTC. Tester strains expressing CYP2C19, CYP2D6 or CYP2E1, could complement the strains developed so far. Acknowledgements M.K. is holder of a PhD fellowship and P.T. is holder of a MS fellowship, both of the PRAXIS XXI Program ŽPortugal.. This work was supported in part by the European Community Sub-Program Science and Technology of the 2nd Support Framework, JNICT Žproject PRAXISrPSAUrCr67-96. and grants from The National Institutes of Health: GM16488-28 ŽR.W.E. and ES07628 ŽC.W.F... References w1x F.P. Guengerich, Metabolic activation of carcinogens, Pharmacol. Ther. 54 Ž1992. 17–61. w2x F.J. Gonzalez, H.V. Gelboin, Role of human cytochrome P450 in the metabolic activation of chemical carcinogens and toxins, Drug Chem. Rev. 26 Ž1994. 165–183. w3x E.C. Miller, J.A. Miller, Searches for ultimate carcinogens and their reaction with cellular macromolecules, Cancer 47 Ž1981. 2327–2345.

81

w4x N.P.E. Vermeulen, Role of metabolism in chemical toxicity, in: C. Ioannides ŽEd.., Cytochrome P450: Metabolic and Toxicological Aspects, CRC Press, Boca Raton, FL, 1996, pp. 29–53. w5x D. Anderson, Genotoxicity assays, in: E. Arinc¸, J.B. Schenkman, E. Hodgson ŽEds.., Molecular Aspects of Oxidative Drug Metabolizing Enzymes, NATO ASI Series, Vol. H90, Springer-Verlag, Berlin, Germany, 1995, pp. 303–396. w6x F.J. Gonzalez, K.R. Korzekwa, Cytochromes P450 expression systems, Annu. Rev. Pharmacol. Toxicol. 35 Ž1994. 369–390. w7x F.P. Guengerich, Enzymology of rat liver cytochrome P450, in: F.P. Guengerich, ŽEd.., Mammalian Cytochromes P450, Vol. 1, CRC Press, Boca Raton, FL, 1987, pp. 1–54. w8x F.P. Guengerich, T. Shimada, Oxidation of toxic and carcinogenic chemicals by human cytochrome P450, Chem. Res. Toxicol. 4 Ž1991. 391–407. w9x F.J. Gonzalez, Human cytochrome P450: problems and prospects, Trends Pharmacol. Sci. 13 Ž1992. 346–352. w10x A.R. Goeptar, H. Scheerens, N.P.E. Vermeulen, Oxygen and xenobiotic reductase activities of cytochrome P450, Crit. Rev. Toxicol. 25 Ž1995. 25–65. w11x L.W. Wormhoudt, J.N.M. Commandeur, N.P.E. Vermeulen, Genetic polymorphisms of human N-acetyltransferase, cytochrome P450, glutathione S-transferase and epoxide hydrolase enzymes, Crit. Rev. Toxicol. 29 Ž1999. 59–124. w12x J.W. Bridges, S.A. Hubbard, Problems in providing an appropriate drug metabolising system for short-term tests for carcinogens, in: G.M. Wialliams, R. Kroes, H.W. Waaijers, K.W. van de Poll ŽEds.., The Predictive Value of Short-term Screening Tests in Carcinogenicity Evaluation, Elsevier, Amsterdam, 1980, pp. 69–88. w13x R. Langenbach, P.B. Smith, C. Crespi, Recombinant DNA approaches for the development of metabolic systems used in in vitro toxicology, Mutat. Res. 277 Ž1992. 251–275. w14x J. Rueff, C. Chiapella, J.K. Chipman, F. Darroudi, I. Duarte Silva, M. Duverger-van Bogaert, E. Fonti, H.R. Glatt, P. Isern, A. Laires, M. Llagosterra, P. Mossesso, A.T. Natarajan, F. Palitti, A.S. Rodrigues, A. Schinoppi, G. Turchi, G. Werle-Schneider, Development and validation of alternative metabolic systems for mutagenicity testing in short-term assays, Mutat. Res. 353 Ž1996. 151–176. w15x F.J. Gonzalez, C.L. Crespi, H.V. Gelboin, cDNA-expressed human cytochrome P450s: a new age of molecular toxicology and human risk assessment, Mutat. Res. 247 Ž1991. 113–127. w16x M. Kranendonk, P. Mesquita, A. Laires, N.P.E. Vermeulen, J. Rueff, Expression of human cytochrome P450 1A2 in Escherichia coli: a system for biotransformation and genotoxicity studies of chemical carcinogens, Mutagenesis 13 Ž1998. 263–269. w17x M. Kranendonk, M. Ruas, A. Laires, J. Rueff, Isolation and prevalidation of an Escherichia coli tester strain for the use in mechanistic and metabolic studies of genotoxins, Mutat. Res. 312 Ž1994. 99–109. w18x M. Kranendonk, F. Pintado, P. Mesquita, A. Laires, N.P.E. Vermeulen, J. Rueff, MX100, a new Escherichia coli tester

82

w19x

w20x

w21x

w22x

w23x

w24x

w25x

w26x

w27x

w28x

w29x

w30x

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83 strain for use in genotoxicity studies, Mutagenesis 11 Ž1996. 327–333. M. Kranendonk, J.N.M. Commandeur, A. Laires, J. Rueff, N.P.E. Vermeulen, Characterisation of enzyme activities and cofactors involved in bioactivation and bioinactivation of chemical carcinogens in the tester strains Escherichia coli K12 MX100 and Salmonella typhimurium LT2 TA100, Mutagenesis 12 Ž1997. 245–254. M. Kranendonk, C.W. Fisher, R. Roda, F. Carreira, P. Theisen, A. Laires, J. Rueff, N.P.E. Vermeulen, R.W. Estabrook, Escherichia coli MTC, a NADPH cytochrome P450 reductase competent mutagenicity tester strain for the expression of human cytochrome P450: comparison of three types of expression systems, Mutat. Res. 439 Ž1999. 287–300. H. Murakami, Y. Yabbusaki, T. Sakaki, M. Shibata, H. Ohkawa, A genetically engineered P450 monooxygenase: construction of the functional fused enzyme between rat cytochrome P450c and NADPH-cytochrome P450 reductase, DNA 6 Ž1987. 189–197. M.S. Shet, C.W. Fisher, P.L. Holmans, R.W. Estabrook, Human cytochrome P450 3A4: enzymatic properties of a purified recombinant fusion protein containing NADPH-P450 reductase, Proc. Natl. Acad. Sci. USA 90 Ž1993. 117448– 117452. A. Parikh, M.J. Gilliam, F.P. Guengerich, Drug metabolism by Escherichia coli expressing human cytochrome P450, Nature Biotechnol. 15 Ž1997. 784–788. J.A.R. Blake, M. Pritchard, S. Ding, G.C.M. Smith, B. Burchell, C.R. Wolf, T. Friedberg, Coexpression of a human P450 ŽCYP3A4. and P450 reductase generates a highly functional monooxygenase system in Escherichia coli, FEBS Lett. 397 Ž1996. 210–214. D.P. Josephy, L.S. Debruin, H.L. Lord, J.N. Oak, D.H. Evans, Z. Guo, M. Dong, F.P. Guengerich, Bioactivation of aromatic amines by recombinant human cytochrome P4501A2 expressed in Ames tester strain bacteria: a substitute for activation by mammalian tissue preparations, Cancer Res. 55 Ž1995. 799–802. 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 Ž1998. 70–74. 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 heterocylic amines, Cancer Res. 58 Ž1998. 1833–1838. C.W. Fisher, D.L. Caudle, C.M. Martin-Wixtrom, R.H. Tukey, R.H. Waterman, R.W. Estabrook, High level expression of functional human cytochrome P450 1A2 in Escherichia coli, FASEB J. 6 Ž1992. 759–764. G. Reigh, H. McMahon, M. Ishizakki, T. Ohara, K. Shimane, Y. Esumi, C. Green, C. Tyson, S.-I. Ninomiya, Cytochrome P450 species involved in the metabolism of quinoline, Carcinogenesis 17 Ž1996. 1989–1996. R.W. Wang, D.J. Newton, T.D. Scheri, A.Y.H. Lu, Human cytochrome P450 3A4-catalysed testosterone 6b-hydroxyl-

w31x w32x

w33x w34x

w35x

w36x

w37x

w38x

w39x

w40x

w41x

w42x

w43x

w44x

ation and erythromycin N-demethylation. Competition during catalysis, Drug Metab. Dispos. 25 Ž1997. 502–507. F.P. Guengerich, In vitro techniques for studying drug metabolism, J. Pharmacokin. Biopharm. 24 Ž1996. 521–533. F.P. Guengerich, Human cytochrome P450 enzymes, in: P.R. Ortiz de Montellano ŽEd.., Cytochrome P450, Plenum, New York, 1995, pp. 473–535. F.P. Guengerich, A. Parikh, Expression of drug-metabolising enzymes, Curr. Opin. Biotechnol. 8 Ž1997. 623–628. R.C.A. Onderwater, A.R. Goeptar, P.R. Levering, R.M.E. Vos, P.N.M. Konings, J. Doehmer, J.N.M. Commandeur, N.P.E. Vermeulen, The use of macroporous microcarriers for the large-scale growth of V79 cells genetically designed to express single human cytochrome P450 isoenzymes and for the characterisation of the expressed cytochrome P450, Protein Expr. Purif. 8 Ž1996. 439–446. B.W. Penman, L. Chen, H.V. Gelboin, F.J. Gonzalez, C.L. Crespi, Development of a human lymphoblastoid cell line constitutively expressing human CYP1A1 cDNA: substrate specificity with model substrates and promutagens, Carcinogenesis 15 Ž1994. 1931–1937. E.M.J. Gillam, Z. Guo, Y.-F. Ueng, H. Yamazaki, I. Cock, P.E.B. Reilly, W.D. Hooper, F.P. Guengerich, Expression of X cytochrome P450 3A5 in Escherichia coli: effects of 5 modification, purification, spectral characterisation, reconstitution conditions and catalytic activities, Arch. Biochem. Biophys. 317 Ž1995. 374–384. S.A. Wrighton, W.R. Brian, M.A. Sari, M. Iwasaki, F.P. Guengerich, J.L. Raucy, D.T. Molowa, M. Vandenbranden, Studies on the expression and metabolic capabilities of human liver cytochrome P450IIIA5 ŽHLP3., Mol. Pharamacol. 38 Ž1990. 207–213. C. Sengstag, H.P. Eugster, F.E. Wurgler, High promutagen activating capacity of yeast microsomes containing human cytochrome P450 1A and human NADPH cytochrome P450 reductase, Carcinogenesis 15 Ž1994. 837–843. T. Aoyama, S. Yamano, P.S. Guzelian, H.V. Gelboin, F.J. Gonzalez, Five of 12 forms of vaccinia virus-expressed human hepatic cytochrome P450 metabolically activate aflatoxin B1, Proc. Natl. Acad. Sci. USA 87 Ž1990. 4790–4793. L.M. Forrester, G.N. Neal, D.J. Judah, M.J. Glancey, C.R. Wol, Evidence for involvement of multiple forms of cytochrome P450 in aflataoxin B1 metabolism in human liver, Proc. Natl. Acad. Sci. USA 87 Ž1990. 8306–8310. C.L. Crespi, B.W. Penman, D.T. Steimel, H.V. Gelboin, F.J. Gonzalez, The development of a human cell line stably expressing human CYP3A4: role in the metabolic activation of aflatoxin B1 and comparison to CYP1A2 and CYP2A3, Carcinogenesis 12 Ž1991. 355–359. D.L. Eaton, E.P. Gallagher, Mechanisms of aflatoxin carcinogenesis, Annu. Rev. Pharmacol. Toxicol. 34 Ž1994. 135–172. H. Yamazaki, Y. Inui, S.A. Wrighton, F.P. Guengerich, T. Shimada, Procarcinogen activation by cytochrome P450 3A4 and 3A5 expressed in Escherichia coli and by human liver microsomes, Carcinogenesis 16 Ž1995. 2167–2170. T. Shimada, F.P. Guengerich, Evidence for cytochrome

M. Kranendonk et al.r Mutation Research 441 (1999) 73–83

w45x

w46x

w47x

w48x

P450NF, the nifedipine oxidase, being the principle enzyme involved in the bioactivation of aflatoxins in human liver, Proc. Natl. Acad. Sci. USA 86 Ž1989. 426–465. T. Shimada, E.M.J. Gillam, T.R. Sutter, P.T. Strickland, F.P. Guengerich, H. Yamazaki, Oxidations of xenobiotics by recombinant human cytochrome P450 1B1, Drug Metab. Dispos. 29 Ž1997. 617–622. Y.-F. Ueng, T. Shimada, H. Yamazaki, F.P. Guengerich, Oxidation of aflatoxin B1 by bacterial recombinant human cytochrome P450 enzymes, Chem. Res. Toxicol. 8 Ž1995. 218–225. E.P. Gallagher, L.C. Wienkers, P.L. Stapleton, K.L. Kunze, D.L. Eaton, Role of human microsomal and human complementary DNA-expressed cytochrome P4501A2 and P4503A4 in the bioactivation of aflatoxin B1, Cancer Res. 54 Ž1994. 101–108. H.S. Ramsdell, D.L. Eaton, Species susceptibility to aflatoxin B1 carcinogenesis: comparative kinetics of microsomal biotransformation, Cancer Res. 50 Ž1990. 615–620.

83

w49x D.J. Newton, R.W. Wang, A.Y.H. Lu, Cytochrome P450 inhibitors, Drug Metab. Dispos. 23 Ž1995. 154–158. w50x T.K.H. Chang, F.J. Gonzalez, D.J. Waxman, Evaluation of triacetyloleandomycin, a-naphtoflavone and diethyldithiocarbamate as selective chemical probes for the inhibition of human cytochrome P450, Arch. Biochem. Biophys. 311 Ž1994. 437–442. w51x W. Tassaneeyakul, D.J. Birkett, M.E. Veronese, M.E. McManus, R.H. Tukey, L.C. Quattrochi, H.V. Gelboin, J.O. Miners, Specificity of substrate and inhibitor probes for human cytochrome P450 1A1 and 1A2, J. Pharmacol. Exp. Ther. 265 Ž1993. 401–407. w52x H. Iwata, K.-i. Fujita, H. Kushida, A. Suzuki, Y. Konno, K. Nakamura, A. Fujino, T. Kamataki, High catalytic activity of human cytochrome P450 coexpressed with human NADPHcytochrome P450 reductase in Escherichia coli, Biochem. Pharmacol. 55 Ž1998. 1315–1325.