ANALYTICAL BIOCHEMISTRY 244, 384–392 (1997)ARTICLE NO. AB969911

A Continuous Spectrophotometric Assay for MonoamineOxidase and Related Enzymes in Tissue Homogenates

Andrew Holt,* Dennis F. Sharman,† Glen B. Baker,‡ and Monica M. Palcic**Department of Chemistry, University of Alberta, Edmonton, Alberta T6G 2G2, Canada; †Department of Pharmacology,University of Cambridge, Tennis Court Road, Cambridge CB2 1QJ, United Kingdom; and ‡Neurochemical Research Unit,Department of Psychiatry, 1E7.44 WMC, University of Alberta, Edmonton, Alberta T6G 2B7, Canada

Received September 20, 1996

oxidoreductase (deaminating) (copper-containing); semi-A continuous, peroxidase-linked spectrophotomet- carbazide-sensitive amine oxidase (SSAO)] (2). While

ric assay is described which is suitable for measuring the physiological functions of most SSAO enzymes re-monoamine and diamine oxidase and semicarbazide- main unknown, mitochondrial MAO, which exists assensitive amine oxidase activities in tissue homoge- two isozymes designated MAO-A and MAO-B (3), regu-nates. In the assay, 4-aminoantipyrine is oxidized and lates tissue levels of amine neurotransmitters and pro-then condenses with vanillic acid to give a red qui- tects the animal from the effects of pharmacologicallynoneimine dye. The absorbance at 498 nm is propor- active dietary amines. MAO-A inhibitors are used clini-tional to the amount of hydrogen peroxide released cally in the treatment of depression (4), while inhibitionin the amine oxidase reaction. The molar absorption of MAO-B may alleviate symptoms and slow the pro-coefficient of the dye at pH 7.6 was 4654 M01 cm01. The gression of Parkinson’s disease (5).method is suitable for use with any amine oxidase sub- The ability of amine oxidase enzymes to convert com-strate which has a higher oxidation–reduction poten-

pounds such as 1-methyl-4-phenyl-1,2,3,6-tetrahydro-tial than does 4-aminoantipyrine. Following preincu-pyridine (6) and allylamine (7) to highly toxic productsbation of rat liver homogenates with selectiveunderlines the potential usefulness of a simple assaymonoamine oxidase (MAO)-A and -B inhibitors, kineticwhich could demonstrate whether or not endogenousconstants were obtained for metabolism of the mixedor environmental substances, or drugs, might be aminesubstrate, p-tyramine. Inhibition of MAO in rat liveroxidase substrates. Such an assay would be particu-homogenates was also measured following administra-larly useful if kinetic constants for metabolism of thetion of the antidepressant, phenelzine. This inexpen-compound could also be obtained, perhaps allowingsive assay which employs reagents with low toxicitymore accurate predictions of the degree of substratecan thus be used to determine the degree of inhibition

of MAO elicited by potential antidepressant and anti- turnover in vivo than might otherwise be the case.parkinsonian agents. Drugs, their metabolites, and en- Presently, a variety of assay protocols exist, both forvironmental toxins can also be screened as possible continuous and discontinuous assays, which allowamine oxidase substrates or inhibitors, and kinetic quantification of amine oxidase activities (see Ref. 8constants for turnover of novel substrates can be de- and references therein). Perhaps the most widely usedtermined. q 1997 Academic Press discontinuous procedure is the radiochemical assay

(see Ref. 9), which relies on the formation of radiola-beled aldehyde during incubation of an appropriatelylabeled amine substrate with MAO or SSAO. However,Mammalian monoamine oxidase enzymes have beenwhen data are obtained by this method, or other discon-classified either as EC 1.4.3.4 [amine: oxygen oxidore-tinuous methods, a number of errors which are oftenductase (deaminating) (flavin-containing); monoamineoverlooked and which may invalidate the results thusoxidase (MAO)1] (1) or as EC 1.4.3.6 [amine: oxygenderived may be introduced (10). The most common ofsuch errors is probably the failure of workers to ensure1 Abbreviations used: MAO, monoamine oxidase; SSAO, semicar-

bazide-sensitive amine oxidase; ABTS, 2,2 *-azino-bis(3-ethylbenzthi-azoline-6-sulfonic acid); vanillic acid, 4-hydroxy-3-methoxybenzoicacid; clorgyline, N-methyl-N-propargyl-3-(2,4-dichlorophenoxy)pro- amine hydrochloride; phenelzine, 2-phenethylhydrazine sulfate; ty-

ramine, 4-hydroxyphenylethylamine hydrochloride.pylamine hydrochloride; pargyline, N-methyl-N-benzyl-2-propynyl-

384 0003-2697/97 $25.00Copyright q 1997 by Academic Press

All rights of reproduction in any form reserved.

AID AB 9911 / 6m25$$$121 12-13-96 14:56:48 aba

SPECTROPHOTOMETRIC ASSAY FOR AMINE OXIDASES 385

that product formation proceeds linearly for the dura- plasma amine oxidase activities and by Holt and Baker(21) to assay porcine kidney diamine oxidase. In thistion of the incubation period, with the result that en-

zyme activities are underestimated and, when the ef- reaction, 4-aminoantipyrine acts as the proton donorin the peroxidase reaction and then condenses with 2,4-fects of amine oxidase inhibitors are under study, the

inhibitor potency may also be underestimated. Thus, dichlorophenol to form a red quinoneimine dye. Theabsorbance, measured at 495 nm, is directly propor-it is generally accepted that continuous assay systems

are to be preferred over discontinuous assay systems tional to the amount of hydrogen peroxide formed dur-ing amine metabolism. While this protocol has provedif accurate kinetic data are required.

The reaction catalyzed by monoamine oxidase en- suitable for the assay of most, if not all enzymes classi-fied as EC 1.4.3.6, no color formation was evident whenzymes can, for most substrates, be summarised as fol-

lows: RCH2NR1R2 / H2O / O2 r RCHO / NHR1R2 / homogenates of, or mitochondria purified from, ratliver or brain were incubated with MAO substrates (A.H2O2. It is apparent that the identity of the aldehyde

formed is dependent upon the identity of the substrate Holt, unpublished work). Thus, there existed no rapid,continuous assay for MAO enzymes which permittedand, while some aldehydes can be detected directly by

spectrophotometry, the technique is often unsuitable the use of tissue homogenates and allowed flexibilityin the choice of substrate.for use with crude tissue homogenates (11). Similarly,

the disappearance of kynuramine can be followed con- In the present study, we have determined the causeof our inability to measure MAO activities by the perox-tinuously, but this and similar protocols are limited

to use with a single substrate (12). Ammonia (NH3) idase-linked method described above and have adaptedthe assay to circumvent the problem. A rapid, quantita-production can be measured continuously with a cou-

pled colorimetric assay (13). However, ammonia is de- tive spectrophotometric assay is thus described whichcan be made using 96-well microtiter plates or cuvettesrived from primary amine substrates (both R1 and R2

are hydrogen) and, while most SSAO enzymes are and which is suitable for continuous measurement ofamine metabolism by MAO-A and -B and SSAO en-thought to accept only primary monoamines as sub-

strates (2), the classical MAO enzymes can also metab- zymes in crude tissue homogenates. The technique canbe used to assess whether novel compounds might beolize secondary and tertiary monoamines. Thus, only

the use of oxygen and the formation of hydrogen perox- amine oxidase substrates or inhibitors, and to deter-mine kinetic constants for such interactions. The clini-ide generally remain independent of the substrate used

or of the enzyme classification. Polarographic determi- cal application of this technique is also demonstrated.nation of oxygen consumption is an accurate and repro-ducible means by which amine oxidase activities might MATERIALS AND METHODSbe measured (14). However, this technique is not suit-

Materialsable for rapid processing of large numbers of samples.Most assays designed to detect hydrogen peroxide are Clorgyline, pargyline, phenelzine, 2,4-dichlorophe-discontinuous coupled assays leading to formation of a nol, vanillic acid, 4-aminoantipyrine, peroxidase (typefluorescent product (see Refs. 8, 15). One direct contin- II, from horseradish), p-tyramine hydrochloride, anduous fluorometric assay for MAO-B has been described benzylamine hydrochloride were purchased fromrecently (16). Relatively high concentrations of purified Sigma Chemical Company (St. Louis, MO). 5-Hy-enzyme were required and the assay is limited to exper- droxy[G-3H]tryptamine creatinine sulfate was pur-iments with one substrate. chased from Amersham International plc (Amersham,

A highly sensitive colorimetric assay has been re- Bucks, UK) and b-[ethyl-1-14C]phenylethylamine hy-ported by Szutowicz et al. (17) to be suitable for the drochloride from Du Pont (Markham, Ontario, Can-determination of MAO activities. However, this peroxi- ada). Ready Safe liquid scintillation cocktail was ob-dase-coupled method, which uses 2,2 *-azino-bis(3- tained from Beckman Instruments Inc. (Fullerton, CA).ethylbenzthiazoline-6-sulfonic acid) (ABTS) as the All other reagents were of analytical grade, where pos-chromogen, suffers from the drawback that ABTS sible.causes substantial inhibition of MAO at concentrations Pea seedling amine oxidase was isolated and purifiedpresent in the assay. Furthermore, the chromogen is as described previously (22). Phenylethylamine oxidaseinsoluble and absorbs only weakly at the relatively neu- from Arthrobacter globiformis (23) was a generous gifttral pH conditions necessary for MAO activity. As a from Professor K. Tanizawa (Osaka University, Japan).result, while this reagent is extremely useful in thediscontinuous assay described by the authors, it is un-

Animalssuitable for use in continuous measurement protocols.A peroxidase-linked colorimetric assay, described Naı̈ve male Sprague–Dawley rats (280–300 g), ob-

tained from the University of Alberta Health Sciencesoriginally by Yamada et al. (18), has been adapted andused by Elliott et al. (19, 20) to measure sheep blood Laboratory Animal Services, were maintained on a 12-

AID AB 9911 / 6m25$$$122 12-13-96 14:56:48 aba

HOLT ET AL.386

h light–dark cycle, during which time they were al-lowed free access to drinking water and a standardlaboratory rat diet. Animals were killed by stunningand decapitation, and livers and brains were dissectedout, washed in ice-cold potassium phosphate buffer (0.2M, pH 7.6), and stored at 0707C until required.

For ex vivo examinations of the effects of the antide-pressant drug phenelzine (2-phenethylhydrazine sul-fate) on rat liver MAO, phenelzine was administeredas a single intraperitoneal injection to male Sprague–Dawley rats (240–300 g) at doses of 40 or 250 mmolkg01 in an injection volume of 1 ml kg01. Control ratsreceived water in place of phenelzine. Animals werekilled by decapitation, 2 h after injection, and liverswere dissected out and stored at 0707C until required.

Effects of 2,4-Dichlorophenol on MAO Activities

Homogenates of brain cortices from three rats wereprepared 1:40 (w/v) in ice-cold potassium phosphatebuffer (0.2 M, pH 7.6) with a Polytron mechanical ho-mogenizer. Radiochemical assays of brain MAO-A and-B activities were done in triplicate, as described by

FIG. 1. The peroxidase-linked continuous assay for amine oxidaseHolt and Baker (21). Assay tubes contained 25 ml ofenzymes. In the reaction, 4-aminoantipyrine acts as the proton donortissue homogenate, 25 ml of water (controls) or aqueous in the peroxidase reaction and then condenses with vanillic acid. The

2,4-dichlorophenol (0.3 mM–1 mM) and 50 ml of an ap- structure of the quinoneimine dye, which absorbs maximally at 498propriate radiolabeled substrate; these were, for MAO- nm, has not been determined but that shown is thought to be the

most likely candidate.A, 5-[3H]hydroxytryptamine (250 mM, sp act 1 mCimmol01), and for MAO-B, [14C]phenylethylamine (50mM, sp act 1 mCi mmol01). Results from control experi-ments indicated that, under the conditions used, the used in kinetic studies of MAO-A activity. Inhibitionrate of formation of products was linear for the dura- of catalase was found to be unnecessary in the presenttion of the 10-min incubation period. Sigmoid inhibition study (see Discussion).curves were fitted to the data using the nonlinear re- For some experiments, MAO was partially purifiedgression facility of GraphPad Prism, version 1.03 by isolation of mitochondria from liver homogenates.(GraphPad Software, San Diego, CA). Liver tissue (5 g) was homogenized 1:40 (w/v) in 0.3 M

sucrose. Following centrifugation at 1000g for 10 min,the supernatant was further centrifuged at 10,000g forPreparation of Hepatic MAO for Spectrophotometric30 min to obtain a crude mitochondrial pellet. The pel-Studieslet was resuspended in 4 ml of 0.3 M sucrose and wasHomogenates of liver were prepared 1:25 (w/v) in layered onto 40 ml of 1.2 M sucrose. A mitochondrialice-cold potassium phosphate buffer, with a Polytron pellet was obtained by centrifugation at 53,000g formechanical homogenizer. Homogenates were centri- 2 h. Following a single wash in potassium phosphatefuged at 1000g and 47C for 15 min and the supernatant, buffer, mitochondria were suspended in 40 ml bufferwithdrawn carefully by syringe, was used as the source and were stored in aliquots of 1 ml, at 0707C, untilof MAO. Aqueous solutions of 50 mM clorgyline (a selec- required for experimentation.tive inhibitor of MAO-A) or 50 mM pargyline (a selective

inhibitor of MAO-B) were then added to aliquots ofSpectrophotometric Measurement of MAO Activityhomogenates at a ratio of 1:100 (v/v) such that final

inhibitor concentrations were 500 nM, and homoge- The peroxidase-linked assay described by Elliott etal. (19) was modified (Fig. 1) by replacing 2,4-dichloro-nates were incubated with inhibitors at 377C for 30

min. A third aliquot of homogenate was incubated with phenol with vanillic acid (4-hydroxy-3-methoxybenzoicacid), which does not inhibit MAO at the concentrationswater in place of inhibitor. Homogenates were stored

at 0207C and were used within 48 h. Activity of MAO- used in the assay. The chromogenic solution preparedfor inclusion in the assay mixture contained vanillicA declined rapidly following homogenization, with a

half-life of approximately 24 h even in frozen homoge- acid (1 mM), 4-aminoantipyrine (500 mM), and peroxi-dase (4 U ml01) in potassium phosphate buffer (0.2 M,nates, and freshly prepared homogenates were thus

AID AB 9911 / 6m25$$$122 12-13-96 14:56:48 aba

SPECTROPHOTOMETRIC ASSAY FOR AMINE OXIDASES 387

pH 7.6). These reagent concentrations were chosen fol- were made in which homogenates, previously treatedwith pargyline, were further incubated with clorgylinelowing detailed preliminary experiments to obtain opti-

mal assay conditions (see Discussion). Chromogenic so- (500 nM), and vice versa, at 377C for 30 min, prior toaddition of tyramine substrate. Six replicate assayslutions were prepared on a daily basis and were kept

at 47C until required. were made in microtiter plate wells, as described above.When assays were made in 96-well polystyrene mi-

crotiter plates (Corning, NY), the wells contained 50 ml Dependence of Reaction Velocity upon Enzymetissue homogenate, 50 ml chromogenic solution, and Concentration200 ml amine substrate, prepared in potassium phos-

A range of volumes of inhibitor-treated homogenatesphate buffer. Blank wells had buffer added in place of(20–50 ml) was made up to total volumes of 50 ml withsubstrate. There was a negligible difference betweenpotassium phosphate buffer and then incubated withblank rates in wells where substrate had been omittedtyramine, at 500 mM or 2.5 mM, to assay MAO-A andand in wells where homogenate had been omitted. Re--B activities, respectively. Five replicate assays wereactant volumes were scaled up to a total volume of 1 mlmade in microtiter plate wells, as described above.when assays were done in spectrophotometer cuvettes.

Reaction mixtures for these latter assays were warmedMolar Absorption Coefficient for the Quinoneimineto 377C prior to the addition of tissue homogenate.

DyeReactions were followed in a THERMOmax mi-croplate reader (Molecular Devices, Menlo Park, CA) A hydrogen peroxide stock solution was standardizedwith a 490-nm filter and a plate chamber temperature by titration with potassium permanganate and a molarof 377C. Absorbance readings were taken every 15 s for absorption coefficient for the quinoneimine dye was30–40 min and initial (maximum) reaction velocities then calculated from the slope of a standard curve,were determined by linear regression of the data (SOF- obtained with hydrogen peroxide. A chromogenic solu-Tmax version 2.32, Molecular Devices). When the reac- tion was prepared containing vanillic acid (10 mM), 4-tions were done in 1-ml glass cuvettes, absorbance in- aminoantipyrine (5 mM), and peroxidase (20 U ml01)creases were followed at 498 nm and 377C in a Hewlett in potassium phosphate buffer (0.2 M, pH 7.6). To mi-Packard 8451A diode array spectrophotometer with a crotiter plate wells were added 50 ml of chromogenicwater-heated cell holder. Measurements were made ev- solution, 50 ml of rat liver homogenate, and 200 ml of aery 5 s for 30–40 min and initial reaction velocities range of concentrations of hydrogen peroxide preparedwere determined automatically by curve smoothing fol- in potassium phosphate buffer, such that the final con-lowed by derivatization of the data. centrations of peroxide ranged from 0 to 467 mM. Reac-

While the assay has been designed to permit multiple tion mixtures were warmed to 377C, and an endpointcontinuous measurements in a plate reader, it can be absorbance reading at 490 nm was made for each well.used discontinuously, to allow large numbers of sam- Data were analyzed using the linear regression facilityples to be measured in a spectrophotometer. Assays are of GraphPad Prism, version 1.03.done in a volume of 1 ml in glass test tubes, as de-scribed above, and reactions are stopped after an ap-

Comparison with an Existing Continuous Assaypropriate time period by plunging the tubes into iceProcedure for MAOand adding phenelzine (10 mM, 10 ml). Samples are

then allowed to reach ambient temperature before Experiments were done to determine if reaction ve-locities obtained by this method were consistent withtransferring to disposable methacrylate cuvettes.those obtained by an alternative protocol. For this, amodification of the procedure described by Tabor et al.Selective Measurement of MAO-A and MAO-B(11) was used, in which the formation of benzaldehydeActivitiesfrom benzylamine was followed directly at 254 nm.Quartz assay cuvettes contained 33 ml mitochondriaIn order to estimate MAO-A and -B activities, liver

homogenates were incubated with the mixed substrate, preparation, 300 ml potassium phosphate buffer (0.2 M,pH 7.6), and 667 ml benzylamine (450 mM in buffer,p-tyramine, at 500 mM to measure MAO-A following

inhibition of MAO-B with pargyline (500 nM), at 2.5 yielding a final assay concentration of 300 mM). Blankscontained buffer in place of enzyme. The increase inmM to measure MAO-B following inhibition of MAO-A

with clorgyline (500 nM), or at 500 mM to measure total absorbance at 254 nm was followed at 377C for 20 min.A molar absorption coefficient of 12,500 M01 cm01 wasMAO when no inhibitor had been included. Substrate

concentrations were chosen with respect to literature used for benzaldehyde (25). In parallel studies, cu-vettes contained 33 ml mitochondria preparation, 133Km values for tyramine metabolism by MAO in rat liver

isolated mitochondria (24). To indicate the effective- ml buffer, 167 ml chromogenic solution, and 667 ml ben-zylamine (450 mM). The increase in absorbance at 498ness of the selective MAO inhibitors, parallel studies

AID AB 9911 / 6m25$$$122 12-13-96 14:56:48 aba

HOLT ET AL.388

nm was followed at 377C for 20 min. Under these condi-tions, the B form of MAO is largely responsible forturnover of benzylamine.

Measurement of Benzylamine Turnover by OtherAmine Oxidase Enzymes

Studies were undertaken to confirm that this assaymethod is suitable for measurement of amine oxidaseenzymes other than MAO. The turnover of benzyl-amine by pea seedling amine oxidase and phenylethy-lamine oxidase from A. globiformis was followed at 254and 498 nm, as described above. Final benzylamineconcentrations were 1.5 mM for the pea seedling en-zyme and 300 mM for the bacterial enzyme, based onrespective Km values of 283 and 45 mM obtained in ourlaboratory by the method of Elliott et al. (19, 20).

FIG. 2. Effects on tyramine metabolism of selective inhibitors ofhepatic MAO-A and -B. Metabolism of tyramine (500 mM for MAO-Kinetic Constants for Tyramine Metabolism by MAO-A and 2.5 mM for MAO-B) was estimated colorimetrically, as de-A and MAO-Bscribed in the text, following preincubation of liver homogenates withpargyline (MAO-A), pargyline followed by clorgyline (A / I), clorgy-Kinetic constants were determined for metabolism ofline (MAO-B), clorgyline followed by pargyline (B / I), water (A /tyramine by MAO-A and -B in rat liver homogenatesB), and water followed by clorgyline plus pargyline (A / B / I).which had been preincubated with pargyline or clorgy- Results show the mean { SE from six assays.

line, respectively. Six replicate assays were made inmicrotiter plate wells and contained 50 ml tissue ho-mogenate, 50 ml chromogenic solution, and 200 ml tyr-

RESULTSamine (final concentrations, 10–80 mM for MAO-A and50–400 mM for MAO-B). The increase in absorbance at

Effects of 2,4-Dichlorophenol on MAO Activities490 nm was followed at 377C for 40 min, as describedabove. Data were analyzed using the linear regression Rat brain homogenates were incubated with thefacility of GraphPad Prism, version 1.03. MAO-A substrate, 5-[3H]hydroxytryptamine, or the

MAO-B substrate, [14C]phenylethylamine, in the ab-Measurement of ex Vivo Enzyme Inhibition Following sence (controls) or presence of a range of concentrations

Administration of Phenelzine of 2,4-dichlorophenol. Rat brain MAO-A was inhibitedby 2,4-dichlorophenol with an IC50 value of 44.1 mM,Following acute administration of vehicle, or of thewhile the compound was a more potent inhibitor ofMAO inhibitor, phenelzine, to rats, livers were homoge-MAO-B, displaying an IC50 value of 12.9 mM.nized 1:25 (w/v) in ice-cold potassium phosphate buffer

(0.2 M, pH 7.6). Homogenates were centrifuged at1000g and 47C for 10 min and the supernatants used

Selective Measurement of MAO-A and MAO-Bas the source of MAO. Individual liver homogenatesActivitieswere then divided into two aliquots and were incubated

with clorgyline or pargyline, as described above, to in- Figure 2 shows the effects of selective inhibitors ofhibit MAO-A or -B activities, respectively. Remaining MAO-A and -B upon tyramine metabolism by rat liverMAO activities were then assayed by incubating 50 ml homogenates. Preincubation with clorgyline or pargy-homogenate with 200 ml tyramine (final concentration line, at 500 nM, reduced MAO-A and -B activities by500 mM for MAO-A and 2.5 mM for MAO-B) in the pres- 97 and 93%, respectively, in homogenates where oneence of 50 ml chromogenic solution and following the or other enzyme activity had previously been inhibitedincrease in absorbance at 490 nm and 377C for 45 min. selectively. Clorgyline and pargyline caused minimalSix replicate assays were made in microtiter plate nonselective inhibition of MAO-B and -A, respectively,wells, as described above. since the sum of initial velocities in MAO-A and-B controls was not substantially lower than that in

Protein Assays samples to which inhibitors had not been added. Theseresults indicate the effectiveness of combining the useProtein contents of homogenates were measured by

the method of Lowry et al. (26), with bovine serum of selective inhibitors with a nonselective substrate inorder to distinguish between MAO subtypes.albumin as standard.

AID AB 9911 / 6m25$$$122 12-13-96 14:56:48 aba

SPECTROPHOTOMETRIC ASSAY FOR AMINE OXIDASES 389

Dependence of Reaction Velocity upon EnzymeConcentration

The initial rates of tyramine metabolism by bothMAO-A and -B were found to increase linearly withincreasing enzyme concentration (Fig. 3).

Molar Absorption Coefficient for the QuinoneimineDye

Figure 4 shows a standard curve for hydrogen perox-ide, with data obtained in the presence of rat liver ho-mogenate. The molar absorption coefficient for the qui-noneimine dye under the experimental conditions, theproduct of the slope of the straight line, the reactionvolume, and the reciprocal of the pathlength, was 4654M01 cm01. This value, which takes account of the pres-ence of hepatic catalase activity, is only 15% lower thanthe molar absorption coefficient obtained in the absence

FIG. 4. Standard curve for hydrogen peroxide. Known amounts ofof liver homogenate (not shown) and suggests that cata- hydrogen peroxide were added to chromogenic solution in the pres-lase does not have a substantial effect on estimations of ence of rat liver homogenate, in volumes identical to those used inthe rate of hydrogen peroxide production by MAO. assays of substrate turnover. Plates were warmed to 377C, and end-

point absorbance readings were made at 490 nm. Results show meanabsorbance readings from six wells. The value of S, the molar absorp-tion coefficient which is a function of the slope of the line, is 4654Comparison with an Existing Continuous AssayM01 cm01 (r2 ú 0.999).Procedure for MAO

When the metabolism of benzylamine by MAO fromisolated mitochondria was measured, initial rates

The respective molar absorption coefficients used were(mmol min01 mg01 protein { SE; n Å 4) were calculated12,500 and 4654 M01 cm01.as 7.7 { 0.4 (l 254 nm) and 9.5 { 0.4 (l 498 nm).

Measurement of Benzylamine Turnover by OtherAmine Oxidase Enzymes

The rate of turnover of benzylamine by pea seedlingamine oxidase, measured at 498 nm, was found to be127% of the value obtained when turnover was mea-sured at 254 nm. When benzylamine metabolism bybacterial phenylethylamine oxidase was examined, theratio was 112%.

Kinetic Constants for Tyramine Metabolism by MAO-A and MAO-B

Figure 5 shows Lineweaver-Burk plots for the metab-olism of tyramine by rat liver MAO-A (top) and MAO-B (bottom). Km and Vmax values, calculated from theabscissal and ordinate intercepts, respectively, were,for MAO-A, 35 mM and 33 nmol h01 mg01 protein, andfor MAO-B, 150 mM and 51 nmol h01 mg01 protein.

Measurement of ex Vivo Enzyme Inhibition FollowingFIG. 3. Dependence of reaction velocity upon enzyme concentra-

Administration of Phenelzinetion. A range of volumes of liver homogenate which had been preincu-bated with clorgyline (n) to inhibit MAO-A or pargyline (s) to inhibit Figure 6 shows inhibition of MAO-A (stippled col-MAO-B were then incubated with tyramine, at 2.5 mM or 500 mM, umns) and -B (hatched columns) measured ex vivo fol-respectively, and amine turnover was estimated colorimetrically, as

lowing acute administration of water (controls) ordescribed in the text. Results show the mean { SE from five assays,and error bars do not exceed symbol size. phenelzine (40 mmol kg01, LOW PLZ; or 250 mmol kg01,

AID AB 9911 / 6m25$$$123 12-13-96 14:56:48 aba

HOLT ET AL.390

HIGH PLZ). Enzyme activities in control homogenates(nmol h01 mg01 protein { SE; n Å 5) were 2.7 { 0.4(MAO-A) and 35.5 { 1.4 (MAO-B). A small degree ofselectivity for phenelzine toward MAO-A was evident,this being consistent with results from other ex vivostudies in which discontinuous assay methods wereused (27).

DISCUSSION

Since hydrogen peroxide is a product formed duringdeamination of almost all amine oxidase substrates,it represents an ideal target for a quantitative amineoxidase assay, and although generally less sensitivethan other detection methods, spectrophotometry per-haps represents that technique which is most widely

FIG. 6. Inhibition of rat hepatic MAO ex vivo following administra-accessible to all laboratories. Thus, we have developedtion of the antidepressant, phenelzine. Rats were administered phen-elzine in a single intraperitoneal injection of 40 mmol kg01 (LOWPLZ) or 250 mmol kg01 (HIGH PLZ), 2 h prior to removal of livers.Liver homogenates were preincubated with clorgyline or pargylinebefore addition of tyramine [500 mM to measure MAO-A (stippledcolumns) or 2.5 mM to measure MAO-B (hatched columns)], andremaining MAO activities are expressed as a percentage of those incontrols, injected with water in place of inhibitor. Results showmeans { SE from five animals.

a continuous spectrophotometric assay for hydrogenperoxide which is straightforward, inexpensive, anduses reagents having low toxicities. MAO and SSAOactivities obtained by this method were comparablewith those obtained by direct measurement of benzal-dehyde production at 254 nm. Loss of benzaldehyde asa result of Schiff’s base formation with benzylaminecauses underestimations of enzyme activities mea-sured by the latter method, a fact reflected in the highervalues regularly obtained using the peroxidase-linkedassay. Similar peroxidase-based methods already exist(18, 19). However, the phenols employed are more toxicand, as we have demonstrated here, can also inhibitMAO with IC50 values significantly lower than the con-centrations required for successful coupling in theassay.

In the present assay, 2,4-dichlorophenol has been re-placed by vanillic acid. One drawback to this is thathepatic catalase activity, which metabolizes hydrogenperoxide and which is inhibited by 2,4-dichlorophenol,remains as a potential source of error. Catalase canalso be inhibited by preincubating homogenates with3-amino-1,2,4-triazole in the presence of ascorbic acid(28). However, ascorbic acid has a lower oxidation-re-duction potential than does 4-aminoantipyrine and isFIG. 5. Lineweaver-Burk plots for the metabolism of tyramine by

rat liver MAO-A (top) and MAO-B (bottom). Initial reaction rate is also a proton donor in the peroxidase reaction, with therepresented by v. Rat liver homogenates were preincubated either consequence that no color formation was evident whenwith clorgyline or pargyline prior to incubation with tyramine (10– homogenates were pretreated in this way (not shown).80 mM for MAO-A or 50–400 mM for MAO-B), and rates of change of

For this reason, neither may ascorbic acid be includedabsorbance were measured as described in the text. Data shown arethe means of readings from six assays (r2 ú 0.998). as an antioxidant in aqueous solutions of amine sub-

AID AB 9911 / 6m25$$$123 12-13-96 14:56:48 aba

SPECTROPHOTOMETRIC ASSAY FOR AMINE OXIDASES 391

strates to reduce auto-oxidation. Ascorbic acid can be A transient, initial increase in absorbance is regu-larly observed which is unrelated to the production ofremoved by ultracentrifugation of homogenates and re-

suspension of pellets and, while this step could easily hydrogen peroxide from the amine substrate and whichmay instead result from an interaction of the qui-be incorporated into the process of isolating mitochon-

dria by sucrose density gradient centrifugation, it noneimine dye with one of the reaction constituents.We have also seen a similar effect when dichlorophenol,would certainly prove inconvenient in studies such as

the ex vivo examination of the effects of MAO inhibitors rather than vanillic acid, was included in the reactionmixture. This phenomenon occurred with variousin a number of animals. However, preliminary experi-

ments (not shown) suggested that the reduction in the amine substrates and with MAO and purified SSAOenzymes, required the presence of all assay constit-rate of color formation due to catalase with the present

method was less than 20% and that inhibition of cata- uents, and was reduced, but not abolished, by limitingthe concentrations of 4-aminoantipyrine and peroxi-lase in most experiments was, therefore, unnecessary.

Like ascorbic acid, some MAO substrates such as 5- dase to the minimum necessary for instantaneous cou-pling. While no blank assay could be designed whichhydroxytryptamine, noradrenaline, and other catechol-

amines also have lower oxidation-reduction potentials would account for the effect, following the transient,initial increase, rates of change of absorbance were di-than does 4-aminoantipyrine (29). Thus, when a novel

compound is to be tested as a potential amine oxidase rectly proportional to the concentration of amine oxi-dase enzyme and so data collection did not commencesubstrate by this method, it would be prudent first to

determine whether or not the compound might act as until after this initial period.Oxygen is required as a second substrate by aminea proton donor in the peroxidase reaction. This is

achieved by adding a slight excess of the compound to oxidase enzymes and the Km of MAO toward oxygencan approach the concentration of dissolved oxygen in50 ml of chromogenic solution made up to 1 ml with

buffer in a spectrophotometer cuvette prior to the addi- air-saturated water (10). Over lengthy incubation peri-ods, depletion of oxygen may thus contribute to a depar-tion of 100 nmol hydrogen peroxide. If no color forma-

tion is evident, the compound is unsuitable for use with ture from linearity for substrate turnover. Further-more, if amine turnover is not measured underthis assay system. However, if color formation occurs,

the test should be repeated in the absence of 4-amino- conditions of saturating oxygen concentrations, Vmax

values may be underestimated. Thus, it may be desir-antipyrine to ensure that color formation is therebyprevented; the oxidized forms of 5-hydroxytryptamine able to oxygenate buffers immediately prior to prepara-

tion of reaction solutions, although subsequent warm-and noradrenaline did form colored complexes, al-though of lower intensity and with absorbance maxima ing of solutions will lead to desolvation of gases and

the resulting bubbles may interfere with absorbanceat wavelengths below 498 nm (not shown). Similarly,some phenolic substrates such as m-tyramine condense readings. In this respect, one should note that some

amines, including tyramine, are relatively unstable inwith the oxidized form of 4-aminoantipyrine to give acolored compound and, at the same time, deplete the warm, oxygenated solutions of neutral pH.

The partial pressure of oxygen in air-saturated waterconcentration of amine substrate. Again, the colorformed is likely to have an absorbance maximum differ- is sufficiently high to saturate SSAO enzymes, and oxy-

genation is unnecessary when these enzymes are beingent from 498 nm and such interactions can be testedfor by excluding vanillic acid from a control experiment. studied. We have successfully used the present assay

protocol to measure amine turnover by bacterial andThe concentrations of 4-aminoantipyrine and vanillicacid present in the assay are sufficient to allow forma- plant SSAO (copper amine oxidase) enzymes. The tech-

nique has also proved invaluable in our laboratoriestion of almost 100 mM quinoneimine dye, which, witha microtiter plate pathlength of 9 mm, will produce an during isolation and purification of porcine kidney di-

amine oxidase and porcine and bovine aortic SSAO en-absorbance change of approximately 0.35. Preliminaryexperiments (not shown) suggested that an enzyme zymes, where it has been used to detect amine oxidase

activities in fractions eluted from anion-exchange andconcentration yielding a rate of change of absorbanceat Vmax of between 0.1 and 0.5 h01 would permit both affinity columns. Triton X-100, which has a high ab-

sorbance in the UV range, is included in such purifica-the use of lower substrate concentrations for the deter-mination of kinetic constants and the examination of tion procedures to extract enzymes from cell mem-

branes. Although it forms a milky suspension with 2,4-the effects of inhibitors, cases in which the rate ofchange of absorbance would be significantly lower than dichlorophenol and causes some cloudiness when pres-

ent in the discontinuous ABTS assay (G. Alton, per-in controls. Reagent concentrations are thus more thansufficient to accommodate this degree of absorbance sonal communication), Triton X-100 does not interact

with vanillic acid. Thus, turnover of amine substrateschange. The detection limit below which absorbancemeasurements became unreliable was found to be near in column fraction samples can be followed continu-

ously at 498 nm, without interference resulting from0.01 h01.

AID AB 9911 / 6m25$$$123 12-13-96 14:56:48 aba

HOLT ET AL.392

9. Lyles, G. A., and Callingham, B. A. (1982) Biochem. Pharmacol.the presence of detergents or high protein concentra-31, 1417–1424.tions.

10. Tipton, K. F., and Singer, T. P. (1993) Biochem. Pharmacol. 46,In summary, we have designed a continuous, coupled1311–1316.

assay method for the detection of hydrogen peroxide to 11. Tabor, C. W., Tabor, H., and Rosenthal, S. M. (1954) J. Biol.measure amine metabolism by MAO and SSAO en- Chem. 208, 645–661.zymes, including diamine oxidase and copper amine 12. Weissbach, H., Smith, T. E., Daly, J. W., Witkop, B., and Uden-oxidases. The assay is suitable for determination of friend, S. (1960) J. Biol. Chem. 235, 1160–1163.

13. Yamada, H., Suzuki, H., and Ogura, Y. (1972) Adv. Biochem.kinetic constants for amine turnover, and for estima-Psychopharmacol. 5, 185–201.tion of inhibitor potencies, either in vitro in tissue ho-

14. Krueger, M. J., and Singer, T. P. (1993) Anal. Biochem. 214,mogenates or ex vivo, following drug administration to116–123.the animal. We have also used the assay to show that

15. Matsumoto, T., Furuta, T., Nimura, Y., and Suzuki, O. (1982)some novel amines, metabolites of fluoxetine (Prozac), Biochem. Pharmacol. 31, 2207–2209.are MAO substrates (A. Holt, unpublished work), dem- 16. Zhou, J. J. P., Zhong, B., and Silverman, R. B. (1996) Anal. Bio-onstrating the potential applications of this method to chem. 234, 9–12.basic and clinical research in the fields of toxicology, 17. Szutowicz, A., Kobes, R. D., and Orsulak, P. J. (1984) Anal. Bio-

chem. 138, 86–94.neurochemistry, and drug metabolism.18. Yamada, H., Isobe, K., Tani, Y., and Hiromi, K. (1979) Agric.

Biol. Chem. 43, 2487–2491.ACKNOWLEDGMENTS 19. Elliott, J., Fowden, A. L., Callingham, B. A., Sharman, D. F., and

Silver, M. (1991) Res. Vet. Sci. 50, 334–339.This work was supported by NSERC Grant OGP3045 to M.M.P.20. Elliott, J., Callingham, B. A., and Sharman, D. F. (1992) Comp.We thank Richard Strel for assistance with ex vivo studies.

Biochem. Physiol. 102C, 83–89.21. Holt, A., and Baker, G. B. (1995) Prog. Brain. Res. 106, 187–

REFERENCES 197.22. McGowan, R. E., and Muir, R. M. (1971) Plant Physiol. 47, 644–1. Blaschko, H. (1974) Rev. Physiol. Biochem. Pharmacol. 70, 83–

648.148.23. Tanizawa, K., Matsuzaki, R., Shimizu, E., Yorifuji, T., and Fu-2. Callingham, B. A., Holt, A., and Elliott, J. (1991) Biochem. Soc.

kui, T. (1994) Biochem. Biophys. Res. Commun. 199, 1096–1102.Trans. 19, 228–233.24. Tipton, K. F., Fowler, C. J., and Houslay, M. D. (1982) in Mono-3. Johnston, J. P. (1968) Biochem. Pharmacol. 17, 1285–1297. amine Oxidase: Basic and Clinical Frontiers (Kamijo, K., Usdin,

4. Laux, G., Classen, W., Sofic, E., Becker, T., Riederer, P., Lesch, E., and Nagatsu, T., Eds.), pp. 87–99, Excerpta Medica, Amster-K. P., Struck, M., and Beckmann, H. (1990) J. Neural Transm. dam.32(Suppl.), 189–195. 25. Suva, R. H., and Abeles, R. H. (1978) Biochemistry 17, 3538–

5. Li, X.-M., Juorio, A. V., and Boulton, A. A. (1996) Prog. Brain 3545.Res. 106, 99–112. 26. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

6. Chiba, K., Trevor, A. J., and Castagnoli, N., Jr. (1984) Biochem. (1951) J. Biol. Chem. 193, 265–275.Biophys. Res. Commun. 120, 574–578. 27. Andree, T. H., and Clarke, D. E. (1982) Biochem. Pharmacol. 31,

825–830.7. Boor, P. J., Hysmith, R. M., and Sanduja, R. (1990) Circ. Res.66, 249–252. 28. Barwell, C. J., and Ebrahimi, S. A. (1994) J. Neural Transm.

41(Suppl.), 35–39.8. Yu, P. H. (1986) in Neuromethods: Neurotransmitter Enzymes(Boulton, A. A., Baker, G. B., and Yu, P. H., Eds.), Vol. 5, pp. 29. Nelson, D. R., and Huggins, A. K. (1974) Anal. Biochem. 59, 46–

53.235–272, Humana Press, Totowa, NJ.

AID AB 9911 / 6m25$$$123 12-13-96 14:56:48 aba