Portal de Periódicos da CAPES

Influence of Flavin Analogue Structure on the Catalytic Activities and Flavinylation Reactions of Recombinant Human Liver Monoamine Oxidases A and B

1999; Elsevier BV; Volume: 274; Issue: 33 Linguagem: Inglês

10.1074/jbc.274.33.23515

1083-351X

J. Richard Miller, Dale E. Edmondson,

Biochemical and Molecular Research

Two riboflavin-deficient (rib 5)Saccharomyces cerevisiae expression systems have been developed to investigate the influence of riboflavin structural alterations on the covalent flavinylation reaction and activity of recombinant human liver monoamine oxidases A and B (MAO A and B). Nineteen different riboflavin analogues were tested with MAO A and nine with MAO B. MAO expression and flavinylation were determined immunochemically with antisera to MAO and an anti-flavin antisera. Expression levels of both MAO A and B are invariant with the presence or absence of riboflavin or riboflavin analogues in the growth medium. Flavin analogues with a variety of seven and eight substitutions are found to be covalently incorporated and to confer catalytic activity. The selectivities of MAO A and MAO B for flavin analogue incorporation are found to be similar, although 8α-methylation of the flavin resulted in a higher level of catalytic activity for MAO B than for MAO A. N (3)-Methylriboflavin and 8-nor-8-aminoriboflavin are not covalently bound as they are not converted to their respective FAD forms by yeast. 5-Carba-5-deazaflavin and 7,8-nor-7-chlororiboflavin are not covalently incorporated into MAO A and do not support catalytic activity. A flavin peptide was isolated from MAO A containing 7-nor-7-bromo-FAD and was demonstrated to be covalently attached to Cys-406 by an 8α-S -thioether linkage by sequence analysis and by matrix-assisted laser desorption ionization time of flight mass spectroscopy. MAO A partially purified from yeast grown on 8-nor-8-chlororiboflavin exhibited an absorption spectrum indicating the covalent flavin is an 8-nor-8-S -thioflavin, suggesting a nucleophilic displacement mechanism that supports the quinone-methide mechanism previously suggested as a general mechanism for covalent flavin attachment. Two riboflavin-deficient (rib 5)Saccharomyces cerevisiae expression systems have been developed to investigate the influence of riboflavin structural alterations on the covalent flavinylation reaction and activity of recombinant human liver monoamine oxidases A and B (MAO A and B). Nineteen different riboflavin analogues were tested with MAO A and nine with MAO B. MAO expression and flavinylation were determined immunochemically with antisera to MAO and an anti-flavin antisera. Expression levels of both MAO A and B are invariant with the presence or absence of riboflavin or riboflavin analogues in the growth medium. Flavin analogues with a variety of seven and eight substitutions are found to be covalently incorporated and to confer catalytic activity. The selectivities of MAO A and MAO B for flavin analogue incorporation are found to be similar, although 8α-methylation of the flavin resulted in a higher level of catalytic activity for MAO B than for MAO A. N (3)-Methylriboflavin and 8-nor-8-aminoriboflavin are not covalently bound as they are not converted to their respective FAD forms by yeast. 5-Carba-5-deazaflavin and 7,8-nor-7-chlororiboflavin are not covalently incorporated into MAO A and do not support catalytic activity. A flavin peptide was isolated from MAO A containing 7-nor-7-bromo-FAD and was demonstrated to be covalently attached to Cys-406 by an 8α-S -thioether linkage by sequence analysis and by matrix-assisted laser desorption ionization time of flight mass spectroscopy. MAO A partially purified from yeast grown on 8-nor-8-chlororiboflavin exhibited an absorption spectrum indicating the covalent flavin is an 8-nor-8-S -thioflavin, suggesting a nucleophilic displacement mechanism that supports the quinone-methide mechanism previously suggested as a general mechanism for covalent flavin attachment. monoamine oxidase succinate dehydrogenase 2-(N -cyclohexylamino)-ethanesulfonic acid yeast riboflavin synthase gene matrix-assisted laser desorption ionization time of flight mass spectroscopy high performance liquid chromatography kilobase pair polymerase chain reaction polyacrylamide gel electrophoresis Monoamine oxidases A and B (MAO A and MAO B)1 (EC 1.4.3.4) are homodimeric flavoenzymes found in the outer membrane of mitochondria of higher eukaryotes and are expressed in a tissue-specific manner (2Weyler W. Hsu Y.P. Breakefield X.O. Pharmacol. Ther. 1990; 47: 391-417Crossref PubMed Scopus (292) Google Scholar). These enzymes catalyze the oxidation of primary, secondary, and tertiary amines to imines with concomitant reduction of oxygen to hydrogen peroxide (3Kearney E.B. Salach J.I. Walker W.H. Seng R. Singer T.P. Biochem. Biophys. Res. Commun. 1971; 42: 490-496Crossref PubMed Scopus (46) Google Scholar). The physiological role of MAO B is to oxidize exogenous and endogenous amines, many of which could mimic the function of neurotransmitters. MAO A oxidizes neurotransmitters such as serotonin, dopamine, and norepinephrine. The structures and mechanisms of these enzymes are currently being investigated. The gene sequences of both MAO A and B from human liver have been determined to be ∼70% identical (4Bach A.W. Lan N.C. Johnson D.L. Abell C.W. Bembenek M.E. Kwan S.W. Seeburg P.H. Shih J.C. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 4934-4938Crossref PubMed Scopus (713) Google Scholar). A covalently bound FAD cofactor is present in either enzyme, attached to a conserved cysteinyl residue (Cys-406 in MAO A and Cys-397 in MAO B) via an 8α-S -cysteinyl flavin linkage (Scheme FS1). Neither the need for covalent FAD attachment nor the mechanism of covalent incorporation is known. Covalent attachment of flavin cofactors has been found in more than 20 proteins that catalyze a variety of reactions in both prokaryotes and eukaryotes (5Edmondson D.E. DeFrancesco R. Muller F. Chemistry and Biochemistry of Flavoenzymes. 1. CRC Press, Boca Raton, FL1991: 73-104Google Scholar). The processes by which flavins are incorporated into proteins appear to differ and to vary in complexity. The simplest known system is the heterodimeric p -cresol-methylhydroxylase fromPseudomonas putida . The enzyme is capable of autoflavinylation in vitro (formation of an 8α-O -tyrosyl-FAD), requiring only the recombinantly produced purified flavin-binding subunit, a separately expressed and purified heme-containing subunit, and FAD (6Kim J. Fuller H. Kuusk V. Cunane L. Chen Z. Matthews F.S. McIntire W.S. J. Biol. Chem. 1995; 270: 31202-31209Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar). Flavinylation of the mitochondrial matrix enzyme succinate dehydrogenase to form 8α-N (3)-histidyl-FAD appears to be more complex than that of p -cresol-methylhydroxylase. In vitro studies indicate that flavinylation requires the presence of other factors including molecular chaperones, ATP, and effector molecules such as succinate, fumarate, or malate (7Robinson K.M. Lemire B.D. J. Biol. Chem. 1996; 271: 4055-4060Abstract Full Text Full Text PDF PubMed Scopus (69) Google Scholar). These additional factors have been suggested to aid in folding the enzyme into a conformation that permits covalent attachment of the flavin (8Robinson K.M. Lemire B.D. J. Biol. Chem. 1996; 271: 4061-4067Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar). Flavinylation of MAO B has been examined by its expression in riboflavin-depleted COS-7 cells (9Zhou B.P. Lewis D.A. Kwan S.W. Abell C.W. J. Biol. Chem. 1995; 270: 23653-23660Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar, 10Zhou B.P. Lewis D.A. Kwan S.W. Kirksey T.J. Abell C.W. Biochemistry. 1995; 34: 9526-9531Crossref PubMed Scopus (25) Google Scholar). Electroporation of these cells with a cDNA encoding MAO B and either FAD or 8α-hydroxy-FAD results in the production of catalytically active enzyme. In the absence of added FAD, the inactive apoenzyme form of MAO B is produced. These results have been suggested to support a flavinylation mechanism for MAO B involving additional factors that would process and activate the 8α position of the flavin prior to covalent attachment (9Zhou B.P. Lewis D.A. Kwan S.W. Abell C.W. J. Biol. Chem. 1995; 270: 23653-23660Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar, 11Decker K. Muller F. CRC Chemistry and Biochemistry of Flavoproteins. 2. CRC Press, Boca Raton, FL1991: 343-375Google Scholar). Flavin analogues that have modifications in the isoalloxazine ring or in the ribityl side chain have been used extensively to address structure-activity and mechanistic questions in flavoproteins containing non-covalently bound flavin coenzymes (12Ghisla S. Massey V. Biochem. J. 1986; 239: 1-12Crossref PubMed Scopus (171) Google Scholar). The application of this approach to those flavoenzymes containing covalently bound flavins has not been extensively investigated due to the technical limitations involved in introducing flavin analogues into a covalent linkage with the protein. The pioneering work of Lambooy and co-workers (13Kim Y.S. Lambooy J.P. Proc. Soc. Exp. Biol. Med. 1978; 157: 466-471Crossref PubMed Scopus (6) Google Scholar, 14Dix B.A. Lambooy J.P. J. Nutr. 1981; 111: 1397-1402Crossref PubMed Scopus (8) Google Scholar, 15Kim Y.S. Lambooy J.P. Int. J. Biochem. 1985; 17: 975-981Crossref PubMed Scopus (3) Google Scholar) first demonstrated the possibility of using flavin analogues as probes of structure and mechanism in this subclass of flavoenzymes. These studies involved feeding flavin analogues with modifications in the 7 and 8 positions of the flavin ring to riboflavin-deficient rats. The activities of MAO and succinate dehydrogenase (SDH) were determined and compared with riboflavin-fed control animals. Incorporation of 8α-methylriboflavin and 7α-methylriboflavin into succinate dehydrogenase and into MAO was shown using 14C-labeled analogues. Some catalytic activity was observed for both enzymes when either flavin analogue or no flavin analogue was present in the diet, although the level of activity depended on which riboflavin analogue was used and which tissue sample was analyzed. This work was limited in that it was not possible to isolate sufficient quantities of enzyme for detailed kinetic and structural analysis. Additionally, the number of analogues that could be tested was limited due to the anti-vitaminic activity of some flavins in whole animals. Later studies (16Oltmanns O. Lingens F. Z. Naturforsch. 1967; 226: 751-754Crossref Scopus (29) Google Scholar) investigating the covalent incorporation of flavin analogues into succinate dehydrogenase used yeast strains deficient in the biosynthesis of riboflavin. This allowed the use of several flavin analogues as probes for their ability to be incorporated and function catalytically in yeast SDH (17Oestreicher G. Grossman S. Goldenberg J. Kearney E.B. Edmondson D.E. Singer T.P. Lambooy J.P. Comp. Biochem. Physiol. 1980; 67B: 395-402Google Scholar). Preparations of yeast succinate dehydrogenase were catalytically active with either 7α-methylriboflavin or 8α-methylriboflavin. The 8α-methylriboflavin analogue was demonstrated to be covalently incorporated into SDH by isolation and analysis of the respective flavin-containing peptide (17Oestreicher G. Grossman S. Goldenberg J. Kearney E.B. Edmondson D.E. Singer T.P. Lambooy J.P. Comp. Biochem. Physiol. 1980; 67B: 395-402Google Scholar). The recent successful expression of the human liver MAO A and MAO B genes in Saccharomyces cerevisiae (18Urban P. Andersen J.K. Hsu H.P. Pompon D. FEBS Lett. 1991; 286: 142-146Crossref PubMed Scopus (32) Google Scholar, 19Weyler W. Titlow C.C. Salach J.I. Biochem. Biophys. Res. Commun. 1990; 173: 1205-1211Crossref PubMed Scopus (48) Google Scholar) prompted us to develop a riboflavin-deficient yeast strain as a system to study the incorporation of flavin analogues as substitutes for the normal flavin in MAO A and B. Such a system would allow experimental approaches to probe the effects of structural alterations of the flavin coenzyme on the catalytic properties of either MAO A or MAO B as well as the structural requirements of the flavin molecule necessary for covalent flavin incorporation into these enzymes. By separately expressing MAO A or MAO B, such a system avoids the difficulties of interpretation of results from the animal system, where both MAO A and MAO B are expressed in most tissues. For those flavin analogues that are covalently incorporated into MAO, new insights into the mechanistic role of the flavin coenzyme in catalysis might be determined. With the yeast expression system it should be possible to isolate sufficient quantities of MAO to allow structural and kinetic analysis of the enzyme forms produced with various flavin analogues and possibly provide new insights into the mechanistic role of the flavin in catalysis. This work describes two systems for the covalent incorporation of flavin analogues with alterations in the isoalloxazine ring or in the ribityl side chain into MAO A or MAO B. Flavinylation is monitored by the expression of catalytically functional enzymes and from the use of flavin-specific antisera as described by Barber et al . (20Barber M.J. Eichler D.C. Solomonson L.P. Ackrell B.A. Biochem. J. 1987; 242: 89-95Crossref PubMed Scopus (28) Google Scholar). Those flavin analogues whose structures prevented covalent incorporation or were not converted to their FAD coenzyme form failed to support any catalytic activity with either enzyme. Conversely, all flavin analogues that were covalently bound to either enzyme did support catalytic activity. Structural analysis of MAO A expressed in the presence of 7-nor-7-bromoriboflavin demonstrated that covalent incorporation occurred at Cys-406 at the 8α position of the flavin. Flavin analogues with 8-nor-8-halogen substituents were incorporated into either enzyme as the 8-nor-8-S -thioflavin. Both MAO A and MAO B exhibit similar specificities for flavin analogue incorporation. Genotypes of all yeast strains are shown in Table I. S. cerevisiae strain W303.1B (MAT aleu 2 his 3 trp 1 ura 3ade 2-1 canR cyr+) was the kind gift of Dr. Philippe Urban (CNRS, Gif-sur-Yvette, France) (18Urban P. Andersen J.K. Hsu H.P. Pompon D. FEBS Lett. 1991; 286: 142-146Crossref PubMed Scopus (32) Google Scholar). RH218 (MAT a trp 1 gal 2 mel mar ciro) was generously provided by Dr. James Salach (Veterans Administration Medical Center, San Francisco) (21Miozzari G. Neiderberger P. Huffer R. J. Bacteriol. 1978; 134: 48-54Crossref PubMed Google Scholar). RM1 (MAT a trp 1 gal 2 mel mar ciro, ura 3Δ::his G) was created from RH218 by targeted disruption of the URA3 locus. RM2 (MAT a leu 2 his 3 trp 1ura 3 ade 2–1 canR rib 5Δ::his G) and RM3 (MAT atrp 1 gal 2 mel mar ciro ura 3Δ::his Grib 5Δ::URA 3) are riboflavin auxotrophic strains derived from W303.1B and RM1, respectively. RM2 yeast were grown in media supplemented with 50 mg/liter riboflavin or flavin analogue during growth and expression of MAO. Media were prepared as described (18Urban P. Andersen J.K. Hsu H.P. Pompon D. FEBS Lett. 1991; 286: 142-146Crossref PubMed Scopus (32) Google Scholar, 19Weyler W. Titlow C.C. Salach J.I. Biochem. Biophys. Res. Commun. 1990; 173: 1205-1211Crossref PubMed Scopus (48) Google Scholar). RM3 yeast cultures were grown and induced in media supplemented with 50 mg/liter riboflavin or flavin analogue. All cultures containing free riboflavin were protected from light at all times to prevent flavin photodegradation and any degradative effects on the cells.Table IGenotypes of S. cerevisiae strains used in this studyStrainGenotypeATCC 225501-aFrom Ref. 53.MATarib 7–3 ade 2RM100MATa rib 7–3ade 2 ura 3W303.1B1-bFrom Ref. 18.MATaleu 2 his 3 trp 1 ura 3ade 2–1 canR cyr+RH2181-cFrom Ref. 21.MATatrp 1 gal 2 mel mar cir 0RM1MATa trp 1 gal 2 mel mar cir 0 ura 3Δ∷his GRM2MATaleu 2 his 3 trp 1 ura 3ade 2–1 canR cyr+ rib 5Δ∷his GRM3MATa trp 1gal 2 mel mar cir 0 ura 3Δ∷his Grib 5Δ∷URA 31-a From Ref. 53Bacher A. Lingens F. J. Biol. Chem. 1971; 246: 7018-7022Abstract Full Text PDF PubMed Google Scholar.1-b From Ref. 18Urban P. Andersen J.K. Hsu H.P. Pompon D. FEBS Lett. 1991; 286: 142-146Crossref PubMed Scopus (32) Google Scholar.1-c From Ref. 21Miozzari G. Neiderberger P. Huffer R. J. Bacteriol. 1978; 134: 48-54Crossref PubMed Google Scholar. Open table in a new tab TheURA 3/ADE 2 plasmids pYeDP60HMAOA and pYeDP60HMAOB, used to express MAO A and MAO B in S. cerevisiae strains W303.1B and RM2, were the generous gift of Dr. Philippe Urban and have been described previously (18Urban P. Andersen J.K. Hsu H.P. Pompon D. FEBS Lett. 1991; 286: 142-146Crossref PubMed Scopus (32) Google Scholar). Higher levels of MAO A expression were achieved using pGPD(G)-2MA3′ut (kindly provided by Dr. James Salach) in strains RH218 (19Weyler W. Titlow C.C. Salach J.I. Biochem. Biophys. Res. Commun. 1990; 173: 1205-1211Crossref PubMed Scopus (48) Google Scholar) and RM3. Disruption of the URA 3 locus to produce strain RM1 was accomplished by transformation of strain RH218 with the 4.6-kbHin dIII fragment of pSR299 (gift of Dr. Sue Jinks-Robertson, Emory University) which contains a deletion mutant of theURA 3 coding sequence. Transformants were selected for their ability to survive in the presence of 5-fluoroorotic acid. Strains RM1 and W303.1B were made auxotrophic for riboflavin through replacement of portions of the promoter and the N-terminal coding region of theRIB 5 gene with a functional copy of the URA 3 gene. This gene disruption cassette was created through polymerase chain reaction amplification of the RIB 5 gene and 1 kb of flanking sequences on either side of the gene. The PCR template was total genomic DNA from strain W303.1B using oligonucleotides RIB5FWD (5′-AACTGTTTGTCTGCACTCTCTAACT) and RIB5REV (5′-TTCAGCGTCCTGAGAAGCATTCGG). The 2.6-kb PCR product was subcloned using the TA cloning kit (Invitrogen) to give vector pCRII-Rib5. Partial digestion of pCRII-Rib5 with Eco RI resulted in a 2.6-kb band that was subsequently cloned into the Eco RI site of pGem7 (Promega) to give pGem7-Rib5. Digestion of pGem7-Rib5 withKpn I and Eco RV removed the promoter elements and 364 base pairs of the N-terminal coding sequence of Rib5 that was replaced with the 5.14-kb Kpn I-Eco RV partial digestion product of pHUKH1 (22Earley M.C. Crouse G.F. Gene (Amst.). 1996; 169: 111-113Crossref PubMed Scopus (10) Google Scholar) (gift of Dr. Gray Crouse, Emory University) to give pΔRib5-KanR. The kanamycin resistance cassette of pΔRib5-KanR was excised by Pst I digestion and the plasmid religated to give pΔRib5. Strains W303.1B and RM2 were transformed with the 6.1-kbXba I-Sac I fragment of pΔRib5 and transformants selected for uracil prototrophy on plates containing riboflavin (50 mg/liter). Riboflavin auxotrophs selected by this procedure were confirmed by PCR of genomic DNA in addition to their inability to grow in the absence of riboflavin. Uracil auxotrophy was recovered in strain RM2 through excision of the URA 3 gene via his G repetitive sequences flanking the URA3 gene (23Alani E. Cao L. Kleckner N. Genetics. 1987; 116: 541-545Crossref PubMed Scopus (769) Google Scholar). Selection for gene excision was accomplished after growth in rich media for several generations followed by plating on 5-fluoroorotic acid containing synthetic media supplemented with riboflavin (50 mg/liter). Riboflavin, tritiated at the 5′-hydroxymethyl (12.6 mCi/mmol), was synthesized as described (24Kekelidze T.N. Edmondson D.E. McCormick D.B. J. Labelled Comp. Radiopharm. 1995; 36: 953-960Crossref Scopus (7) Google Scholar). RM2 yeast were grown to late logarithmic phase in the presence of 20 mg/liter of the tritiated cofactor. Total non-covalently bound flavin coenzymes were isolated (25Fazekas A.G. Kokai K. Methods Enzymol. 1971; 18B: 385-398Crossref Scopus (45) Google Scholar) and chromatographed on reverse-phase HPLC (24Kekelidze T.N. Edmondson D.E. McCormick D.B. J. Labelled Comp. Radiopharm. 1995; 36: 953-960Crossref Scopus (7) Google Scholar) using authentic unlabeled standards (riboflavin, FMN, and FAD) as controls. Fractions were collected, and radioactivity in each fraction was determined by scintillation counting. Cytosol (60–70 mg of protein in 8 ml) was isolated from lysates of W303.1B or RM2 yeast as described (18Urban P. Andersen J.K. Hsu H.P. Pompon D. FEBS Lett. 1991; 286: 142-146Crossref PubMed Scopus (32) Google Scholar). Small molecules were separated from protein material on a Bio-Gel P2-DG (Bio-Rad) gel filtration column (2.5 × 20 cm) equilibrated with 25 mm Tris base, pH 7.4, 1 mmEDTA, 0.6 m sorbitol. Fractions (5.4 ml) were collected and analyzed for fluorescence (λex 267 nm, λem525 nm). Those exhibiting fluorescence were purified and concentrated using a C18 Sep-Pak cartridge (Waters Associates). Apoglucose oxidase was prepared from 80 mg (total weight) of theAspergillus niger holoenzyme (Sigma type V) as described (26Sanner C. Macheroux P. Ruterjans H. Muller F. Bacher A. Eur. J. Biochem. 1991; 196: 663-672Crossref PubMed Scopus (41) Google Scholar) with the following changes. Apoenzyme eluting from the desalting column was collected in 3-ml fractions into 9 ml of 0.5 mpotassium phosphate, pH 7.0. Fractions containing significant absorbance at 280 nm were pooled and concentrated to 10 ml followed by overnight dialysis against 4 liters of 50 mm potassium phosphate, pH 6.8. The final solution (1 mg/ml protein) was stored on ice and was stable for several weeks. The apoenzyme exhibited less than 1% of the glucose oxidase activity of the FAD-reconstituted enzyme. Reconstitution assays utilized a modification of the peroxidase-coupled MAO assay described by Holt et al. (27Holt A. Sharman D.F. Baker G.B. Palcic M.M. Anal. Biochem. 1997; 244: 384-392Crossref PubMed Scopus (220) Google Scholar) and were conducted with either commercial FAD or total flavins purified from yeast cell lysates. Apoglucose oxidase (5 μg) was added to a FAD solution or a solution of flavins purified from yeast in 50 mm potassium phosphate, pH 7.0, with 0.1% (w/v) bovine serum albumin. After overnight incubation at room temperature, samples were assayed by hydrogen peroxide formation upon glucose addition. Assays were conducted at 30 °C in 50 mm potassium phosphate, pH 7.0, 0.5 m d-glucose, 1% (w/v) bovine serum albumin, 4 mm aminoantipyrine, 8 mm vanillic acid, and 40 μg/ml peroxidase (Sigma). The rate of formation of a colored complex between peroxidase-oxidized 4-aminoantipyrine and vanillic acid (λmax 498 nm E a = 4654m−1 cm−1) (27Holt A. Sharman D.F. Baker G.B. Palcic M.M. Anal. Biochem. 1997; 244: 384-392Crossref PubMed Scopus (220) Google Scholar) was determined and equated to the rate of glucose oxidase catalyzed hydrogen peroxide formation. MAO A and MAO B were expressed inS. cerevisiae strain RM2 containing plasmids pYeDP60HMAOA or pYeDP60HMAOB as described (18Urban P. Andersen J.K. Hsu H.P. Pompon D. FEBS Lett. 1991; 286: 142-146Crossref PubMed Scopus (32) Google Scholar) with the addition of 50 mg/liter riboflavin during growth. Cells were washed with distilled water prior to incubation in galactose-containing medium. Mitochondria were isolated as described (18Urban P. Andersen J.K. Hsu H.P. Pompon D. FEBS Lett. 1991; 286: 142-146Crossref PubMed Scopus (32) Google Scholar) with the addition of protease inhibitors (2 μg/ml aprotinin, 0.5 μg/ml leupeptin, 1 μg/ml pepstatin A, 0.1 mm benzamidine, and 0.1 mm sodium metabisulfite) prior to spheroplast lysis. Solubilization of MAO from outer mitochondrial membranes was performed by dilution of total mitochondrial protein to 2 mg/ml with solubilization buffer (50 mm Tris-HCl, pH 7.4, 2 mm EDTA, 0.6m sorbitol, 25% glycerol) followed by the addition of an equal volume of 0.8 mg/ml digitonin (Sigma) in solubilization buffer. Samples were incubated on a tube rotator at 4 °C for 30 min and centrifuged at 9000 × g for 15 min at 4 °C to remove insoluble material. Supernatants were decanted and centrifuged at 252,000 × g for 45 min at 4 °C producing a bright yellow pellet that contained highly enriched MAO. This pellet was resuspended by homogenization in solubilization buffer and diluted with the same to a protein concentration of 0.2 mg/ml. Expression of MAO A in S. cerevisiae strain RM3 was conducted in two 12-liter fermenters (New Brunswick Scientific) as described (19Weyler W. Titlow C.C. Salach J.I. Biochem. Biophys. Res. Commun. 1990; 173: 1205-1211Crossref PubMed Scopus (48) Google Scholar) with the addition of 50 mg/liter riboflavin or flavin analogue during growth and induction. Protein concentrations in either crude yeast mitochondrial fractions or in purified protein preparations were determined using a slight modification of the method of Bearden (28Bearden J.C. Biochim. Biophys. Acta. 1978; 553: 525-529Crossref Scopus (405) Google Scholar). Suspensions of crude mitochondria or digitonin-solubilized protein fractions were diluted with sodium hydroxide (final concentration, 10 mm) prior to colorimetric protein determination. This treatment resulted in more reproducible protein determinations due to more efficient protein solubilization. The kynuramine oxidase activity of MAO A was determined fluorometrically at 30 °C as described (29Krajl M. Biochem. Pharmacol. 1965; 14: 1684-1686Crossref PubMed Scopus (381) Google Scholar) with the exception that 2% (w/v) reduced Triton X-100 was added to the assay buffer. Benzylamine oxidase activity of MAO B was determined at 30 °C using a peroxidase-linked spectrophotometric assay (27Holt A. Sharman D.F. Baker G.B. Palcic M.M. Anal. Biochem. 1997; 244: 384-392Crossref PubMed Scopus (220) Google Scholar) and was measured at 1 mm O2. This condition gave more reproducible assay values since this O2 concentration is approximately four times the K mO2 of bovine liver MAO B (30Walker M.C. Edmondson D.E. Biochemistry. 1994; 33: 7088-7098Crossref PubMed Scopus (143) Google Scholar). Rabbit polyclonal antisera to recombinant human liver MAO A (anti-MAO antibody) purified from RH218 S. cerevisiae was produced by HTI Bio-Products, Santa Ysabel, CA. This antiserum cross-reacts with recombinant human liver MAO B as well and was used for detection of either MAO A and MAO B. Rabbit antisera specific for covalently bound flavins (anti-flavin antibody) were obtained using the haptenN -6-(6-aminohexyl)-FAD conjugated to bovine serum albumin (a gift from Dr. Mike Barber, University of South Florida College of Medicine, Tampa, FL) as described (20Barber M.J. Eichler D.C. Solomonson L.P. Ackrell B.A. Biochem. J. 1987; 242: 89-95Crossref PubMed Scopus (28) Google Scholar). Rabbit injections and serum bleeds were done by abV Immune Response, Inc. (Derry, NH). SDS-PAGE separation of proteins from digitonin-solubilized mitochondria was conducted according to Laemmli (31Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (212382) Google Scholar) using 7.5% polyacrylamide gels. Proteins were transferred to nitrocellulose membranes using a Bio-Rad Mini Trans-Blot apparatus in blot buffer (25 mm Tris base, 192 mmglycine, 20% (v/v) methanol) for 1 h at 100 V. For blots using the anti-MAO antisera, the membrane was blocked with 3% gelatin in TTBS (137 mm NaCl, 2.7 mm KCl, 25 mm Tris base, pH 8.0, 0.2% Tween 20) for 30 min at room temperature. Anti-flavin blots were blocked with WBT (20 mmTris base, pH 7.4, 0.3 m NaCl, 0.05% Tween 20) for 30 min at room temperature and washed in TTBS. Primary antibodies were diluted in antibody buffer (1% (w/v) gelatin in TTBS). Antisera to MAO A was diluted 1:5000 for use in anti-MAO blots, and the anti-flavin antisera was diluted 1:1250. Antibody solutions were added to the blots and allowed to incubate at room temperature for 1 h. Blots were washed twice in TTBS and incubated with either goat anti-rabbit horseradish peroxidase-conjugated secondary antibody (Sigma, for the anti-MAO blots) or goat anti-rabbit alkaline phosphatase-conjugated secondary antibody (Sigma, for the anti-flavin blots). Both secondary antibodies were diluted 1:3000 in antibody buffer. The anti-MAO and anti-flavin blots were developed using 3,3′-diaminobenzidine tetrahydrochloride and 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium, respectively. The anti-MAO A antiserum could detect a minimum of 2 ng of MAO A and 10 ng of MAO B. The anti-flavin antiserum recognized either covalently bound FAD or FMN on Western blots and exhibited a sensitivity to MAO A that was 5–10-fold lower than the anti-MAO antibody. (See SchemeFS2 for structures of the riboflavin analogues used in this study.) The following analogues were the generous gifts of Dr. John Lambooy (University of Maryland, Baltimore, MD): 7-nor-7-chlororiboflavin, 8-nor-8-chlororiboflavin, 7-nor-7-bromoriboflavin, 8-nor-8-bromoriboflavin, 7α-methylriboflavin, 8α-methylriboflavin, 7α,8α-dimethylriboflavin, 7-nor-7-bromo-8α-methylriboflavin, 7α-methyl-8-nor-8-bromoriboflavin, 7-nor-7-chloro-8α-methylriboflavin, and 7α-methyl-8-nor-8-chlororiboflavin. Additional flavins were provided by Dr. Sandro Ghisla (University of Konstanz) as follows: 8-nor-8-fluororiboflavin, 7-nor-7-chloro-8-nor-8-chlororiboflavin, and 8-nor-8-aminoriboflavin. Dr. D. B. McCormick of this department provided initial quantities of N (3)-methylriboflavin. The analogue 5-deaza-5-carbariboflavin was previously synthesized in this laboratory. Additional quantities of 8-nor-8-chlororiboflavin and 7-nor-7-bromoriboflavin were required for large scale growths of yeast auxotrophic for riboflavin. These analogues were synthesized as described by Lambooy (32Lambooy J.P. Methods Enzymol. 1971; 18B: 437-447Crossref Scopus (12) Google Scholar) using 3-chloro-4-methylaniline (Aldrich) and 3-methyl-4-bromoaniline (Lancaster), respectively, as starting materials. 2′-Deoxyriboflavin was synthesized as described (33Murthy Y.V. Massey V. J. Biol. Chem. 1995; 270: 28586-28594Abstract Full Text Full Text PDF PubMed Scopus (18) Google Scholar). The synthesis of th