Amine oxidase activity in commercial preparations of bovine serum albumin

Life Sciences, Vol. 32, pp. 635-643 Printed in the U.S.A.

Pergamon Press

AMINE OXIDASE ACTIVITY IN COMMERCIAL PREPARATIONS OF BOVINE SERUM ALBUMIN Alia S. AI-Naji and David E. Clarke Department of Pharmacology College of Pharmacy University of Houston Houston, TX 77004 USA (Received in final form October 22, 1982)

Summary Amine oxidase activity has been identified in commercial samples of bovine serum albumin (BSA). Benzylamine, phenylethylamines and to a lesser extent, indoleamines, were found to be substrates. The amine oxidase activity was inhibited by semicarbazide and was virtually absent in electrophoretically purified samples. Kinetic analysis of benzylamine deamination and experiments utilizing mixed substrates indicate that more than one catalytic activity may be involved. The results show that amine deamination should be considered as a potential source of error in experiments employing high concentrations of commercially available B S A preparations. This would be of particular importance for in vitro studies with dopamine since this amine was found to b e - d ~ n a t e d at a rapid rate. Introduction We report that commercial samples of BSA contain amine oxidase activity. Strong indications of oxidase activity in samples of BSA was first reported by Hirsch (1) and led to the discovery of polyamine deamination (spermine oxidase) in bovine plasma (2). In the present study, we show that several aromatic monoamines, such as dopamine, can act as substrates. These observations have important experimental implications since the biological properties of aromatic monoamines are studied frequently in vitro where commercial samples of BSA are used as a constituent of the ~ s a y , culture or perfusion media. Under these experimental conditions, the oxidase activity in commercial samples of BSA can be expected to introduce serious artifactual errors when utilizing amines which act as substrates. Methods and Materials Stability and preparation of BSA. Commercial crystalline or powdered sampl-6~---6Y-]~S~i--were stored aY-4~Z-and were made-up in 1 mM potassium phosphate buffer (pH 7.8) for assay. Solid samples retained stable amine oxidase activity for up to one year (longest time tested) while the dissolved samples were stable for at least one month (longest time tested). Exposure of solid or dissolved samples to air at room temperature for 24hr did not affect the specific activity of the amine oxidase.

0024-3205/83/060635-09503.00/0 Copyright (c) 1983 Pergamon Press Ltd.

636

Oxidase Activity in Bovine Serum Albumin

Vol. 32, No. 6, 1983

Enzymatic ~ . Activities were determined radiochemically (3,4). In this procedure, t~le radio-labelled deaminated products are extracted from an aqueous solution into an organic phase by acidification. Aliquots of BSA solution (40 mg/ml) were incubated with 50 ,1 of labelled substrate (10 ~Ci/pmole) prepared in 0.2 M phosphate buffer, pH 7.8. If necessary, water was added to give a final volume of 100 ~1. Unless indicated otherwise, inhibitors were preincubated with the BSA solution for 20 min at 37°C prior to the addition of substrate. Quoted concentrations represent final concentrations in the assay tubes. All reactions were run in triplicate and terminated by cooling the tubes on ice followed by the addition of 10 ,1, 3N HC1. Blank values were obtained by adding 10 ,1, 3R HC1 just after the addition of substrate. The deaminated products were extracted in 600 ,1 of toluene: ethyl acetate (1:1 saturated with water) and a 400 ,1 aliquot taken for liquid scintillation spectrometry with quench correction. None of the drugs used altered the extraction efficiency of the deaminated metabolites which ranged from 77 percent with dopamine to 96 percent with benzylamine. Quoted specific activities are corrected for recovery. Km and Vmax determinations. Resulting double reciprocal plots were analyse-e-d-~ computer program using linear regression analysis and the method of Wilkinson (5). Little difference between the derived Km and Vmax values were noted providing sufficient data points were generated. Values quoted in Results were derived by linear regression. Chemicals. BSA samples were purchased from Sigma Chemical Company, St. Louis, MO, USA and Calbiochem-Behring Corporation, La Jolla, CA, USA. The catalog and lot numbers are given in Table I of the Results. [14C] Benzylamine hydrochloride (12.5 mCi/mmole) was purchased from ICN Pharmaceuticals, Irvine, CA, USA. [14C]Dopamine hydrochloride (44.8 mCi/mmole), [14C]tyramine hydrochloride (50.7 mCi/mmole) and [14C]5hydroxytryptamine binoxalate (51.5 mCi/mmole) were purchsed from New England Nuclear, Boston, MA, USA. [SH]Tryptamine hydrochloride (0.71 Ci/mmole) was purchased from Amersham, Arlington Heights, IL, USA. The specific activities of the labelled compounds were adjusted to 10 pCi/pmole with the respective unlabelled salts. All other compounds used were of analytical grade where possible. (-)-Deprenyl was generously donated by Dr. J. Roth, State University of New York at Buffalo, Buffalo, NY, USA. Results Table I gives the specific activities for benzylamine deamination by various samples of BSA and a single sample of human serum albumin (sample 8, Table I). The samples were obtained from Sigma Chemical Company and Calbiochem-Behring Corporation. All samples are fraction V as prepared by the general methods of Cohn et al. (6) and Cohn et al. (7), but vary in purity (all are at least 95 percent albumin), p H - a n K precise method of purification. Purification by electrophoresis (samples 7 and 8) showed only very low deaminating activity, suggesting that a contaminant, rather than albumin itself, is involved. The deamination of benzylamine was found to be linear with reaction time, BSA concentration and obeyed Michaelis-Menten kinetics. From four separate experiments, the mean apparent Km ( ~ M ) and Vmax (pmol benzylamine deaminated/mg protein/min) values were: 132.5 -+ 2.4 and 91.6 _+ 6.0, respectively. These values were obtained using sample 5 (Table I) and benzylamine concentrations of 16.7, 20, 25, 33.3, 50, 100 and 200 pM.

Vol. 32, No. 6, 1983

0xidase Activity in Bovine Serum Albumin

TABLE

637

I

Deamination of Benzylamine by Commercial Preparations of Bovine Serum Albumin Sample Number

Source

Catalog Number

Lot Number

1 2 3 4 5 6

Sigma Sigma Sigma Sigma Sigma Calbiochem

A-8022 A-9647 A-4503 A-4503 A-4503 12659

I09C-0400 97C-01531 37C-0354 68C-0208 78C-0202 802915

72 8*+

Calbiochem Calbiochem

12657 126658

903565 902948

Specific Activity i.I 7.4 40.2 21.5 58.5 77.0

± ± ± ± ± ±

0.35 0.26 1.43 0.67 1.20 1.53

0.2 ± 0.07 0.4 ± 0.06

Activity was assayed for 5 min with benzylamine (500 pM) using i mg bovine serum albumin and is expressed as pmoles deaminated/mg protein/ min ± the standard error of the mean f%r three experiments, each run in triplicate. *electrophoretically purified, human serum albumin.

Unless stated otherwise, all subsequent experiments were done using sample 5 since our original findings were made using this sample. Figure 1 shows that benzylamine deamination was inhibited in a concentration-dependent fashion by semicarbazide following a 20 rain preincubation prior to the addition of substrate. The inhibitory action of semicarbazide was partially time-dependent since concentrations of 0.1 and 1.0~M produced only 50 percent of the inhibitory effect shown in Figure 1 when added at the same time as benzylamine (no preincubation). Several other compounds were tested for inhibitory activity at 1.0~JM using a 20 min preincubation period. In these experiments, benzylamine deamination was not altered by potassium cyanide or the acetylenic monoamine oxidase (MAO) inhibitor (-)-deprenyl but was inhibited completely by cuprison and the hydrazine-containing MAO inhibitor, phenelzine (data not shown). Various amines were tested for inhibitory activity toward the deamination of benzylamine (67.5 ~M; about half its Km concentration) and the results are given in Table II. Of these amines, dopamine, 2-phenylethylamine, tyramine, tryptamine and 5-hydroxytryptamine were tested at 250 ~M to determine whether they acted as substrates for the oxidase. The results, compared with those for benzylamine, are given in Table III. Dopamine was found to be deaminated at a rate equal to benzylamine, whereas the related compounds, 2-phenylethylamine and tyramine, were less active. By comparison, the indoleamines, tryptamine and 5-hydroxytryptamine, were poor substrates. Where tested, all deamination was inhibited by preincubation with semicarbazide (1.0raM). Since benzylamine and dopamine were deaminated at the most rapid rates (Table III), mixed substrate experiments were undertaken to examine the kinetic interaction between these two amines. Table IV shows that unlabelled dopamine produced competitive inhibition of [14C]benzylamine deamination and a Ki value of 139.6 pM was obtained for dopamine. In the reverse

638

Oxidase Activity in Bovine Serum Albumin

Vol. 32, No. 6, 1983

,oo 8O .~

60

~

20 0

10

8

6

4

2

-LOG M SEMICARBAZIDE

FIG. 1 Inhibition by semicarbazide of benzylamine deamination. Activity remaining was assayed for 5 rain with benzylamine (10 ~M) after preincubation of 0.5 mg bovine serum albumin with semicarbazide for 20 rain at 37°C. Each point is the mean + the standard error of the ratio (percentage) for three experiments each assayed using triplicate determinations.

experiment, however, [14C]dopamine deamination was not influenced by unlabelled benzylamine, except for a slight Km shift at the highest concentrat_ion of benzylamine (150 pM)*. The results given in Table IV suggest the involvement of more than one catalytic activity and an experiment using a much wider range of benzylamine concentrations (25, 33.3, 50, 100, 200, 250, 333, 500, 666, 750 and 1000 pM) revealed low (199.1 ~M) and high (603.4 pM) Km values for benzylamine (Fig. 2). In view of this finding it was necessary to determine whether higher concentrations of benzylamine (around its high Km concentration) would now show inhibition toward dopamine deamination. Table V shows that 400 and 800pM concentrations of benzylamine acted competitively toward [14C]dopamine deamination. A Ki value for benzylamine of 563.5 pM was obtained from these data. This value is not inconsistent with the high Km value obtained for benzylamine deamination in Figure 2 (603.5 ~M). *In these and subsequent experiments it was necessary to use another bottle of BSA sample 5 (Table 1). The new sample gave a higher Km value for benzylamine even though the sample carried the same "lot number."

Vol. 32, No. 6, 1983

Oxidase Activity in Bovine Serum Albumin

639

TABLE II Effect of Amines on the Deamination of Benzylamine by Bovine Serum Albumin Amine (l.0mM)

Percent Activity

Control (benzylamine alone) Monoamines: Dopamine Tyramine Octopamine Phenylethylamine N-methylphenylethylamine Amphetamine Tryptamine 5-Hydroxytryptamine Diamines: Cadaverine Putrescine Polyamines: Spermine Spermidine Peptide: Lysyl vasopressin

100 23.8 25.5 30.8 40.2 97.4 94.7 34.8 62.0 91.5 104.2 32.2 33.0 85.3

Activity was assayed for 5 min with benzylamine (67.5 NM) using 1 mg bovine serum albumin. The amines were added at the same time as benzylamine and the reactions were run in triplicate.

TABLE Ill Specific Activities for Amine Deamination by Bovine Serum Albumin and Inhibition by Semicarbazide Amine Benzylamine Dopamine Phenylethylamine Tyramine Tryptamine 5-Hydroxytryptamine

Specific Activity* 42.73 41.43 17.62 10.93 2.66 0.43

± ± ± ± ± ±

0.62 0.18 0.15 0.12 0.15 0.03

Semicarbazide percent inhibition I00 i00 100 100 100 ND

Activity was assayed for 5 min with the respective amines (250 ~M) using 1 mg bovine serum albumin. Semicarbazide (l.0mM) was preincubated with the bovine serum albumin for 20 min at 37° prior to the addition of substrate. *pmoles deaminated/mg protein/min ± the standard error for six determinations. ND = not determined.

640

Oxidase Activity in Bovine Serum Albumin

Vol. 32, No. 6, 1983

TABLE IV Mixed Substrate Experiments for the Deamination of Benzylamine and Dopamine by Bovine Serum Albumin Substrate

Competing Amine

[14C]Benzylamine " " "

None Dopamine (25 NM) Dopamine (50 NM) Dopamine (100 NM)

178.2 257.5 262.4 298.9

85.1 97.0 91.0 87.6

[14C]Dopamine

None Benzylamine (25 pM) Benzylamine (75 NM) Benzylamine (150 ~M)

135.7 124.2 128.8 148.9

57.7 51.4 51.1 51.3

" "

Km*

Vmax**

A c t i v i t i e s were assayed for 5 min with [z4C]benzylamine (25, 33.3, 50, 100 and 200 pM) and [z4C]dopamine (25, 33.3, 50, 100, and 150 NM) using 1 mg bovine serum albumin. The competing, non-radioactive amine, was added at the same time as the labelled amine. A Ki value of 139.6 NM was obtained for dopamine against benzylamine deamination but no reliable Ki value could be derived for the reverse interaction. *NM. **pmoles deaminated/mg protein/min.

10

151

8

>

6 4 2

I

-0.5

o

0:5

I

I

2

;

1IS (.uM)x 10 - 2

FIG. 2 Double reciprocal plot of benzylamine deamination. Abscissa: reciprocal of benzylamine concentration (pM); ordinate: reciprocal of velocity (v) of enzyme reaction in arbitrary units. Activity was assayed for 5 rain with benzylamine (25, 33.3, 50, 100, 200, 250, 333, 500, 666, 750, and 1000 pM) using 1 mg bovine serum albumin. Each point is the mean of three determinations. Inset: Eadie-Scatchard plot of v divided by the substrate concentration (s) versus v.

Vol. 32, No. 6, 1983

Oxidase Activity in Bovine Serum Albumin

641

TABLE V Kinetic Constants for the Deamination of Dopamine by Bovine Serum Albumin in the Absence and Presence of Benzylamine Substrate

Competing Amine

[14C]Dopamine

None Benzylamine (400 pM) Benzylamine (800 pM)

"

Km* 140.7 285.4 420.0

Vmax** 58.2 57.6 59.8

Activity was assayed for 5 min with [14C]dopamine (25, 33.3, 50, 100 and 200 NM) using 1 mg bovine serum albumin. Non-radioactive benzylamine and [14C]dopamine were added at the same time. A Ki value of 563.5 pM was obtained for benzylamine. *pM. **pmoles deaminated/mg protein/min. Discussion In 1953, Hirsch (1) suspected amine oxidase activity in commercial samples of BSA and subsequently the enzyme spermine oxidase was discovered in bovine and sheep plasma (2,8). This enzyme is now believed to be copper-containing and to oxidize the primary amino group of certain aliphatic and aromatic amines, including benzylamine (1,8-11). Since the work of Hirsch (1), no further studies concerning deamination by commercial samples of BSA have been reported. The present work shows that several aromatic monoamines can act as substrates and indicates that more than one catalytic site may be involved. The deamination does not stem from albumin itself, since electrophoretically purified samples were virtually devoid of activity (Table I). According to the manufacturers, the majority of impurities in BSA preparations are protein globulins and lipid. It seems, therefore, that the deaminating properties derive from these contaminants which, in the samples studied, amount to no more than 5 percent of the total material. Although we have only tested BSA samples purchased from Sigma and Calbiochem, the preparative methods of Cohn et al. (6,7) are used widely. Thus, amine oxidase activity should be suspected as a contaminant in all commercial samples of BSA (fraction V) where Cohn's method forms the basis of preparation. During the course of the present experiments it was discovered that benzylamine exhibits both high (603pM) and low Km values, the latter ranging from 132 to 199 pM, depending upon the experiment and BSA sample used. Since Table II characterizes benzylamine deamination at 67.5 pM, these results relate primarily, but not exclusively, to the low Km site. The data show that none of the benzylamine deamination is attributable to a MAO (E.C. 1. 4. 3. 4), diamine oxidase (E.C. 1. 4. 3. 6.) or lysyl oxidase. Benzylamine deamination by MAO would have been inhibited by N-methylated and alphamethylated phenylethylamines, as well as by the potent MAO inhibitor, (-)-deprenyl (10,12). Similarly, diamine oxidase would have been susceptible to inhibition by both cadaverine and putrescine (10). Lysyl oxidases are excluded on the grounds that lysyl vasopressin was virtually without effect (13). Furthermore, according to Shieh et al. (13) benzylamine is not a substrate for lysyl oxidase, even though some nonpeptidyl amines can act as

642

Oxidase Activity in Bovine Serum Albumin

Vol. 32, No. 6, 1983

substrates (14). Recently, it was reported that purified bovine serum amine oxidase (spermine oxidase) does possess a weak ability to oxidize peptidylbound lysine (15). Thus, the small influence of lysyl vasopressin, if real, may be reflective of this property. The involvement of spermine oxidase is indicated by the marked inhibitory effect of spermine and spermidine. From Table II, however, it can be seen that dopamine proved to be somewhat more active than the polyamines. The inhibition of benzylamine deamination by semicarbazide and phenelzine is consistent with the known properties of bovine plasma spermine oxidase (10,16,17). Although both compounds are carbonyl-trapping agents, it cannot be assumed that the activity is pyridoxal-dependent (16,17). Hirsch (1) and Tabor et al. (8) showed a concentration-dependent inhibition of bovine and sheep sp~rm-lne oxidase with cyanide but found no effect with EDTA (8). In the present study, cuprison, the avid copper chelating agent, produced complete inhibition whereas cyanide was inactive. These data are puzzling since cyanide is also believed to work via the copper component of plasma amine oxidase (18). However, Lindstrom and Petterson (19) have suggested that cuprison can act via Schiff-base formation, rather than by metal chelation. It is possible, therefore, that the copper component of the amine oxidase in commercial samples o f BSA is rendered insensitive to cyanide due, perhaps, to modifications sustained during the preparation of the albumin. The kinetic studies involving mixed substrates indicate that although dopamine binds to the high affinity site for benzylamine, dopamine itself is not deaminated at this site. Whereas dopamine acted as a competitive inhibitor of benzylamine deamination with a Ki of 139.6 ~M, its own deamination was not inhibited by the same range of benzylamine concentrations (Table IV). Rather, dopamine appears to be deaminated at the high Km site for benzylamine since higher concentrations of benzylamine (around its high Km value) were required to inhibit dopamine deamination (Table V). Furthermore, the derived Ki value for benzylamine (563.5 ~M) is in close agreement with its Km value (603.4 ~M) for the high Km site. While two sites seem to be the most satisfactory explanation for the kinetic data, it is not known whether these two sites are located on a single enzyme or whether two distinct enzymes are involved. Although the latter possibility seems most likely (17), as yet, the two activities have not been separated. The present results have relevance to experimental biology. BSA is used in biological experiments such as enzyme assays (20) and as a constituent in perfusion media (21,22). Concentrations of 70 to 80 mg/mi BSA are quite usual for the perfusion of organs in vitro. These concentrations of BSA are at least 70 to 80 times higher t-~an--Those used in the present experiments and would be anticipated, therefore, to give rise to marked deamination in studies employing susceptable amines. For example, there is much current interest in the renal actions of dopamine, yet the medium used for isolated perfused kidneys contains BSA (22). Deamination of exogenously added or endogenously released dopamine by the contaminating oxidase could lead to misconceptions and misinterpretations concerning the physiological and/or pharmacological importance of this amine. Although the use of semicarbazide to inhibit deamination (Fig. 1) or electrophoretically purified BSA (Table 1) would largely circumvent this problem, semicarbazide (and other hydrazines) are not specific in their actions and the expense of electrophoretically purified BAS may prove prohibitive. In this regard, Andree and Clarke (23) utilized polyvinylpyrrolidone instead of BSA, and obtained viable, long-lasting isolated perfused rat brain preparations when analysing the metabolic fate of benzylamine.

Vol. 32, No. 6, 1983

Oxidase Activity in Bovine Serum Albumin

643

References i. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

17. 18. 19. 20. 21. 22. 23.

]. G. HIRSCH. J. Exp. Med. 97, 327-343 (1953). J.G. HIRSCH. J. Exp. Med. 97, 345-355 (1953). R. E. McCAMAN, M. W. McCKIVIAN, ]. M. HUNT, AND M. S. SMITH. J. Neurochem. 12, 15-23 (1965). B.A. CALLINGTIAM and R. LAVERTY. J. Pharm. Pharmac. 25, 940-947 (1973). G . N . WILKINSON. Biochem. ]. 80, 324-332 (1961). E . J . COHN, L. E. STRONG, W.-]::. HUGHES, JR., D. L. MULFORD, J. N. ASHWORTH, M. MELIN, and H. L. TAYLOR. J. Am. Chem. Soc. 6_88, 459-475 (1946). E. J. COHN, W. L. HUGHES, JR., and J. H. WEARE. J. Am. Chem. Soc. 69, 1753-1761 (1947). C. W_TABOR, H. TABOR, and S. M. ROSENTHAL. J. biol. Chem. 208, 645-661 (1954). RT-. B. RUCKER and B. L. O'DELL. Biochem. biophys. Acta. 235, 32-43 (1971). H. BLASCHKO. Rev. Physiol. Biochem. Pharmac. 70, 83-148 (1974). F. BUFFONI. Pharmac. Res. Commun. 12, 101-114 (19-8-0). ]. KNOLL and K. MAGYAR. Adv. Bioc-[iem. Psychopharmac. 5_, 393-398 (1972). J. J. SHIEH, R. TAMAYE, and T. YASUNOBU. Biochem. biophys. Acta. 377, 229-238 (1975). PT-.C. TRACKMAN and H. M. KAGAN. J. biol. Chem. 254, 7831-7836 (1979). O. ODA, T. MANABE, and T. OKYAMA. J. Biochem. 8__9_9, 1317-1323 (1981). K. WATANABE, R. A. SMITH, M. INAMASU, and K. T. YASUNOBU, Adv. Biochem. Psychopharmac. 5, 107-117 (1972). K. T. YASUNOBU, H. ISHIZA-'KI, and N. MINAMIUTA. Molec. cell. Biochem. 13, 3-29 (1976). R. BARKE--R, N., BODEN, G., CALEY, S. C., CHARLTON, R., HENSON, M. C., HOLMES, I. D., KELLY, and P. F., KNOWLES. Biochem. J 177, 289-302 (1979). A. LINDSTR~YM and G., PETTERSON. Eur. J. Biochem. 4__88, 229-236 (1974). M. M-O. HUH and A. J . , FRIEDHOFF. I. biol. Chem. 454, 299-308 (1979). R. K. ANDJUS, K., SUHARA, and H. A., SLOVITER. J. appl. Physiol. 22, 1033-1039 (1967). J. C. SF-FRAY and A. S., KARUZA. J. Physiol., Lond. 299, 45-54 (1980). T. H. ANDREE and D. g., CLARKE. Biochem. Pharmac. 3_0, 959-965 (1981).