Molecular Cloning, Sequencing, and Expression inEscherichia coliof Mouse Flavin-Containing Monooxygenase 3 (FMO3): Comparison with the Human Isoform

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 347, No. 1, November 1, pp. 9–18, 1997 Article No. BB970322

Molecular Cloning, Sequencing, and Expression in Escherichia coli of Mouse Flavin-Containing Monooxygenase 3 (FMO3): Comparison with the Human Isoform1 J. Greg Falls,*,2 Nathan J. Cherrington,* Kieran M. Clements,† Richard M. Philpot,‡ Patricia E. Levi,* Randy L. Rose,* and Ernest Hodgson*,3 *Department of Toxicology, North Carolina State University, Box 7633, Raleigh, North Carolina 27695; †Department of Entomology, North Carolina State University, Box 7613, Raleigh, North Carolina 27695; and ‡Laboratory of Signal Transduction, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709

Received April 3, 1997, and in revised form July 30, 1997

The sequence of mouse flavin-containing monooxygenase 3 (FMO3) was obtained from several clones isolated from a mouse liver cDNA library. The nucleotide sequence of mouse FMO3 was 2020 bases in length containing 37 bases in the 5* flanking region, 1602 in the coding region, and 381 in the 3* flanking region. The derived protein sequence consisted of 534 amino acids including the putative flavin adenine dinucleotide and NADP/ pyrophosphate binding sites (characteristic of mammalian FMOs) starting at positions 9 and 191, respectively. The mouse FMO3 protein sequence was 79 and 82% identical to the human and rabbit FMO3 sequences, respectively. Mouse FMO3 was expressed in Escherichia coli and compared to E. coli expressed human FMO3. The FMO3 proteins migrated with the same mobility (Ç58 kDa) as determined by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and immunoblotting. The expressed FMO3 enzymes (mouse and human forms) were sensitive to heat and reacted in a similar manner toward metal ions and detergent. Catalytic activities of mouse and human FMO3 were high toward the substrate methimazole; however, in the presence of trimethylamine and thioacetamide, FMO-dependent methimazole oxidation by both enzymes was reduced by greater than 85%. Other

substrates which inhibited methimazole oxidation were thiourea and thiobenzamide and to a lesser degree N,N-dimethylaniline. When probed with mouse FMO3 cDNA, FMO3 transcripts were detected in hepatic mRNA samples from female mice, but not in samples from males. FMO3 was detected in mRNA samples from male and female mouse lung, but FMO3 message was not detected in mouse kidney sample from either gender. Results of immunoblotting confirmed the tissue- and gender-dependent expression of mouse FMO3. q 1997 Academic Press Key Words: flavin-containing monooxygenase 3; FMO3; FMO3 sequence; FMO3 expression.

The microsomal flavin-containing monooxygenases (FMOs, EC 1.14.13.8)4 catalyze the flavin adenine dinucleotide (FAD)-, NADPH-, and O2-dependent oxidation of numerous xenobiotics containing nitrogen, sulfur, phosphorous, or selenium heteroatoms (1–3). Although primarily involved in the detoxication process, activation of compounds to a more reactive chemical species may occur and ultimately elicit a toxic response (4, 5). In contrast with the numerous exogenous compounds identified as substrates for the FMOs, relatively few endogenous substrates are known; examples of these

1

Sequence data from this article have been deposited in the GenBank Libraries under Accession No. U87147 (mouse FMO3). 2 Present address: Department of Radiation Oncology, Duke University Medical Center, Box 3433, Durham, NC 27710. 3 To whom reprint requests should be addressed. Fax: (919) 5157169.

4 Abbreviations used: FAD, flavin adenine dinucleotide; FMO, flavin-containing monooxygenase; SSC, standard sodium citrate; IPTG, isopropyl b-D-thiogalactoside; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine; pJL-FMO3m, pJL-mouse FMO3; pJL-FMO3h, pJL-human FMO3; ORF, open reading frame.

9

0003-9861/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

arca

10

FALLS ET AL.

include cysteamine (6, 7) and the cysteine S-conjugates (8, 9). The dietary compound trimethylamine is converted to its corresponding N-oxide by FMO; however, individuals deficient in FMO (presumably FMO3) may develop trimethylaminuria, a genetic disorder resulting in the excretion of the malodorous free amine (10, 11). Purified FMO was first obtained from pig liver microsomes (12), and its catalytic mechanism has been extensively characterized. Subsequent purification procedures were developed for the rat (13), mouse (14), rabbit (15, 16), guinea pig (17), and macaque (18). Based on modulation of FMO activity due to the effect of metal ions on partially purified rabbit lung and liver FMO, it was proposed that these tissues might possess different forms of FMO (19). In addition, Ohmiya and Mehendale (20–22) observed differences in the FMO activity of rabbit liver and lung microsomes toward the substrates N,N-dimethylaniline (DMA), imipramine, and chlorpromazine. More conclusively, it was demonstrated that the purified lung and liver FMOs (subsequently designated FMO2 and FMO1, respectively) from rabbit were catalytically and immunochemically distinct (16, 23). Sequencing of the corresponding cDNAs yielded evidence of distinctly different but related genes (24). To date, five FMO isoforms (designated FMO1–FMO5) have been identified by amino acid or cDNA sequencing, each represented by a single gene (for reviews, see 25–28). Orthologous genes share at least 80% amino acid identity, whereas homologous FMOs are 52–57% identical. Highly related forms (greater than 98% identity) within a single species are the result of allelic variation (29). In addition to the well-documented species and tissue-dependent expression of FMOs, endogenous factors such as developmental status or gender affect expression. Based on mRNA expression, the primary isoform expressed in adult human liver appears to be FMO3; however, FMO1 appears to be the predominant form expressed in human fetal liver (30). The previously reported gender differences in mouse hepatic FMO activity (31, 32) are now known to be related to differences in isozyme expression. FMO expression in mouse liver can be female-predominant (FMO1), female-specific (FMO3), or gender-independent (FMO5) (33). We have recently shown that testosterone is the primary sex steroid responsible for the sexually dimorphic expression of mouse liver FMOs 1 and 3 (34). Although FMO has been purified from mouse liver (14), kidney (35), and lung (36), no complete amino acid or cDNA sequences have been published. Consequently, the purpose of this study was to obtain the cDNA sequence of mouse FMO3, to utilize a cDNArecombinant expression system for identifying similarities in catalytic properties between the mouse and hu-

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

man isoforms, and to examine tissue- and gender-dependent expression of mouse FMO3. MATERIALS AND METHODS Construction of the cDNA library. Livers were removed from adult female CD-1 mice (Charles River Laboratories, Raleigh, NC), snap frozen in liquid nitrogen, and then stored at 0807C until analysis. Total RNA was isolated using acid guanidine thiocyanate:phenol:chloroform extraction (37), followed by selection of mRNA using oligo(dT)–cellulose (38). A cDNA library was constructed using the l ZAP Express Vector kit (Stratagene Cloning Systems). Following selection of cDNA ranging in size from 1.5 to 6.5 kb, XhoI–EcoRI linkers were utilized for the insertion of cDNA into l ZAP Express. Screening of the cDNA library. The female mouse liver cDNA library was screened with a 3 * fragment (720 bases) of rabbit FMO3 cDNA obtained by digesting the full-length cDNA with the restriction enzymes HindIII–EcoRI. The fragment (50 ng/labeling reaction) was radiolabeled with [a-32P]dCTP (3000 Ci/mmol, DuPont NEN) using a random-primed labeling kit (Boehringer-Mannheim), and unincorporated nucleotides were removed with Stratagene push columns. The labeled probe was then added directly to the hybridization solution at a concentration of 106 cpm/ml. Plaque lifts were first prehybridized for 2 h at 257C and then hybridized in the same solution of 61 SSC, 100 mg/ml salmon sperm DNA, 41 Denhardt’s solution, 0.5% SDS, and 50% formamide (hybridization solution). Because of differences in FMO3 homology among species, hybridizations were performed under low stringency conditions (377C overnight). Filters were washed for 15 min each in the following sequence: 11 SSC, 0.1% SDS at 377C; 0.11 SSC, 0.1% SDS at 457C; and 0.11 SSC, 0.1% SDS at 607C. Exposure to Kodak X-ray film was for 24 h at 0807C. Clones of interest were isolated and screened to purity. To determine if the identified clones were of sufficient length to contain the coding region, PCR of the isolated l phage (containing insert) was performed (primers were BK reverse and T7). Plasmids from putative full-length clones were excised with ExAssist helper phage from bacteriophage as the pBK phagemid (subcloning step) and then transfected into the XLOLR Escherichia coli strain. Plasmid DNA was then prepared for sequencing. Nucleotide sequencing of clones. End sequencing of three clones, clones 1, 2, and 7, was performed in our laboratory using the universal primers T3 and T7 and the dideoxy-mediated chain termination method (39). In addition, each strand of clones 2 and 7 (the two largest clones) was completely sequenced by the DNA Sequencing Core Laboratory, Interdisciplinary Center for Biotechnology Research at the University of Florida (Gainesville, FL). Clone 6 was end sequenced by the DNA Sequencing Core Laboratory. Construction of E. coli expression vectors. The vector utilized was pJL-2, a derivative of pKK233-2 (Pharmacia, LKB Biotechnology, Inc.) which has been used successfully in the E. coli expression of several FMO isoforms (40–42). In pJL-2, the pBR322 origin of replication has been replaced with the one from pUC and a translation enhancer sequence inserted between the ribosome binding site and the initiation codon (43). Additionally, an EcoRV cloning site has replaced the PstI site, limiting possible cloning sites to XbaI, NcoI, EcoRV, and HindIII (42). To prepare the vector for insertion of the FMO3 cDNA, the restriction enzymes XbaI and HindIII were used to cut pJL-2. For the mouse FMO3 clone 7, high-fidelity PCR was utilized to add an XbaI restriction site 5* of the start codon (sense primer: 5*-GAAAAGTCTAGATGAAGAAGAAAGTGGCCA-3*) and a HindIII site 3* of the stop codon (antisense primer: 5*-TCTCAGAAGCTTTAGAGGGACTCACGATCA-3*). The product was then cut with XbaI and HindIII restriction enzymes. Because the mouse FMO3 clone contained an internal XbaI site, sequential ligation of the two products from the

arca

SEQUENCING AND EXPRESSION OF MOUSE FMO3 restriction digestion was performed. First, the larger 3 * fragment (961 bases) was ligated into the XbaI/HindIII digested pJL-2. Next, the smaller 5* fragment (644 bases) was ligated into pJL-2 containing the partial insert. To determine the orientation of the second insert, an Eco1090I site (at base 118) was cut with the appropriate enzyme. Plasmid cDNA obtained from pJL-mouse FMO3 (pJL-FMO3m) was completely sequenced. Expression in E. coli. E. coli XL-1 Blue cells were transformed with nonrecombinant pJL-2 (pJL) or pJL-FMO3m and then plated onto LB plates containing 100 mg/ml ampicillin. E. coli JM109 cells transformed with pJL-human FMO3 (pJL-FMO3h) were also plated onto LB–ampicillin plates. Single colonies from each transformation were grown overnight at 377C in 5 ml of LB–ampicillin (50 mg/ml; LB-Amp). These cultures were added to 500 ml of LB-Amp and grown at 377C (3 h) until an OD600 reading of 0.5. The cultures were cooled to room temperature, IPTG was added to a final concentration of 1 mM, and then the cultures were grown overnight at 307C with shaking at 150 rpm. The following day, subcellular fractions were prepared as described by Lawton and Philpot (44) and Itagaki et al. (41). Analysis of cDNA-expressed proteins. The particulate fractions (20 mg) of E. coli transformed with pJL, pJL-FMO3m, or pJL-FMO3h were electrophoresed on a 7.5 % SDS–polyacrylamide gel (3% stacking gel) under denaturing conditions (45) and stained with Coomassie blue. In addition, particulate fractions of expressed FMO3 proteins, human liver microsomes, and mouse liver microsomes (CD-1 strain, male and female) were subjected to SDS–PAGE, transferred to nitrocellulose (46), and then reacted with an anti-rabbit FMO3 antibody (40). Incubations of blots were similar to those described by Venkatesh et al. (36) except an anti-goat IgG secondary antibody was used (Sigma Chemical Co.). Mouse liver microsomes were prepared as previously described (47). FMO activity was determined by monitoring methimazole oxidation (48) at 412 nm (377C, pH 8.4) with a Shimadzu double-beam spectrophotometer. Sample and reference cuvettes contained a 1-ml reaction mixture of 0.1 M Tricine/KOH, pH 8.5, 0.1 mM EDTA, 0.06 mM 5,5*-dithiobis(nitrobenzoic acid), 0.2 mM dithiothreitol, 0.1 mM NADPH, 30 to 60 mg particulate fraction, and 1.0 mM methimazole (sample cuvette only). To access temperature stability, particulate fractions were first heated for 0, 1, 2, 3, and 5 min at 457C and then transferred to an ice-cold Eppendorf tube prior to determining FMO activity. For examining the effects of varying pH on FMO activity, the Tricine buffer was adjusted accordingly with KOH. Additions of sodium cholate and MgCl2 (final concentrations of 1% and 100 mM, respectively) were made to the sample and reference cuvettes prior to the determination of the reaction rate, whereas additions of noctylamine (3 mM) and other substrates were added to the sample and reference cuvettes subsequent to the addition of methimazole (i.e., methimazole oxidation was monitored for 2 min first). Kinetics of methimazole metabolism were determined from results obtained by the addition of increasing amounts of methimazole (2.5, 5, 15, 35, 100, 300, 1000, and 3000 mM final concentrations) to the sample cuvette. Protein concentrations were determined by the method of Bradford (49) with Bio-Rad reagents. Flavin content (FAD and FMN) of E. coli particulate fractions was determined fluorometrically by the method of Faeder and Siegel (50). Analysis of mouse FMO3 mRNA and protein tissue levels. The livers, kidneys, and lungs of adult male and female CD-1 mice (Charles River Laboratories, Raleigh, NC) were removed and processed as previously described. Five micrograms of each mRNA sample (prepared as described above) was electrophoresed in a 1% agarose gel containing formaldehyde (51), followed by transfer onto a Hybond N/ membrane (Amersham) with fixation by NaOH. The blot was incubated in prehybridization solution (61 SSC, 250 mg/ml salmon sperm DNA, 51 Denhardt’s, and 0.1% SDS) at 47C for 2 h and then reacted with a mouse FMO3 full-length cDNA probe (labeling procedure described above) in hybridization solution overnight at

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

11

657C. Blots were washed sequentially for 15 min in the following solutions: (i) 11 SSC, 0.1% SDS at 457C, (ii) 0.11 SSC, 0.1% SDS at 457C, and (iii) 0.11 SSC, 0.1% SDS at 657C, and then subjected to autoradiography for 48 h. Microsomes were prepared from tissue samples as described previously (47), and immunoblotting was performed as described above. The antibody used was prepared against expressed rabbit FMO3.

RESULTS

Cloning of FMO3 from a mouse liver library. The mouse liver library (Ç225,000 plaques) was screened with a 3* cDNA fragment from rabbit FMO3. Because the 3* ends of FMOs are the least conserved (27), this fragment was expected to hybridize more specifically to mouse FMO3, reducing the possibility of detecting other mouse liver FMO forms. More than 50 plaques were identified as positives, and 10 were selected to undergo further screenings. Following tertiary screenings, eight clones were identified as putative FMO3 clones. Primers 5* and 3* of the multiple cloning site, BK reverse and T7, respectively, were used in a PCR to determine the relative insert sizes of these clones. Three clones had insert sizes of at least 1.9 kb, suggesting that they might contain the open reading frame (ORF). Based on the ORFs, rabbit and human FMO3 cDNAs were 1593 and 1596 nucleotides, respectively (30, 40). Sequencing of cDNAs encoding mouse FMO3. Both strands of the two largest clones, designated clones 2 and 7, were sequenced entirely, and the smaller clone (clone 1) was end sequenced. The mouse sequence (clone 7) was consistent with the primary structure of rabbit and human FMO3, and according to the accepted nomenclature system for FMOs (25), this newly described sequence represents the mouse ortholog of FMO3. The complete nucleotide sequence of mouse FMO3 (Fig. 1) consisted of 2020 bases, 1602 in the coding region, 37 in the 5* flanking region, and 381 in the 3* flanking region. The sequence of clone 7 is reported here as mouse FMO3 and has been entered in the GenBank Libraries under Accession No. U87147. Clone 2 contained a 64-base insert (beginning 414 nucleotides downstream from the translation start site) which was not present in clone 7 or clone 1. The 64base sequence showed no homology to any of the known FMO forms or to any flavoproteins. In addition, translation of the encoded protein (including the 64-base sequence) was predicted to terminate after 138 amino acids. Except for this 64-base insert, clone 2 was identical to clone 7. To determine if this sequence was a cloning artifact or a possible intron, genomic DNA was amplified by PCR using primers designed from regions 5* and 3* of this sequence. In addition, clone 7 was amplified, and the products of each reaction were resolved on an agarose gel. Results indicate that this sequence is extraneous DNA (data not shown); how-

arca

12

FALLS ET AL.

FIG. 1. The complete nucleotide sequence of mouse FMO3. The sequence contains 37 bases of 5* flanking region, 1602 bases of coding region, and 381 bases of 3* flanking region. The start (ATG) and stop (TGA) codons and the consensus 3 * polyadenylation signal (AATAAA) are underlined and in boldface. The mouse FMO3 sequence has been entered in the GenBank Libraries under Accession No. U87147.

ever, its origin is unknown. An artifact of cloning has been reported for FMO4 (40). End sequencing verified clone 1 as mouse FMO3, and no ambiguities were found between the portion of clone 1 that was sequenced and the sequence of clone 7 (the sequence reported here). The nucleotides surrounding the mouse FMO3 trans-

consists of a coding region (534 amino acids) which is 82 and 79% identical to that of rabbit and human FMO3, respectively, and 52–56% identical to other FMO forms. In addition, amino acid positions 19 and 191 of the mouse FMO3 contain the consensus sequence (GxGxxG/A) for the putative FAD and NADP

SEQUENCING AND EXPRESSION OF MOUSE FMO3

13

FIG. 2. The complete derived amino acid sequence for mouse FMO3. Differences with the rabbit and human FMO3 sequences are shown. The putative pyrophosphate-binding sequences (GxGxxG/A) are underlined. The human sequence derived from cDNA (HFMO3) is from Dolphin et al. (GenBank Z47552, 1995) and the rabbit sequence also obtained from cDNA is from Burnett et al. (40). Conservation of amino acids in all three forms is indicated by an asterisk (consensus line).

(data not shown). The FMO3 sequences contain strongly predicted regions of hydrophobicity at the NH2 and COOH termini and at four internal sites (including one around position 195 and three sites between positions 300 and 500). As observed with most FMOs, the C-terminus tends to be hydrophobic, but is not highly conserved. Beginning at amino acid position 512, the mouse sequence (allowing for gaps) is only 60 and 59% identical to rabbit and human FMO3, respectively. The mouse FMO3 protein has a calculated molecular weight of 60,515 Da and a theoretical pI of 8.64, whereas the calculated human FMO3 protein was 60,047 Da and pI 8.3 (30).

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

Expression of FMO3 in E. coli. SDS–PAGE and Coomassie blue staining (Fig. 3A) of the 100,000g particulate fractions of E. coli transformed with pJL (lane 2), pJL-FMO3m (lane 3), or pJL-FMO3h (lane 4) revealed a high level of FMO3 expression in the pJLFMO3m and pJL-FMO3h samples. The pJL-FMO3m and pJL-FMO3h transformed samples showed a band which corresponded to a molecular weight of Ç58 kDa. This protein was absent in the particulate fractions from the pJL sample (transformed with vector alone). FAD content of the particulate fractions was determined by the method of Faeder and Siegel (50). Based on a 1:1 molar ratio of FAD to FMO, and based on

arca

14

FALLS ET AL.

lecular weight observed in the male and female mouse microsomal samples appears to be FMO5. Previous results in our laboratory indicated that the anti-rabbit FMO3 antibody, in addition to detecting mouse FMO3, detected mouse FMO5 (data not shown); however, this antibody does not detect human or rabbit FMO5 (41). Catalytic properties of the expressed enzymes. Various physical and catalytic properties of the mouse and human FMO3 proteins expressed in E. coli were examined using methimazole (1.0 mM) as a substrate. Data presented represent the averages of two determinations. FMO activity of the expressed enzymes was decreased in a time-dependent manner in response to elevated temperature (Fig. 4A). Following heating at 457C for 5 min, only 20% of mouse and human FMO3 activity remained, similar to results obtained for hu-

FIG. 3. Analysis of expressed FMO3. (A) Samples (10 mg of protein) of the 100,000g particulate fractions from IPTG-induced cultures of E. coli were electrophoresed on polyacrylamide gels in the presence of SDS and stained with Coomassie blue. Lane 2 contained E. coli transformed with nonrecombinant pJL vector; lane 3, transformation with pJL-mouse FMO3; and lane 4, transformation with pJL-human FMO3. Lane 1 shows the molecular weight standards of myosin (205 kDa), b-galactosidase (116 kDa), phosphorylase b (97 kDa), bovine plasma albumin (66 kDa), ovalbumin (45 kDa), and carbonic anhydrase (29 kDa). The arrowhead indicates the expressed FMO3 (approximately 58 kDa). (B) Immunoreactivity of expressed proteins and liver microsomes toward anti-rabbit FMO3 antibody. Samples [lane 1, expressed rabbit FMO3 (0.2 mg); lane 2, expressed human FMO3 (2 mg); lane 3, expressed mouse FMO3 (2 mg); lane 4, human liver microsomes (6 mg); lane 5, male mouse liver microsomes (5 mg); lane 6, female mouse liver microsomes (5 mg); and lane 7, expressed mouse FMO3 (4 mg)] were electrophoresed on polyacrylamide gel in the presence of SDS, transferred to nitrocellulose, reacted sequentially with the primary and secondary antibodies, and then visualized with BCIP/NBT. The arrowhead points to proteins detected at Ç58 kDa.

the calculated molecular weight of FMO3 (60.5 kDa), increases in flavin content related to pJL-recombinant transformation were 532 pmol/mg protein for pJLFMO3m and 590 pmol/mg protein for pJL-FMO3h. This corresponded to FMO3 expression levels of 3.1 and 3.5% for the total E. coli particulate fraction of mouse and human samples, respectively. The mobility and immunoreactivity (toward antirabbit FMO3 antibody) of the expressed proteins were then compared with samples of human liver microsomes, mouse liver microsomes, and rabbit-expressed FMO3 (Fig. 3B). FMO3 protein (Ç58 kDa) was detected in all the samples examined except those of male mouse. The minor cross-reacting protein of higher mo-

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

FIG. 4. Properties of expressed FMO3. (A) Effect of elevated temperature on FMO activity of mouse and human FMO3 expressed in E. coli. Samples were heated at 457C for 0, 1, 2, 3, or 5 min, stored on ice, and then assayed for activity by monitoring methimazole (1.0 mM) oxidation. (B) Effect of changes in pH on FMO activity of mouse and human FMO3 expressed in E. coli as determined by methimazole (1.0 mM) oxidation. Data presented represent averages of two determinations, and the results differed by less than 10%.

arca

15

SEQUENCING AND EXPRESSION OF MOUSE FMO3

FIG. 5. Effects of various treatments on the metabolism of methimazole (1.0 mM) catalyzed by mouse and human FMO3 expressed in E. coli. Treatments included heating at 457C for 5 min, 1% sodium cholate (cholate), 100 mM magnesium chloride (MgCl2), 3 mM noctylamine (Oc-NH2), and 1 mM imipramine. Data presented represent averages of two determinations, and the results differed by less than 10%.

man FMO3 (55) and rabbit FMO3 (40). In contrast, FMO2 has been reported to be refractory to changes in activity under these conditions (56). Differences in FMO activity as a function of pH (ranging from 7.0 to 11.5) were then examined. The pH optima of mouse and human FMO3 were 9.0 and 8.5, respectively (Fig. 4B). In comparison, rabbit FMO1 had a pH optimum of 9.0, and rabbit FMO2 exhibited a pH optimum of 9.5 (56). Based on tertiary amine N-oxygenation, Lomri et al. (55) reported that human FMO3 had a pH optimum of approximately 10. However, the true pH optimum may be 8.5 since deprotonation of the amines probably influenced the apparent pH (55). A difference was observed in the effects of sodium cholate (1%) and MgCl2 (100 mM) on FMO activity (Fig. 5). Mouse FMO3 was more responsive to the addition of MgCl2 (42% increase) and sodium cholate (18% increase) than human FMO3 (marginal increase with both). However, rabbit FMO3 was inhibited by both sodium cholate and MgCl2 at the concentrations indicated (40). A primary alkylamine, n-octylamine, was examined for its effect on FMO activity (Fig. 5). With mouse FMO3, a slight increase in FMO activity was observed (8%); however, human FMO3 activity was inhibited by greater than 50%. The effect of n-octylamine on rabbit FMO3 was reported to be negligible (40); however, n-octylamine has been demonstrated to significantly inhibit human FMO3-mediated metabolism (41, 55). Chlorpromazine and imipramine, substrates for FMO1 but not FMO2 (44), and other known FMO substrates were examined for their ability to modulate metabolism of methimazole catalyzed by the expressed

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

enzymes (Table I). For these assays, the activity of the expressed enzymes was monitored for 2 min, followed by the addition of the compounds to the sample and reference cuvettes and the re-recording of the reaction rate. Results [reported as percentage inhibition of FMO-catalyzed methimazole (1.0 mM) oxidation] indicate that thioacetamide was an excellent inhibitor of both mouse and human FMO3 activity. Strong inhibition by trimethylamine, a dietary xenobiotic, was also demonstrated. Other substrates having an observable inhibitory effect on human and mouse FMO3-mediated methimazole metabolism were thiourea, thiobenzamide, and, to a lesser degree, N,N-dimethylaniline. The significant inhibition of thiourea and thiobenzamide toward metabolism of phenothiazine derivatives has been reported for expressed human FMO3 (55). Chlorpromazine (0.1 mM) and imipramine (1.0 mM) showed little or no effect on methimazole metabolism. Using methimazole as a substrate, the kinetics of FMO activity were determined for the mouse-expressed FMO3. A double-reciprocal plot of the data revealed linear kinetics, and the Km for methimazole was 56.2 mM and the Vmax was 21.7 nmol/min/mg (average of two independent determinations). In comparison, rabbit-expressed FMO3 exhibited a Km for methimazole near 30 mM and a Vmax of 21.6 (40). FMO3 expression in mouse liver, lung, and kidney. Samples (5 mg) of mouse hepatic, pulmonary, and renal mRNA were analyzed with a full-length mouse FMO3 cDNA probe (Fig. 6A). In agreement with previous findings which utilized a rabbit FMO3 probe (33), no transcripts were detected in liver samples from males;

TABLE I

Inhibition by FMO Substrates on Methimazole Oxidation Catalyzed by cDNA-Expressed FMO3a FMO activity (% of control) Substrate

(mM)

Mouse

Human

Methimazole / Thioacetamide / Trimethylamine / Thiourea / Thiobenzamide / N,N-Dimethylaniline / Chlorpromazineb / Imipramine

1.0 1.0 1.0 1.0 1.0 1.0 0.1 1.0

100 0 10 22 29 59 112 115

100 9 14 28 24 72 96 109

a Effect of substrate addition on methimazole oxidation. The rate of methimazole oxidation was monitored for 2 min, and then the substrates were added to the sample and reference cuvettes, and the reaction rate was recorded for 3 min. b 1.0 mM chlorpromazine interfered with the methimazole assay. Data presented represent the averages of two determinations, and the results differed by less than 10%.

arca

16

FALLS ET AL.

ples is apparently cross-reactive toward FMO5 (addressed above). DISCUSSION

FIG. 6. Representative blots demonstrating tissue- and gender-dependent expression of mouse FMO3. (A) Analysis of messenger RNA. Samples (5 mg) were separated by electrophoresis on a 1% formaldehyde/agarose gel and transferred to a Hybond N/ membrane. The membrane was hybridized with [a-32P]dCTP-labeled mouse FMO3 probe, and bands were detected by autoradiography. Sample loadings correspond to the following lanes: male (lane 1) and female (lane 2) mouse liver; male (lane 3) and female (lane 4) mouse kidney; and male (lane 5) and female (lane 6) mouse lung. The arrowhead indicates a transcript size of Ç2.3 kb. Transcript sizes were determined by comparison with RNA kilobase markers (Life Technologies). (B) Analysis of FMO3 protein levels detected by an anti-rabbit FMO3 antibody. Microsomal samples (lanes 1–6, 3 mg protein) were prepared from CD-1 mice and loaded as follows: lane 1, female lung; lane 2, male lung; lane 3, female kidney; lane 4, male kidney; lane 5, female liver, and lane 6, male liver. Lane 7 contained 0.1 mg of rabbit FMO3 expressed in E. coli (positive control). The arrowhead points to proteins detected at Ç58 kDa. Immunoblotting was performed as described in the legend to Fig. 3B.

however, samples from females showed a band of high intensity at 2.3 kb. FMO3 mRNA was also detected in lung samples, but no gender differences were evident; however, FMO3 was not detected in the kidney samples. In support of the Northern blot data, Western blot analysis (Fig. 6B) confirmed the tissue- and genderdependent expression of FMO3. Bands corresponding to a protein of Ç58 kDa were detected in samples from the lung (males and females) and liver (females only), but not the kidney. The minor band of higher molecular weight observed in the male and female hepatic sam-

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

A cDNA library prepared from female mouse liver mRNA was screened with a radiolabeled 3* fragment (bases 1360–2080) of rabbit FMO3 cDNA. Since the 3* ends of FMOs are the least conserved (27), this segment was chosen to selectively detect the mouse FMO3 isoform. A cDNA clone encoding a protein of 534 amino acids was identified and found to be 79 and 82% identical to human (30) and rabbit FMO3 (40, 57), respectively. In addition, the presence of putative FAD and NADP/ binding sites and the percentage identity of mouse FMO3 compared to the other FMO forms (i.e., FMO1, FMO2, FMO4, and FMO5) confirm that the reported sequence is the mouse FMO3 ortholog. Only one consensus polyadenylation signal was evident in mouse FMO3 cDNA, consistent with the Northern blot results showing one band at 2.3 kb in the mouse liver samples (females only). Multiple polyadenylation signals have been reported for the FMO3 sequence of human (54) and rabbit (40); however, consistent with our results, the detection of only one band in human (54) and rabbit (40) mRNA samples was reported. Generally, the derived amino acid sequence of the mouse FMO3 is similar to that of rabbit and human FMO3. An exception occurs in the hydrophobic region of the C-terminus, a region which diverges in peptide sequence among all FMO forms (25). Based on computer-generated predictions of transmembrane regions, it was previously speculated that the C-terminal region of FMOs was involved in membrane binding; however, Lawton and Philpot (44) found that deletion of this region did not affect functionality since the expressed mutant rabbit FMO2 protein remained membrane bound and active. Differences in catalytic properties of the FMO isozymes have been reported. Expressed rabbit FMO1 metabolizes chlorpromazine and imipramine, and its activity is activated by primary aliphatic amines (e.g., n-octylamine). Conversely, FMO2 metabolizes n-octylamine, but not imipramine or chlorpromazine. However, one report suggests that expressed rabbit FMO3 has a lower affinity for chlorpromazine than expressed rabbit FMO1 (40). Additionally, n-octylamine had virtually no effect on the inhibition of methimazole oxidation of expressed rabbit FMO3 (40). N-oxide formation of a phenothiazine derivative by cDNA-expressed human FMO3 was effectively inhibited by the addition of 4.5 mM n-octylamine, suggestive of primary amine metabolizing capability by human FMO3 (55). Based on our results and previous reports (40, 41), inhibition of methimazole oxidation by n-octylamine appears to

arca

SEQUENCING AND EXPRESSION OF MOUSE FMO3

be species-specific and may represent a functional difference relating to primary amine metabolism and/or activation among orthologous FMO3 isozymes. For example, our results showed that the expressed mouse FMO3 enzyme was slightly activated by n-octylamine, but the expressed human FMO3 enzyme was strongly inhibited by this compound. The results of expressed human FMO3 toward n-octylamine are in agreement with those of Itagaki et al. (41). For the expressed mouse and human enzymes, FMO activity was increased by metal ions and detergents; however, activity of the expressed rabbit FMO3 enzyme was inhibited (40). Itagaki et al. (41) have reported an inhibition of the expressed human FMO3 activity toward metal ions and detergents, whereas for our expressed human FMO3 enzyme, we observed relatively no change in activity. Based on reports by Nagata et al. (58) on rabbit FMO2 and pig FMO1 and on findings of Lomri et al. (55), it has been proposed that human FMO3 may be more similar to rabbit FMO2 than to pig FMO1 in terms of substrate specificity (26). Our results indicate that expressed mouse FMO3 is more similar to rabbit FMO2 than to rabbit FMO1 in terms of imipramine or chlorpromazine inhibition of methimazole metabolism, but more like rabbit FMO1 with respect to the effect of n-octylamine. However, the response of expressed human FMO3 toward n-octylamine was similar to that reported for cDNA-expressed rabbit FMO2 (44). Considerable tissue-, species-, age-, and gender-dependent expression has been reported for the various FMOs (26–28). Results presented here indicate that regulation of mouse FMO3 expression is controlled by several different mechanisms. For instance, FMO3 expression in the liver is under strict hormonal control (33, 34); however, FMO3 expression in the male and female lung is not dependent on gender, and FMO3 expression in the kidney (of either gender) was not evident. The absence of FMO3 in the kidney is interesting since this finding contrasts with a previous study which reported, based on mRNA levels, expression of FMO3 in mouse kidney (40). One explanation for this discrepancy may reside in the use of a cross-species probe and the possible detection of multiple FMO forms (i.e., rabbit cDNA hybridized to mouse mRNA); however, this does not seem to be the case. Previously, we have used rabbit probes to examine hepatic FMO mRNA expression in the mouse, and multiple transcripts were not detected. In addition, a rabbit FMO3 cDNA fragment (720 bp of the 3* end) was used successfully to identify the mouse FMO3 cDNA clones reported in this study. Furthermore, FMO3 was not detected in kidney microsomal samples on Western blots reacted with an anti-FMO3 antibody, and the mouse FMO3 cDNA probe did not detect a transcript in the kidney. Thus, based on our results, it appears that FMO3 is either

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

17

not present in mouse kidney or is expressed at very low levels or its expression is strain-specific. Of particular interest was the strong inhibition of methimazole oxidation following the addition of trimethylamine. FMO3 is thought to be the isozyme responsible for the human disease trimethylaminuria (28). Based on our results using human- and mouseexpressed FMO3, inhibition of methimazole oxidation by trimethylamine (and other FMO substrates) was similar. In addition, FMO3 mRNA and protein expression is high in the female mouse liver (33) as well as mRNA expression in the adult human liver (30). However, the male mouse liver expresses high levels of FMO1, but does not express FMO3 (33), which is similar to the FMO isozyme profile found in the human fetal liver (30). Based on previous reports demonstrating the relatively low hepatic mRNA expression of FMO3 in several species (40), from a toxicological standpoint the female mouse may be a more appropriate animal model relevant to humans. Coupled with in vitro metabolism studies using expressed FMO isozymes, in vivo studies using male and female mice should aid in predicting the mechanisms of FMO-mediated xenobiotic metabolism in humans. In conclusion, mouse FMO3 has been sequenced and the derived amino acid sequence is 82 and 79% identical to the rabbit and human forms, respectively. In addition, mouse FMO3 expression is both tissue and gender dependent. Recombinant protein expression studies presented here indicate that, with the exception of n-octylamine, mouse and human FMO3 inhibition profiles are similar toward many FMO substrates (i.e., thioacetamide, trimethylamine, thiourea, thiobenzamide, and N,N-dimethylaniline). We are currently investigating the metabolism of other xenobiotics by the mouse- and human-expressed FMO3 isozymes. ACKNOWLEDGMENTS This project was supported in part by PHS Grant ES00044. The authors thank Lila Overby (NIEHS, Research Triangle Park, NC) for the production of antibodies and pJL-FMO3h. Yan Cao (NCSU, Raleigh, NC) is appreciated for his technical support.

REFERENCES 1. Ziegler, D. M. (1993) Annu. Rev. Pharmacol. Toxicol. 33, 179– 199. 2. Hodgson, E., Blake, B. L., Levi, P. E., Genter, M. B., Mailman, R. B., Lawton, M. P., and Philpot, R. M. (1995) in Molecular Aspects of Oxidative Drug Metabolizing Enzymes: Their Significance in Environmental Toxicology, Chemical Carcinogenesis and Health (Arinc, E., Schenkman, J. B., and Hodgson, E., Eds.), pp. 225–236, Springer Verlag, Berlin. 3. Hodgson, E., Rose, R. L., Adams, N. H., Deamer, N. J., Genter, M. B., Venkatesh, K., and Levi, P. E. (1995) in Molecular Aspects of Oxidative Drug Metabolizing Enzymes: Their Significance in Environmental Toxicology, Chemical Carcinogenesis and Health

arca

18

4.

5. 6.

7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 23.

24. 25.

26. 27.

28.

FALLS ET AL. (Arinc, E., Schenkman, J. B., and Hodgson, E., Eds.), pp. 461– 478, Springer Verlag, Berlin. Ziegler, D. M. (1988) in Microsomes and Drug Oxidations (Minors, J., Birkett, D. J., Drew, R., and McManus, M., Eds.), pp. 297–304, Taylor and Francis, London. Hodgson, E., Rose, R. L., Ryu, D.-Y., Falls, J. G., Blake, B. L., and Levi, P. E. (1995) Toxicol. Lett. 82/83, 73–81. Ziegler, D. M., Poulsen, L. L., and Duffel, M. W. (1990) in Microsomes, Drug Oxidations, and Chemical Carcinogenesis (Coon, M., Conney, A., Estabrook, R., Gelboin, H., Gillette, J., and O’Brian, P., Eds.), Vol. II, pp. 637–645, Academic Press, New York. Ziegler, D. M. (1980) in Enzymatic Basis of Detoxication (Jakoby, W. B., Ed.), Vol. I, pp. 201–227, Academic Press, New York. Elfarra, A. A. (1995) Chem.–Biol. Interact. 96, 47–56. Duescher, R. J., Lawton, M. P., Philpot, R. M., and Elfarra, A. A. (1994) J. Biol. Chem. 269, 17525–17530. Al Waiz, M., Ayesh, R., Mitchell, S. C., Idle, J. R., and Smith, R. L. (1987) Clin. Pharmacol. Ther. 42, 588–594. Shelley, E. D., and Shelley, W. B. (1984) J. Am. Med. Assoc. 251, 253–255. Ziegler, D. M., and Mitchell, C. H. (1972) Arch. Biochem. Biophys. 150, 116–125. Kimura, T., Kodama, M., and Nagata, C. (1983) Biochem. Biophys. Res. Commun. 110, 640–645. Sabourin, P. J., Smyser, B. P., and Hodgson, E. (1984) Int. J. Biochem. 16, 713–720. Williams, D. E., Hale, S. E., Muerhoff, A. S., and Masters, B. S. S. (1984) Mol. Pharmacol. 28, 381–390. Tynes, R. E., Sabourin, P. J., and Hodgson, E. (1985) Biochem. Biophys. Res. Commun. 126, 1069–1075. Yamada, H., Yuno, K., Oguri, K., and Yoshimura, H. (1990) Arch. Biochem. Biophys. 280, 305–312. Sadeque, A. J. M., Thummel, K. E., and Rettie, A. E. (1993) Biochim. Biophys. Acta 1162, 127–134. Devereux, T. R., Philpot, R. M., and Fouts, J. R. (1977) Chem.– Biol. Interact. 15, 277–287. Ohmiya, Y., and Mehendale, H. M. (1981) Pharmacology 22, 172–182. Ohmiya, Y., and Mehendale, Hh;1285. Williams, D. E., Ziegler, D. M., Nordin, D. J., Hale, S. E., and Masters, B. S. S. (1984) Biochem. Biophys. Res. Commun. 125, 116–122. Lawton, M. P., Gasser, R., Tynes, R. E., Hodgson, E., and Philpot, R. M. (1990) J. Biol. Chem. 265, 5855–5861. Lawton, M. P., Cashman, J. R., Cresteil, T., Dolphin, C. T., Elfarra, A. A., Hines, R. N., Hodgson, E., Kimura, T., Ozols, J., Phillips, I. R., Philpot, R. M., Poulsen, L. L., Rettie, A. E., Shephard, E. A., Williams, D. E., and Ziegler, D. M. (1994) Arch. Biochem. Biophys. 308, 254–257. Cashman, J. R. (1995) Chem. Res. Toxicol. 8(2) 165–181. Lawton, M. P., and Philpot, R. M. (1995) in Reviews in Biochemical Toxicology (Hodgson, E., Bend, J. R., and Philpot, R. M., Eds.), pp. 1–27, Toxicology Communications, Raleigh, NC. Phillips, I. R., Dolphin, C. T., Clair, P., Hadley, M. R., Hutt, A. J., McCombie, R. R., Smith, R. L., and Shepard, E. A. (1995) Chem.– Biol. Interact. 96, 17–32.

AID

ABB 0322

/

6b41$$$221

09-29-97 13:19:11

29. Nikbakht, K. N., Lawton, M. P., and Philpot, R. M. (1992) Pharmacogenetics 2, 207–216. 30. Dolphin, C. T., Cullingford, T. E., Shephard, E. A., Smith, R. L., and Phillips, I. R. (1996) Eur. J. Biochem. 235, 683–689. 31. Wirth, P. J., and Thorgeirsson, S. S. (1978) Biochem. Pharmacol. 27, 601–603. 32. Duffel, M. W., Graham, J. M., and Ziegler, D. M. (1980) Mol. Pharmacol. 19, 134–139. 33. Falls, J. G., Blake, B. L., Cao, Y., Levi, P. E., and Hodgson, E. (1995) J. Biochem. Toxicol. 10, 171–177. 34. Falls, J. G., Ryu, D.-Y., Cao, Y., Levi, P. E., and Hodgson, E. (1997) Arch. Biochem. Biophys. 342, 212–223. 35. Venkatesh, K., Levi, P. E., and Hodgson, E. (1991) Biochem. Pharmacol. 42, 1411–1420. 36. Venkatesh, K., Blake, B., Levi, P. E., and Hodgson, E. (1992) J. Biochem. Toxicol. 7, 163–169. 37. Chomcynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156– 158. 38. Celano, P., Vertino, P. M., and Casero, R. A. (1993) Biotechniques 15, 26–28. 39. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463–5467. 40. Burnett, V. L., Lawton, M. P., and Philpot, R. M. (1994) J. Biol. Chem. 269(19), 14314–14322. 41. Itagaki, K., Carver, G. T., and Philpot, R. M. (1996) J. Biol. Chem. 271, 20102–20107. 42. Overby, L. H., Buckpitt, A. R., Lawton, M. P., Atta-Asafo-Adjei, E., Schulze, J., and Philpot, R. M. (1995) Arch. Biochem. Biophys. 317, 275–284. 43. Pernecky, S. J., Larson, J. R., Philpot, R. M., and Coon, M. J. (1993) Proc. Natl. Acad. Sci. USA 90, 2651–2655. 44. Lawton, M. P., and Philpot, R. M. (1993) J. Biol. Chem. 268, 5728–5734. 45. Laemmli, U. K. (1970) Nature 227, 680–685. 46. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci USA 76, 4350–4354. 47. Blake, B. L., Rose, R. L., Mailman, R. B., Levi, P. E., and Hodgson, E. (1995) Xenobiotica 25, 377–393. 48. Dixit, A., and Roche, T. E. (1984) Arch. Biochem. Biophys. 233, 50–63. 49. Bradford, M. M. (1976) Anal. Biochem. 72, 248–254. 50. Faeder, E. J., and Siegel, L. M. (1973) Anal. Biochem. 53, 332– 336. 51. Rave, N., Crkvenjakov, R., and Boedtker, H. (1976) Nucleic Acids Res. 6, 3559–3567. 52. Kozak, M. (1987) Nucleic Acids Res. 15, 8123–8133. 53. Breathnack, R., and Chambon, P. (1981) Annu. Rev. Biochem. 50, 349–383. 54. Lomri, N., Gu, Q., and Cashman, J. R. (1992) Proc. Natl. Acad. Sci. 89, 1685–1689. 55. Lomri, N., Yang, Z., and Cashman, J. R. (1993) Chem. Res. Toxicol. 6, 425–429. 56. Lawton, M. P., Kronbach, T., Johnson, E. F., and Philpot, R. M. (1991) Mol. Pharmacol. 40, 692–698. 57. Ozols, J. (1991) Arch. Biochem. Biophys. 290, 103–115. 58. Nagata, T., Williams, D. E., and Ziegler, D. M. (1990) Chem. Res. Toxicol. 3, 372–376.

arca