Eur. J. Biochem. 83, 131 - 135 (1978)

Active-Site Titration of Pig-Plasma Benzylamine Oxidase Anders LINDSTROM and Gosta PETTERSSON

Avdelningen for Biokemi, Kemicentrum, Lunds Universitet

(Received August 31, 1977)

1. Titration of benzylamine oxidase with benzylamine under anaerobic conditions shows that full reduction of the enzymic 470-nm chromophore is obtained on the addition of one mole of substrate per mole of enzyme. Concomitantly, one mole of benzaldehyde per mole of enzyme is produced.

2. A single prosthetic group interacting with carbonyl reagents can be detected on titration of benzylamine oxidase with phenylhydrazine. Titration data reported to indicate a higher content of prosthetic groups were obtained under conditions where equilibration between enzyme and phenylhydrazine is insufficiently complete.

3 . It is concluded that pig-plasma benzylamine oxidase contains a single catalytically active site. This means that the two copper atoms present in the enzyme may be structurally or functionally different.

Pig-plasma benzylamine oxidase, which catalyzes the oxidative deamination of various monoamines, contains copper and a prosthetic group interacting strongly with carbonyl reagents [l - 31. We have shown previously that the absorbance changes and concomitant inhibition of enzyme activity associated with the binding of hydrazine derivatives to this prosthetic group can be used for active-site titration of the enzyme [4]. The titration data reported by us provided evidence that benzylamine oxidase contains a single reactive prosthetic group per enzyme molecule. The latter result was recently questioned by Buffoni and Ignesti, who concluded from similar experiments that there are three reactive prosthetic groups per enzyme molecule [ 5 ] . These controversial results led to different inferences about the number of copper atoms (two according to us; one according to Buffoni and Ignesti) being present per active site of the enzyme, which introduces a corresponding uncertainty con- cerning the functional role of enzyme-bound copper in the catalytic mechanism.

In relation to magnetic resonance spectroscopic studies of ligand binding to copper in benzylamine oxidase it became of importance to resolve the above controversy. The data reported in the present in- vestigation show that the apparently conflicting ob- servations made on active-site titration of benzylamine

Enzyme. Benzylamine oxidase or amine oxygen oxidoreductase (deaminating) (EC 1 4 3 6 )

oxidase with phenylhydrazine can be attributed mainly to differences in the experimental conditions used. Independent determinations of the number of cata- lytically active sites have now been made by anaerobic titration of the enzymic 470-nm chromophore with benzylamine, as well as by examination of the stoichio- metry of catalytic benzaldehyde production under anaerobic conditions. The results obtained provide clear evidence that pig-plasma benzylamine oxidase contains a single reactive prosthetic group.

EXPERIMENTAL PROCEDURE

Materials

The preparation of homogeneous benzylamine oxidase from pig plasma, as well as methods for deter- mination of protein concentration and enzyme activity, have been described previously [6]. Enzyme prepara- tions used in the present investigation were at least 95 pure according to specific activity determina- tions.

Other reagents used were commercial samples of analytical grade. [7-'4C]Benzylamine . HCl (2.36 Ci/ mol) was obtained from Mallinckrodt (St Louis, U .S .A,).

Anaerobic Experiments Unless otherwise stated, experiments reported in

the present paper were performed at 25 "C in 0.1 M phosphate buffer pH 7.0.

z 32 Titration of Benzylamine Oxidase

Anaerobic conditions (less than 1 pM oxygen) were obtained by the nitrogen-flushing technique described previously [7]. Enzyme solutions being deoxygenated contained 1 :,( (v/v) of polyethylene- glycol 6000 to prevent denaturation of the protein. Under such conditions no detectable (less than 3 "/,) loss of enzyme activity occurred during the deoxy- genation procedure. Control experiments established that the presence of 1 polyethyleneglycol in the reaction solutions had no inhibitory or activating effect per ,re on enzyme activity, as determined by steady-state kinetic techniques at different concentra- tions of substrate.

Anaerobic titrations were performed in a spectro- photometric observation cell similar to that described by Reinhammar [8]. The cell was equipped with a side-arm, and was sealed with a gas-tight rubber membrane through which substrate could be added under anaerobic conditions using a Hamilton syringe. In a typical titration experiment the cell contained initially about 100 nmol of enzyme in 2.0 ml of de- oxygenated phosphate buffer. A deoxygenated buffer solution of the substrate was then added in 5O-pl portions, and the absorbance of the reaction solution was recorded 5 min after each addition of substrate. When required, perchloric acid was tipped in from the side-arm to denature the enzyme. All absorbance measurements were made with an Aminco DW2 UVjVIS spectrophotometer.

Benzaldch I&> Production

The catalytic production of ''C-labelled benzal- dehyde from [7-'4C]benzylamine in the presence of benzylamine oxidase was assayed by the method of Oi et (11. [9]. Reactions were performed in the absence of oxygen as described for the anaerobic titration experiments. After complete titration of the enzyme with an excess of radioactively labelled substrate, 0.5 nil of 70°0 perchloric acid was tipped in from the side-arm to stop the reaction and to acidify the reac- tion solution. The aqueous reaction solutions (2.8 ml) were then extracted with 10 ml of ethylacetate. The radioactivity of the ethylacetate extracts was deter- mined by standard counting techniques in a Searle mark 111 model 6880 liquid scintillation system.

Plzenylhpilrurincl Titrations

Spectrophotonietric active-site titrations of benzyl- amine oxidase with phenylhydrazine at 430 nm were performed as described previously [4]. Freshly pre- pared solutions of phenylhydrazine were added in 5O-yl portions to 2 ml of a 0.1 M phosphate buffer solution (pH 7.0) containing varied initial concentra- tions of enzyme (0.2- 12 pM). These concentrations of enzyme are lower than those (20-40 yM) pre-

viously recommended by us for active-site titrations [4] and were used to examine the effects of slow equilibration between enzyme and phenylhydrazine. Reaction solutions were invariably allowed to equili- brate for 10 min after each addition of phenylhydra- zine.

Copper Determination5

Copper determinations reported in the present paper were performed with a Unicam SP90 atomic absorption spectrophotometer.

RESULTS

Anaerobic Titration of Benzylamine Oxrdase with Substrate

Pig-plasma benzylamine oxidase exhibits a weak absorption band centered at 470 nm [I]. This absorp- tion band disappears during catalysis on reduction of the enzyme by amine substrates, and reappears on reoxidation of the enzyme by molecular oxygen [lo-121. Since reduction of the enzyme by amine substrates is known to be essentially irreversible [7], an estimate of the number of catalytically active sites can be obtained spectrophotometrically at 470 nm by titration of the enzyme with benzylamine under anaerobic conditions.

Fig. 1 shows the results obtained in such an active- site titration of benzylamine oxidase at pH 7. The addition of less than equimolar amounts of benzyl- amine to the enzyme solution led to an approximately proportional decrease of the 470-nm absorbance (measured 5 min after each addition of substrate and corrected for dilution effects). Full reduction of the enzymic 470-nm chromophore was obtained with 1.0 ('~0.1) mol of substrate per mol of enzyme; the addition of a further two equivalents of substrate did not produce any significant additional absorbance changes. The absorption coefficient for the 470-nm chromophore calculated from the data in Fig. 1 was 2400 M-' ern-', which agrees well with the value reported previously [ 1 1 3 .

The same results, indicating the presence of a single catalytically reactive 470-nm chromophore per protein molecule, were obtained on titration of the enzyme with benzylamine at five different pH values ranging from 6.8 to 9.1.

Production of Benzaldehyde under Anaerobic Conditions

Quantitative determination of the amount of benzaldehyde catalytically produced from benzyl- amine in the presence of benzylamine oxidase was performed by isotope techniques similar to those described by Oi er a/ . [9]. Benzylamine oxidase

A. Lindstrom and G. Pettersson 133

0.1 1 I I I 0 1 2 3

Benzylarnine added (equiv.) Fig. 1. Spectrophotometric titration oJ henzjlumine oxidase with substrate under anaerobic conditions. The enzymic 470-nm chromo- phore was reduced by titrating 128 nmol of enzyme in 2 ml 0.1 M phosphate buffer (pH 7.0) with 50.~1 portions of 0.8 m M benzyl- amine. Absorbances (A470) have been corrected for the dilution effect and refer to a volume of 2 ml

(76 nmol) was titrated anaerobically at pH 7 using l4C-labelled benzylamine with a specific activity of 4920 counts min-' nmol-'. Full reduction of the enzymic 470-nm chromophore was obtained on the addition of an equimolar amount of benzylamine. After the addition of a further two equivalents of ''C-labelled substrate the reaction solution was al- lowed to stand for 10 min, when perchloric acid was tipped in to denature the enzyme. The amount of 14C-labelled benzaldehyde formed was then estimated by liquid scintillation counting of an ethylacetate extract of the acidified reaction solution. As shown in Table 1 (expt l), the ethylacetate extract contained 31 % of the radioactivity added. This corresponds to 0.92 mol of benzaldehyde produced per mol of enzyme (uncorrected value).

To determine the amount of unreacted benzyl- amine extracted from the reaction solution, the experi- ment was repeated in the absence of enzyme. The ethylacetate extract thus obtained contained only 1 % of the radioactivity added (Table 1, expts 2 - 3).

This indicates that at least 97% of the radioactivity extracted in the main experiment derives from benzaldehyde formed during the anaerobic reduction of enzyme by substrate.

To determine the efficiency of the benzaldehyde extraction procedure, the experiment was repeated under aerobic conditions. In this case the enzymic reaction was allowed to proceed for 120 min (sufficient to yield a more than 99% complete conversion of benzylamine into benzaldehyde [7]) before acid was tipped in. The ethylacetate extract thus obtained contained 93 - 95 % of the radioactivity added (Table 1, expts 4- 5).

After correction for the observed efficiency of the extraction procedure (1 % for benzylamine; 94 % for benzaldehyde) the calculated amount of benzaldehyde formed in the main experiment becomes 0.96 mol per mol of enzyme. Considering the precision of the experi- ments, this result leads us to conclude that one mole of benzaldehyde per mole of enzyme is catalytically produced during the anaerobic reduction of benzyl- amine oxidase by substrate.

Copper Content of Benzylamine Oxiduse

Enzyme preparations used in our studies of benzylamine oxidase during the last three years have been routinely assayed for copper by atomic absorp- tion spectrometry. Summing these analyses up, we find that the mean copper content of six different preparations with a purity exceeding 95 % corresponds to the presence of 2.1 ( f 0.2) copper atoms per protein molecule, calculated using a molecular weight of 196000 for the enzyme [1,2]. This result agrees with that reported previously by us [13], as well as with that obtained by Buffoni et al. [14] and by Boden et ul. [15]. Consequently, the suggestion [5] that benzyl- amine oxidase contains three active sites, each in- volving one copper, is inconsistent with the observed copper content of the enzyme.

Less extensively purified enzyme preparations (60- 90 % pure) have usually been found to exhibit

Table 1. Catalytic production of'benzaldehyde.from benzylamine under anaerobic conditions [7-'4C]Benzylamine (4920 counts min-' nmol-') was reacted with benzylamine oxidase in 0.1 M phosphate buffer (pH 7.0) for 20 min (absence of oxygen) or 120 min (presence of oxygen), I4C-labelled benzaldehyde being extracted with ethylacetate after acidification of the reaction solution

Expt Amounts of reactants Total radioactivity Fraction - extracted

oxygen enzyme benzylamine added extracted - ~~

m M nmol counts min-' % ~~~~ ~ ~ ~~ -~

0 76 228 1 122000 344000 31 0 0 152 748 000 7 200 1 0 0 228 1 122000 11 000 1 0.25 76 76 374000 348 000 93 0 25 76 228 1122000 1068000 95

I34 Titration of Bcnzqlamine Oxidase

Phmylhydra-ine added (equiv)

Fig. 2. , S / ~ ( ' ~ ~ / ~ ( I / ) / / ~ I I ~ ) / I ~ ~ ~ ~ I ( riwu/ion o/ hcnzylaiiiiizr o\-iriu.ve i v i f h / ,k( .r i~. / /7~.r lr .rr_i l lc . ~ r t 430 m i . Varied concentrations of enzyme ( c ~ ) litrated with phcnylti!draiine i n 0.1 M phosphate buffer pH 7.0. Absorbance changes (1 were measured 10 min after each addition of phcnylhydrarine. and have been corrected for the dilution effect. Arrows indicate final absorbance changes after the addition of approximately 6 equivalents of phenylhydrazine

an up to 40' higher copper-to-protein ratio. This can probably be attributed to the presence of cerulo- plasmin and or non-specifically bound copper as iinpuritie\.

Titration qf ' Bctiryluminc Oxiduse with Plirii,1,Ili,~,lh.ci,-inc

The irreversible binding of phenylhydrazine to benzylaniinc oxidase leads to the appearance of an absorption band centered at 430 nm [4]. Fig.2 shows the 430-nm absorbance changes (corrected for dilu- tion') observed on titration of benzylamine oxidase with phenylh>,drazine at different initial concentra- tions of enzyme. Reaction solutions were allowed to equilibrate for 10 min after each addition of phenyl- hydrazine, which agrees with the conditions used by us and by Wuffoni and Ignesti in previous titration studies [4.5]. As indicated by the data shown in Fig. 3, a reaction time of 10 min is sufficient to obtain a satisfactorily complete equilibration as long as the concentration of unreacted enzyme exceeds 3 - 5 pM. At lower concentralions of unreacted enzyme the equilibration with phenylhydrazine becomes signif- icantly incomplete. resulting in curvature of the cor- responding portion of the titration plots (Fig. 2).

d O'ol i 4 3

" 0 10 20 30 40 5L\ 60 Time (mini

Fig. 3. Time-dependence of the C ~ q U f / f h l ' U f i i J / l /ielwcvn her1:.1./u/f7inc~ onidase undphi~nylhy~lrri-irzr. Absorbance chanpcs at 430 nm record- ed after the addition of half an equivalent of phenylhydrazinc t o varied concentrations of enzyme (cI ) in 0.1 M phosphate buffer pH 7.0. Arrows indicate absorbance levels obtained after B reaction time of 3 h

The initial enzyme concentrations used in the titration experiments of Buffoni and Ignesti [5] ranged from 0.3 to 0.7 pM. Under such conditions non-linearity due to incomplete equilibration extends over the entire titration curve, which makes it vir- tually impossible to obtain a reliable estimate of the equivalence point (lower curve in Fig.2). This non- linearity may be difficult to detect in view of the small absorbance changes involved and, when ne- glected, leads to an overestimate of the amount of phenylhydrazine required for stoichiometric forma- tion of the phenylhydrazone derivative.

When initial enzyme concentrations above I0 pM are used in the titration experiments, the precision of the absorbance measurements is considerably in- creased. Furthermore, non-linearity due to insufficient equilibration becomes insignificant except near the equivalence point, and this point can be reliably estimated by extrapolation of the linear portions of the titration curve (upper curve in Fig.2). The equi- valence point thus obtained from the data in F i g 2 corresponds to 1 .O (rt 0.15) mol of phenylhydrazine per In01 of enzyme, confirming the results of our previous investigation [4].

We have shown that the appearance of the 430-nm absorption band parallels inactivation of the enzyme by phenylhydrazine [4]. The kinetic active-site titra- tions reported by Buffoni and Ignesti [5], therefore, are open to the same basic criticism as their spectro- photometric titrations. In both cases an overestimate of the actual number of active sites is obtained due to incomplete equilibration between enzyme and inhibi- tor under the conditions used.

A. Lindstrom and G. Pettersson 135

DISCUSSION

The data reported in the present investigation provide a reasonable explanation for the apparently conflicting results obtained in previous phenylhydra- zine titrations of pig-plasma benzylamine oxidase. There is no inconsistency between our experimental observations and those of Buffoni and Ignesti [ 5 ] . However, kinetic and spectrophotometric active-site titrations performed as described by Buffoni and Ignesti will lead to an overestimate of the number of reactive prosthetic groups in the enzyme, unless cor- rections are introduced to account for incomplete equilibration between enzyme and phenylhydrazine. Active-site titrations carried out under conditions where no such corrections are required provide clear evidence that there is a single reactive prosthetic group in pig-plasma benzylamine oxidase [4] (cj: Fig. 2).

Independent and more direct evidence that benzyl- amine oxidase contains only one reactive prosthetic group is provided by the observation that a single equivalent of substrate is required to fully reduce the enzymic 470-nm chromophore under anaerobic con- ditions (Fig. 1). The presence of a single catalytically active center can be inferred also from the observation that one mole of aldehyde per mole of enzyme is catalytically produced under such conditions (Table 1).

The 470-nm absorbance changes occurring during the catalytic action of benzylamine oxidase are known to reflect a rapid reduction of the enzyme by the amine substrate, and a subsequent slow reoxidation of the reduced enzyme species by molecular oxygen [lo - 121. Examination of the transient-state kinetics of catalytic aldehyde formation under aerobic conditions has led to the suggestion that the aldehyde product is released prior to the rate-limiting reoxidation process [16]. Direct evidence in favour of such a reaction mecha- nism is provided by the present observation that aldehyde production with the expected stoichiometry takes place during the anaerobic reduction of enzyme by substrate. In this respect the present results agree with those obtained for related enzymes such as the amine oxidases from beef plasma [9], pig kidney [17], and pea seedlings [18]. This lends support to the idea that copper-containing amine oxidases operate by a common basic reaction mechanism.

The present investigation establishes that a single prosthetic group is catalytically active in pig-plasma benzylamine oxidase. Further, it confirms previous reports that the enzyme contains two copper atoms per protein molecule [I3 - 151, which means that there is no 1 : 1 correspondence between the copper content and the number of catalytically active sites. In this respect pig-plasma benzylamine oxidase appears to

differ from the copper-containing diamiiie oxidase from pig kidney [19]. Application of the active-site titration techniques described here to some additional related enzyme systems might clarify whether the presence of two copper atoms per active site is a general characteristic of copper-containing amine oxidases or only exceptionally encountered.

Evidence has been reported indicating that at least one of the two copper ions in pig-plasma benzylamine oxidase has a catalytically essential function [20]. Experiments aiming at deciding whether or not the two copper ions exhibit distinguishable structural or functional properties are in progress.

This work was supported by grants from the Swedish Natural Science Research Council.

REFERENCES

1.

2.

3.

4.

5. 6.

7.

8.

9.

10.

31.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Buffoni, F. & Blaschko, H. (1964) Proc. R. Soc. Lond. B Bid. Sri. 16 1, 1 53 -- 1 67.

Blaschko, H. & Buffoni, F. (1965) Proc. R. Soc. Lond. B Biol. Sci. 163,45 - 60.

Buffoni, F. (1968) in Pyridoxal Catalysis: Enzyme.r and Model Systems (Snell, E. E.. Braunstein, A. E., Severin, E. S. & Torchinsky, Y. hi., eds) pp. 363-374, Interscience, New York.

Lindstrom, A. & Pettersson, G. (1973) Eur. J . Biochem. 34,

Buffoni, F. & Ignesti, G. (1975) Biochem. J . 145, 369-372. Lindstrom, A,, Olsson, B. & Pettersson, G. (1974) Eur. J . Bio-

chenz. 42, 177- 3 82. Lindstrom, A, , Olsson, B. & Pettersson, G. (1973) Eur. J .

Bioclietn. 35, 70 - 77. Reinhammar, B. R . M . (1972) Biochim. Biophys. Acta, 275,

Oi, S., Inamasu,M. & Ydsunobu, K. T. (1970) Biochemi.stry,

Lindstrom, A., Olsson, B., Pettersson, G. & Szymanska, J .

Olsson, B., Olsson, J. & Pettersson, G. (1976) Eur. J . B iochm.

Olsson, B., Olsson, J. & Pettersson, G. (1977) Eur. J . Biochem.

Lindstroni, A. & Pettersson, G. (1974) Eur. J . Biochem. 48,

Buffoni, F., Della Corte, L. & Knowles, P. F. (1968) Biochem.

Boden, N., Charlton, S. C., Holmes, M. C. 81 Knowles, P. F. (1973) Biochem. Soc. Trans. 1, 1008.

Lindstrom, A., Olsson, B. & Pettersson, G. (1974) Eur. J . Bio- chern. 42, 377-381.

Finazzo Agro, A,, Rotilio, G., Costa, M. T. & Mondovi, B. (1969) FEBS Lett . 4, 31 -32.

Yamnsaki, E. F.. Swindell, R . & Reed, D . J. (1970) Biochem- i.c.try, Y, 1206-1210.

Mondovi, B., Costa, M . T., Finazzi A g o , A. & Rotilio, G. (1 968) in Pjiridoxd Catalysis: Enzymes and Model Systenzs (Snell, E. E., Braunstein, A. E., Severin, E. S. & Torchinsky, Y. M., eds) pp. 403-423, Interscience, New York.

Lindstrom, A,, Olsson, B. & Pettersson, G. (1974) Eur. J . Bioclimi. 48. 237 - 243.

564-568.

245 - 259.

Y, 3378-~3383.

(1974) Eur. J . Biochem. 47, 99- 105.

71,375-382.

74, 329-335.

229 - 236.

J . 106, 575 - 576.

A. Lindstrijm and G. Pettersson, Avdelningen for Biokemi, Kemicentrurn, Lunds Universitet, Box 740, S-220 07 Lund, Sweden