,Io,irnol of Neuroc/zeinistryLippincoit—Raven Publishers, Philadelphia© 1996 International Society for Neurochemistry

Influence of C Terminus on MonoamineOxidase A and B Catalytic Activity

Kevin Chen, Hsiu-Fei Wu, and Jean Chen Shih

Department of Molecular Pharmacology and Toxicology, School of Pharmacy,University’ of Southern California, LOS Angeles, California, U.S.A.

Abstract: Monoamine oxidase (MAO) A and B play im-portant roles in the metabolism of neurotransmitters anddietary amines. The domains important for enzyme speci-ficities were studied by construction of chimeric MAOA/B enzymes. Exchange of the N-terminal 45 amino acidsof MAOA with the N-terminal 36 residues of MAOB (chi-meric enzymes B36A and A45B) resulted in the same sub-strate and inhibitor sensitivities as the wild-type MAOAor B. Thus, the N terminus may not be responsible forMAOA or B enzyme specificities. When MAOB C-terminalresidues 393—520 were replaced with MAOA C-terminalresidues 402—527 (chimeric B393A) catalytic activity wasnot detectable. Chimeric B393A consists of eight residueswith different charges, three less proline residues (458,476, and 490), and one additional proline at 518 com-pared with wild-type MAOB. These differences may haveinduced conformational changes and affected MAOBcatalytic activity. Thus, the C terminus of MAOB is criticalfor maintaining MAOB in an active form. It is interestingthat when the C terminus of MAOA was switched withMAOB (chimeric A402B), little effect was observed onMAOA catalytic activity. This new information is valuablefor further studies of the structure and function relation-ship of this important enzyme. Key Words: Monoam-ine oxidase—Conformation—Activity—Mitochondria—Transfection.J. Neurochem. 66, 797—803 (1996).

Monoamine oxidase (MAO) (EC 1.4.2.4), a FAD(flavin-adenine dinucleotide)-containing enzyme (Naraet al., 1966; Erwin and Hellerman, 1967), has beenused as a marker for the outer mitochondrial membrane(Tipton, 1967; Greenawalt and Schnaitman, 1970).The two types of MAO (A and B) differ in their sub-strate preferences, inhibitor specificities, tissue expres-sion, and cell distribution (Johnson, 1968; Knoll andMagyar, 1972; Fowler et al., 1978, Grimsby et al.,1990). Human placenta tissue contains predominantlyMAOA (Egashira, 1976), whereas platelets (Donnellyand Murphy, 1977), lymphocytes (Bond and Cundall,1977), and chromaffin cells (Youdim et al., 1984)express primarily MAOB. To understand tissue- andcell-specific expression for MAOA and B, the promot-

ers for both MAOA and B genes have been studied(Zhu et al., 1992). It is interesting that even thoughthere is ‘—~60%sequence identity between the MAOAand B promoters, the transcription factor binding sitesare different (Zhu et al., 1992). The location of MAOApromoter elements has been confirmed, but differentinitiation sites and repressors were found in differentcells (Denney et at., 1994). Gel retardation and DNaseI footprinting showed that Sp I is an important tran-scription factor for MAOA expression and the MAOAcore promoter exhibits bidirectional activity (Zhu etat., 1994).

MAOA preferentially oxidizes serotonin (5-hy-droxytryptamine, 5-HT) and is irreversibly inactivatedby low concentrations of the acetylenic inhibitor clor-gyline. MAOB preferentially oxidizes phenylethylam-inc (PEA) and henzylamine and is irreversibly inacti-vated by low concentrations of pargyline and deprenyl.Dopamine, tyraminc, and tryptamine are substrates forboth forms of MAO.

The deduced amino acid sequences from humanliver MAOA and B cDNAs show that the A and Bforms share 70% sequence identity (Bach et al., 1988)and are the products from different genes. The cDNAsfor human liver MAOA and MAOB encode proteinsof 527 and 520 amino acids, respectively. Both MAOAand B cDNA can be expressed in COS cells as catalyti-cally active enzymes (Lan et al.. l989a), indicatingthat the substrate or inhibitor sensitivities of MAOAor B activity is contained within a single polypeptide.The polypeptide exists as either a monomer or a homo-dimer but not as a heterodimer. Furthermore, the de-duced amino acid sequences of human brain and plate-let MAOB are identical to human liver, demonstrating

Received April 18, 1995: revised manuscript received August 18,995; accepted August 25. 1995.

Address correspondence and reprinl requests to Dr. J. C. Shth atDepartment of Molecular Pharmacology and Toxicology, School ofPharmacy, University of Southern California, Los Angeles, CA90033, U.S.A.

Abbreviations used: DMEM. Dulhecco~smodified Eagle medium:5-1-IT, 5-hydroxytryptamine or serolonin: MAO, monoamine oxi-dase; PEA, phenylethylamine: SDS. sodium dodecyl sulfate.

797

798 K. CHEN ET AL.

that platelet MAOB may be used as a marker of brainenzyme in clinical studies (Chen et at., 1993).

MAOA and B have strikingly similar genomic struc-tures; i.e., both consist of 15 exons and exhibit identicalexon—intron organizations, suggesting that they are de-rived from duplication of a common ancestral gene(Grimsby et al., 1991). They are closely linked andhave been mapped between bands Xpl 1.23 and Xp22.lof the X chromosome (Ozelius et al., 1988; Lan et al.,l989b). Both MAOs are deleted in some Norrie dis-ease patients (Lan et al., l989b; Sims et al., 1989), andrecently, MAOA deficiency was found to be associatedwith impulsive aggressive behavior in a Dutch family(Brunner et al., l993a,h). Thus, MAO deficiency maybe associated with behavioral aberrations.

Using site-directed mutagenesis, the functional im-portance of specific cysteine residues in each form ofMAO was demonstrated (Wu et al., 1993). There arenine cysteines in MAOA and B and each of them wasmutated to serine. The results of this study show that,in addition to the FAD binding site (Cys406 in MAOAand Cys397 in MAOB), one cysteine residue, Cys374,plays an important role in MAOA catalytic activity,whereas two cysteine residues, Cys ~ and Cys3~t’,areimportant for MAOB catalytic activity. These cysteineresidues may be important because when they are mu-tated to serine in COS cells, catalytic activity was notobserved.

In this study, four chirneric MAOA/B enzymes wereconstructed and characterized to further delineate re-gions that are important for MAOA and B catalyticactivity.

MATERIALS AND METHODS

Enzymes and chemicalsRestriction enzymes, Klenow Fragment of DNA polymer-

ase I, T4 DNA ligase, and phenol were purchased fromBethesda Research Laboratories (BRL, Gaithersburg, MD,U.S.A.), and New England Biolabs (Beverly, MA, U.S.A.).RNase A, DNase I, dithiothreitol, and isopropyl-~-D-thio-galactopyranoside (IPTG) were purchased from BoehringerMannheim Biochemicals (Indianapolis, IN, U.S.A.). 5-Bromo-4-chloro-3-indoIyl-~-u-gaIactopyranoside (X-gal)was purchased from ICN Biochemicals (Costa Mesa, CA.U.S.A.). dNTPs were purchased from Pharmacia Biotech-nology (Piscataway, NJ. U.S.A.). SeaKem medium agarosewas purchased from FMC Bioproducts (Rockland, ME,U.S.A.). Biotrans nylon membranes were obtained from ICNPall Biosupport (East-Hills, NY, U.S.A.). The site-directedmutagenesis kit was obtained from Amersham Corporation(Arlington Heights, IL, U.S.A.). Bactotryptone and Bactoyeast extract were purchased from DIFCO Laboratories (De-troit, MI, U.S.A.). Cesium chloride was from Cabot Inc.(Revere, PA, U.S.A.). DNA sequencing kit was from UnitedStates Biochemicals (Cleveland, OH, U.S.A.). [‘4C1-Phenylethylamine ([iaC I PEA) (58 mCi /mmol) and 5- [3HI -

hydroxytryptamine ([3H]5-HT) (26.7 Ci/mmol) were ob-tained from Du Pont-NEN (Wilmington, DE, U.S.A.). [a-32PIdATP was from ICN Radiochemicals (Irvine, CA.U.S.A.). For western blots, biotinylated secondary antibody,

avidin, and hiotinylated horseradish peroxidase were pur-chased from Pierce Laboratories (Rockford, IL, U.S.A.). Allother chemicals were obtained from Sigma (St. Louis. MO.U.S.A.). The pECE vector was kindly provided by Dr. JimOu (University of Southern California).

Construction of A45B and B36A chimeric plasmids

Site—directed mutagenesis was performed to create anX;nal restriction enzyme site at nucleotide 206 of MAOAcDNA. A45B (MAOA N-terminal residues 1—45 and MAOBC-terminal residues 37—520) and B0,A (MAOB N-terminalresidues 1—36 and MAOA C-terminal residues 46—527)were then generated by the tise of the common XmaI site inMAO cDNAs (Bach et al., 1988).

To generate the chimeric B0A plasmid, a 1.7-kb XmoI—LcoRl MAOA eDNA fragment (encoding C-terminal resi-dues 46—527), was first ligated to a 184-hp Xnzal—HindlllMAOB eDNA fragment (encoding N-terminal residues I —

36). The restAting chimeric construct was then ligated to a2.9-kb HindIII— EcoRl digested pECE plasmid. The 1.7-kbMAOA fragment was isolated from an M13 mpI8-MAOAcDNA clone and the 184-hp MAOB fragment was isolatedfrom a Bluescript SK MAOB eDNA clone. Both clones wereconstructed previously in our laboratory (Bach et al., 1988).

The A45B chimeric plasmid was generated by ligation of a2.3-kb X.’nal—XhaI fragment (isolated from a pECE-MAOBeDNA clone that contains C-terminal residues 37—520) toa 7.3-kb XmaI— Xhal fragment (isolated from an Ml 3 nip 18-MAOA eDNA clone that encodes N-terminal residues I --

45). The 9.6-kb construct was digested with EcoRI. produc-ing the 2.5-kb fragment, which was then subcloned into theEcoRl site of pECE.

Construction of A402B and B393A chimericplasmids

MAOA-pECE and MAOB-pECE plasmids that were pre-viously constructed in our laboratory (Lan et al., l989a)were digested with Scat to generate four DNA fragments.These four fragments were used to make chimeric A402B(MAOA N-terminal residues 1—402 and MAOB C-terminalresidues 393—520) and chimeric B:s1A (MAOB N-terminalresidues 1—393 and MAOA C-terminal residues 407—527).A4112B was produced by the ligation of the 2.1-kb fragment(encoding MAOA N-terminal residues 1—402) to the 3.25-kb fragment (encoding MAOB C-terminal residues 393—520). B19~Awas constructed by ligation of the 2.7-kb Irag-ment (encoding MAOA C-terminal residues 402—527) tothe 2.15-kb fragment (encoding MAOB N-terminal residues1—393).

Double strand DNA sequencingTo verify the identity of each chimeric plasmid, chimeric

pECE-MAOA or B constructs were purified from large-scalepreparations. These plasmids were then sequenced accordingto a double-stranded DNA sequencing method (ProntcgaProtocols and Application Guide, Promega. 1990)

High-efficiency calcium phosphate transfectionAfter the identities of chimeric constructs were confirmed

by sequencing, they were transiently transfected into COScells by a high-efficiency calcium phosphate transfectionmethod (Chen and Okayama, 1987). COS cells (5 )< l0~)were seeded in 100-mm Petri dishes, incubated overnight in10 ml of Dulbecco’s modified Eagle medium (DMEM). andsupplemented with 10% fetal bovine serum, penicillin, andstreptomycin. Fifty micrograms of plasmid DNA was mixed

J. Neurochem., Vol. 66, No. 2, 1996

CARBOXY TERMINAL IS CRITICAL FOR MAOB ACTIVITY 799

with 0.5 ml of 0.25 M CaCI2, 0.5 ml of 2x BES (50 mMBES: N,N-bis ( 2-hydroxyethyl ) -2-aminoethanesulfonic acid,280 niM NaCI, and 1.5 mM Na2HPO4, at pH 6.9), thenincubated for 15 mm at room temperature. One milliliter ofcalcium phosphate—DNA solution was added drop by dropto the plate of cells and the mixture was swirled gently andincubated for 15—24 h at 37°Cunder 3% CO,. The mediumwas removed and the cells were rinsed once with DMEMmedium, refed, and incubated for 24 h at 37°C with 5%CO2. After 24 h, the cells were trypsinized and harvestedby centrifugation at 1,000 g for 3 mm. The pelleted cellswere used for MAO assay.

Determination of MAOA and B catalytic activity,kinetic parameters, and IC511

Approximately 5 x i0~transfected COS cells were ho-mogenized in 250 p1 of 50 mM sodium phosphate buffer(pH 7.4). A 40-~ilaliquot of the homogenates (—~8X l0~cells) was added to a 1-mi assay mixture containing 50 mMsodium phosphate buffer at pH 7.4 and 100 p.M [

3Hl5-HTbr MAOA or 10 p.M I ‘4C IPEA for MAOB. The reactionmixture was incubated at 37°C for 20 mm and terminatedby adding 0.1 ml 6 M HCI. When 5-HT was used as asubstrate, the reaction product was extracted with 6 ml ofethyl acetate/benzene (1:1). When PEA was used as a sub-strate, the reaction product was extracted with 6 ml of tolu-ene. Four milliliters of organic layer was withdrawn andmixed with 5 ml of scintillation fluid from NationalDiagnos-tics. Theradioactivity of the reactionproduct wasdeterminedby liquid scintillation spectrometry.

I ‘4C I PEA and [3H]5-HT concentrations ranging from Ito 10 and from 10 to 100 pM, respectively, were used todetermine the kinetic parameters K~,and V~

05.The apparentK,, and ~ values were calculated from the linear interceptand the slope of double-reciprocal Lineweaver-Burk plots.Inhibition to the expressed wild-type and mutant MAOA orB by deprenyl and clorgyline was done by preincubation ofvarious concentrations of inhibitors with whole-cell homoge-nates (-~8X i0~cells) at 37°C for 30 mm before addingthe substrate. The concentration of deprenyl or clorgylinerequired for 50% inhibition of MAO activity (IC50) wascalculated from the x intercept of Hill plots. These experi-ments were repeated a minimum of three times.

Western blot analysis of chimeric and wild-typeMAO transfected cells

One hundred micrograms of transfected mutants and wild-type whole-cell homogenates were denatured by boiling for5 mm in sample buffer 12% sodium dodecyl sulfate (SDS),65 mM Tris buffer, pH 6.8, 1% ~-mercaptoethanol, 1%bromophenol blue, and 10% glycerol I~then subjected to 1.5mm SDS-polyacrylamide gel electrophoresis in 4% stacking,10% slab gels (Laemmli, 1970) at 20 mA. The current wasincreased to 40 mA after the samples had passed throughthe stacking gels.

The proteins ill the gels were transferred overnight electro-phoretically to nitrocellulose filters (Schleicher & Schuell,Inc.. Keenc, NH, U.S.A.) at 25°C at 50 V (Towbin et al.,1979). Immunological detection of bound MAO was per-formed by modifying the protocol of Young and Davis(1983). All incubations were performed at roomtemperaturewith gentle agitation.

Nitrocellulose filters were blocked with a mixture com-posed of 135 p1 of normal goat serum in 10 ml of 10 mMTris-HCI, pH 7.5, 0.9% NaC1, and 3% Tween 20 for 4 h.

FIG. 1. Map of the chimeric MAOA/B polypeptides. The mapswere prepared based on the nucleotide sequence ofthe chimericcDNA clones described in Materials and Methods. MAOA proteinsequences are depicted by clear, unfilled horizontal bars; MAOBsequences are depicted by hatched bars. Restriction enzymesites used for constructing the chimeric enzymes (Xmal andScal) are indicated above MAOA and MAOB.

Filters were first rinsed for 15 mm twice with TBSC 110mMTris-HCI, pH 7.5, 0.9% NaCI, and 3% (wt/vol) nonfatdry milk (Carnation Co., Los Angeles, CA, U.S.A.)I, incu-bated overnight in TBSC with a 500x dilution of rabbit anti-MAOB serum as described previously (Wu et al., 1993),and washed for 7 mm twice with TBSC. The filters werethen incubated with biotinylated secondary antibody, 45 p1of antibody in 10 ml of TBST (10 mM Tris-HC1, pH 7.5,0.9% NaCI, and 0.05% Tween 20) for 4 h, and washed for7 mm with TBST twice. After the avidin and biotinylatedhorseradish peroxidase complex (Pierce) was reconstituted,the filters were incubated for 30 mm. Then the filters werewashed for 7 mm three times in 10 mM Tris-HCI, pH 7.5,0.9% NaCl, and overlaid with a high-sensitive enzygraphicweb (IB1, New Haven, CT, U.S.A.) for the color reactionof horseradish peroxidase.

RESULTS

To determine which region(s) of MAOA and Bis responsible for substrate specificity and inhibitorsensitivity, four chimeric plasmids were constructed.These plasmids were transfected into COS cells andthe enzymatic properties analyzed. The structures ofthese chimeric enzymes are shown in Fig. 1.

Substrate specificity and inhibitor sensitivity ofchimeric enzyme B36A and A45B transfected inCOS cells

As shown in Table 1, the Kn, value of 5-HT of chi-meric B36A (110.9 ~ l0~ M) was similar to that ofwild-type MAOA (119.3 x l0~ M). The sensitivityof B36A to inhibitors was similar to wild-type MAOA,

J. Neuroc’he,n., Vol. 66, No. 2, 1996

800 K. CHEN ET AL.

TABLE 1. Kinetic parameter.s and IC51, values of wild-type and chuineric mutant.v expressed in COS cells

Substrate IC,,, (M)

[‘HIS-UT I ‘4C]PEA [‘HIS-UT [‘4C]PEA

EnzymeK,,, K,,

(10—” M) V,.,,” (10 “ lul) V,,,,,,”Deprenyl Clorgylinc Deprenyl Clorgyline(l0” M) (l(’’’ M) (tO M) (10 M)

MAOAMAOBB,,,AA

45BA4~2BB595A

119.3 ± II 36.8 ±0.34.6 ±0.6 34.4 ±0.4

110.9 ± 1.1) 34.1 ±0.3 — —

— — 3.6 ±0.4 24.2 ±0.231.0 ±t).4 11.9 ±0.1 —

— —-

1.5 ±0.1 3.7 ±0.42.7 ±0.3 4.0 ±0.4

2.3 ±0.2 3.0 ±0.5 — —

— — 3.6 ±0.4 6.)) ±)).52.5 ±0.3 3.0 ±0.4 — —

—

K,,, and V,5.,, values of the wild-type and chimcric enzymes were determined as described in Materials and Methods.[‘

4C]PEAranging from 1 to 10 pM and l’HlS-HT ranging from 10 to 100 pM were used as substrates. The proteinconcentration of the cell homogenate was determined by Lowry assay (Lowry et al., 1951). The IC,,, values of wild-typeand mutant enzymes were determined in the presence of various concenti’ations of deprcnyl and clorgyline. The K,,, and IC,,,values are the mean ±SD of three determinations. —, not detectable.

“V,.,. is expressed as nanomoles per 20 mm per milligram of protein.

when 5-HT was used as the substrate as well. TheIC

511 values of B36A for deprenyl (2.3 x l0~ M) andclorgyline (3.0 x 10_to M) also resembled that ofwild-type MAOA IC59 values (deprenyl, 1.5 x 100M; clorgyline, 3.7 )< tO ‘° M). No PEA oxidationwas detected when MAOA or chimeric B5~,5Awastransfected into COS cells.

The K,~and IC511 values for A45B were similar towild-type MAOB (Table I). The K,,, value of PEA forA45B (3.6 x l0~M) was similar to that of wild-typeMAOB (4.6 ~ to

5’M). The IC50 values of A45B for

deprenyl (3.6 ~ iO~ 41) and clorgyline (6.0 ~ l0~M) were comparable with that of wild-type MAOB(deprenyl, 2.7 >< l0~41; clorgyline, 4.0 ~ i0~ 41).5-HT oxidation was not found when MAOB or chi-meric A45B was transfected into COS cells.

When using deprenyl and clorgyline on wild-typeMAOA and B, the IC511 values were similar to that ofthe enzymes derived from human brain homogenates.This result suggested that the catalytic properties ofthe transfected MAOA and B were similar to the en-dogenous MAOA and B in human brain. Thus, it waspossible to use these transfected MAOs as a modelsystem to study their structure—function relationship.

The exchange between the N-terminus 45 aminoacids of MAOA and the N-terminal 36 amino acids ofMAOB (resulting in chimeric B56A and A45B) did notaffect the kinetic parameters of MAOA or B. Thisresult indicated that the N terminus might not be criti-cal for substrate and inhibitor specificities.

Substrate specificity and inhibitor sensitivity ofchimeric enzymes A402B and B393A

Chimeric A402B exhibited substrate and inhibitorsensitivities similar to that of wild-type MAOA (TableI). When 5-HT was used as substrate, the IC511 valuesof A402B for deprenyl (2.5 x l0~’41) and clorgyline(3.0 x lO~~M) were similar to wild-type MAOA(deprenyl, 1.5 >< 10641; clorgyline, 3.7 x l0’°M).

However, the K,,, value for 5-HT of A4112B (31 X 10 ~‘

41) was lower by approximately one-third when com-pared with wild-type MAOA (118 >< l0

6M). There-fore, the replacement of the MAOA C-terminal 126amino acids by the MAOB C-terminal 128 amino acidsappeared to affect the conformation of the enzymeto favor the binding of 5-HT. When COS cells weretransfected by MAOA or chimeric A

492B, no PEA oxi-dation was detected. No activity was observed withB595A when either 5-HT or PEA was used as substrate(Table 1).

Western blot analysis of wild-type and chimericA/B enzymes

Western blot analysis was performed to investigatewhether the lack of MAOB activity in transfectedB395A COS cells was due to nonexpression. A poly-clonal anti-MAOB antiserum was applied to blots con-taining chimeric enzymes and positive and negativecontrols (Fig. 2). When expressed in COS cells, thechimeric enzymes A45B (lane 6) and B195A (lane 7)and wild-type MAOB enzyme (lane 5) migrated tothe same molecular mass as the MAOB enzyme fromhuman brain homogenate (~60,0O0 Da; lanes 2—4)and were recognized by the rabbit MAOB antiserum.The intensity of the bands was similar between B595Aand A45B, suggesting the amounts of the chimeric en-zymes expressed were comparable. COS cells itself(lane I ) did not contain the 60,000-Da MAOB protein.as seen in human brain and the chimeric enzymes.Because the 60,000-Da protein could only representthe transfected MAOB chimeric, the lack of catalyticactivity seen in B395A enzyme was not due to nonex-pression.

DISCUSSION

The deduced amino acid sequences of MAOA andB cDNA shared ~70% homology (Bach et al., 1988).

I. Neur,,,heni., Vol. 66, No. 2, /996

CARBOXY TERMINAL IS CRITICAL FOR MAOB ACTIVITY 801

FIG. 2. Western blot of wild-type and chimeric enzymes ex-pressed in COS cells. Homogenates were made from varyingamounts of transfected COS cells or tissues. The amount ofprotein present was determined by Lowry assay (Lowry et al.,1951). MAOB was visualized by using a specific anti-MAOBantiserum. See text for details. Values on the left indicate molec-ular mass in kilodaltons. Lane 1, 100 pg COS cells; lane 2, 85pg of human brain homogenate; lane 3, 450 pg of human brainhomogenate + 100 pg COS cells; lane 4, 85 pg of human brainhomogenate + 100 pg COS cells; lane 5, wild-type MAOB; lane6, A45B; lane 7, B393A.

Within the polypeptides, two regions were conserved.One was located in the N terminus where MAOA hadan additional nine unique amino acids when comparedwith MAOB. The other nonconservative region existedat the C terminus starting from residue 450 of MAOA.Chimeric MAOA/B enzymes were constructed inwhich (I) the N-terminal 45 amino acids of MAOAand the N-terminal 36 amino acids of MAOB werereciprocally exchanged, and (2) the C-terminal 126amino acids of MAOA and the C-terminal 128 aminoacids of MAOB were reciprocally exchanged.

The results showed that the N-terminal 45 aminoacids of MAOA and 36 amino acids of MAOB mightnot be involved with substrate and inhibitor specifici-ties of MAOA and B, respectively. This result wasconsistent with a previous report in which the MAOA,MAOB, and N-terminal chimeric enzymes weretransfected in human embryonic kidney 293 cells anddifferent ligands and inhibitors were used (Gottowiket al., 1993). An AMP-binding site was found nearamino acids 15—29 and 6—20 in MAOA and B, respec-tively. The sequences of these binding sites have beenshown to he similar to other flavoproteins (Bach et at.,

1988). The data from this study suggested that theseregions were important for both MAOA and B catalyticactivity but not critical for their substrate and inhibitorspecificity. Recently, it has been shown that G1u34 inhuman MAOB was essential for catalysis (Kwan etal., 1995).

This study shows that when the 126 amino acids ofC-terminus MAOA were replaced by the 128 aminoacids of C-terminus MAOB, this chimeric enzyme(A

495B, Table 1) exhibited enzymatic properties simi-lar to wild-type MAOA, whereas MAOB activity wasnot detected. Therefore, it was unlikely that the 126amino acids of the C terminus of MAOA contributedto the inhibitor specificities of MAOA. This result alsoindicated that the C terminus (128 residues, 393—520)of MAOB had no effect on MAOA substrate and inhib-itor specificity.

In contrast, when 128 amino acids from the C termi-nus of MAOB were replaced by 126 amino acids fromthe C terminus of MAOA, this chimeric enzyme(B39IA, Table 1) showed no detectable enzymatic ac-tivity when using either 5-HT or PEA as substrate.The comparison between the C-terminal amino acidresidues of MAOA and B revealed that there werenumerous differences between the two enzymes(Fig. 3).

The following amino acid changes in chimeric B595Aresulted in a charge difference: MAOB histidine (43 Ito MAOA lysine (440), MAOB histidine (452) toMAOA asparagine (461 ). MAOB histidine (485) toMAOA asparagine (494), MAOB histidine (512) toMAOA tyrosine (521). MAOB aspartic acid (460) toMAOA Iysine (469), MAOB valine (470) to MAOAlysine (479), MAOB lysine (514) to MAOA tyrosine(523), and MAOB glycine (515) to MAOA lysine(524). These charge differences might have alteredthe conformation of the chimeric and obliterated itscatalytic activity (Table 1). It is interesting that whenthe complementary charge differences were introducedinto the chimeric MAOA, A4112B catalytic activity andsubstrate inhibitor specificity were not affected (TableI). This result confirmed the view that the active siteof MAOB may be more restricted structurally thanMAOA (Kalir et al., 1981).

in addition, the following proline residues were cx-

FIG. 3. Comparison of MAOA and B carboxy-terminal amino acids involved in the chimeric constructs. Protein sequences of humanwild-type MAOA (HMAOA), MAOB (HMAOB; Bach et al., 1988), and trout MAO (TMAO; Chen et al., 1994) are aligned. Italicizedboldface letters depict hydrophobic residues that may be involved in mitochondrial targeting. Italicized lowercase letters depict the C-terminal residues that may be responsible for mitochondrial targeting after deletion of the last 24 C-terminal residues. The deleted 24C-terminal residues are underlined (Weyler, 1992). (“), Charge differences; (#), proline residue differences.

.l.,v,’,,r,,,/,e,,,., Vol. 66. N,,. 2, /996

802 K. CHEN ET AL.

changed: MAOB prolines (458, 476, and 490) wereexchanged with, respectively, MAOA threonine(467), glutamic acid (485), and serine (499); MAOAproline (527) was exchanged with MAOB valine(5 18) (Fig. 3). Because the conformational flexibilityof proline was low and may affect the three-dimen-sional structure of the enzyme, it is tempting to specu-late that these four proline residues in the C terminalmay contribute to the rigidity of MAOB and thus werecrucial for MAOB activity.

It has been suggested that both hydrophobic andpositively charged residues at the C terminus wereimportant for mitochondrial targeting (Nguyen et al.,1993). The targeting sequence for MAOB to mito-chondria has been shown to be in the 29 C-terminalamino acids (492—520) (Mitoma and Ito, 1992).However, when the 24 C-terminal amino acids (504—527) of MAOA (Fig. 3, underlined) were deleted andexpressed in yeast, the truncated enzyme was targetedto mitochondria and exhibited catalytic activity(Weyler, 1992). The new carboxy terminus of thetruncated enzyme had nine hydrophobic (494—502)and one positively charged residue (503), shown inFig. 3 as italicized and lowercased. The structural char-acteristics deemed important in mitochondrial tar-geting still existed, suggesting that the new C terminusmight still be able to target MAOA to the mitochon-drion. However, the possibility that additional mito-chondrial targeting sequence(s) could be located inother regions of the enzyme could not be ruled out.

A novel trout liver MAOA was cloned recently inwhich the C terminal was 27 residues shorter thanhuman MAOA (501—527). The trout MAO still ex-hibited catalytic activity when either 5-HT or PEA wasused as substrate (Chen et al., 1994). The carboxyterminus of trout MAO is composed of 12 residues(437—448) that were hydrophobic (Fig. 3, boldface).There was no positively charged residue present in thisregion of the enzyme. Therefore, in this case, onlyhydrophobic, not positively charged residues might beimportant in mitochondrial targeting. These results alsosuggested that the mitochondrial targeting sequence ofMAOA and B was structurally and functionally com-plex. Because there were 16 and 17 hydrophobic aminoacids in the C-terminal end of MAOA and B, respec-tively (Fig. 3, boldface), switching the C terminus ofMAOA and B might not have affected mitochondrialtargeting.

In summary, this study showed that the 128 C-termi-nal amino acids (393—520) might be critical to MAOBcatalytic activity. Furthermore, this study suggestedthat the following amino acids, histidine (431, 452,485, and 512), aspartic acid (460), valine (470), ly-sine (514), glycine (515), and proline (458, 476, and490), might be critical for either appropriate conforma-tion or substrate binding of MAOB. This new informa-tion is valuable for further studies of the structure andfunction relationship of this important enzyme.

Acknowledgment: We thank Debra Kline, Ph.D.. for

helping to edit the manuscript. This study was supported bygrants ROI MH37020 and R37 MH39085 (MERIT Award)and Research Scientist Award KO5 MH00796 from the Na-tional Institute of Mental Health. Support from the Boyd andElsie Welin Professorship is also appreciated.

REFERENCES

Bach A. W. J., Lan N. C., Johnson D. L., Ahell C. W.. BembenekM. E., Kwan S. W., Seeburg P. H.. and Shih J. C. (I 988 (eDNAcloning of human liver moiloamine oxidase A and B: molecularbasis of differences in enzymatic properties. Proc. Nail. Acod.Sci. USA 85, 4934—4938.

Bond P. A. and Cundall R. L. (1977) Properties of monoamineoxidase (MAO) in human blood platelets, plasma. lymphocytesand granulocytes. Clin. C/tim. Acw 80, 317—326.

Brunner U. G., Nelen M. R., van Zandvoort P., Abeling N. 0. 0. M..van Gennip A. H.. Wolters E. C., Kuiper M. A., Ropers H. H..and van Oost B. A.) 1993a) X-linked borderline mental retarda-tion with prominent behavioral disturbance: phenotype, geneticlocalization, and evidence for disturbed monoamine metabo-lism. Am. J. Hum. Genet. 52, 11)32— 1039.

Brunner H. G., Nelen M.. Breakebield X. 0., Ropers H. H.. and vanOost B. A. (1993/)) Abnormal behavior associated wtth a poutmutation in the structural gene for monoamine oxidase A ci-ence 262, 578—580.

Chen C. and Okayarna H. (1987) High-efficiency transformation s,fmammalian cells by plasinid DNA. Mo/. Cell. Biol. 7, 27452752.

Chen K., Wu H. F., and Shih J. C. (1993) The deduced aminoacid sequences of human platelet and frontal cortex monoamineoxidase B are identical. J. Neurocheun. 61, l87—19().

Chen K., Wu H-F., Grimsby J., and Shih J. C. 11994) Cloningof a novel monoamine oxidase eDNA from trout liver. Mol.Pharrnacol. 46, 1226— 1233.

Denney R. M., Sharma A.. Dave S. K., and Waguespack A. (1994)A new look at the promoter of the human monoamine oxtdaseA gene: mapping transcription initiation sites and capacity todrive luciferase expression. J. Neurochem. 63, 843—856.

Donnelly C. H. and Murphy D. L. (1977) Substrate- and inhibitor-related characteristics of human platelet monoamine oxidase.Biochem. Phar,nacol. 26, 853—858.

Egashira T. (1976) Studies on monoamine oxidase. XVIII. Enzy-matic properties of placental monoamine oxidase. ./pn. .1. P/tar-macol. 26, 493—500.

Erwin V. 0. and Hellerman L. ( 1967 ) Mitochondrial monoamineoxidase: I. Purification and characteri,ation ofthe bovine ktdnc~enzyme. I. Biol. C/tern. 242, 4230—4238.

Fowler C. J., Callingham B. A., Mantle T. J., and Tipton K. F.(1978) Monoamine oxidase A and B: a useful concept. Mol.Pharmacol. 27, 97—101.

Gottowik J., Cesura A. M., Malherbe P., Lang 0., and Prada M. D.(1993) Characterization of wild-type and mutant forms of hu-man monoamine oxidase A and B expressed in a mammalian

cell line. EBBS Lell. 317, 152—156.Greenawalt J. W. and Schnaitman C. 1197(1) An appraisal (If the

use of monoamine oxidase as an enzyme marker for the outermitochondrial membrane. J. Cell Biol. 46, 173—179.

Grimsby J., Lan N. C., Neve R.. Chen K.. and Shih J. C. (1990)Tissue distribution of human monoamine oxidase A and BmRNA. J. Neurochem. 55, 1166—1169.

Grimsby J., Chen K., Wang L.-J., Lan N. C., and Shih J. C. (1991)Human monoamine oxidase A and B genes exhibit identicalexon intron organization. Proc. Nail. Acae/.Sci. USA 88, 3637—3641.

Johnson J. P. (1968) Some observations upon a new inhibitor ofmonoamine oxidase in human brain. Biochen,. Phormacol. 17,1285— 1297.

Kalir A.. Sabbagh A., and Youdim M. B. H. (1981 ) Selective acety-

J. Neurochern., Vol. 66, No. 2. /996

CARBOXY TERMINAL IS CRITICAL FOR MAOB ACTIVITY 803

lenic ‘‘suicide’’ and reversible inhibitors of monoam ie oxidasetype A and B. Br. J. Pluarmacol. 73, 55—64.

Knoll J. and Magyar K. (1972) Some puzzling pharmacologicaleffects of monoamine oxidase inhibitors. Ac/i’. Bioche,n. P,sv-c/topliarmacol. 5, 393—4(18.

Kwan S. W.. Lewis D. A., Zhou B. P., and Ahell C. W. (1995)Characterization of a dinuclcotide-binding site in monoamineoxidase B by site—directed mutagenesis. Arch. Bioc/ient. Rio-

Pkv,s.~116,385—391.Laemmli U. K. (1970) Cleavage of structural proteins during the

assembly of the head of bacteriophage ‘l’4. Nature 227, 680—685.

Lan N. C., Chen C., and Shih J. C. (I989a) Expression of functionalhuman monoamine oxidase A and B cDNAs in mammaliancells. J. Neurochem. 52, 1652— 1654.

Lan N. C., Heinzrnann C., Gal A.. Klisak I., Orth U., Lai U., GrimshyJ., Sparkes R. S.. Mohandas T.. and Shih J. C. (1989h) Humanmonoamine oxidase A and B genes map to Xp 11.23 and aredeleted in a patient with Norrie disease. Geaonuic,s 4, 552—559.

Lowry 0. H., Rosebrough N. J.. Farr A. L., and Randall R. J. (1951Protein measurement with the Folin phenol reagent. I. Rio!.C/tern. 193, 265—275.

Mitoma J. and Ito A. (1992) Mitochondrial targeting signal of ratliver monoamine oxidase B is located at its carboxy terminus.

J. Bloc/tern. ( Tokyo) 111, 20—24.Nara S., lgaue I., Gomes B., and Yasunohu K. T. (1966) The pros-

thetic groups of animal amine oxidases. Biochem. Biopluv,v. Rex.Common. 23, 324—328.

Nguyen M.. Millar D. G., Yong V. W., Korsnieyer S. J.. and ShoreG. C. (1993) Targeting of Be 1-2 to the mitochondrial outermembrane by a COOH-terminal signal anchor sequence. J. Biol.C’hern. 268, 25265—25268.

Ozelius L., Usu Y.-P.. Bruns G.. Powell J. F., Chen S.. Weyler W.,

Utterback M., Zucker D., Haines Trofatter J. A., ConneallyP. M., Gusella J. F.. arid Breakefield N. 0. (1988) Humanmottoatiline oxidase gene 1 MAOA): chromosome position(Xp21-pl I) and DNA polymorphism. Genooiic,s 3, 53—58.

Sims K. B., de Ia Chapelle A., Norio R., Sankila E.-M., HsuY.-P. P., Rinehart W. B.. Corey T.J., Ozelius L., Powell J. F.,Bruns G., Gusella J. F., Murphy D. L.. and Breakefield X. 0.

1989) Monoamine oxidase deficiency in males with an X chro-mosoilte deletion. Neuron 2, 1069— 1076.

Tipton K. F. (1967) The sub-mitochondrial localization of mono-amine oxidase in rat liver and brain. Biochirn. Biophv,s. Ada135, 910—920.

Towbin H., Stachelin T., and Gordon J. (1979) Eleetrophoretie trans-fer of proteins from polyaerylamide gels to nitrocellulose sheets:procedure and some applications. Proc. Nat!. Acacl.Sci. USA76, 4350—4354.

Weyler W. (1992) Functional expression of C-terminally truncatedhuman monoamine oxidase type A in Saccharomvce.s ceretu.t-iae. .1. Neural. Tran,srn. 41) Suppl.) .3 — 15.

Wu H. F., Chen K., and Shih J. C. (1993) Site-directed nsutagenesisof monoansine oxidase A and B: the role of eystines. Mol.Pharrnacol. 43, 888—893.

Youdiin M. B. H.. Banerjee D. K., :tnd Pollard H. (1984) Isolatedchromaffin cells from adrenal medulla contain primarily mono-amine oxidase B. Science 224, 619—620.

Young R. A. and Davis R. W. (1983) Efficient isolation of genesby using antibody probes. Proc Nail. Acac/. Sci. USA 80, 1194—1198.

Zhu Q.-S., Grimshy J.. and Shih J. C. 11992) Promoter organizationof human monoamitse oxidase (MAO) A and B genes. J. Neu-ro.sci. 12, 4437—4446.

Zhu Q.-S., Chen K., and Shih J. C. (1994) Bidirectional promoterof human nionoanhine oxidase A (MAOA ) controlled by tran-scription factor SpI. .1. Neuro,sci. 14, 7393—7403.

.1. Neuro, he,,,., Vol. 66, Nc~.2, 1996