J. Biochem. 86, 97-104 (1979)

Substrate Specificity and Reaction Mechanism

of Putrescine Oxidase

Masato OKADA, Seiichi KAWASHIMA, and Kazutomo IMAHORI

Department of Biochemistry, Faculty of Medicine, The University of Tokyo, Bunkyo-ku, Tokyo 113

Received for publication, January 9, 1979

Putrescine oxidase [EC 1.4.3.4] of Micrococcus rubens oxidizes many kinds of synthetic poly

amines: triamines (spermidine types), tetramines (spermine types), and N-substituted putres

cines. Polyamines possessing terminal 4-aminobutylimino groups in their structures were

more active as substrates. Putreanine was oxidized at a rate comparable to that of putrescine,

and was converted to 1-pyrroline and 8-alanine. Activities and Km values for polyamines

were affected by the substituent attached to the 4-aminobutylimino group of the polyamine,

and especially by its methylene chain length. It was also found that two types of oxidation

occurred in the oxidation of polyamines by putrescine oxidase. When the moieties attached

to the 4-aminobutylimino groups in polyamines were less hydrophobic, these polyamines were

oxidized at the secondary amino groups to form 1-pyrroline. Polyamines which contained a

hydrophobic substituent attached to the 4-aminobutylimino group were oxidized at the ter minal primary amino group of the 4-aminobutylimino moiety to form ammonia. N,N•Œ-

Bis(4-aminobutyl)-1,3-diaminopropane ([‡U,4-3-4]) and N-(4-aminobutyl)-N•Œ-(3-aminopropyl)

1,3-diaminopropane ([‡U,4-3-3]) were oxidized to form 1-pyrrolinium salt derivatives as a result

of oxidation of the terminal primary amino groups. It was concluded that the essential struc

ture for substrates of putrescine oxidase is a 4-aminobutylimino group (NH2(CH2),NH-).

Putrescine oxidase [EC 1.4.3.4], putrescine: which catalyze the degradation of spermidine (deaminating) (flavin containing), also oxidize spermine (4-10). from Micrococcus rubens has a high Swain and DeSa proposed a model for the specificity for amines (1). This oxidized of this enzyme based on studies of the putrescine, cadaverine, spermidine, and 1,6 interactions between the enzyme and competitive diaminohexane (1-3), but is exceptional in that it inhibitors: the enzyme has an anionic point and a does not oxidize spermine (1), since many hydrophobic region in the active site (3). How ever, this model failed to explain why putrescine oxidase oxidizes spermidine, but not spermine. Abbreviations: Me, methyl; Et, ethyl; n-Pr, n-propyl; The substrate specificity for polyamines other n-Bu, n-butyl; n-Am, n-amyl; [‡T, x-y], NH2(CH2) x- NH(CH2)yNH2; [‡V, x-y-z], NH2(CH2)xNH(CH2) than diamines, spermidine and spermine has not y- NH(CH2)zNH2; [‡V, R], NH2(CH2)4NH-R; [‡W, X], examined. In order to characterize the properties

NH2(CH2)4NH(CH2)2-X. of the active site in more detail, it is necessary to

Vol. 86, No. 1, 1979 97 98 M. OKADA, S. KAWASHIMA, and K. IMAHORI

examine the substrate specificity for many poly 12 N HCl : H2O (60: 20: 20, v/v) (15). The amines. amines were detected by spraying 0.1 % ninhydrin

In this paper, we report the activities of in acetone. IR spectra were taken with a Hitachi putrescine oxidase towards many kinds of synthetic EPI-G2 spectrophotometer, and NMR spectra polyamines (triamines, tetramines, and N-sub with a Hitachi Perkin-Elmer R-20A spectrometer stituted putrescine) and the structures of their at 90 MHz using tetramethylsilane as an external oxidation products. Based on these results, the standard. Liquid chromatographic analyses of essential structures of polyamines for activity as polyamines and their oxidation products were substrates, as well as the binding properties of the carried out using a Hitachi 835 amino acid ana

enzyme, are discussed. lyzer in the buffer system with suitable modification

(16). Activity of Putrescine Oxidase towards Poly MATERIALS AND METHODS amines-The maximal velocity (Vmax) and pH

Materials-Putrescine oxidase of Micrococcus optimum of putrescine oxidase for polyamines

rubens was purified to homogeneity by an affinity were determined from the amount of hydrogen

chromatographic procedure (11). The specific peroxide evolved using the coupling assay method activity of the enzyme used in this study was with peroxidase and o-dianisidine (17) in 0.1 M

34.7 ƒÊmol/min/mg protein. Peroxidase and cata borate buffer at pH 7.5-10.5.

lase were purchased from Boehringer, Mannheim, Km Values for Polyamines-Km values for

Germany. Putrescine (free base and hydro polyamines were estimated from the competitive chloride salt), spermidine trihydrochloride, and inhibition by 1,8-diaminooctane of the oxidation

spermine tetrahydrochloride were the products of of the polyamines. First, Km for putrescine was

Nakarai Chemicals Ltd., Japan. Other diamines, determined from Lineweaver-Burk plots based on dimethylenetriamine, trimethylenetetramine, and the above assay method. Next, the Ki value for

N-(3-aminopropyl)-1,3-diaminopropane were pur 1,8-diaminooctane, which is a competitive inhibitor

chased from Tokyo Chemical Industry Co., Japan. of putrescine oxidase (3), was determined (Ki=

Other polyamines were synthesized in our labo 9.2•~10-6M). The oxidation velocities for poly ratory (12). Other chemicals were of the highest amines in the presence (vi) and absence (v) of

grade available. 1,8-diaminooctane at various concentrations (i) 1-(3-Hydroxy-l-oxoinden-2-yl)-1-pyrrolinium were measured, then Km values for the polyamines

Hydroxide, Inner Salt ([V])-This compound were calculated from v/vi vs. i plots based on the

(13) was prepared by the condensation reaction of equation: v/vi=1+i/Ki[Km/(Ki+s)], where Ki is

proline with ninhydrin according to the method the inhibitor constant of 1,8-diaminooctane and of Grassmann and Arnim (14). NMR (in CDCl3) s is the concentration of polyamine.

S 9.76 (s, 1H, imine proton), 8.00 (2d, 4H J=2Hz, Determination of 1-Pyrroline-l-Pyrroline

indene ring protons), 5.43 (t, 2H, J=8Hz, 5-H), formed by the oxidation of polyamines was deter

3.60 (m, 2H, 3-H), 2.80 (quin, 2H, J=8Hz, 4-H). mined according to the method described by

Holmstedt et al. (18, 19), except that a value of

2.1 •~ 103 M-1 • cm-1 was used for the molar extinc

tion coefficient of the adduct between 1-pyrroline

and o-aminobenzaldehyde at 435 nm (18). The reaction mixture contained 0.5 ml of 0.1 M borate

(pH 9.0), 10 pl of 50 mM polyamine (0.50 ƒÊmol), and 20 ƒÊl (ca. 1.0 unit) of putrescine oxidase. After

incubation for 16 h, 0.1 ml of 0.6 M phosphate

(pH 5.5) was added to the incubation mixture. After 1 h, the absorbance at 435 nm was mea

Measurement-TLC was carried out on sured. Avicel SF cellulose (Funakoshi Pharmaceutical Determination of Ammonia-Ammonia was Co.) in the following solvent system : 1-propanol: determined with an amino acid analyzer using

J. Biochem. SUBSTRATE SPECIFICITY OF PUTRESCINE OXIDASE 99

ammonium sulfate as a standard. Nessler's reagent TABLE I. Vmax, Km, and pH optimum values of

(20) was not applicable in this case because a putrescine oxidase for polyamines, including diamines. The reaction mixtures (0.555-0.570 ml) contained 2.5 precipitate was formed in some cases. The ƒÊ mol of polyamines, and 0.06-1.10 units of purified enzymatic reactions were carried out by methods putrescine oxidase. Vmax and Km were determined by similar to that described for the determination of the methods described in " MATERIALS AND 1-pyrroline. When the reaction stopped, 5011 METHODS." Vmax values are relative ratios based aliquots were withdrawn from the incubation on the velocity of putrescine oxidation at pH 9.0. mixtures and diluted to 550 pl with water. These

diluted mixtures were used as samples for the

amino acid analyzer.

RESULTS

Activities of Putrescine Oxidase towards Poly

amines-Vmax, pH optimum, and Km for various

polyamines were determined, as summarized in Table I; most of the polyamines examined were

active as substrates, and polyamines possessing

4. aminobutylimino groups were more active. This

agrees with the model which Swain and DeSa

proposed for the active site of putrescine oxidase

(3): the length of the 4-aminobutylimino group corresponds to the distance from the active (cata

lytic) point to the anionic point which has a binding

role. It is interesting that the activity for [IV,

COOH] (putreanine) (21) is comparable to that

for putrescine, in spite of its higher Km value.

So far, no enzymes other than putrescine oxidase

have been reported that can degrade putreanine.

On the other hand, Km values for polyamines

tend to decrease with increasing methylene chain

length in their structures. This tendency is clear when Km values for the members of the [‡T,4-y]

or [‡V,R] group are compared. This is comparable

to the known tendency that Ki values for com

petitive inhibitors of putrescine oxidase decrease as the methylene chain length increases (3), sug

gesting the existence of a hydrophobic interaction between the enzyme and the substituent on the

4-aminobutylimino group in polyamines.

Formation of 1-Pyrroline and Ammonia on

Polyamine Oxidation by Putrescine Oxidase Putrescine oxidase oxidizes putrescine to form

1-pyrroline, ammonia, and hydrogen peroxide,

and converts spermidine to 1-pyrroline and 1,3

diaminopropane, generating hydrogen peroxide ND: Not determined because the formation of H2O2 (1, 2). However, it has not been clarified whether all the [I,4-y], [‡U,4-y-4], [‡V,R], and [IV,X] types was observed only for a short initial period. Other

of polyamines are also oxidized to form 1-pyrroline polyamines, [I, 2-2], [‡U, 2-2-2], [‡U, 4-8-4], and [‡U, 6-6-6] were not active as substrates. as in the case of spermidine.

Vol. 86, No. 1, 1979 100 M. OKADA, S. KAWASHIMA, and K. IMAHORI

The amounts of 1-pyrroline and ammonia did not give 1,6-diaminohexane or 1,8-diamino

formed by polyamine oxidation are shown in octane, respectively, as judged by TLC and liquid

Table ‡U. Based on these results we can classify chromatographic analysis. These results evidently

polyamines into the following three groups. (1) indicate that oxidation by putrescine oxidase tends The oxidation of group I is 1-pyrroline to occur at one of the primary amines as far as

only ([‡T,4-2], [I,4-3], [‡W,COOH]). (2) The oxida these two polyamines are concerned.

tion product of group 2 is ammonia only ([‡T,4-6], Oxidation Products of Putreanine-As shown

[‡T,4-8], [‡U,x-y-z], [‡V,n-Bu], [‡V,n-Aml). (3) The in Table ‡U, putreanine ([‡W,COOH]) gave 1

oxidation product of group 3 are both 1-pyrroline pyrroline stoichiometrically as one of the products and ammonia. In all cases, the sum of ammonia on enzymatic oxidation. The other could be

and 1-pyrroline is equivalent to the amount of identified as ƒÀ-alanine by liquid chromatographic

the substrate (0.5 ƒÊmol), within experimental analysis. This analysis was carried out using the

error. amino acid analyzer in a way similar to that for

The reaction mechanism will be described in the determination of ammonia, described above. detail in the following section. However, it should The retention times were 30.58 and 30.86 min for

be added that the oxidation of [I,4-6] or [‡T,4-8] ƒÀ-alanine and the product, respectively, and the

difference was within the range of experimental

error. TABLE II. Amounts of 1-pyrroline and ammonia Oxidation Product of [‡U,4-3-4]-Although formed by the enzymatic oxidation of polyamines con the results in Table II indicate that [‡U,4-3-4] is taining 4-aminobutylimino groups. The amounts of 1-pyrroline and ammonia were determined by the deaminated at one of the two terminal primary methods described in " MATERIALS AND METH amines by putrescine oxidase, the mechanism of

ODS." The amount of polyamines used in these deter the reaction is not clear, since the nature of the minations was 0.50 ƒÊmol. other product was not determined. In order to

Fig. 1. Elution profile of the oxidation products of [‡U

, 4-3-4] from a Dowex 50W-X4 column. The pro cedure is described in the text. Fractions (11.6 ml) were collected. Aliquots (50 pl) were taken and

monitored by the ninhydrin method (22), followed by measurement of the absorbance at 570 nm. Peaks I and II correspond to NH3 and the oxidation product, respectively.

J. Biochem. SUBSTRATE SPECIFICITY OF PUTRESCINE OXIDASE 101 elucidate this point, we attempted to identify the ing to peak II in Fig. 1 were collected, HCl and

oxidation product of [‡U,4-3-4]. A large-scale water were removed by evaporation, and the incubation was carried out with 88 mg of [‡U,4-3-4], residue was further dried in an evacuated desiccator

0.2 mg of catalase, and 50 units of putrescine over H2S04 and NaOH to yield a pale yellow

oxidase in 10 ml of 0.1 M borate buffer, pH 9.0. syrup (56 mg). The IR spectrum of this compound

This incubation was continued at 25°C with is shown in Fig. 2. This spectrum shows bands

stirring while the oxidation was followed by at 2200 and 1660 cm-1 which are characteristic

TLC. After 24 h, the ninhydrin-positive starting of an immonium compound (23, 24). material (R f 0.20, [‡U,4-3-4]) was converted to a The NMR spectrum of this product is shown

product (R f 0.58). This incubation mixture was in Fig. 3 together with that of [‡U,4-3-4] (the directly applied to a Dowex 50W-X4 column starting material). The NMR spectrum shows

(H+ form, 1.4 •~ 32 cm). The column was washed a signal at a low chemical shift (3, 9.27). This with 200 ml of water, then the products were type of signal is also observed in the case of [V],

eluted with a linear gradient of HCl from 0.0 to which is a 1-pyrrolinium salt derivatives (8 9.76).

6.0 M in a total volume of 1,000 ml. The elution The other peaks could be assigned as shown in

profile is shown in Fig. 1. Fractions correspond Fig. 3b. The oxidation product of [‡U,4-3-4] was

Fig. 2. IR spectrum of the oxidation product of [‡U, 4-3-4] (syrup film).

Fig. 3. NMR spectra of [‡U, 4-3-4] (a) and its oxidation product (b) in D2O at 90 MHz.

Vol. 86, No. 1, 1979 102 M. OKADA, S. KAWASHIMA, and K. IMAHORI

Fig. 4. NMR spectra of [‡U, 4-3-3] (a) and its oxidation product (b) in D2O at 90 MHz.

identified as the salt of a 1-pyrrolinium derivative as shown in Fig. 3b. It can be concluded that

[‡U,4-3-4] is oxidized only at the primary amino group. Oxidation Product of [‡U,4-3-3]-The experi mental results described above suggest that putres Fig. 5. Mechanism for the formation of the 1-pyrroli cine oxidase oxidizes one of the primary amines nium salt in the oxidation of polyamines by putrescine of tetramines. However in the case of asymmetric oxidase. tetramines such as [‡U,4-3-3], the enzyme might oxidize either one of the primary amines pref a cyclization reaction of this type has not been erentially. We therefore attempted to isolate reported previously. For example, Smith (9, 25) the oxidation product of [‡U,4-3-3]. A large-scale presumed that 1-(3-aminopropyl)-2(or 3)-pyrroline incubation was carried out as in the case of [‡U, was formed from spermine by polyamine oxidase 4-3-4], except that 100 mg of [‡U,4-3-3] was used in barley. However, on the basis of our experi as the starting material. After incubation for mental results, it seems reasonable to assume that 24 h, the ninhydrin-positive starting material (R f the oxidation product is not 1-(3-aminopropyl) 0.17, [‡U,4-3-3]) was converted to a product (R f 2 (or 3)-pyrroline, but the 1-(3-aminopropyl)-1 0.53). A pale yellow syrup (58 mg) was obtained. - pyrrolinium salt. This compound gave an IR spectrum analogous to that shown in Fig. 2. The NMR spectrum is DISCUSSION shown in Fig. 4b together with that of [‡U,4-3-3]

(the starting material). The oxidation product Putrescine oxidase cleaves putrescine oxidatively of [‡U,4-3-3] was identified as a salt of 1-pyrrolinium to produce 1-pyrroline, ammonia, and hydrogen derivative, as shown in Fig. 4b, indicating that the peroxide. It was reported that the enzyme also primary amine at the 4-aminobutyl group was cleaves the C-N bond of spermidine to form selectively oxidized in preference to that at the 1-pyrroline and 1,3-diaminopropane (2). How 3-aminopropyl group. ever, since the number and variety of substrates These salts of 1-pyrrolinium derivatives are so far tested are quite limited, it is difficult to presumably formed by intramolecular cyclization understand the substrate specificity and mode of of aldehyde derivatives, as shown in Fig. 5. This action of this enzyme. Thus we have synthesized 1-pyrrolinium salt formation is interesting, because many kinds of diamines, triamines, and tetramines

J. Biochem. SUBSTRATE SPECIFICITY OF PUTRESCINE OXIDASE 103

and tested them as substrates for the enzyme. As rest of the molecule gives the 1-pyrrolinium deriva summarized in Table I, polyamines which serve tive. Thus, this bond should face the catalytic as substrates of the enzyme generally contain a point and the imino group of the 4-aminobutyl 4-aminobutyl group. Thus, spermine [11,3-4-3] imino moiety presumably binds to the anionic cannot act as a substrate, but even N-alkylputres point. This kind of binding will be called reverse cine is oxidized by this enzyme. These observa binding.

tions are in contrast with the report that plasma We can now explain the results for polyamines

acts preferentially on the 3-amino of class 3. The binding of these polyamines can

propylimino moiety (26). Next, we examined the occur in two modes: correct and reverse binding. reaction mechanism of putrescine oxidase by The finding that the sum of 1-pyrroline and ammo analyzing the reaction products of polyamines nia is constant and equal to the molar amount of containing a 4-aminobutyl moiety. The results the substrate supports this view. Among this are summarized in Table ‡U. As described above, class, some polyamines may favor correct binding we can classify these polyamines into three groups. and some reverse binding.

(1) Polyamines which give 1-pyrroline but not The problem is thus how the enzyme can ammonia as the reaction product, (2) polyamines determine the mode of binding. Firstly we will which give ammonia but not 1-pyrroline, and (3) consider polyamines of [I,x-y] type in Table U.

polyamines which give both 1-pyrroline and It is clear that as the number of methylene groups ammonia as reaction products. It is interesting attached to the 4-aminobutylimino moiety increases that in all cases the sum of 1-pyrroline and ammonia the triamines tend to bind reversely. Such a is constant and equal to the molar amount of the tendency will appear if we suppose that there is a substrate. hydrophobic region to the left of the anionic We will next consider these reactions in terms point, as shown in Fig. 6. The same is true for of the active site model of Swain and DeSa (3) polyamines of [III,R] type. As the number of with some modifications, as shown in Fig. 6. In methylene groups attached to the 4-aminobutyl this model, the anionic point and catalytic point imino group increases, reverse binding becomes are arranged so as to match the dimensions of the more favorable. Polyamines of [II,4-y-4] type 4-aminobutyl group. This explains why this which have long hydrophobic chains beside the enzyme oxidizes polyamines which contain a 4-aminobutylimino moiety naturally bind reversely. 4-aminobutyl group. On the other hand, putreanine, which has a We will consider polyamines of class 1; for hydrophilic moiety attached to the 4-aminobutyl instance, spermidine is oxidized to form 1-pyrroline imino group, shows correct binding. Moreover, and 1,3-diaminopropane (2). We have shown as shown in Table I, the Km values of all polyamines that the reaction products of putreanine are 1 decrease as the hydrophobicity of the moiety

- pyrroline and ƒÀ-alanine. It is supposed that attached to the 4-aminobutylimino group increases.

[1,4-2] is oxidized to form 1-pyrroline and diamino These results are consistent with the existence of ethane. All of these compounds are cleaved a hydrophobic region in the active site of the oxidatively at the C-N bond of the 4-amino present enzyme. butylimino group. This bond must face the In conclusion, the present results show that catalytic point. Thus, the primary amino group polyamines containing a 4-aminobutyl group are of the 4-aminobutyl moiety should bind to the preferred substrates of this enzyme. The active anionic point (Fig. 6). Binding in this manner

will be called correct binding.

Next we will consider polyamines of class 2.

One reaction product was ammonia. However,

as shown in Figs. 3 and 4, the other oxidation

products of [11,4-3-4] and [11,4-3-3] are 1-pyr rolinium derivatives. In reactions of this type,

only one bond attached to the primary amino Fig. 6. Schematic model of the active site of putrescine oxidase. group of the 4-aminobutyl moiety is split, and the

Vol. 86, No. 1, 1979 104 M. OKADA, S. KAWASHIMA, and K. IMAHORI site of the present enzyme contains a hydrophobic 11. Okada, M., Kawashima, S., & Imahori, K. (1979) region, an anionic point, and a catalytic point, J. Biochem. 85, 1225-1233 as shown in Fig. 6. As a result of the arrangement 12. Okada, M., Kawashima, S., & Imahori, K. (1979) of these points or regions, the mode of substrate J. Biochem. 85, 1235-1243 binding is selected. Polyamines containing a 13. Johnson, A.W. & McCaldin, D.J. (1958) J. Chem. hydrophobic moiety attached to the 4-amino Soc. 817-822 14. Grassmann, W. & Arnim, K.V. (1934) Annalen butylimino group tend to show reverse binding, 509,288-303 while those with a less hydrophobic moiety give "correct" binding 15. Dubin, D.T. & Rosenthal, S.M. (1960) J. Biol. . The bond attached to the Chem. 235, 776-782 terminal amino group or to the imino group of 16. Marton, L.J. & Lee, P.L.Y. (1975) Clin. Chem. 21, the 4-aminobutylimino moiety is oxidatively cleaved 1721-1724 by the enzyme in the case of correct or reverse 17. Guidotti, G., Colombo, J.P., & Foa, P.P. (1961) binding, respectively. Anal. Chem. 33,151-153 18. Jakoby, W.B. & Fredericks, J. (1959) J. Biol. Chem. 234,2145-2150 REFERENCES 19. Holmstedt, B., Larsson, L., & Tham, R. (1961) 1. Yamada, H., Tanaka, A., & Ogata, K. (1965) Biochim. Biophys. Acta 48, 182-186 Agric. Biol. Chem. 29, 260-261 20. Bessman, S.P. & Bessman, A.N. (1955) J. Clin. 2. Adachi, 0., Yamada, H., & Ogata, K. (1966) Agric. Invest. 34, 622-636 Biol. Chem. 30,1202-1210 21. Kakimoto, Y., Nakajima, T., Kumon, A., Matsu 3. Swain, W.F. & DeSa, R.J. (1976) Biochim. Biophys. oka, Y., Imaoka, N., Sano, I., & Kanazawa, A. Acta 429, 331-345 (1969) J. Biol. Chem. 244, 6003-6007 4. Hasse, K. & Schuhrer, K. (1962) Biochem. Z. 336, 22. Moore, S. & Stein, W.H. (1954) J. Biol. Chem. 211, 20-34 907-913 5. Padmanabhan, R. & Kim, K. (1965) Biochem. 23. Witkop, B. (1954) J. Amer. Chem. Soc. 76, 5597 Biophys. Res. Commun. 19,1-5 5599 24. Bellamy, L.J. (1954) in The Infra-red Spectra of 6. Hill, J.M. (1967) Biochem. J. 104,1048-1055 Complex Molecules pp. 223-231, John Wiley & Sons, 7. Tabor, C.W. & Kellogg, P.D. (1970) J. Biol. Chem. 245,5424-5433 Inc., New York 25. Smith, T.A. (1972) Phytochemistry 11, 899-910 8. Holtta, E. (1977) Biochemistry 16, 91-100 26. Tabor, C.W., Tabor, H., & Bachrach, U. (1964) 9. Smith, T.A. (1970) Biochem. Biophys. Res. Commun, J. Biol. Chem. 239, 2194-2203 41,1452-1456 10. Yamada, H. & Yasunobu, K.T. (1962) J. Biol. Chem. 237, 1511-1516

J. Biochem.