Gaba shunt in developing soybean seeds is ... - Wiley Online Library

PHYSIOLOGIA PLANTARUM 94: 219-228, 1995

Cpw

Printed in Denmark - alt rif-ht>. reserved

Gaba shunt in developing soybean seeds is associated with hypoxia Barry J. Shelp, Craig S. Walton, Wayne A. Snedden, Lucie G. Tuin, Ivan J. Oresnik and David B. Layzell

Shelp, BJ., Walton.C.S..Snedden,W.A..Tuin.L.G.,Oresnik,I.J.andLayzell.D.B. 1995. Gaba shunt in developing soybean seeds is associated with hypoxia. - Physiol. Plant. 94: 219-228. In the present study we investigated the proposal that Ihe )'-aminobutyrate (Gaba) shunt in developing soybean (Glycine max [L.] Merr.) seeds is associated with hypoxia. The ontogeny and pH profile of enzymes associated with glutamate metabolism (glutamate decarboxylase [EC 4.1.1.15]. Gaba transaminase [EC 2.6.1.19], succinic semialdehyde dehydrogenase [EC 1.2.1.16], glutamate dehydrogenase {EC 1.4.U], glutamate:oxaloacetate transaminase [EC 2.6.1.1], glutamate:pyruvate transaminase [EC 2.6.1.2] and 2-cxoglutarate dehydrogenase complex [EC 1.2.4.2]) and hypoxia (aicohol dehydrogenase [ADH, EC 1.1.1.1] and pyruvate decarboxylase [PDC, EC 4.1.1.1]) were determined in cotyledons, nucellus and seed-coat tissues. Gaba-shunt enzymes were ubiquitous in the developing seed. Activities of enzymes catalyzing glutamate-C entry into the Krebs cycle via 2-oxoglutarate were generally greater than those of Gaba-shunt enzymes. In cotyledons, the activity of ADH increased throughout seed development (up to 72 days after anthesis [DAA]), whereas PDC was static during early development, then increased. In contrast, the activities of ADH and PDC in maternal tissues (nucellus and seed coat) were initially high, then declined dramatically after .37 DAA. The adenylate energy charge (AEC) = ([ATP] + 0.5 [ADP])/ ([ATP] + ]ADP] + [AMP]) of soybean seeds from fruits (37 DAA) frozen in situ was low (0.67±0.01) compared to the AEC of adjacent pod tissue (0.82±0.04) and cotyledons exposed to air (0.84±0.01). A 60-min time-course study showed that the rate of ]U-'^C]-glutamate catabolism by an intact excised cotyledon at 37 DAA was markedly lower at 8 and 0% O: than at 21%; the pool size of ['•'C]-Gaba was unaffected. The data indicated that: (i) Gaba-shunt activity is not a response to limited glutamate deamination/transamination: (2) the soybean seed is hypoxic; and (3) the relative partitioning of glutamate-C through glutamate decarboxylase is increased by hypoxia. Key words - Adenylate energy charge, Gaba shunt, glutamate decarboxylase, Glycine max. hypoxia. Krebs cycle, stress. B.J. Shelp icorresponding author). C.S. Walton, W.A. Snedden andL. G. Tuin. Dept of Horticultural Science and Interdepartmental Plant Physiology Program. Univ. of Giielph. Guelph, ON. Canada N1G2W1; I.J. Oresnik and D.B. Layzell. Dept of Biology. Queen's Univ., Kingston. ON. Canada K7L3N6.

Introanction Developing seeds must receive their nutrients for growth and development from the mother plant (Pate 1984, Thome 1985), The major form of N delivered to a soybean embryo is generally helieved to he glutamine (Rainbird et al, 1984), hut Hsu et al, (1984) suggested that glutamine is quantitatively converted to glutamate in the

nucellus, which separates the seed coat from the embryo, Balance sheets for the utilization of amino compounds in maturing fruits of cowpea (Peoples et al. 1985) and in developing seeds of soybean (Micallef and Shelp 1989a) indicate that many other amino acids present in seed protein are synthesized in situ. Thus, glutamate and/or glutamine are probably involved in a number of biosynthetic reactions leading to the synthesis of amino

Received 7 September, 1994; revised 17 February, 1995 Physiol.Planl. 94, 1995

219

CO2

Fig. 1. Some metabolic routes associated with glutamate metaboli.sm and hypoxia.

acids. Indeed, soybean cotyledons cultured in vitro possess the capability to synthesize all amino acids required for protein synthesis when glutamine is the only nitrogen source (Thompson et al. 1977), It is also believed that glutamate is produced in the soybean cotyledon as a means of recycling arginine-derived C and N (Micallef and Shelp 1989b), Glutamate is involved in a number of metabolic pathways, including the production of glutamine via glutamine synthetase and 2-oxoglutarate via transamination (glutamateioxaloacetate transaminase [GOT] or glutamate:pyruvate transaminase [GPT]) and deamination (GDH) reactions (Fig, 1). Glutamate can also be directly decarboxylated via glutamate decarboxylase (GDC) to y-aminobutyrate (Gaba), which is then converted to succinic semialdehyde and succinate by Gaba transaminase (Gaba-T) (in vivo amino acceptor uncertain) and succinic semialdejiyde dehydrogenase (SSADH), respectively (Bown and Shelp 1989, Satya Narayan and Nair 1990), These last three reactions constitute the Gaha shunt and permit glutamate-C to enter the Krebs cycle, while bypassing GDH and 2-oxoglutarate dehydrogenase (OGDH) reactions. Recently, Tuin and Shelp (1994a) found that ['"C]glutamate supplied to an intact soybean cotyledon excised at 37 days after anthesis (DAA) is rapidly metabolized via multiple routes, including deamination, trans220

amination and decarboxylation. Labeled Gaba is recovered from metabolism of [U-"Cl-glutamate, but not [l-'"'C]-glutamate. demonstrating its production by decarboxylation of the carbon one of glutamate. The in situ flux of glutamate through glutamate decarboxyiase, as e.stimated from the changing specific activity of ['""CjGaba during the early metabolism of [U-'*C]-glulamate, is comparable to the rate of direct glutamate incorporation into protein (L.G, Tuin and B.J. Shelp, unpublished data). Despite the prominence of the Gaba shunt, its physiological role in developing soybean cotyledons in particular, and in plants in general, remains uncertain. The accumulation of Gaba is associated with a variety of stress conditions including hypoxia, cold and insecl and mechanical damage (Bown and Shelp 1989, Satya Narayan and Nair 1990, see references therein), as well as HVglutamate symport (Chung et al. 1992, Snedden et al. 1992), Hypoxia causes cyto.solic acidification (Roberts el al. 1984. Kurkdjian and Guem 1989, Yoshida 1994), which in tum can stimulate GDC activity (Streeter and Thompson 1972a, Snedden et a!, 1992) and Gaba synthesis (Crawford et al. 1994). Furthermore, hypoxia decreases the [NAD]/[NADH] ratio (Raymond et. al. 1987) which may restrict the flow of glutamate-C into the Krebs cycle. Research by Sinclair (1988) suggests that the diffusion of O; through tissues of developing soybean seeds limits metabolism. If that is the case, Gaba synthesis may be a response to hypoxia. In the present study we have addressed the following questions: (1) Do enzymes responsible for 2-oxogiutarate metabolism limit entry of glutamate-C into the Krebs cycle? (2) Is the developing soybean seed operating under hypoxic conditions? (3) Is the relative partitioning of glutamate-C into the Gaba shunt increased by hypoxia^ This was achieved by determining: the amino acid composition of cotyledons, nucellus and seed-coat tissues, together with the ontogeny and pH response of some enzymes associated with hypoxia, the Krebs cycle and the Gaba shunt; the adenylate energy charge (AEC) of seeds attached to or detached from the mother plant; and the in situ metabolism of [U-'''C]-glutamate by detached cotyledons in response to various Oi concentrations. Abbreviations - ADH, alcohol dehydrogenase [EC l.t.1.1]; AEC, adenylate energy charge: DAA, days after anthesis: Gaba, y-aminobutyrate; Gaba-T, Gaba transaminase (EC 2.6.1.19): GDC, glutamate decarboxylase (EC 4.1.1.15): GDH, glutamate dehydrogena.se (EC 1.4.12): GOGAT, glutamine:2-oxoglutarate transaminase (EC 1.4.1.14), GOT, glutamate:oxaloacetate transaminase (EC 2.6.1.1); GPT, glutamate:pyruvate transaminase (EC 2.6.1.2): GS, glutamine synthetase (EC 6.3.1,2); LDH, lactate dehydrogenase (EC 1.1.1.27); OGDH, 2-oxoglutarate dehydrogenase complex (EC 1.2.4.2); PDC, pyruvate decarboxylase (EC 4.1.1,1); SSA, succinic semialdehyde; SSADH, succinic semialdehyde dehydrogenase (EC 1,2,1.16).

Physiol. Plan; 94, 1995

Materials and methods Plant material Non-nodulated soybean [Glycine max (L.) Merrill cv. Maple Arrow] plants were grown in Promix (Premier Brands, Inc., Red Hill, PA, USA), five per 9-1 pot, in a naturally lighted greenhouse with day/night temperatures of 23/I7°C, During the winter months natural lighting was supplemented by high-intensity sodium vapor lamps (model SDN AGRO 430 W 2/3, Philips Lighting Co., Somerset, NJ, USA), yielding a photosynthetic photon flux density of 100 \imol m~- s~' at pot level and a 16-h light/8-h dark period. Each pot was watered twice weekly with 1 1 of a one-quarter strength Hoagland's nutrient solution (Hoagland and Amon 1950) containing 8 mM N and watered other days with tap water. Preparation and assay of cell-free extracts Fruits were harvested at random and developing seeds were removed. The cotyledons (the embryonic axis was discarded), nucellus (peeled from inner surface of seed coat) and seed coat were collected and either frozen in liquid Ni within 2 min after harvest or used directly for preparation of cell-free extracts. Freshly harvested material was used for the determination of SSADH and GabaT activities, which are .sensitive to freezing in liquid N:. The sample number in each pooled replicate differed depending on the tissue weight and number needed to produce a 1- to 2-g sample. Tissue was ground with a chilled mortar and pestle in five volumes of 0.2 M potassium phosphate buffer (pH 7.5), containing 0.1 mM ethylenediaminetetracetic acid, 1 mM dithiothreitol, 0,1 mM pyridoxal phosphate, 5 mM 2-mercaptoethanol and 5% (v/v) glycerol. The homogenate was centrifuged at 23000 s for 15 min. Two ml of the supernatant were desalted using a Pharmacia (Baie d'Urfe, QC, Canada) Sephadex G-25 column (1,6 x 8 cm). All operations were conducted at 4°C. The optimum substrate, pH and cofactor requirements were identified for each enzyme using cotyledonary tissue (C,S, Walton 1993. Thesis, Univ, of Guelph, Guelph, ON, Canada), All assays were conducted at 30°C within 2 h of desalting the cell-free extracts (about 3 h after tissue harvest). All spectrophotometric assays to measure the reduction of NAD or the oxidation of NADH at 340 nm were conducted on a Beckman (Mississauga, ON, Canada) DU-64 spectrophotometer connected to a circulating water bath. Enzyme kinetics were tabulated and plotted using a Beckman Kinetic Soft-Pak module. Enzyme activities are presented as nmol tissue"' min"' calculated by dividing the activity in nmol [g fresh weight]"' min"', by the average sample weight. Automatic derivatization with ortho-pthalaldehyde and reverse-phase high performance liquid chromatography (HPLC), using the buffers and column described by Oaks et al. (1986), were used to separate the amino acids produced in enzyme assays. Physiol. Planl. 94. 1995

Glutamate decarboxylase (GDC) activity was determined as [l-"C]-glutamate-dependent production of '••CO, by the method of Snedden et al, (1992). Gaba transaminase activity was determined as the Gaba-dependent production of alanine and glutamate, respectively, with pyruvate or 2-oxoglutarate as amino acceptors, using a modified method of Satya Narayan and Nair (1986a), A 1-ml assay mixture contained 100 mMTrisHCI (pH 8.2), 2 mM Gaha, 20 \iM pyridoxal phosphate, either 10 mM pyruvate or 10 mM 2-oxoglutarate and desalted extract. The reaction mixture was incubated for 3 h, then sulfosaiicylic acid was added to a final concentration of 60 mM and the precipitate was removed by centrifugation. The supernatant was neutralized with 1 M NaOH, then diluted with HPLC-grade H2O for amino acid analysis. The complete reaction mixture minus Gaba or the oxo-acid served as controls. Succinic semialdehyde dehydrogenase (SSADH) was assayed by a method modified from Satya Narayan and Nair (1989). A 1-ml assay mixture contained 100 mJW 3-([l,l-dimethyl-2hydroxy-ethy!3amino)-2-hydroxy-propanesulfonic acid buffer (pH 9.5), 75 \iM succinic semialdehyde, 0,5 mAf NAD, 14 mM 2-mercaptoethanol and desalted extract. Glutamate dehydrogenase (GDH) was measured in the deaminating direction by a method modified from Me Kenzie et al. (1981). A 1 -ml assay mixture contained 100 mMTris-HCI (pH 8.8), 40 mM glutamate, 0.25 mM NAD and desalted extract. The reaction rate was measured before and after the addition of glutamate. The 2-oxoglutarate dehydrogenase complex (OGDH) was determined by the method of Karam and Bishop (1989), using (2-hydroxyethyl)imino-tris[hydroxymethy!]methane buffer (pH 6.5). Glutamate:oxalo acetate transaminase (GOT) and glutamate;pyruvate transaminase (GPT) were both measured by the production of glutamate using methods modified from Ireland and Joy (1990). For GOT, the 1-ml assay mixture contained 100 mMTris-HCI (pH 8.0), 15 mM aspartate, 20 mM 2-oxoglutarate, 40 fiM pyridoxal phosphate and desalted extract. For GPT, the 1-ml assay mixture contained 100 tnM Tris-HCI (pH 8.0), 10 mM alanine, 6.25 mM 2-oxoglutarate and desalted extract. The complete reaction mixture minus the amino acid or oxo-acid served as controls for these transaminases. The reactions, initiated by extract addition, proceeded for 30 or 60 min and were terminated by sulfosaiicylic acid addition. The products were determined by HPLC. Alcohol dehydrogenase (ADH) was measured by the method of Kimmerer (1987), Pyruvate decarboxylase (PDC) was measured by coupling the production of acetaldehyde to NADH oxidation by ADH as described by Morrell et al, (1990). For determination of protein, 100 ^t! of desalted extract were mixed with 1 ml of 12% (w/v) cold trichloroacetic acid and held at 4°C overnight. The protein precipitate was pelleted by centrifugation for 2 min at 10000 g, then dissolved in 0,4 M NaOH. Protein was measured by the method of Bradford (1976) with bovine serum albumin as the standard.

221

B

1 I 1

A^ OGDH

'

i SSADH

Gaba-T

(1

J-

D 600

} 4 200 -

• 1 16 23 30 37 44 51 5S 65 72

23

30

37

44

51

58

65

72

23

30

37

-44 51

65

72

Days after anthesis Fig. 2. Development and in vitro enzyme activities (nmol cotyledon"' min ') of cotyledons from developing soybean seeds. Panel A shows seed fresh weight (•), cotyledon fresh weight (A) and protein content (O); panel B shows activities of GDC (•), SSADH (O) and OGDH (A); panel C shows activity of Gaba-T with either pyruvate [•] or 2-oxoglutarate (D) as amino acceptors; panel D shows activities of GDH (•), GPT (A) and GOT (O); and panel E shows activities of PDC (O) and ADH (•). Enzyme and protein data are the means ±SE (n = 3 pooled samples); fresh weights are the means ±SE (n= 10 individual seeds). Development study One group of seeds (10 pots per group) was planted every 7 days. Plants were checked twice weekly and new flowers tagged to allow precise determination of fruit age at harvest. Fruits from plants of different ages (30-72 DAA) were harvested on the same day, and a random sampling of developing seeds was conducted. Tissues to be analyzed for soluble amino acid composition using Picotag-HPLC (Tuin and Shelp 1994a) were frozen in liquid NT within 2 min after fruit excision. Enzyme assays were conducted on tissues either freshly harvested or frozen in liquid N2, then stored at -20°C until required. For each tissue, the pooled sample was derived from at least 100 seeds, although the actual number depended on tissue weight. From this pooled sample, three subsamples, each consisting of individual tissues from at least 15 seeds, were taken for determinations of enzyme activity, amino acid composition, protein and fresh weights. In addition, some fruits were dipped into liquid N2 while attached to the plants; they were then removed from the plant and shattered in a mortar and pestle, and the cotyledons and nucellus/seed coat were separated and retained for amino acid analysis, Adenylate analysis In the ftrst study, fruits were harvested at random from plants 37 DAA, and developing seeds were removed and pooled. Four subsamples of three seeds each were placed in 10-ml syringe barrels covered with aluminum foil and connected in series by tubing and 18-gauge hypodermic needles to allow continuous gassing (900 ml min"') with 222

O2 concentrations ranging from zero to 21 % (balance N;). After 5 min and with the gas stream still flowing, the contents of each syringe barrel were rapidly dumped into liquid N;. AI! operations were conducted at room temperature. In another study, fruit tissues were prepared in two ways. In the first (seeds designated as detached), developing seeds were immediately frozen upon excision from an attached fruit; each seed served as a replicate (n = 4). In the second (seeds designated as attached), the attached fruit was frozen in liquid N, and shattered, then the pod and developing seeds were separated; each fruit served as a replicate (n= 10). The tissues were extracted with perchloric acid and pool sizes of the adenylates ATP, ADP and AMP were determined by assays that coupled adenylate use to NADH or NADPH oxidation/reduction as described by Oresnik and Layzell (1994),

In situ [U-''*C]-glutainate metabolism Detached cotyledons (minus the embryonic axis) weighing approximately 100 mg were randomly selected from fruits at about 30-37 DAA and maintained at 4°C for approximately 30 min. They were then held at the mouth of a Wheaton 34-ml serum vial (containing a CO2 trap) heing sparged with 0, 8 or 21% O2 (balance N,) and injected, as described by Tuin and Shelp (1994a), with 2 \i\ (29,6 KBq) of a purified [U'*C]-glutamate (Amersham, Oakville, ON, Canada) solution without added carrier (1,5 mM giutamate, 9,99 TBq mol"') or with carrier (901 mM, 11,1 GBq mol"'). Each cotyledon was lowered Physiol. Plant. 94, 1995

Tab. I. Fresh weight, protein content and in vitro enzyme activities (nmol nucellus ' min ') of nucellus from developing soybean seed. Enzyme and protein data are the means ± SE (n = 3 pooled samples). Fresh weights are the means ± SE (n = 10 individual seeds). NA, not assayed. Enzyme/Parameter

Fresh weight (mg) Protein (mg) Glutamate dehydrogenase Glutamate decarboxylase Gaba:pyruvate transaminase Gaba;2-oxoglutarate transaminase Succinic semialdehyde dehydrogenase Glutamate;oxaloacetate transaminase Glutamate;pyruvate transaminase Alcohol dehydrogenase Pyruvate decarboxylase

Days after anthesis 37

51

58

72

3.0±0.2 0.043 + 0.004 1.4 ±0.2 1.00 ±0.02 0.06±0.01 0.004±0.001 0.075 ±0.001 1.55±0.01 K).3±l.! 4.0 ±0.3 0.32 ±0.05

4,5 ±0.6 0,047 ±0.005 2.2 ±0.1 1.11 ±0.20 NA NA NA 1.55 ±0.05 11.5 ±0.1 2.3 ±0.3 0.50 ±0.05

5.6±0.3 0.045 ±0.004 3.1 ±0.2 2.03 ±0.30 0.18 ±0.07 0.018 ±0.009 NA 1.90 ±0.08 157±0.5 NA 0.66 ±0.05

2.0 ±0.4 0.044 ±0.004 1.1 ±0.1 0.13±0.01 NA NA NA 0.28 ±0.07 4.8±0.1 2.0 ±0.1 0.28 ±0.07

into its respective serum vial, which was stoppered and incubated in the dark at 30°C for 20, 40 or 60 min. Distribution of '""C among the ethanol-soluble products (amino and organic acids) of [U-'''C]-glutamate metabolism was determined by ion-exchange or reverse-phase HPLC, coupled to on-line radioisotope detection (Tuin and Shelp 1994a), Using published procedures (Tuin and Shelp 1994a), the data were corrected for fU-"C]-g!utamate that did not penetrate the cotyledon.

Results Seed development: growth and enzyme activities Soybean seed development was monitored up to 72 DAA. Maximum seed and cotyledon fresh weights were attained by 65 DAA (Fig. 2). Seed development was slower than that in a previous study from our lab (Micallef and Shelp 1989a), probably as a result of suboptimal sunlight during the fall growing season. In the present study, harvest maturity was about 79 DAA, compared to

72 DAA found previously. Cotyledon fresh weight and protein content steadily increased with seed development, Nucellus fresh weight approximately doubled during the period from 37 to 58 DAA, then declined at 72 DAA (Tab. 1), Over the same periods, seed-coat fresh weight remained static, then declined (Tah. 2). The protein content of the nucelius remained steady with development (Tab. 1), whereas that of the seed coat declined slightly (Tab. 2). The Gaba-shunt enzymes (GDC, Gaba-T, SSADH) in cotyledons generally increased with development (Fig, 2). The maximum GDC activity was 4.5 nmol cotyledon ' min"' at 65 DAA. Gaba-T activity with pyruvate as amino acceptor was 1.2-1,6 times more than activity with 2-oxoglutarate throughout development; the maximum activity of Gaba-T with pyruvate was about 0,5 nmo! cotyledon"' min"' at 58 DAA. SSADH exhibited its maximum rate of 15 nmol cotyledon"' min"' also at 58 DAA. GDH activity throughout development was at least 10 times the activities of Gaba-shunt enzymes, exhibiting a maximum rate of 115 nmol cotyledon"' min"'. The OGDH activity (maximum rate of about 18 nmol cotyle-

Tab. 2. Fresh weight, protein content and in vitro enzyme activities (nmol [seed coat]"' min"') of seed coat from developing soybean seeds. Enzyme and protein data are the means ±SE (n = 3 pooled samples). Fresh weights are the means ±SE (n= 10 individual seeds). NA, not assayed; ND, not detected. Enzyme/Parameter

Fresh weight (nng) Protein (mg) Glutamate dehydrogenase Glutamate decarboxylase Gaba;pyruvate transaminase Gaba:2-oxoglutarate transaminase Succinic semialdehyde dehydrogenase 2-Oxoglutarate dehydrogenase complex Glutamate:oxaloacetate transaminase Glutamate;pyruvate transaminase Alcohol dehydrogenase Pyruvate decarboxylase

Physiol. Planl. 94. 1995

Days after anthesis 37

51

58

72

42±6 0.56±0.05 24±1 1.^7±2 1.9 ±0.5 0.90±0.16 0.20 ±0,04 1.90 ±0.05 24±1 105±4 14±l 12±2

45±9 0.49 ±0.05 34±6 54±3 NA NA 0.65±0.13 2.23 ±0.05 23±4 97±11 17±1 6±1

48±1 0.38±0.04 23 ±7 39 ±2 2.0 ±0.1 0.77 ±0.07 1.13±0,22 ND 21 ±3 93 ±6 9±0 5±i

33 ± 1 0.41 ±0.04 26 ±4 25±2 NA NA 0.90±0.18 0.62 ±0,05 12±1 35 ±0 7±1 l±0

223

100 -

£ 80 -

1 •» I

-

-i

20 -

1

1

\\ /i4 \A// /

% \ \

\ \ ^

steady during development with activity being about 2 0 100% of GDC activity and about 12-24 times the activity of OGDH, The activities of GOT and GPT declined slightly with development and were of the same order of magnitude as GDC and GDH activities, ADH and PDC activities also declined slightly but ranged from only 4-28% of GDC activity. The activities of Gaba-shunt enzymes from cotyledons showed very different pH profiles; the pH optima for GDC, Gaba-T and SSADH were about 5,8, 8.2 and 9,4, respectively (Fig. 3). Gaba;2-oxoglutarate transaminase was more sensitive to changing pH than Gaba;pyruvate transaminase. Amino acid composition of tissues of developing seed

Fig. 3. Activities of GDC (D), Gaba-T with either pyruvate (O) or 2-oxoglutarate (•) as amino acceptor, and SSADH (A) as a function of pH. Assays were conducted on extracts from 37DAA cotyledons. Overlapping buffers (100 vnM) were used to cover the entire pH range; Pyridine-HCI, 4.5-5.3; 2-(N-morpholino)ethanesulfonic acid, 5,3-6.0; Bis-Tris, 6.0-7.0; Tris, 7.08,5; 3-(|l,l-dimethyl-2-hydroxy-ethyl]amino)-2-hydroxy-propanesulfonic acid, 8.5—10.5. Data are the means of three measurements from pooled samples. Maximum activities (nmol min ' [g fresh weight]"') were; GAD, 29; Gaba-T, 2.1 with pyruvate, 1.8 with 2-oxoglutarate; SSADH, 28. d o n ' min"') remained relatively steady throughout development, GOT and GPT activities generally increased over the developmental period. Maximum rates were about 240 and 130 nmol cotyledon"' min"', respectively, ADH activity increased dramatically with development; maximum activity was about 700 nmol cotyledon"' min"' at 72 DAA, PDC activity remained relatively constant during early development, but increased about four times during the latter half of the developmental period, to an activity of about 160 nmol cotyledon"' min"'. Activities of Gaba-shunt enzymes in the nucellus increased up to 58 DAA, then declined (Tab. 1), The maximum rate of GDC activity was about 11 times for the Gaba;pyruvate-T activity. Gaba-T activity with pyruvate as amino acceptor was 9-14 times more than activity with 2-oxoglutarate during development. SSADH activity at 37 DAA was similar to Gaba;pyruvate-T activity. With development the activities of GDH, GPT and GOT showed trends similar to, and were similar or greater than, activities of the Gaba-shunt enzymes, ADH and PDC activities were slightly higher and lower than GDC activity, respectively. In seed coat, Gaha-shunt enzymes exhibited distinctly different developmental profiles (Tab, 2), GDC showed its highest enzyme activity at 37 DAA (137 nmol [seed coat]"' min"') and suhsequently declined by about 80%, In contrast, Gaba-T activity was much lower, but constant over time; activity with pyruvate as amino acceptor was 1,1-1,6 times higher than with 2-oxoglutarate, SSADH activity was initially lower than Gaba-T activity but by 58 DAA was similar to it, GDH activity was relatively 224

The composition of the major free amino acids in three tissues of developing soybean seeds at 37 DAA is shown in Tab. 3. In cotyledons, the glutamate pool was about 11 times the Gaba pool, whereas in the nucellus and seed coat, respectively, the glutamate pool was 65 and 11% of the Gaba pool. These data were obtained from fruits which were detached and dissected, and the tissues visK frozen within 2 min. The glutamate;Gaba ratio of cotyledons of attached fruit was similar to that found with cotyledons from these detached fruit (C.S. Walton and B.J. Shelp, unpublished data). However, it was not possible to separate the nucellus and seed coat from fruit that was frozen while attached to the plant, and in the combined tissue the Gaba pool was about two times higher than the glutamate pool. The relative amino acid composition of cotyledons or nucellus/seed coat was sitnilar at 37 and 58 DAA (data not shown).

Adenylate pools in detached or attached seeds: response to oxygen The adenylate ratios, but not the total adenylate pool, of detached developing seeds (37 DAA) were significantly affected by the oxygen concentration in the surrounding environment (data not shown). The calculated AECs were 0.84 ±0.01,0.72 ±0.03,0,53 ±0,02,0,54 ±0,01 and 0,44±0.03 for 21, 10, 5, 2.5 and 0% O., respectively. Tab. 3. Major free amino acids in three tissues of the developing soybean seed (nmol tissue"') at 37 DAA. Data are the ±SE (ri = 3 pooled samples). Tissue

Amino acid

Asparagine/glycine Histidine Glutamate Aspartate Serine Gaba Glutamine Total

Cotyledon

Nucellus

Seed coat

1512±!80 I 090 ±107 800 ±28 346 ±78 I59±5O 74 ±14 33 ±4 6628±1615

37±5 I0±2 15±3 13±! 18±5 23±5 11±1 240 ±40

431 ±135 402±110 71±23 137±3O 229 ±62 64O±I8O 232 ±58 3 333 ±970

Ptiysiol. Plant. 94. 1995

Tab. 4. Pool sizes (nmol |g fresh weight]"') of ATP, ADP and AMP in either seed and pod of fruit (37 DAA) that was frozen in liquid Nl while attached to the plant (attached), or in seed rapidly detached from an attached fruit and frozen (detached). AEC calculated as (iATPl + 0.5 |ADP])/(|ATP] + (ADP] + |AMP]); "attached seed" included embryonic axis, nucellus and seed coat. The data represent the means ±sr-: (n = 4 seed.s and 10 fruits for detached and attached, respectively). Tissue

Seed Pod Seed

Attached Attached Detached

AEC

Adenylate pool

Method ATP

ADP

AMP

85.6± 7.8 36.7± 27 99.5 ±15.0

94.3 ±8.5 ll.4±34 56.1 ±8.1

15.1 ±1.4 5.2 ±1.4 14,9 ±2.4

Developing seeds briefly exposed, or not exposed, to air had AEC values of 0.75 ±0.03 and 0.67±0.01, respectively (Tab. 4); the AEC of the pod not exposed to air was 0.82 ±0.04, If oxygeti availability is the pritnary factor responsible for regulating AEC, these data suggest that the oxygen concentration within the attached fruit would be between 5 and 109^^.

In situ [U-'''C]-glutanaatc metabolism by the cotyledon: response to oxygen The response of |U-'''C]-glutatnate metaboli.sm to various O2 treatments was investigated over a 60-min time course (Tab. 5). When additional carrier glutamate was not included with the injection of radioisotope, the recoveo' of '••C as glutamate generally decreased with time and was negatively related to the O; concentration. At 20 min, for example, approximately 6 and 36% less glutamate was metabolized at 8 and 0"% O;, respectively, than at 21%, The pool size of | '•'Cl-Gaba, which had already reached its maximum by this time, was not significantly affected by Ol concentration: however, the rate of its subsequent decline was apparently positively related to O; concentration. Similar results were found when 900 nmol of glutamate were added with |U-'''C]-g!utamate, but the pool size of ['''C]-Gaba was increased markedly.

Total 188 ±12 53.3±5,9 170±22

0.67 ±0.01 0.82 ±0.04 0.75±0.03

Discussion Characterization of Gaba-shunt enzymes Enzymes capable of catalyzing the reactions involved in conversion of glutamate to succinate via Gaba and SSA were found throughout development in the cotyledon, nucellus and seed coat of developing soybean seeds. All three enzymes of the Gaba shunt are also found in soybean-cotyledon callus (Tokunaga et al. 1976), radish leaves (Streeter and Thompson 1972b) and potato tubers (Satya Narayan and Nair 1986a), The pH and kinetic properties of GDC, Gaba-T (with the exceptions noted below) and SSADH were similar to those described previously (Streeter and Thompson 1972b, Galleschi et al, 1985, Satya Narayan and Nair 1985, 1989, Yamaura et al. 1988). Throughout development, Gaba-T activity with saturating substrates was about 0.5-, 11- and 1-fold higher with pyruvate than with 2-oxoglutarate in cotyledons, nucellus and seed coat, respectively. Other studies have reported that Gaba-T isolated from peanut-seedling cotyledons (Dixon and Fowden 1961), radish leaves (Streeter and Thompson 1972a), soybean leaves (Wallace et al. 1984), wheat embryo (Galleschi et al. 1985) and potato tuber (Satya Narayan and Nair 1986a) is 2-19 times more active with pyruvate as the amino acceptor. These results suggest that Gaba-T from plant sources has

Tah. 5. Recovery of "C as glutamate and Gaba from the dark metabolism of [U-'-'Cl-glutamate (supplied without or with additional carrier) hy detac'hed cotyledons (30-37 DAA) as a function oftime and O; concentration. Data represent the means ± SE (n = 3 or 4); note thalthe pool sizes are three orders of magnitude different between the two glutamate treatments. O: (%)

0 8 21

Phvsiol. Plant. •M.1995

Glutamate carrier (nmol)

Time (min)

20 40 60 20 40 60 20 40 60

1.5

900

f'*Cl-Glutamate I ''C]-Gaba (pmol cotyledon

['*C]-Gaba [»C]-Glutamate (ntnol cotyledon -')

899 ±6 744 ±151 536 ±138 621 ±78 407 ±62 464 ±,17 562 ±102 474 ±86 321 ±65

193 ±2 202 ±15 145 ±62 226 ±38 137 ± 37 1H±21 196±28 105 ± 1 78±9

580 ±1,32 286 ±85 198 ±35

146 ±25 329 ±174 134 ±58

225

a higher specificity for pyruvate than 2-cxoglutarate, but further information on the relative affinity of Gaba-T for pyruvate and 2-oxoglutarate is needed. It is noteworthy that in organisms other thati higher plants, such as mouse (Schousboe et al, 1973), Rhizobium trifolii (Freney and Gibson 1975) and Saccharomyces (Ramos et al. 1985), Gaba-T is specific for 2-oxoglutarate only. The pH optimum of Gaba-T with both amino acceptors was about 8.2, although Gaba:2-oxoglutarate-T activity was more sensitive to pH than Gaba:pyruvate-T. Other researchers reported pH optima of 8.9 for GabaipyruvateT from radish leaves (Streeter and Thompson 1972a), 8.3 for Gaba:2-oxoglutarate-T from isolated clover-nodule bacteroids (Freney and Gibson 1975) and 8,5 for an unknown Gaba-T from wheat embryo (Galleschi et al. 1985), No other study has compared pH profiles for Gaba-T with pyruvate and 2-oxoglutarate, The differential sensitivity to pH and the variation in specificity between tissues towards pyruvate and 2-oxoglutarate provide some support for the hypothesis that Gaba-T, like other plant transaminases (Ireland and Joy 1990, Good and Muench 1992), exists as multiple isoforms in soybean seeds. With pyruvate as amino acceptor the carbon skeleton of glutamate, after conversion to Gaba and SSA, enters the Krebs cycle as succinate with the amino group being used to form alanine. In contrast, the use of 2-oxoglutarate as amino acceptor creates a shunt in which the amino group of glutamate is constantly recycled to glutamate while the carbon skeleton enters the Krebs cycle.

consuming Gaba was correlated with in vivo Gaba accumulation. Other studies reported ratios of 23/1/9 for potato tuber (Satya Narayan and Nair 1986b) and 20/1/2 for radish leaves (Streeter and Thompson 1972b). With the exception of only OGDH in seed coat, the activities of all enzymes involved in metabolism of 2oxoglutarate (GDH, OGDH, GOT and GPT) were similar to or greater than the activity of GDC in tissues of developing seeds. Therefore, the data suggest that factors other than low enzyme levels in alternative pathways are responsible for controlling the flux of glutamate-C to the Krebs cycle via the Gaba shunt. The ontogeny of glutamine synthetase (GS) and NADH-specific glutamine:2-oxoglutarate transaminase (GOGAT), the enzymes that assimilate ammonia or convert glutamine into glutamate, also differed between tissues of developing soybean seeds (C.S. Walton and B.J. Shelp, data not shown), GOGAT activity in cotyledons declined after 37 DAA, whereas it increased in both maternal tissues. GS activity in cotyledons initially was low, then increased 10 times during later development, whereas in both maternal tissues it was higher during early development. Glutamate and glutamine were, respectively, major and minor components of the amino acid pool of cotyledons at 37 DAA. This relationship was reversed in the seed coat, suggesting that glutamine was transformed into glutamate between the two tissues. Therefore, the present study supports the suggestion by Hsu et al. (1984) that glutamine supplied via the phloem from the mother plant is quantitatively converted to glutamate by GOGAT in the nucellus, which lies between the seed coat and the embryo.

Ontogeny of enzymes associated with glutamate catabolism and ammonia assimilation The accumulation of Gaba in response to various stress conditions is well documented (Bown and Shelp 1989, Satya Narayan and Nair 1990). Indeed, a 30-fold increase in Gaba content of soybean leaves has been observed after only 5 min of cold shock (Wallace et al, 1984), The precautions taken in this study minimized, as much as possible, changes in amino acid composition or enzyme activity during sample collection. The ontogeny of Gaba-shunt enzymes differed between tissues of the developing soybean seeds. In cotyledons, the ratio of GDC/Gaba:pyruvate-T/SSADH activities, measured in vitro under optimal pH, was 13/1 /2 i, a value which did not change substantially during development, Gaba was a minor free amino acid in cotyledons at 37 DAA, In contrast to cotyledons, the maternal tissues (nucellus and seed coat) had a predominance of GDC activity. The activity ratios were 15/1/1 and 697/9/1 for nucellus and seed coat at 37 DAA, respectively. The relative activity of seed-coat GDC declined somewhat during development, giving an activity ratio of 34/2/1 at 58 DAA, At 37 DAA, Gaba was the second most abundant amino acid in the nucellus and the major free amino acid in seed coat. Thus, in tissues of developing soybean seed the in vitro activity ratio of enzymes producing and 226

Ontogeny of enzymes associated with hypoxia and adenylate energy charge Enzymes associated with hypoxia were present in all tissues investigated in the developing soybean seed. !n cotyledons, the activity of ADH increased throughout development, whereas PDC was static during early development, then increased. In contrast, the activities of ADH and PDC in maternal tissues were initially high, then declined to low values with respect to other enzymes. ADH and lactate dehydrogenase (LDH, another enzyme associated with hypoxia) are expressed in both embryos and seed coats of developing Phaseolus seeds (Boyle and Yeung 1983, Yeung and Blackman 1987). In that species, ADH and LDH activities in the embryo increase as the seed develops, whereas those in the seed coat decline as the embryo enters the maturation phase. Thus it appears that cotyledons of developing soybean seeds were hypoxic throughout development, whereas the maternal tissues were transiently hypoxic during early development. In the present study, we also determined the AEC as an indicator of cellular metabolic status. The AEC value of developing soybean seeds (37 DAA) in fruits frozen in situ was O.67±0.01, a value lower than those found in Physiol. Planl.

adjacent pod tissue (0,82±0.04) and in soybean seed frozen immediately after excision from attached fruit (0.75 ±0.03) and much lower than those typical of aerobic tissues (O,84±O.O1; >O,8O, Pradet and Raymond 1983), Furthermore, the AEC value of excised seeds was positively related to the concentration of Oi in the surrounding environment. Values in the literature for AEC of soybean seeds include 0.55 for pod, 0.24 for seed coat and 0.44 for cotyledons during mid pod-fill (Fader and Koller 1984) and 0.78 for seeds at mid pod-fill to less than 0,10 at maturity (Quebedeaux 1981), In these studies a leisurely approach to seed harvesting, which is likely to disrupt adenylate pools, was used (Oresnik and Layzell 1994). in contrast, the rapid-freezing methods used here minimized disturbance-induced artifacts, and therefore the adenylate values are believed to be reliable measures. These results, together with the presence of ADH and PDC, are consistent with the hypothesis that the metabolism of intact developing soybean seeds operates under O; limitation. Decreasing Oi availability rapidly (within 20 min) reduced the in situ rate of [U-'''C]-glutamate catabolism by cotyledons without affecting the ["C]-Gaba pool suggesting that the relative partitioning of glutamate through GDC was enhanced. Decreasing O2 also resulted in a slower rate of disappearance of this ['''C]-Gaba pool over the next 40 min and could have contributed to Gaba accumulation over a longer time. There was no evidence that the activities of Gaba-shunt enzymes changed during a 3-h incubation at 2 1 % O. (C.S. Walton and B.J. Shelp, data not shown), although changes might have occurred during the period when the cotyledons were stored at 4''C on ice. Nevertheless, the data indicate that GDC activity and Gaba accumulation were affected by the hypoxic environment found in intact seeds. Acknowledgments - The authors thank A.M. Deschene for technical assistance. This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada. References Bown, A. & Shelp, B.J. 1989. The metabolism and physiological roles of 4-aminobutyric acid. - Biochem. Life Sci. Adv. 8: 21-25. Boyle, S. A. & Yeung, E C . 1983. Embryogeny of Phaseolus: Developmental pattern of iactate and alcohol dehydrogenases. - Phytochemistry 22: 2413-2415. Bradford, M,M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. - Anal. Biochem. 72: 248254. Chung, I., Bown, A. W. & Shelp, B.J. 1992. The production and efflux of 4-aminobutyrate in isolated mesophyll ceils. Plant Physiol. 99: 659-664. Crawford, L.A., Bown, A.W., Breitkreuz, K B . & Guinel, F. 1994. The synthesis of y-aminobutyric acid In response to treatments reducing cytosolic pH. - Plant Physiol. 104: 86,5-871. Dixon, R . O . D . & Fowden, I, 1961. y-Aminobutyric acid metabolism in plants: Part 2. Metabolism in higher plants. Ann. Bot. 25: 513-5.30. Fader, G . M . & KoHer. R.H. 1984. Relationships between respil(! Physioi. Plam. 94. I

ration rate and adenylate and carbohydrate pools of the soybean fruit. - Plant Physiol. 75: 694-699. Freney, J. R. & Gibson. A. H. 1975. Accumulation of 4-aminobutyrate in Trifolium subterraneum root nodules: Effect of Rhizobium trifolii strain on aminotransferase activity. Aust. J. Plant Phj-sio). 2: 66,^-668. Galleschi, L., Lazzeri, G.. Spano, C , Capocchi, A. & Roris, C. 1985. Detection of y-aminobutyric acid transaminase activity in wheat embrvos. - Biochem. Physiol. Pflanz. 180: 629-633. Good. A.G. & Muench, D.G. 1992. Purification and characterization of an anaerobicallv induced alanine aminotransferase from barley roots. - Plant Physiol. 99: I52(H525. Hoagland, D.R. & Amon, D.I. 1950. The water culture method of growing plants without soil. - Calif. Agric. Exp. Stn. Circ. 347. Hsu, F.C., Bennet, A.B. & Spanswick, R.G. 1984. Concentrations of sucrose and nitrogenous compounds in the apoplast of developing soybean seed coats and embryos. - Plant Physiol. 75: 181-186. Ireland, R,J. & Joy, K.W. 1990. Aminotransferases. - tn Methods in Plant Biochemistry' (P.M. Dey and J . B . Harbome, eds). Vol. 3, Enzymes of Primary Metabolism, pp. in-lib. Academic Press. Toronto. JSBN 0-12-461013-7. Karam. G.A. & Bishop. S.H. 1989. a-Ketoglutarate dehydrogenase from cauliflower mitochondria: Preparation and reactivity with substrates. - Phytochemistry 28: 3291-3292. Kimmerer, T. W. 1987. Alcohol dehydrogenase and pyruvate decarboxylase activity in leaves and roots of Eastern cottonwood {Popuhis deltoides Bartr.) and soybean f.GI\cine max L.). - Plant Physiol. 84: 1210-1213. Kurkdjian, A. &. Guem, J. 1989. Intracellular pH: measurement and importance in cell activity. - Annu. Rev. Plant Physiol. 40: 271-303. McKenzie, E.A.. Copeland, L. & Lees, E.M. 1981. Glutamate dehydrogenase activity in developing soybean seed: Kinetic properties of three fomns of the enzyme. - Arch. Biochem. Biophys. 212: 298-305. Micallef. B.J. & Shelp, B.J. 1989a. Arginine metabolism in developing soybean cotyledons. I. Relationship to nitrogen nutrition. - Plant Physiol. 90: 624-