Oxygenation of (3Z)-alkenal to (2E)-4-hydroxy-2-alkenal in soybean seed (Glycine max L.)

BB

ELSEVIER

Biochimica et Biophysica Acta 1303 (1996) 83-91

Biochi ~mie~a et BiophysicaA~ta

Oxygenation of ( 3 Z) -alkenal to ( 2 E) -4-hydroxy-2-alkenal in soybean seed( Glycine max L.) Hitoshi Takamura a,1 Harold W. Gardner b,, a Department of Food Science and Nutrition, Nara Women's University, Nara 630, Japan b National CenterforAgricultural Utilization Research, ARS, USDA, 1815 N. University Street, Peoria, IL 61604, USA

Received 4 March 1996; accepted 15 May 1996

Abstract (3Z)-Alkenals, such as (3Z)-hexenal and (3Z)-nonenal, are produced from polyunsaturated fatty acids via lipoxygenase and hydroperoxide lyase catalysis, but in soybeans (Glycine max L.) (3Z)-alkenals have a fleeting existence. In this study it was shown that soybean seeds possess two pathways that metabolize (3Z)-alkenals. One is a soluble (3Z):(2E)-enal isomerase that transformed (3Z)-hexenal and (3Z)-nonenal into the corresponding (2E)-alkenals. The other was a membrane-bound system that converted (3Z)-hexenal and (3Z)-nonenal into (2 E)-4-hydroxy-2-hexenal and (2 E)-4-hydroxy-2-nonenal, respectively. The latter conversion was shown to absorb 0 2 with a pH optimum of 9.5. Little effect observed with lipoxygenase inhibitors suggested that oxidation was not catalyzed by lipoxygenase. Instead, a specific (3Z)-alkenal oxygenase was implicated in forming intermediate alkenal hydroperoxides. Hydroperoxide-dependent peroxygenase (epoxygenase) is known to reduce hydroperoxides to their corresponding hydroxides and is also known to be inhibited by hydrogen peroxide preincubation. Consequently, intermediate 4-hydroperoxy-2-alkenals could be observed after inhibiting hydroperoxide-dependent peroxygenase by preincubation with hydrogen peroxide. Because 4-hydroxy-2-alkenals are potent toxins, these compounds may be produced as nonvolatile plant defensive substances. Keywords: Hexenal; 4-Hydroxy-2-hexenal; 4-Hydroxy-2-nonenal; Lipoxygenase pathway; Nonenal; Peroxygenase; (Glycine max L.)

1. Introduction (2 E)-4-Hydroxy-2-nonenal (HNE) was first observed to be generated in biological systems as a result o f N A D P H dependent, iron-stimulated peroxidation of lipids by rat liver microsomes [1]. Intense interest in ( 2 E ) - 4 - h y d r o x y 2-alkenals has been maintained principally because this

Abbreviations: HEX, (3Z)-hexenal; t-HEX, (2E)-hexenal; HHE, (2E)-4-hydroxy-2-hexenal; NON, (3Z)-nonenal; t-NON, (2E)-nonenal; HNE, (2E)-4-hydroxy-2-nonenal; HPNE, (2E)-4-hydroperoxy-2-nonenal; Mes, 2-(N-morpholino)ethanesulfonic acid; Epps, 3-[4-(2-hydroxyethyl)-l-piperazinyl]propanesulfonic acid; Ches, N-cyclohexyl-2aminoethanesulfonic acid; Pipes, 1,4-piperazinediethanesulfonic acid; BTP, 1,3-bis[tris(hydroxymethyl)methylamino]propane; ETYA, 5,8,11,14-eicosatetraynoic acid; NDGA, nordihydroguaiaretic acid; SHAM, salicyl hydroxamate; O-TMS, O-trimethylsilyloxy; GC, gasliquid chromatography; GC-MS, gas-liquid chromatography-mass spectrometry. * Corresponding author. Fax: + 1 (309) 6816686; e-mail: [email protected]. I Research completed while H.T. was on sabbatical leave at the National Center for Agricultural Utilization Research, Peoria, IL, USA.

class of compounds show toxicity in mammalian systems [2]. Possible reasons for toxicity include depletion of glutathione protection [3], modification of protein histidine residues [4], as well as reaction with protein lysines [5] and sulfhydryls [6], and alkylating guanine residues [7]. It is believed that HNE and (2 E)-4-hydroxy-2-hexenal (HHE) generally are derived from polyunsaturated fatty acids, n - 6 and n - 3, respectively [2], but more detailed investigations of their origins are few [8,9]. Recently, a pathway for the biosynthesis o f HNE has been demonstrated in plants. In soybeans ( G l y c i n e m a x L.), a lipoxygenase product, (9 S, 10 E, 12 Z)-9-hydroperoxy10,12-octadecadienoic acid, was converted into HNE and 9-oxononanoic acid [10]. Since the expected products were (3Z)-nonenal (NON) and 9-oxononanoic acid from hydroperoxide lyase action on the 9-hydroperoxide, an oxidation of N O N was proposed. Such an oxidation of both N O N and (3Z)-hexenal (HEX) was found to be catalyzed by microsomes from V i c i a f a b a seeds [9]. A two-step enzymic process was identified which apparently involved sequential action of a '(3Z)-alkenal oxygenase' and a hydro-

0005-2760/96/$15.00 Copyright © 1996 Published by Elsevier Science B.V. All rights reserved. PII S0005-2760(96)00076-8

84

14. Takamura, H. W. Gardner / Biochimica et Biophysica Acta 1303 (1996) 83-91

peroxide-dependent epoxygenase (termed 'peroxygenase' in this communication). In the present study, a membrane-bound enzymic system that oxidized NON and HEX into HNE and HHE, respectively, was identified and characterized in soybean seeds. It was also shown that a soluble isomerase in soybean competed for (3Z)-alkenal substrate by converting them into (2 E)-alkenals.

M sodium phosphate (pH 6.7). This low-speed supernatant was then centrifuged at 200 000 X g for 60 min at 4°C by using Beckman Model L8-70M ultracentrifuge. The highspeed supernatant was used as a 'soluble fraction.' The pellet was resuspended in the same phosphate buffer by using a glass homogenizer. The suspension was centrifuged again with the same conditions. The pellet obtained was used as the 'membrane preparation.' Protein content was determined by BCA method as described previously [ 13].

2. Materials and methods

2.4. Formation of HHE and HNE by preparations 2.1. Materials O-Benzylhydroxylamine hydrochloride, tetracosane, (3Z)-hexen-l-ol, (3Z)-nonen-l-ol, salicylhydroxamic acid (SHAM), triphenylphosphine, and propyl gallate were purchased from Aldrich (Milwaukee, WI). Nordihydroguaiaretic acid (NDGA), Pipes, Mes, Epps, Ches, and BTP were from Sigma (St. Louis, MO). 5,8,11,14-Eicosatetraynoic acid (ETYA) was from Cayman (Ann Arbor, MI).

2.2. Preparation of aldehydes HEX and NON were prepared from 6 mmol of their respective alcohols, (3Z)-hexen-l-ol and (3Z)-nonen-l-ol, by the method of Corey and Suggs [11] using the buffered alternative (60 mg sodium acetate). Evaporation of solvent from the crude product was carefully done to avoid losses until about 1 ml diethyl ether/dichloromethane remained. This residue was applied to a column (1.5 cm i.d.) containing 12 g SilicAR CC4 (Mallinckrodt) packed with pentane. The product aldehydes were eluted sequentially with 120 ml each of 1% and 2.5% diethyl ether in pentane. (3Z)hexenal eluted between 230 and 280 ml, and (3Z)-nonenal eluted between 210 and 220 ml. Solvent was carefully evaporated, and the residual aldehydes taken up in methanol as a 80 mM solution for use or storage at - 2 0 ° C . The overall yield of (3Z)-hexenal was 11% in 95% purity (2.2% was (2E)-hexenal). The yield of (3Z)-nonenal was 23% in 92% purity (2.1% was (2E)-nonenal). HHE and HNE were prepared as described previously [12].

2.3. Enzyme preparations Soybean (Glycine max L. var. Century) seeds were powdered with dry ice by using a coffee grinder. The powdered soybean was soaked (0.1 g / m l ) in 0.1 M Ches buffer (pH 9.0) in an ice bath for 10 min, homogenized with Polytron for 30 sec, then filtered through 4 layers of cheese cloth. The filtrate was centrifuged at 9300 X g for 15 min at 4°C giving a low-speed supernatant that was used directly as an enzyme source in some experiments. In other experiments a high-speed membrane and supernatant were prepared using the low-speed supernatant prepared as described above, except the homogenizing buffer was 0.1

The low-speed supernatant prepared with 0.1 M Ches buffer (pH 9) was used as an enzyme after 10-fold dilution with the same Ches buffer. The high-speed soluble fraction was diluted in 0.1 M sodium phosphate buffer (pH 6.7). The high-speed membrane preparation was resuspended in the same phosphate buffer in the time-dependent formation study, or resuspended in water then adjusted to 0.1 M concentration with the appropriate buffer in other studies. Heat inactivation was carded out in a hot water bath at 80°C for 15 min. To prevent volatilization of aldehyde substrate, incubations were completed in closed systems. In the time-dependent study, an aliquot (2 ml) of the various preparations was preincubated in a 6 ml teflonstoppered Reacti-flask TM (Pierce, Rockford, IL) at 25°C for 3 - 5 min. Then, HEX or NON were added to the vial with a microsyringe to give a concentration of 2 mM with the low-speed supernatant and 1 mM with the high-speed fractions. The reaction mixture was incubated at 25°C with stirring. At timed intervals, 0.2 ml aliquots were taken with a syringe through a valve in the teflon stopper, and added to a mixture of 0.4 ml methanol and 0.4 ml O-benzylhydroxylamine solution (25 mM O-benzylhydroxylamine in 100 mM Pipes, pH 6.5) to both terminate the reaction and form O-benzyloximes of the aldehydes. After a 10 min incubation at room temperature, 20/xg tetracosane internal standard was added, and the O-benzyloximes were extracted with 1 ml chloroform. The formation of products as a function of pH or hydrogen peroxide addition at two pH values were modified to include twice the concentration of Pipes (pH 6.5), buffer in the O-benzylhydroxylamine reagent to counteract a pH shift caused by the various buffers.

2.5. Oxygen consumption by soybean membrane preparation Oxygen consumption was measured using an oxygen electrode (Oxygraph, Gilson) equipped with a 2.4 ml cell. Incubations were essentially the same as described above for determination of product formation. Oxygen consumption was recorded and the initial rate (/xmol oxygen/ml) was calculated after correcting for the blank comprised of buffer only plus substrate. Appropriate amounts of inhibitors were added as 0.1-0.5 M methanolic solutions and

85

H. Takamura, H. W. Gardner/Biochimica et Biophysica Acta 1303 (1996) 83-91

preincubated for 5 rain prior to adding the substrate. Addition of the same quantity of only methanol served as the control.

2.6. Formation of 9,10-epoxystearic acid from oleic acid by soybean membrane preparation Oleic acid (1 mM) was incubated with 0.4 ml membrane preparation (3.5 mg protein/ml) in 0.1M buffer (Mes at pH 6.5 or Ches at pH 9.5) in the presence of 5 mM hydrogen peroxide. After 5 min incubation at 25°C, the reaction was terminated by adding 1 ml chloroform/methanol (2:1), and 0.4 ml of 0.25 M citrate buffer (pH 3.7) was added to adjust the pH to about 4. The fatty acids were extracted into chloroform, and 40 /zg tetracosane internal standard was added.

2.7. Chromatographic methods The chloroform extracts of O-benzyloximes of aldehydes were dried under a nitrogen flow, and then trimethylsilyloxy (O-TMS) derivatives were prepared for gas-liquid chromatography (GC) analysis using chlorotrimethylsilane, hexamethyldisilazane, and pyridine as described previously [9]. After 10 min incubation at room temperature, the reagents were removed under a nitrogen flow and the sample was dissolved in 100 /~1 2,2,4-trimethylpentane. GC analysis was carried out with a Hewlett-Packard Model 5890 gas chromatograph equipped with a SPB-1 capillary column (30 m X 0.32 mm; film thickness, 0.25 /xm) from Supelco (Bellefonte, PA). O-Benzyloximes of various aldehydes (hydroxyl groups were converted to O-TMS groups) were separated by temperature programming from 160 to 200°C at 4°C/min followed by 200 to 300°C at 10°C/min. Injector and detector temperatures were 260 and 310°C, respectively. Helium was used as the carrier gas at an initial flow rate of 2 ml/min. The retention times of O-benzyloxime derivatives (syn and anti isomers) were as follows: HEX, 3.08 and 3.18 min; t-HEX, 3.55 and 3.66 min; HHE (O-TMS), 6.06 and 6.71 min; NON, 6.00 and 6.31 min; t-NON, 7.20 and 7.65 min; HNE (O-TMS), 10.24 and 11.15 min; tetracosane (internal standard), 14.81 min. GC-MS of these O-benzyloxime derivatives were similar to those described previously [9]. The GC response factors were determined from weighed mixtures of various compounds which had been isolated by preparative thinlayer chromatography as described previously [9]. The response factors (by weight) based on tetracosane (1.0) were: O-benzyloxime of HEX and t-HEX, 0.86; O-benzyloxime of HHE, 0.80 (weight prior to derivatizing as O-TMS), O-benzyloxime of NON and t-NON, 0.91; Obenzyloxime of HNE, 0.85 (weight prior to derivatizing as O-TMS). To analyze 9,10-epoxystearic acid, the lipid extract was methylated with diazomethane. After evaporating the reagent, the residue was dissolved in 100 /~1 2,2,4-trimeth-

ylpentane. GC analysis was carried out as described above. Fatty acid methyl esters were separated by temperature programming from 160 to 270°C at 10°C/min. Injector and detector temperatures were 260 and 280°C, respectively. The retention times of methyl 9,10-epoxystearic acid and tetracosane were 8.94 and 10.08 min, respectively. The response factor (by weight) of methyl 9,10epoxystearic acid based on tetracosane was 0.73 [9]. Thin-layer chromatography of free aldehydes was completed using Silica Gel 60 F254 precoated plates from E. Merck, Darmstadt, Germany (20 cm × 20 cm X 0.25 mm). The developing solvent was ethyl ether/hexane (3:2 v / v ) . Non-destructive detection of (2E)-4-hydroperoxy-2-nonenal (HPNE) and HNE was done by viewing of absorbance under ultraviolet light. HPNE and HNE fractions were scraped from plates and extracted with ethyl ether. Isolated HPNE was reduced with 2 mg triphenylphosphine in 0.5 ml ether ether for 1 hr. For visualizing all aldehydes destructive spraying with 0.4% 2,4-dinitrophenylhydrazine in 2 N HC1 was used.

3. Results

3.1. Oxidation and isomerization of NON by soybean low-speed supernatant O-Benzylhydroxylamine was used to trap aldehydes as their O-benzyloximes, which are more chemically stable, easily recovered, and less volatile. The O-benzyloxime assay was found to be reliably reproducible in routine assay of aldehydes used in this study. When the low-speed supernatant (10-fold diluted) was assayed for activity with NON, HNE was formed in about 18% molar yield after 15 min incubation (Fig. 1). After 15 min, NON had completely disappeared and t-NON comprised only 3% of the

,~, 0.4-

0.20"1"lO ~ ' ~ ' /

5

10

15

Time (min) Fig. 1. Time-dependent formation of (2E)-4-hydroxy-2-nonenal from (3Z)-nonenal by soybean 9300X g supernatant. NON (2 mM) was incubated with untreated supernatant (D), heat-inactivated supernatant (•), or supernatant preincubated 5 rain with 20 mM H202 (O) at 25°C in 0.1 M Ches buffer (pH 9). Protein content was 2.5 mg/ml. Aldehydes were analyzed by GC of their O-benzyloximes,and hydroxyl groups were converted to O-TMS ethers.

86

H. Takamura, H.W. Gardner / Biochimica et Biophysica Acta 1303 (1996) 83-91

molar yield (data not shown). Interestingly, a heat-inactivated sample gave no HNE, proving that it was formed enzymically (Fig. 1), but after 15 min only 5% N O N remained and the level of t-NON was roughly comparable to the reaction with undenatured enzyme (data not shown). W e interprete the lack of total aldehyde recovery, especially with heated enzyme, to reaction with proteins to form Schiff bases or adducts with sulfhydryls. Because a preincubation with 5 m M H 2 0 2 was reported to inhibit the reduction of the initial product, HPNE, into H N E in broad bean ( V i c i a f a b a , L.) preparations [9], we tested preincubation with 5 m M H202 in the low-speed supernatant. However, 5 m M H202 had less effect possibly because of catalase activity present, observed by foaming when H202 was added. After increasing the preincubation concentration to 20 mM, a definite inhibition of HNE formation was observed (Fig. 1). Subsequently, the H2Oz-inhibited reaction was scaled up to 10 ml to attempt to isolate the putative HPNE precursor of HNE (5 min reaction time). The CHC13-extracted products of this reaction (obtained by adding 30 ml C H C 1 3 / C H 3 O H (2:1, v / v ) ) were treated several different ways. A portion was saved for TLC, another was saved for TLC after reducing with triphenylphosphine (a specific hydroperoxide-reducing reagent), and a third was separated by TLC followed by scraping an

upper and a lower UV-absorbing bands. Half the upper band, the putative HPNE, was saved for TLC, and the other half was reduced with triphenylphosphine. All samples were then separated on the same TLC plate with H N E standards (Fig. 2). The results conclusively show that the upper band is transformed by triphenylphosphine into HNE either in the total product mixture or the TLC isolate. Reactions without H2Oz-preincubation showed only the lower UV-absorbing band (HNE) and heat-inactivated enzyme showed neither UV-absorbing band (data not shown). Although one upper band was scraped (spot 2 in Fig. 2), an additional 2,4-dinitrophenylhydrazine reacting spot appeared at a higher Rf in the TLC isolate with much greater intensity compared to the original mixture (compare lane C with lane E, Fig. 2). Although not demonstrated directly, it is likely the new spot is due to free radical decomposition of HPNE during manipulation of the sample. Judging from its migration on TLC and its reddish-orange stain with 2,4-dinitrophenylhydrazine, it is likely that this compound is 4-oxo-2-nonenal, which was observed as a decomposition product of HPNE (see below). After repeating the H202-preincubation experiment, the lower and the upper bands were individually subjected to G C - M S directly, and as O-TMS, and O - b e n z y l o x i m e / O - T M S derivatives. It was confirmed that the lower band was HNE, while the

front-

2

originA

B

C

D

A

E

F

Fig. 2. TLC separation of products obtained from incubation of (3Z)-nonenal with soybean 9300 × g supernatant preincubated with H202 to inhibit hydroperoxide-dependentperoxygenase. Ten ml of diluted low-speed supernatant (0.1 M Ches, pH 9; 2.5 mg protein/ml) was preincubated 5 min with 20 mM H202 at 25°C and then incubated with 2 mM NON for 5 min at 25°C. Products were extracted into CHCI3 by addition of 30 ml CHCI3/CH3OH (2:1, v/v). The CHC13 extract was evaporated carefully to avoid loss of aldehydes. The product mixture was divided into portions for different treatments to be subsequently separated by the TLC shown (developing solvent; ethyl ether/hexane 3:2, v/v). (A) 25 /zg HNE standard; (B) spot 1 previously isolated from TLC and rechromatographed (equivalent to three-fifths of reaction mixture); (C) spot 2 previously isolated from TLC and rechromatographed (equivalent to three-tenths of reaction mixture); (D) same as (C) except isolate was reduced with triphenylphosphine; (E) total reaction products (one-fifth of reaction mixture); (F) same as (E) except mixture was reduced with triphenylphosphine.

H. Takamura, H. W. Gardner/Biochimica et Biophysica Acta 1303 (1996) 83-91

87

3.2. Oxidation and isomerization of (3Z)-alkenals by soybean membrane preparation and soluble fraction

upper band decomposed into several GC peaks. The principal decomposition product from the upper band gave mass spectra consistent with 4-oxo-2-nonenal, [ m / z (percentage relative intensity, ion structure)]: 154 (1, M+); 139 (2, M+-CH3); 125 (100, M+-CHO); 98 (90, M( C H z ) 3 C H 3 + H+); 83 (69, M + - ( C H 2 ) 4 C H 3 ) ; 70 (47); 55 (85); 43 (54). Another less abundant decomposition product was HNE. As observed with the alkoxyl radical conversion of other monoene allylic hydroperoxides, such as 10-hydroperoxy-8-octadecenoic acid [ 14] and 12,13-epoxy9(11)-hydroperoxy- 10(9)-octadecenoic acids [ 15], 4-oxo2-nonenai and HNE would be among the major products expected from decomposition of HPNE. The relative percentage of oxoene to hydroxyene products may vary according to the method used to generate alkoxyl radicals [15]. GC-MS analysis of O-benzyloximes/O-TMS derivatives from the upper band showed a similar pattern of decomposition products.

Since HNE-forming activity in V. faba was membrane bound [9], the low-speed supernatant (in 0.1 M phosphate, pH 6.7) was subjected to 200000 X g to obtain high-speed membrane and soluble fractions. HEX and NON were readily converted to HHE and HNE by the membrane preparation from soybean seeds as assayed by their O-benzyloximes (Fig. 3). This conversion was confirmed by GC and GC-MS analyses and compared to standards. This converting activity was higher for NON to HNE than for HEX to HHE. The soluble fraction had much less hydroxylating activity, but instead the principal activity was isomerization of HEX and NON to t-HEX and t-NON, respectively. It is also shown that heat-inactivated membrane preparation or buffer does not afford HHE or HNE (Fig. 3). These results indicate that soybean membrane 0.2

0.2 A

m e~ e!

e~ ¢.q

@

0.1

,= 0.1

,6

._ ~...,,_..---------~ 0.0

0.0 0

5

10

15

0

Time (min)

0.2

5 10 Time (min)

15

5

15

0.2 B

D

m

i

0.1

0.1

dl

0.0

0.0 0

5 10 Time (min)

15

10

Time (min)

Fig. 3. Time-dependent formation of (2E)-4-hydroxy-2-alkenals and (2E)-alkenals by soybean 200000 x g membrane and soluble fractions. HEX or NON (1 mM) was incubated with membrane preparation (O), heat- inactivated membrane preparation (O), soluble fraction (A), or buffer only ( X ) at 25°C in 0.1 M sodium phosphate buffer (pH 6.7). Protein content was 4 m g / m l . Aldehydes were analyzed by GC of their O-benzyloximes; hydroxyl groups were additionally converted to their O-TMS ether. The mean coefficient of variation were 11.2, 14.9, 10.3, and 9.2% for HHE (A), t-HEX (B), HNE (C), and t-NON (D), respectively.

H. Takamura, H. W. Gardner / Biochimica et Biophysica Acta 1303 (1996) 83-91

88

0.04

0.03

~ O.02 <

,~ ~ ' ~ ' O ~ ~

0.01

v

~ "I~"

OF

0.00 5

I

I

I

I

I

6

7

8

9

10

11

pH Fig. 4. pH-dependent formation of HHE and HNE by soybean 200000 × g membrane preparation. In 0.1 M buffer at various pH values (Mes at pH 5.5-6.5, Epps at pH 7.5-8.5, Ches at pH 9.5, sodium bicarbonate at pH 10.5), 1 mM HEX ((3) or NON ( 0 ) was incubated with membrane preparation (5 mg protein/ml for HEX or 1 mg protein/ml for NON) at 25°C for 1 min. (2E)-4-Hydroxy-2-alkenals were analyzed by GC of their O-TMS ethers of O-benzyloximes. The mean coefficient of variation were 11.9 and 13.9% for HHE and HNE, respectively.

p r e p a r a t i o n c o n t a i n s the e n z y m e s y s t e m w h i c h c o n v e r t s ( 3 Z ) - a l k e n a l s to ( 2 E ) - 4 - h y d r o x y - 2 - a l k e n a l s a n d p r e f e r s N O N to H E X as t h e substrate. Fig. 4 s h o w s the p H - d e p e n d e n t f o r m a t i o n o f H H E a n d HNE by soybean membrane preparation. For both HHE a n d H N E f o r m a t i o n , the a c t i v i t y w a s l o w at acidic p H a n d the o p t i m a l p H w a s 9.5.

3.3. pH-Dependent oxygen consumption by soybean membrane preparation A p a t h w a y w a s p r o p o s e d f r o m N O N to H N E w i t h b r o a d b e a n (Vicia faba L.) m e m b r a n e in w h i c h the first step is o x i d a t i o n o f N O N b y ( 3 Z ) - a l k e n a l o x y g e n a s e [9], a n d it w a s s h o w n a b o v e t h a t the first p r o d u c t o f N O N

0.20

;

"~ 0.15

{.

\

//

.--

'Qx \

z/ o "5 0.10

\~

z //

E

,,~

< 0.05

0.00

C ~

5

•

I

6

7

I

I

I

8

9

10

11

pH Fig. 5. pH-dependent oxygen consumption by soybean 200000 X g membrane preparation. In 0.1 M buffer at various pH values (Mes at pH 5.5-6.5, BTP at pH 7.0-8.3, Ches at pH 9.0-10.0, sodium bicarbonate at pH 10.2-10.6), 1 mM HEX (O) or NON ( 0 ) was incubated with membrane preparation (0.5-1.0 mg protein/ml) at 25°C.

H. Takamura, H. W. Gardner / Biochimica et Biophysica Acta 1303 (1996) 83-91

89

Table 1 Oxygen consumption by either linoleic acid or (3Z)-alkenals in the presence of various lipoxygenase inhibitors at two pH values a Inhibitor

Substrate HEX

Control Hydrogen peroxide ETYA SHAM Propyl gallate NDGA

NON

Linoleic acid

pH 6.5

pH 9.5

pH 6.5

pH 9.5

pH 6.5

pH 9.5

6.0 23.6 3.3 2.2 2.3 2.8

30.4 41.2 35.3 34.3 ND b ND

36.5 43.0 22.1 19.0 3.4 17.7

63.4 67.3 68.1 79.4 ND ND

92.2 7.8 0.0 11.6 0.0 0.0

36.2 6.0 1.1 17.7 ND ND

a Membrane preparations (0.4 to 3.2 mg protein/ml) were preincubated with various lipoxygenase inhibitors (5 mM hydrogen peroxide, 1 mM ETYA, 5 mM SHAM, 5 mM propyl gallate, or 1 mM NDGA) at 25°C for 5 min in 0.1 M Mes at pH 6.5 or 0.1 M Ches at pH 9.5, then either HEX, NON, or linoleic acid was added (all at 1 mM final concentration). b Not determined because of decomposition of propyl gallate and NDGA at pH 9.5.

oxidation was HPNE. This reaction consumes molecular oxygen and theoretically may be monitored by oxygen electrode. Fig. 5 shows the pH-dependent oxygen consumption by a soybean membrane preparation. It is indicated that both HEX and NON were oxidized and the optimal pH was around 9-10, which was similar to that of HHE and HNE formation. This result suggests that (3Z)alkenal oxygenase is involved with HHE and HNE formation.

propyl gallate in the presence of NON. At pH 9.5 inhibition was observed with linoleic acid in the presence of H202, SHAM, or ETYA, but not with either NON or HEX (Table 1). Moreover, the oxygen consumption with linoleic acid substrate was higher at pH 6.5 than at pH 9.5 (Table 1), unlike the optimum of (3Z)-alkenal oxygenase around pH 9 - 1 0 (Fig. 5). These results suggest that (3Z)alkenal oxygenase is distinguished from lipoxygenase activity.

3.4. Inhibition of oxygen consumption by various lipoxygenase inhibitors

3.5. Effects of hydrogen peroxide on the formation of HHE and HNE by soybean membrane preparation

Since soybean is known to contain membrane-bound lipoxygenase activity [16] as well as at least three soluble isozymes, lipoxygenase-1, -2, and -3 [17], it was envisioned that lipoxygenase could be causing the (3Z)-alkenal oxygenase activity. Thus, several lipoxygenase inhibitors [see review 18] were tested for their effect on oxygen consumption after a 5 min preincubation. When linoleic acid was used as the substrate at pH 6.5, 5 mM hydrogen peroxide, 5 mM SHAM, 1 mM ETYA, 5 mM propyl gallate, and 1 mM NDGA were effective inhibitors of oxygen consumption (Table 1). However, with NON or HEX at pH 6.5 H202 proved to be stimulatory, and the others showed much less inhibition, with the exception of

Hydrogen peroxide can be a source of oxygen for a hydroperoxide-dependent peroxygenase (epoxygenase) when simultaneously added with unsaturated substrates (one group uses the term 'peroxygenase' [19] and another 'epoxygenase' [20]). On the other hand, hydroperoxidedependent peroxygenase is strongly inhibited by preincubation with H202 [20]. In the formation of (2E)-4-hydroxy-2-alkenal by Vicia faba both a (3Z)-alkenal oxygenase and a hydroperoxide-dependent peroxygenase were proposed to be involved [9]. Simultaneous addition of H 2 0 2 at pH 6.5 greatly stimulated HHE formation at pH 6.5, but preincubation was similar to the controls (Table 2). HHE formation at pH 9.5 and HNE formation at both

Table 2 Effects of hydrogen peroxide on the formation of HHE and HNE by soybean membrane preparation at two pH values a Formation of HHE (mM)

Control Preincubation Simultaneous addition

Formation of HNE (mM)

pH 6.5

pH 9.5

pH 6.5

pH 9.5

0.013 _+ 0.001 0.016 ___0.002 0.039 ztz 0.000

0.123 _+ 0.003 0.031 + 0.001 0.052 _+ 0.001

0.227 + 0.012 0.039 + 0.002 0.079 + 0.001

0.202 _+ 0.009 0.042 _+ 0.003 0.078 + 0.002

a HEX or NON (1 mM) was incubated with membrane preparation at 25°C for 5 rain in 0.1 M Mes at pH 6.5 or 0.1 M Ches at pH 9.5. Protein content was 5 m g / m l . Control was run without hydrogen peroxide. Hydrogen peroxide (5 mM) was added 5 min prior to incubation or simultaneously. (2 E)-4-Hydroxy-2-alkenals were analyzed by GC of their O-TMS O-benzyloximes.

90

H. Takamura, H.W. Gardner/ Biochimica et BiophysicaActa 1303 (1996) 83-91

Table 3 Formation over time of 9,10-epoxystearic acid (mM) from oleic acid by soybean membrane preparation at two pH values a pH Time (min) 6.5 9.5

1

5

15

0.056 0.002

0.125 0.002

0.238 0.002

a Oleic acid (1 mM) was incubated with 5 mM hydrogen peroxide and membrane preparation at 25°C for 5 min in 0.1 M Mes at pH 6.5 or 0.1 M Ches at pH 9.5. Protein content was 3.5 mg/ml. 9,10-Epoxystearic acid was analyzed by GC of its methyl ester.

pH values were inhibited both by preincubation and simultaneous addition of H 2 0 2, but simultaneous addition was significantly less inhibitory. 3.6. Formation of 9,10-epoxystearic acid from oleic acid by soybean membrane preparation Formation of 9,10-epoxystearic acid from oleic acid, which is catalyzed by hydroperoxide-dependent peroxygenase, was observed after incubation with a soybean membrane preparation at pH 6.5 in the presence of 5 mM H 2 0 2 (Table 3). However, little epoxide was formed at pH 9.5.

4. Discussion

Previously, it was observed that a soybean homogenate catalyzed the conversion of (9S, lOE,12Z)-9-hydroperoxy10,12-octadecadienoic acid into HNE, t-NON, and 9oxononanoic acid [10]. This implied action of hydroperoxide lyase followed by a further conversion of NON into HNE and t-NON. The possibility was strengthened when the conversion of NON into HNE was directly

m

HOQ~

m

Lipoxyg....

OH

e~)'-O2 "

demonstrated in a broad bean seed membrane preparation [9]. In the latter study, evidence was presented to show the probable involvement of a (3Z)-alkenal oxygenase and a hydroperoxide-dependent peroxygenase in the conversion to HNE. In the present study it was shown unequivocally that in soybean HNE is derived from NON and that the activity primarily resides in the membrane fraction. The instability of the activity to heat suggested an enzymic catalysis. A similar, but less active, conversion of HEX to HHE also was catalyzed by the soybean membranes. On the other hand, a (3Z):(2 E)-enal isomerase activity was found in the soluble fraction (high-speed supernatant) that transformed NON and HEX into t-NON and t-HEX, respectively. The enal isomerase has been known for many years, such as the one characterized in cucumber fruit [21]. Until now, the presence of the isomerase in soybean seed has not been directly demonstrated. Its activity probably explains why only the (2E)-alkenals are universally observed as volatile aldehydes in soybean seed preparations, instead of the primary products of hydroperoxide lyase activity, the (3Z)-alkenals. In addition, it should be noted that t-NON formation was relatively small in low-speed supematant adjusted to pH 9, possibly because there was greater competitive activity in HNE-forming activity, but it is also possible that the isomerizing activity is lower at higher pH. In the conversion of NON and HEX to HNE and HHE, respectively, it was shown that 0 2 w a s consumed as a function of pH at a rate largely parallel to the rate of product formation as a function of pH. These O2-uptake data illustrated both the previously unobserved consumption of 0 2 by the (3Z)-alkenal oxygenase and also suggested that the oxidation may be the rate-limiting step. The pH optimum was about 9.5 for both NON and HEX, an optimum not shared with membrane-bound lipoxygenase activity.

~

O

Lip°xygenase~ 02 "

-

-

-

~

-

--O

-

-

~

~

(9Z)-12-Oxo-9-dodecenoic acid

~OH

~

~Jn

OH

-

Hydroperoxidelyase~k~ -

H OH

lyase~

nydroperoxide

9-Oxononanoic acid

o

02~Alkenal oxygenase

H

)

Peroxygenase

02~ Alkenaloxygenase

H

)

Peroxygenase o

(2E)-4-Hydroxy-2-hexenal

(2E)-4-Hydroxv-2-nonenal

Fig. 6. Formation of (2E)-4-hydroxy-2-alkenalsfrom a-linolenic acid (left) and linoleic acid (right) by soybean seed enzymes.

H. Takamura, H.W. Gardner / Biochimica et Biophysica Acta 1303 (1996) 83-91

Beside the disparity in pH optima for the (3Z)-alkenal oxygenase and lipoxygenase, the response of the two activities to typical lipoxygenase inhibitors [18] was considerably different. At pH 9.5 the inhibitors did not inhibit 0 2 uptake by N O N or HEX, but inhibition with the lipoxygenase substrate, linoleic acid, ranged from 51 to 97%. At pH 6.5, the inhibitor hydrogen peroxide was 92% effective with linoleic acid, but was stimulatory with HEX and possibly NON. Other inhibitors at pH 6.5 ranged between 100% to 87% inhibition with linoleic acid, but were notably less inhibitory with N O N and HEX. Taken as a whole these data were highly suggestive that (3Z)-alkenal oxygenase and lipoxygenase are separate enzymes. Previously, it was shown that a hydroperoxidedependent e p o x y g e n a s e / p e r o x y g e n a s e in Vicia f a b a membranes was playing a dual role of forming HNE through both reduction of HPNE (from (3Z)-alkenal oxygenase oxidation of NON) and via rearrangement of 3,4epoxynonanal formed by hydroperoxide O-transfer to NON [9]. It is known from other workers that a hydroperoxidedependent peroxygenase exists in soybean membranes [19]. In the present study, experiments with both preincubation and simultaneous addition of 5 m M H 2 0 2 gave information about the possible involvement of hydroperoxidedependent peroxygenase in the transformation of N O N and HEX to H N E and HHE, respectively. It is known that Vicia f a b a hydroperoxide-dependent peroxygenase is strongly inhibited by a 5 min preincubation with H 2 0 2, but simultaneous addition of H 2 0 2 and an unsaturated substrate supports epoxidation [20]. Although H 2 0 2 preincubation did not inhibit 0 2 uptake in the presence of N O N or HEX, H 2 0 2 did generally inhibit HNE and HHE formation indicating that the (3Z)-alkenal oxygenase was operative but not reduction of the hydroperoxide. Direct evidence was obtained for the formation of HPNE in H 2 0 2inhibited low-speed supernatant, but without H 2 0 2 only HNE was formed. The exception was at pH 6.5 with a simultaneous addition of H 2 0 2 and HEX, in which case HHE production increased three-fold. Preincubation under these conditions had no effect on HHE formation. In the case of this one condition, it seemed that the increased HHE formation was either due to a four-fold increase in precursor 4-hydroperoxy-2-hexenal formation as indicated by 0 2 uptake or by participation of peroxygenase. Significant epoxidation of oleic acid with H 2 0 2 and soybean membrane homogenate was observed at pH 6.5 (Table 3), which confirms the possibility of the participation o f peroxygenase at this pH. It can be seen in Table 2 that in each test, preincubation with H 2 0 2, compared with simultaneous addition, resulted in about double the inhibition of HHE or H N E formation. Previously, with a broad bean membrane preparation it was noted that 1802-1abeled lipid hydroperoxide added simultaneously with N O N inhibited HNE formation, even though significant 180-label was transferred to the hydroxyl of H N E [9]. It seems possible

91

that O-transfer from hydroperoxide to unsaturated substrates by peroxygenase is tightly coupled with hydroperoxide formation, and concentrations o f hydroperoxides greater than metabolic amounts possibly inactivate the enzyme much like that observed with H 2 0 2 preincubation. The pathway from polyunsaturated fatty acid to 4-hydroxy-alkenals initiated by lipoxygenase activity is summarized in Fig. 6 and constitutes the second plant species in which the pathway has been identified. It is plausible that this enzymic system occurs more widely than previously realized. A function for the pathway is not fully known, but it probably plays a role in plant defense [22].

Acknowledgements W e thank Marilyn J. Grove for expert technical assistance.

References [1] Benedetti, A., Comporti, M. and Esterbauer, H. (1980) Biochim. Biophys. Acta 620, 281-296. [2] Esterbauer, H., Schauer, R.J. and Zollner, H. (1991) Free Radic. Biol. Med. 11, 81-128. [3] White, J.S. and Rees, K.R. (1984) Chem. Biol. Interact. 52, 233-241. [4] Uchida, K. and Stadtman, E.R. (1992) Proc. Natl. Acad. Sci. USA 89, 4544-4548. [5] Szweda, L.I., Uchida, K., Tsai, L. and Stadtman, E.R. (1993) J. Biol. Chem. 268, 3342-3347. [6] Uchida, K. and Stadtman, E.R. (1992) Proc. Natl. Acad. Sci. USA 89, 5611-5615. [7] Winter, C.K., Segall, H.J. and Haddon, W.F. (1986) Cancer Res. 46, 5682-5686. [8] Grein, B., Huffer, M., Scheller, G. and Schreier, P. (1993) J. Agric. Food Chem. 41, 2385-2390. [9] Gardner, H.W. and Hamberg, M. (1993) J. Biol. Chem. 268, 68716977. [10] Gardner, H.W., Weisleder, D. and Plattner, R.D. (1991) Plant Physiol. 97, 1059-1072. [11] Corey, E.J. and Suggs, J.W. (1975) Tetrahedron Lett. 2647-2650. [12] Gardner, H.W., Bartelt, R.J. and Weisleder, D. (1992) Lipids 27, 686-689. [13] Smith, P.K., Krohn R.I., Hermanson, G.T., Mallia, A.K., Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J. and Klenk, D.C. (1985) Anal. Biochem. 150, 76-85 [14] Labeque, R. and Marnett, L.J. (1987) J. Am. Chem. Soc. 109, 2828-2829. [15] Gardner, H.W. (1989) Free Radic. Biol. Med. 7, 65-86. [16] Macfi, F., Braidot, E., Petrussa, E. and Vianello, A. (1994) Biochim. Biophys. Acta 1215, 109-114. [17] Axelrod, B., Cheesbrough, T.M. and Laakso, S. (1981) Methods Enzymol. 71,441-451. [18] Vick, B.A. (1993) in Lipid metabolism in plants (Moore, T.S., Jr., ed.), pp. 167-191, CRC Press, Boca Raton, FL. [19l Bl6e, E. and Schuber, F. (1990) J. Biol. Chem. 265, 12887-12894. [20] Hamberg, M. and Fahlstadius, P. (1992) Plant Physiol. 99, 987-995. [21] Phillips, D.R., Matthew, J.A., Reynolds, J. and Fenwick, G.R. (1979) Phytochemistry 18, 401-404. [22] Vaughn, S.F. and Gardner, H.W. (1993) J. Chem. Ecol. 19, 23372345.