A Comparison of Oleic Acid Metabolism in the Soybean - NCBI

Plant Physiol. (1986) 81, 41-44 0032-0889/86/81/0041 /04/$0 1.00/0

A Comparison of Oleic Acid Metabolism in the Soybean (Glycine max [L.J Merr.) Genotypes Williams and A5, a Mutant with Decreased Linoleic Acid in the Seed' Received for publication November 1, 1985 and in revised form January 6, 1986

BARRY A. MARTIN* AND ROBERT W. RINNE Pioneer Hi-Bred International, Inc., Plant Breeding Division, Department of Research Specialists, P.O. Box 85, Johnston, Iowa 50131-0085 (B.A.M.); and United States Department ofAgriculture, Agricultural Research Service, University of Illinois, Urbana, Illinois 61801 (R.W.R.) 18:1, 43 mol% 18:2, and 4 mol% 18:3 (5). While there is some evidence that soybean oleic acid content may be controlled by two or more genes (6) there have been no reports on the biochemical changes accompanying selection for altered oil composition. In higher plants 18:2 is synthesized by desaturation of 18:1 (1, 12, 18). Early studies suggested that oleoyl-CoA was the substrate for this desaturation (18, 19). More recent investigations have demonstrated that safflower microsomes supplied with oleoylCoA initially esterify 18:1 to PC and 18: 1-PC is the substrate for the desaturation reaction (12, 14). Crude homogenates from soybean cotyledons also esterify exogenous oleoyl-CoA to PC by the action of lyso PC 2-acyltransferase before desaturation to 18:2 (15). In addition, the activity of an 'acyl exchange protein' has been postulated to account for formation of linoleoyl-CoA from the 18:2-PC (16). In this report we describe optimization of the reactions that are involved in acylation of 18: 1-CoA to PC and in the desaturation of 18:1 to 18:2. We also compare these reactions in the soybean genotype Williams and in the mutant A5 which contains 2-fold more 18:1 than Williams.

ABSTRACT The metabolism of oleoyl coenzyme A (CoA) was examined in developing seed from two soybean (Glycine max [L.] Merr.) genotypes: Williams, a standard cultivar and A5, a mutant containing nearly twice the oleic acid (18:1) content of Williams. The in vitro rates of esterification of oleoyl-CoA to lysophosphatides by acyl-CoA: lysophosphatidykcholine acyltransferase was similar in both genotypes and lysophosphatidylethanolamine was a poor substrate. Crude extracts desaturated exogenous

I1-'4Cjdioleoyl phosphatidylcholine at 14% of the rate achieved with 11'4CIoleoyl-CoA, and 50 micromolar lysophosphatidylcholine. The desaturase enzyme also required NADH for full activity. Extracts from Williams contained 1.5-fold more oleoyl phosphatidylcholine desaturase activity, on a fresh weight basis, than did A5 and appeared to have a similar affinity for oleoyl-CoA. There was 1.2- to 1.9-fold more linoleic acid (18:2) in phosphatidylcholine from Williams than from A5, measured at two stages of development, but both genotypes had a similar distribution of fatty acids in the one and two positions. Phosphatidylethanolamine in A5 contained relatively more linoleic acid (18:2) in the one position than did Williams. The increased oleic acid (18:1) content in A5 appeared to be a result of decreased rates of 18:1 desaturation of oleoyl-phosphatidylcholine in this genotype.

MATERIALS AND METHODS Growth Conditions. Seed of genotype Williams was obtained from R. L. Bernard, USDA/ARS, University of Illinois and A5 was obtained from W. R. Fehr, Iowa State University. Seeds were grown in the greenhouse in solution culture as described previously (10). Nutrient solution was changed every 7 d. Photoperiod and temperature were 13 h 28°C/l 1 h 24°C, day/night. Irradiance was 300 to 400 uE m2 s-' (400-700 nm). Enzyme Extraction. Immature seed were taken 25 and 35 DAF. The testa and embryonic axes were removed and the cotyledons were chilled to 4°C. The cotyledons were homogenized with a mortar and pestle in 10 volumes of buffer containing 50 mM Hepes (pH 7.2) and 40 mM Na-ascorbate and filtered through Miracloth. The crude filtrates were used directly for assays. Glycerolipid Extraction. Cotyledons were chilled to 4°C and were homogenized in CHC13/methanol (2:1 v/v) in a Polytron. The homogenates were vacuum filtered through Whatman No. 1 paper and the residue was rinsed with methanol. The remaining plant material was removed from the filter paper and homogenized and filtered as before. The combined filtrates were made 1:1:0.9 (v/v/v) CHC13/methanol/H20. After vortexing and centrifugation the CHC13 phase was dried in vacuo and glycerolipids were dissolved in 1 ml CHCl3/methanol (2:1 v/v) and stored at -200C.

The fatty acid composition of oil from mature soybean (Glycine max [L.] Merr.) seed is important because stability of the oil and flavor of products made from it are affected by its linolenic acid (18:3)2 content (3). Most commercial soybean varieties have similar oil composition and contain: 10 to 11 mol% palmitic acid (16:0), 3 to 4 mol% stearic acid (18:0), 18 to 22 mol% oleic acid (18:1) 56 to 60 mol% linoleic acid (18:2), and 7 to 8 mol% 18:3 (8). Treatment of seed with ethyl methanesulfonate has been used to generate mutant stains with altered oil composition. One of these mutants, A5, contains 40 mol% ' Cooperative investigations of the Agricultural Research Service, United States Department of Agriculture, and Department of Agronomy, University of Illinois, Urbana, IL 61801 and supported in part by Grant No. 80465 from the research foundation of the American Soybean Association. 2Abbreviations: 18:3, linolenic acid; 16:0, palmitic acid; 18:0, stearic acid; 18:1, oleic acid; 18:2, linoleic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; DAF, days after flowering; CHAPS, 3-[(3-

cholamidopropyl)dimethyl ammonio]- I -propanefulfonate. 41

42

MARTIN AND RINNE

Lipid Analysis. The glycerolipid fractions were applied to adsorbasil plus 1 TLC plates which were developed in acetone:petroleum ether (75:25 v/v) followed by CHCl3:methanol:glacial acetic acid:H20 (65:15:10:3.5). Lyso-PC, PC, phosphatidyl ethanolamine (PE) and lyso-PE were identified by cochromatography with authentic standards. The PC, PE, lyso-PC, lyso-PE, and free fatty acids were eluted from the plates and fatty acid composition was determined by GC as described elsewhere (7). Stereospecific analysis of PC was accomplished by addition of 0.5 to 1.0 mg of PC or PE to tubes, drying under a stream of N2 gas, then solubilizing by vortexing in 0.1 ml of 10% CHAPS. After the phospholipids were solubilized 0.9 ml of 50 mM Bicine (pH 8.3) and 10 units of phospholipase A2, from Naja naja venom, was added. This mixture was shaken at 150 rpm for 3 h. The reaction was stopped by addition of 2 ml CHCl3/methanol (1:1 v/v). The CHC 13 phase was taken to dryness under N2 gas. Lyso-PC, PC, lyso-PE, PE, and free fatty acids were separated and quantified as described earlier. Lyso-PC and lyso-PE obtained from this method were used as substrates for the lyso-PC 2-acyltransferase studies. Enzyme Assay. Oleic acid desaturase activity was measured by incubation of plant extracts either with 32 ,M [ I -'4C]oleoyl-CoA, 51 uCi/Mmol, (NEN or Amersham) in 50 mm Hepes (pH 7.2), 20 mm Na-Ascorbate, 50 MM lyso-PC, and 1 mM NADH. Alterations or additions to this mixture are noted in the appropriate legends. The assay volume was 200 Ml. After 1 h assays were stopped by addition of 1 ml CHCl3:methanol (1:1 v/v) then, 150 IA of H20 and 50 Al of 0.5 N NaOH was added to hydrolyze Table I. Effects of Various Additions and Deletions from the Basic

Plant Physiol. Vol. 81, 1986

0 x

_> .) cs .

2.5

co

E 20

c

E c

10

1.0 Time (h)

2.0

FIG. 1. Activity of 18:1 desaturase (A) and lyso-PC 2-acyltransferase (0) over 2 h period. These assays were performed using a crude extract

from Williams seed 21 DAF. Note the difference in vertical scale for the two enzymes.

acylthioesters in the methanol:water phase and form a biphasic solution. The assays were stored at -20C overnight to allow saponifiOleic Acid Desaturase Assay cation of the acylthioesters. The mixure was neutralized by Alteration of the Percent of addition of 50 Ml of 0.5 N HCI then vortexed and centrifuged at Basic Assay Control 2000g for S min. The methanol:H20 layer was discarded and the CHC13 layer was dried under a stream of N2 gas at 45C. The Nonea 100 fatty acids were methylated by addition of 0.5 ml of 0.5 N NaOH (-) Lysophosphatidyl choline 30 in methanol kept anhydrous by addition of SA molecular sieves. (+) 0.1 mm n-Propyl gallate 61 After 30 min, 0.5 ml of 0.5 N HCl and 0.5 ml CHC13 was added (-) NADH + FMNH2 5 and the mixtures were vortexed and centrifuged at 2000g for 5 (-) NADH + NADPH 60 min. The methanol:H20 layer was discarded and the CHC13 was (+) 1% Octylglucoside 40 dried under a stream of N2 gas. The fatty acid methyl esters (+) 1% CHAPS 56 were dissolved in methanol and separated by reverse phase (-) Oleoyl CoA chromatography on a Waters C-18 radial compression column. (+) Phosphatidylcholine [1-'4C]dioleoyl 14 The eluant was methanol/H20 (92:8 v/v). The flow rate was 2.5 (+) 10 mg/ml BSA 100 ml min-'. Peaks were detected by absorption at 206 nm. The *Complete assay system contained; 32 uM [1-'4C]oleoyl-CoA, I mM 18:3-me, 18:2-me, and 18:1-me peaks were collected. RadioacNADH, 50 mm Hepes (pH 7.2), 50 Mm lysophosphatidylcholine, 20 mM tivity was measured by addition of tritosol (4) cocktail and radioactivity was counted in a Packard 3330 scintillation counter. Na-ascorbate and crude extracts from Williams. Lyso-PC 2-acyltransferase activity was measured using the same reaction mixture. Assays were stopped after 5 min by Table II. Rates of Oleic Acid Desaturation in A5 and Williams Seed at addition of 1.0 ml CHC13:methanol (1:1 v/v) and 0.3 ml water 25 and 35 DAF Using Concentrations of Oleoyl CoA Near the was added. The mixture was then vortexed and centrifuged at Km (8 gM) and the V,,m. (32 gM) 2000g for 5 min. The methanol:H20 layer was discarded and the CHC13 phase was dried under a stream of N2 gas. The lipids Variety Age Oleoyl-CoA Activity were applied to a TLC plate and phospholipids were separated as described earlier. The PC or PE spot was scraped into 10 ml DAF JM nmol 18:2 h-' g' fresh wt of toluene cocktail (toluene, 12 g PPO, 0.5 g POPOP); radioacA5 25 8 1.25 + 0.06a tivity was measured by Packard 3330 scintillation counter. ± Williams

2.32 0.40 7.97 ± 0.89 12.36 ± 0.40 35 8 1.74 ± 0.16 Williams 2.27 ± 0.17 A5 32 11.08 ± 1.40 Williams 16.27 ± 1.59 aStandard error of the mean of three replicates, two samples per replicate. A5 Williams A5

32

RESULTS The desaturase assay, described previously (7) was investigated to find factors which increased or decreased activity (Table I). Use of 20 mM Na-ascorbate in the grinding buffer and assays made our results more reproducible but was not strictly necessary for activity in crude extracts. Exogenous lyso-PC was necessary to obtain full activity and losses of activity from 40 to 50%

OLEIC ACID METABOLISM IN SOYBEAN Table

III.

Comparison of Lyso-PC 2 Acyltransferase Activity in A5 and Williams Using Lyso PC and Lyso PE, as the Substrates

Variety A5 Williams A5 Williams A5 Williams A5 Williams

Substrate , M 25 Lyso PC

nmol P-lipid h-' g' fresh wt 364.0 458.0 438.0 490.0 10.0 14.0 13.0 18.0

100 Lyso PE

25

100

occurred when it was omitted. Activity in its absence was probably due to endogenous lyso-PC (-5 gM) in the crude extracts. NADH was the best electron donor we examined, and was necessary for full activity. Octylglucoside and CHAPS were the only detergents examined which we could add to the assays and still retain activity. Addition of Triton X-100, Na-cholate, Nadeoxycholate, and Tween 20 all resulted in complete loss of activity (data not shown). When 32 AM[l-'4CJdioleoyl PC (51 uCi Mmol-') was solubilized in CHAPS and substituted for oleoyl CoA in the assays, only 14% of the CHAPS control activity was detected. This result was the same whether crude extracts, microsomal pellets, or CHAPS solubilized enzyme was used in the assay and was similar to results obtained with safflower microsomes (16). In our experiments with soybean we found addition

43

of BSA to assays was unnecessary in contrast to reports by workers with safflower desaturase (14). This difference may be due to the abundant amounts of soybean storage protein in our extracts. When 0.1 mM n-propyl gallate was added to the assays to inhibit lipoxygenase activity and thus reduce oxidation of 18:2, a 40% inhibition of desaturase activity was observed. In these assays no reduction of 18:1 CoA acylation to PC was observed. Full activity of 18:1 desaturase was saturated by 100 AM 02 (data not shown) which is similar to the O2 requirement of soybean stearoyl-ACP desaturase (13), and oleoyl PC desaturase from safflower cotyledons (2). Rates of 18:1 desaturation averaged 35% lower in extracts from A5 compared to Williams extracts at 25 and 35 DAF (Table II). Rates of 18:1 desaturation were determined using 8 and 32 ,uM oleoyl CoA since the desaturases in similar extracts from safflower seed and potato tubers were reported to have Km values for oleoyl-CoA of 10 (17) to 25 iM (1), respectively. The rate observed at 8 gM oleoyl-CoA was 15% of that obtained at 32 Mm oleoyl CoA. The ratio of desaturase activities at 8/32 gM oleoyl CoA was similar for both genotypes and the mean ratio from both ages was 0.16. When total desaturase activities from Williams and A5 were compared at different times after flowering the activity ofWilliams was 1.3- to 1.9-fold greater than observed in A5. No attempts were made to calculate a Km for desaturation of oleoyl CoA since this reaction has been shown to be preceded by acyl CoA-lyso PC 2 - acyltransferase (12, 14). Previous Km values

Table IV. Molecular Species Analysis of PC and PE Extractedfrom Soybean Cotyledons 25 and 35 DAF Values represent a mean from three experiments. Ratio 18:1 Position ( a or (Total) RI/R2a Genotype Age 16:0 16:1 18:0 18:1 18:2 18:3 Mmol (18:2 + 18:3) GLb mol % g Phosphatidylcholine 1.12 1.49 7.9 35.7 27.9 3.9 1 23.9 0.7 25 A5 1.6 38.1 46.5 4.3 0.75 2 6.2 3.3 4.6 4.7 36.9 37.3 4.1 PC 15.0 2.0 1.68 0.67 9.2 25.2 29.3 8.2 27.7 0.4 WilliamsI 0.40 0.9 26.9 61.0 7.0 2 3.3 0.9 5.0 26.0 45.2 7.6 4.8 PC 15.5 0.7 1.60 2.29 6.1 51.4 20.9 1.5 1 19.9 0.7 35 A5 1.43 2 0.5 57.2 37.0 2.9 1.7 0.7 3.3 54.3 29.0 2.1 7.6 PC 10.8 0.5 1.70 0.34 26.7 0.6 11.2 15.5 37.6 8.4 WilliamsI 0.20 1.2 15.6 72.3 6.5 2 3.3 1.1 PC 6.2 15.5 54.9 7.5 8.0 15.0 0.9 Phosphatidylethanolamine 4.2 18.5 27.0 11.1 1 37.4 1.8 1.5 35.5 48.0 6.4 2 7.2 1.4 3.1 2.9 27.0 37.5 8.8 PE 22.2 1.6 4.3 13.4 22.0 16.0 1 43.1 1.2 Williams 2 4.8 2.3 1.4 23.6 59.3 8.6 3.3 PE 2.8 18.5 40.7 12.4 24.0 1.6 1 31.8 0.4 3.3 32.1 27.6 4.8 35 A5 1.0 59.5 31.7 2.7 2 3.6 1.5 5.2 PE 2.2 45.8 29.7 3.6 17.7 1.0 8.0 25.4 16.2 1 5.9 43.6 0.9 Williams 1.3 14.4 71.9 6.3 2 4.3 1.8 5.6 PE 3.6 11.2 48.7 11.2 14.0 1.3 ' R l/R2 = the ratio R = (18:1/18:2+18:3) of position I divided by R of position 2.

A5

25

0.49 1.65

0.30

0.35 0.35

1.00

0.99

0.57

1.73 0.19 0.18 I

1.06

Glycerolipid.

44

MARTIN AND RINNE

Plant Physiol. Vol. 81, 1986

measured for the desaturase (1, 17) using 18:1 CoA as a substrate

18:1-PC was such a poor substrate for the enzyme. Our attempts to obtain saturation kinetics for 18:1 desaturase were also complicated by sigmoidal kinetics. When the amount of endogenous PC in the reaction mixture was accounted for, 18:1 desaturase utilized endogenous 18:1-PC at only 20 to 30% of the rate achieved with 18: l-CoA and lysoity than desaturase activity, (note the different scales in Fig. 1). PC (Table I). This stimulation of desaturase activity by lyso-PC Similar results were obtained with preparations from A5 cotyle- is further evidence that oleoyl-PC is the substrate for the desatdons (data not shown). Soybean microsomal lyso-PC 2-acyltrans- urase reaction in soybean cotyledons. ferase was saturated by about 50 gM oleoyl CoA and had an The difference in fatty acid composition of PE molecular apparent Km for oleoyl-CoA of 11 Mm (data not shown). These species in these genotypes warrants further investigation since it results are similar as was obtained with safflower microsomes may indicate other differences in fatty acid metabolism between (9). The lysophospholipid substrate specificity of the acyltrans- these two genotypes. ferase was also examined, since 18:1 PE has been proposed as an LITERATURE CITE!;D alternate substrate for the desaturase (11, 17). The rates with lyso-PC were about 40-fold higher than those achieved with lyso- 1. ABDELKADER AB, A CHERIF, C DEMANDRE, P MAZLIAK 1973 The OleylPE (Table III). This data combined with results from in vivo coenzyme-A desaturase of potato tubers. Enzymatic properties, intracellular location and induction during "aging" of tuber slices. Eur J Biochem 32: labeling experiments (20) lends support to the notion that oleoyl155-165 PC is also the preferred substrate for 18:1 desaturase in soybeans 2. BROWSE J, CR SLACK 1983 The effects of temperature and oxygen on the rates as for safflower seeds cotyledons reported ( 14). of fatty acid synthesis and oleate desaturation in safflower (Carthaurus Extraction of PC and PE from A5 and Williams seed, 25 and tinctorius). Biochim Biophys Acta 753: 145-152 35 DAF, indicated that both genotypes had similar amounts of 3. DuTroN JH, CR LANCASTER, CD EVANS, JC COWAN 1951 The flavor problem of soybean oil VI 1 1. Linolenic acid. J Am Oil Chem Soc 28: 115-118 these phospholipids (Table IV). Stereospecific analysis revealed U 1975 Tritosol: A new scintillation cocktail based on Triton X-100. differences in both the amounts of polyunsaturated fatty acids 4. FRICKE Anal Biochem 63: 555-558 in the one and two positions of PC and PE as calculated by 18:1/ 5. HAMMOND EG, WR FEHR 1983 Registration of A5 germplasm line of soybean. Crop Sci 23: 192 18:2 + 18:3 = R and the ratio of polyunsaturated fatty acids BA, BF CARVER, JW BURTON, RF WILSON 1982 Inheritance of fatty between the one and two positions as calculated by R 1/R2. 6. MARTIN acid composition in soybean seed oil. Soybean Genet Newsl 10: 89-92 While PC from Williams was more unsaturated than A5 the 7. MARTIN BA, ME HORN, JM WIDHOLM, RW RINNE 1984 Synthesis, composirelative distribution of 18:1, 18:2, and 18:3 in the one and two tion and location of glycerolipids in photoautotrophic soybean cell cultures. Biochim Biophys Acta 796: 1146-1154 positions was similar in these genotypes. The R I/R2 of PE BA, RW RINNE 1985 Relationship between fatty acid composition of indicated a greater proportion of the polyunsaturated fatty acids, 8. MARTIN vegetative and reproductive structures of six soybean genotypes. Crop Sci 18:2 and 18:3, in the one position than in the two position in 25: 1055-1058 of an acylcomparison to Williams. These results indicate a difference in 9. MOREAU RA,A.PK STUMPF 1982 Solubilization and characterization coenzyme O-Lysophospholipid acyltransferase from the microsomes of acylation or acyl exchange during PE biosynthesis in A5 comdeveloping safflower seeds. Plant Physiol 69: 1293-1297 pared to Williams. 10. ROSENBURG LR, 1984 A technique to impose maturation on developing were probably affected by the kinetic properties of the acyltransferase. In our desaturase assays acyltransferase incorporated oleoyl CoA into PC at a linear rate for 15 min, while the desaturase produced 18:2 at a linear rate for up to 2 h (Fig. 1). In addition, there was at least 30-fold more acyltransferase activ-

soybean seed. Masters thesis. University of Illinois, Urbana

DISCUSSION The 18:2 content of PC and PE in Williams cotyledons was nearly 2-fold that in A5, a similar result as was reported for total oil composition (5). Two possible mechanisms for this alteration in 18:1 desaturation to 18:2 are: (a) a difference in supply of 18:1 to the desaturation enzyme or (b) a reduced activity of the desaturase enzyme. Williams and A5 contained similar levels of PC and PE (Table IV) and similar activities of acyl-CoA; lyso-PC 2-acyltransferase (Table III). In addition the acyltransferase from both genotypes preferred lyso-PC as a substrate. These experiments indicated that during the desaturase assays there should be no significant difference in the supply of 18:1 in the form of ['4C]18:1-PC to the desaturase enzyme. The activity of 18:1 desaturase in A5 was 35% lower than in Williams on either a fresh weight (Table II) or per seed (data not shown) basis. These data indicate that the primary lesion in the A5 mutant was lower 18:1 desaturase activity. Our comparison of desaturase from A5 and Williams at 8 and 32 Mm oleoyl CoA indicated that alterations in the supply of 18:1-CoA were proportionally represented in the rates of 18:2 formation. It was impossible to make a precise comparison of the kinetic properties of the desaturases from these two genotypes since exogenous

11. SANCHEZ J, PK STUMPF 1983 The effect of WY14643 and lysophospholipids on phospholipid metabolism and desaturation by maturing safflower seed

microsomes. Plant Physiol 725: 20 12. SLACK CR, PG ROUGHAN, J BROWSE 1979 Evidence for an oleoyl phosphatidylcholine desaturase in microsomal preparations from cotyledons of safflower (Carathamus tinctorius) seed. Biochem J 179: 649-656 13. STUMPF PF, RJ PORRA 1976 Lipid biosynthesis in developing and germinating soybean cotyledons. Arch Biochem Biophys 176: 63-70 14. STYMNE S, LA APPLEQVIST 1978 The biosynthesis of linoleate from oleoylCoA via oleoyl-phosphatidylcholine in microsomes of developing safflower seeds. Eur J Biochem 90: 223-229 15. STYMNE S, LA APPELQVIST 1980 The biosynthesis of linoleate and a-linolenate in homogenates from developing soybean cotyledons. Plant Sci Lett 17: 287294 16. STYMNE S, AK STOBART, G GLAD 1983 The role of the acyl-CoA pool in the synthesis of polyunsaturated 18-carbon fatty acids and triacylglycerol production in the microsomes of developing safflower seeds. Biochim Biophys Acta 752: 198-208 17. TREMOLIERES A, D DRAPIER, JP DUBACQ, P MAZLIAK 1980 Oleoyl-Coenzyme A metabolism by subcellular fractions from growing pea leaves. Plant Sci Lett 18: 257-269 18. VIJAY IK, PK STUMPF 1970 Fat metabolism in higher plants. XLVI. Nature ofthe substrate and the product of oleoyl coenzyme desaturase from Carthamus tinctorius. J Biol Chem 246: 2910-2917 19. VIJAY IK, PK STUMPF 1971 Fat metabolism in higher plants. XLVIII. Properties of oleoyl coenzyme A desaturase of Carthamus tinctorius. J Biol Chem 247: 2910-2917 20. WiLSON RF, HH WEISSINGER, JE BUCK 1980 Involvement of phospholipids in polyunsaturated fatty acid synthesis in developing soybean cotyledons. Plant Physiol 66: 545-599