Properties of a multimeric protein complex from chloroplasts possessing potential activities of...

Eur. J. Biochem. 162,423-431 (1987) 0 FEBS 1987

Properties of a multimeric protein complex from chloroplasts possessing potential activities of NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase and phosphoribulokinase Sylvia NICHOLSON, John S. EASTERBY and Roy POWLS Department of Biochemistry, University of Liverpool

(Received August 4/0ctober 8, 1986) - EJB 86 0841

A homogeneous multimeric protein isolated from the green alga, Scenedesmus ohliquus, has both latent phosphoribulokinase activity and glyceraldehyde-3-phosphate dehydrogenase activity. The glyceraldehyde-3- phosphate dehydrogenase was active with both NADPH and NADH, but predominantly with NADH. Incubation with 20 mM dithiothreitol and 1 mM NADPH promoted the coactivation of phosphoribulokinase and NADPH- dependent glyceraldehyde-3-phosphate dehydrogenase, accompanied by a decrease in the glyceraldehyde-3-phosphate dehydrogenase activity linked to NADH. The multimeric enzyme had a Mr of 560000 and was of apparent subunit composition 8G6R. R represents a subunit of M , 42000 conferring phosphoribulokinase activity and G a subunit of 39000 responsible for the glyceraldehyde-3-phosphate dehydrogenase activity. On SDS-PAGE the Mr-42000 subunit comigrates with the subunit of the active form of phosphoribulokinase whereas that of M,- 39 000 corresponds to that of NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase.

The multimeric enzyme had a szo,w of 14.2 S. Following activation with dithiothreitol and NADPH, sedimenting boundaries of 7.4 S and 4.4 S were formed due to the depolymerization of the multimeric protein to NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase (4 G) and active phosphoribulokinase (2 R). I t has been possible to isolate these two enzymes from the activated preparation by DEAE-cellulose chromatography.

Prolonged activation of the multimeric protein by dithiothreitol in the absence of nucleotide produced a single sedimenting boundary of 4.6 S, representing a mixture of the active form of phosphoribulokinase and an inactive dimeric form of glyceraldehyde-3-phosphate dehydrogenase.

Algal thioredoxin, in the presence of 1 mM dithiothreitol and 1 mM NADPH, stimulated the depolymerization of the multimeric protein with resulting coactivation of phosphoribulokinase and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase. Light-induced depolymerization of the multimeric protein, mediated by reduced thioredoxin, is postulated as the mechanism of light activation in vivo. Consistent with such a postulate is the presence of high concentrations of the active forms of phosphoribulokinase and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase in extracts from photoheterotrophically grown algae. By contrast, in extracts from the dark-grown algae the multimeric enzyme predominates.

There is now considerable evidence that the activities of several enzymes of the Calvin cycle are regulated by light [l]. The first enzyme demonstrated to be activated in vivo by light was NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase [2]. The effect of light was shown to be reversed by a subsequent period of darkness. The stimulatory effect of light was also observed in isolated spinach chloroplasts [3 - 41. Incubation of the isolated chloroplast enzyme with a number of effector molecules (e.g. NADPH and ATP) stimulated the NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activity [5 - 71. The effect of the metabolites was potentiated by dithiothreitol[4,5,7], suggesting that effector-mediated activation operates in conjunction with a reductive control mechanism involving reduced thioredoxin

Correspondence to R. Powls, Department of Biochemistry, Uni- versity of Liverpool, P.O. Box 147, Liverpool, England L693BX

Enzymes. Glyceraldehyde-3-phosphate dehydrogenase (NADP) (EC 1.2.1.13); glyceraldehyde-3-phosphate dehydrogenase (NAD) (EC 1.2.1.12); phosphoribulokinase (EC 2.7.1 .I 9).

[8]. The stimulation of the NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activity of the spinach chloroplast enzyme was correlated with enzyme depolymerization [6,7].

The glyceraldehyde-3-phosphate dehydrogenase isolated from the green alga, Scenedesmus obliquus, has similar properties to the spinach enzyme. It is of high molecular mass, has activity predominantly linked to NADH and can be activated by incubation with a number of chloroplast metabolites in the presence of dithiothreitol [9 - 111. Induction of NADPH- dependent glyceraldehyde-3-phosphate dehydrogenase activity is associated with depolymerization to a tetrameric form of the enzyme. Phosphoribulokinase also exhibits reversible activation by light [12- 141. Two forms of the enzyme have been isolated from spinach which differ in their molecular mass and also in their requirement of thioredoxin for activity [15]. Further similarity between the behaviour of phosphoribulokinase and NADPH-dependent glyceraldehyde-3- phosphate dehydrogenase from spinach is also apparent in the effect of organic solvents on their activation [16].

424

We have recently isolated two forms of phosphoribulokinase from S. obliquus, a low-molecular-mass active form and a high-molecular-mass regulatory form of the enzyme characterised by latent phosphoribulokinase activity [I 71. Activation of the regulatory form of the enzyme by incubation with dithiothreitol was accompanied by its depolymerization to the low-molecular-mass form of the enzyme. This was a completely analogous situation to the algal glyceraldehyde-3- phosphate dehydrogenase. Like the regulatable form of glyceraldehyde-3-phosphate dehydrogenase, the latent phosphoribulokinase was unstable when chromatographed on DEAE-cellulose. In both cases this instability could be prevented by the inclusion of NAD in the chromatographic buffers [17,18].

It would now appear that the numerous similarities between the regulatory forms of phosphoribulokinase and glyceraldehyde-3-phosphate dehydrogenase are not purely fortuitous. The regulatory forms of both algal enzymes are in fact present in a single high-molecular-mass multimeric protein which, on incubation with effector molecules, depoly- merizes to produce the active forms of phosphoribulokinase and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase. We now wish to report the properties of this multimeric protein and discuss its relevance to the light activation of the two enzymes in algae and higher plants. A prelimi- nary report of some related findings has appeared previously ~ 9 1 .

MATERIALS AND METHODS

Growth of the alga, Scenedesmus obfiquus (Cambridge 276/6a), assay of the latent (dithiothreitol-dependent) and active forms of phosphoribulokinase, preparation of algal thioredoxin and molecular mass determination by sedimentation equilibrium analysis were as described previously [17]. All s values quoted are sz0 ,w values and are not extrapolated to zero protein concentration. The isolation of NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase and the assay of glyceraldehyde-3-phosphate dehydrogenase activity have also been described [9].

SDS-PAGE was as described by Laemmli [20] using an LKB 2001 vertical gel apparatus. Protein was stained with Coomassie blue.

Purification of the multimeric enzyme

The first three buffers used were composed of potassium phosphate pH 7.5 of various ionic strengths: buffer 1, ionic strength I = 1.0 M; buffer 2, I = 25 mM; buffer 3, I = 50 mM. Buffer 4 was 50 mM Tris/HCl pH 7.5; buffer 5 , 50 mM potassium phosphate pH 7.2; buffer 6, 200 mM potassium phosphate pH 7.2. All buffers contained mercaptoethanol (0.5 ml/l) and all, with the exception of buffer 1, contained 10% (v/v) glycerol and 75 ml/l NAD.

Three batches of heterotrophically grown algae re- suspended in buffer 1, each containing about 100 g wet weight of algae in a total volume of 150 ml, were used for each preparation.

Mercaptoethanol(50 pl) was added to each batch prior to cell breakage. Each batch of cells was broken by grinding with glass beads (150 ml, 0.25-0.30 mm diameter) for 10 min in a Dynomill (type KDL, W. A. Bachofen, Basle, Switzerland). Efficient cooling was achieved by continuously circulating

the breaking chamber with ice/salt mixture at - 5°C. After breakage, the three batches were bulked and glass beads separated from the algal extract by filtering through a sintered-glass funnel. The dark green filtrate was centrifuged at 48 000 x g for 15 min to remove the insoluble debris of unbroken cells and large chloroplast fragments. Th,e green supernatant was then centrifuged at 194000 x g for 4 h to remove the remaining small chloroplast fragments. After filtering through glass-wool the supernatant was dialysed overnight against 5 1 buffer 2. The dialysed extract was divided into three equal portions and each was applied to a column (30 x 2.5 cm) of DEAE-cellulose previously equilibrated with buffer 3. Proteins were eluted with a linear salt gradient of phosphate formed by mixing 220 ml buffer 1 incorporating 10% glycerol with 220 in1 of buffer 3. A flow rate of 30 ml/h was used and fractions of 6.5 ml were collected, commencing when the gradient was applied to the column. The fractions were assayed for phosphoribulokinase (after incubation with dithiothreitol) and NADH-dependent glyceraldehyde-3- phosphate dehydrogenase activities, protein (Azso) and conductivity. Those with the highest activities of both phosphoribulokinase and NADH-dependent glyceraldehyde-3- phosphate dehydrogenase were bulked and concentrated by ultrafiltration to 2 ml, prior to application to a column (2.6 x 83 cm) of Ultrogel AcA34. The column was eluted with buffer 4 at a flow rate of 27 ml/h. Eluted fractions (5 ml) with the greatest activities of both phosphoribulokinase and NADH-dependent glyceraldehyde-3-phosphate dehydrogenase activities were bulked and applied directly to a hydroxyapatite column (1.6 x 28 cm) pre-equilibrated with buffer 5. After washing with this buffer, the enzyme was eluted with a linear salt gradient made by mixing 500 ml buffer 6 with 500 ml buffer 5. The eluted fractions (5 ml) were assayed and those with the highest activities of both phosphoribulokinase and NADH-dependent glyceraldehyde-3-phosphate dehydrogenase were bulked as the multimeric enzyme.

Enzyme activation

For studying the time-course of activation, the multimeric enzyme (6 units of latent phosphoribulokinase activity) was incubated with dithiothreitol (8 pmol), NADPH (0.4 pmol), Tris/HCl pH 7.5 (40 pmol) in a total volume of 400 p1 at 30°C. At specified times aliquots (10 pl) were removed from the incubation and their phosphoribulokinase, NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase and NADH- dependent glyceraldehyde-3-phosphate dehydrogenase activities were determined. A similar incubation was used to study the effect of thioredoxin on the time-course of activation, but in this case algal thioredoxin (100 pl) was present in the incubation and the concentration of dithiothreitol was varied. The final volume was always 400 pl.

Thioredoxin eluted from the Ultrogel AcA54 column was detected by its ability to costimulate phosphoribulokinase and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activities at low concentrations of dithiothreitol. In this case the incubation mixture was composed of dithiothreitol (0.2 pmol), NADPH (0.2 pmol), Tris/HCl pH 7.5 (20 pmol), the fraction from AcA54 (100 pl) and the multimeric enzyme (1.3 latent units of phosphoribulokinase activity) in a total volume of 200 pl. After 30 min incubation, aliquots (10 pl) were taken for the measurements of phosphoribulokinase activity. Further aliquots were removed after 32 min incubation and their NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activity was determined.

425

t

1 1 1 1 1 1 1 I I I 1

h A ' 140 1

1 2 0 6

100 5

80 4

0

60 3 : a 40 2

20 1

0 0 4 8 12 16 20 24 28 32 36 40 44 48 52

14

12

10 - II) I 8 5

0.6 2 4

20 0.5

16 0.4

12 0.3 0 oa

8 0.2;

4 0.1

0 0 8 16 24 32 40 48 56 64 12 80 84

t

a 1 0

0

FRACTION

Fig. 1 . Copurificution of the high-M, form of latent phosphoribulokinase und NADH-dependent glyreraldehyde-3-phosphuie dehydrogenuse. Elution profiles from the chromatography of extracts of heterotrophically-grown algae. (A) DEAE-cellulose chromatography; (B) gcl-filtration on Ultrogel AcA34; (C) chromatography on hydroxyapatite. ( 0 ) Protein ( A z s o ) ; (0) dithiothreitol-induced phosphoribulokinase activity; (m) NADH-dependent glyceraldehyde-3-phosphate dehydrogenase activity; ( A ) conductivity

For an ultracentrifuge investigation, the multimeric enzyme ( 2 mg/ml) in 120 mM Tris/HCl pH 7.5 was incubated with 18 mM dithiothreitol and 1 mM NADPH for 4 h. The sedimentation behaviour of the preparation was then examined in a Beckman model E ultracentrifuge. The activated enzymes present in the preparation were isolated by DEAE- cellulose chromatography using buffers containing NAD (75 mg/l). When dithiothreitol alone was used for activation, the multimeric enzyme (3.8 mg/ml) in 120 mM Tris/HCl pH 7.5 was incubated with 18 mM dithiothreitol. After prolonged incubation at 4°C the sedimentation behaviour of the preparation was examined.

RESULTS AND DISCUSSION

Isolation and characterisation

The latent form of phosphoribulokinase from the green alga, S. obliquus, has been purified to homogeneity [17]. As well as having latent phosphoribulokinase activity the preparation also had glyceraldehyde-3-phosphate dehydrogenase

activity. This was linked to both NADPH and NADH, identical to that of the high-molecular-mass regulatory form of glyceraldehyde-3-phosphate dehydrogenase (D enzyme) previously isolated from this alga [9]. The latent phosphoribulokinase and glyceraldehyde-3-phosphate dehydrogenase activities would appear to be the property of a single protein, consistent with the finding that both enzyme activities are lost during the chromatography of the algal extracts on DEAE- cellulose. The loss of both activities could be prevented by the incorporation of NAD in the chromatographic buffers [17,18].

During the purification of latent phosphoribulokinase from the heterotrophically grown algae (Fig. 1, Table l), the cytoplasmic NADH-specific glyceraldehyde-3-phosphate dehydrogenase was separated from the chloroplast enzyme at the DEAE-cellulose step. However, no further separation of the latent phosphoribulokinase and the remaining NADH- dependent glyceraldehyde-3-phosphate dehydrogenase activity occurred throughout the gel-filtration and hydroxyapatite steps of the purification procedure (Fig. 1). This further emphasised the distinction between the cytoplasmic glyceral-

426

-1.7

Table 1. Copurification of phosphoribulokinase and NADH-dependent glyceraldehyde-3-phosphate dehydpogenase activities of the multimeric enzyme

Stage of purification Protein Dithiothreitol-induced phosphoribulokinase NADH-dependent glyceraldehyde-3-phosphate dehydrogenase

activity specific yield purifica- activity specific yield purification activity tion activity

mg units units/mg % -fold units units/mg % -fold - - 48 000 x g supernatant 12375 17970 1.45 - - 2640 0.21

Dialysed 194000 x g supernatant 3 675 6863 1.87 38 1.3 2436 0.66 92 3.1

DEAE-cellulose fractions 262 3418 13.04 19 9.0 302 1.15 11 5.5 AcA34 fractions 17.6 1370 77.6 7.6 53.5 130 7.4 5 35.2 Hydroxyapatite fractions 5.0 789 157 4.4 108 60 11.9 2 56.7

I I I I 1 - ' . 6 t i

dehyde-3-phosphate dehydrogenase and the chloroplast phosphoribulokinase/glyceraldehyde-3-phosphate dehydrogenase complex. There was an apparent substantial loss of activity, measured in the phosphoribulokinase assay, on removal by centrifugation of the membrane fraction. This was only partially accounted for by loss of membrane-bound ATPase activity and could imply appreciable specific binding of phosphoribulokinase to the thylakoid membranes.

As both the kinase and dehydrogenase activities were associated with a single homogeneous protein, it was necessary to resolve a discrepancy between the M , reported previously for the algal glyceraldehyde-3-phosphate dehydrogenase and the complex now isolated. The regulatory form of glyceraldehyde-3-phosphate dehydrogenase has been shown to have a M , of 550000 [9], whilst for the latent form of phosphoribulokinase a M , of 470000 has been reported [17]. For a homogeneous sample of the multimeric enzyme eluted from hydroxyapatite and possessing both latent phosphoribulokinase and NADH-dependent glyceraldehyde-3-phosphate dehydrogenase activities a z-average molecular weight of 560000 was obtained by low-speed sedimentation equilibrium measurements (Fig. 2). SDS-PAGE showed the multimeric enzyme to be composed of two different polypeptides of M , 39000 and 42000 (Fig. 3). Based upon an M , of 560000 and the subunit M , values determined by SDS-PAGE, possible subunit compositions for the multimeric protein are 8G6R, 7G7R or 6G8R. R represents the subunit of M , 42000 expected to

Fig. 3.SD.S-PAGE. From left to right: track 1, the multimeric protein with latent phosphoribulokinase and NADH-dependent glyceraldehyde-3-phosphate dehydrogenase activity; track 2 marker proteins from top: phosphorylase b, bovine serum albumin, ovalbumin, carbonic anhydrase, trypsin inhibitor (soya), lysozyme (runs with front); track 3 active phosphoribulokinase from DEAE-cellulose chromatography of activated multimeric enzyme; track 4 active phosphoribulokinase isolated from algae; track 5 markers as track 2; track 6 NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase from DEAE-cellulose chromatography of the activated multimeric enzyme; track 7 NADPH-dependent glyceraldehyde-3- phosphate dehydrogenase isolated from algae

confer phosphoribulokinase activity since it comigrates on SDS-PAGE (Fig. 3) with the subunit of the small, active form of phosphoribulokinase isolated from photoheterotrophically grown S.obliquus [17]. The G subunit would be responsible for glyceraldehyde-3-phosphate dehydrogenase activity as this comigrates on SDS-PAGE with the subunit of the NADPH- dependent glyceraldehyde-3-phosphate dehydrogenase (T en-

427

zyme), also isolated from the light-grown alga [9]. Scanning at 590 nm of the multimeric protein stained with Coomassie blue after SDS-PAGE has always given an intensity ratio of staining between 8:5 and 8: 5.5 for the G : R subunits, consistent with an 8G6R composition. Since the intensity of staining with Coomassie blue does depend upon the amino compositions of the two subunits, it may be purely fortuitous that this ratio is close to that expected for an 8G6R subunit composition. However, as shown later, the multimeric protein, on activation, dissociates to a tetrameric NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase (4G) and a dimeric phosphoribulokinase (2R). This would imply an 8C6R subunit composition for the multimer.

It is possible that the previous report of 470000 M , for the latent form of phosphoribulokinase [17] was due to the isolation of an 8G4R species formed by the loss of a phosphoribulokinase dimer. All recent determinations of M , have consistently given values in the range 550000- 560000 cor- responding to an 8G6R composition.

I I I I I I 1 I I I > c

2 0

2 "

TIME(mln1

Fig. 4. Coactiwtion of' phosphorihulokinase and NADPH-dependent glyccrul~~ii~d~-3-phosphate dehydrogenase activity of the rnultimeric enzyme. Activation was performed with- 20 mM dithiothreitol and 1 mM NADPH as described in Materials and Methods. (0 ) NADPH- dependent glyceraldehyde-3-phosphate dehydrogenase activity; ( W ) NADH-dependent glyceraldehyde-3-phosphate dehydrogenase activity; (0) phosphoribulokinase activity

Mechanism qf activation Incubation of the large multimeric protein with 20 mM

dithiothreitol and 1 mM NADPH produced marked increases in the activities of both phosphoribulokinase and NADPH- dependent glyceraldehyde-3-phosphate dehydrogenase, whereas there was an accompanying decrease in the glyceraldehyde-3-phosphate dehydrogenase activity linked to NADH (Fig. 4). Changes in molecular size occurring during activation at high dithiothreitol concentrations were investigated by sedimentation velocity analysis in the ultracentrifuge. Prior to activation the multimeric enzyme was characterised by a s ~ ~ , of 14.2 S. Following the activation with dithiothreitol and NADPH this boundary disappeared and was replaced by two more slowly sedimenting boundaries of 7.4 S and 4.4 S (Fig. 5). The 7.4-S boundary is characteristic of that of the NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase [9], whereas it is now apparent that the active form of phosphoribulokinase is responsible for the 4.4-S boundary rather than an inactive dimer of glyceraldehyde-3-phosphate dehydrogenase as previously suggested [ 1 I].

Activation by dithiothreitol in the absence of nucleotide promoted only a transient stimulation of the NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activity [l 11. Prolonged incubation with dithiothreitol produced a preparation characterised in the ultracentrifuge by a single boundary sedimenting at 4.6 S (Fig. 5). This boundary must now be considered to be due to a mixture of the active dimeric form of phosphoribulokinase as well as an inactive diiner of glyceraldehyde-3-phosphate dehydrogenase.

To confirm that the activation by dithiothreitol and NADPH was promoting depolymerization, an activated sample was chromatographed on DEAE-cellulose with all the buffers containing NAD. The multimeric protein was no longer present, it had been replaced by a form of glyceraldehyde-3-phosphate dehydrogenase with activity linked to NADPH as well as the active form of phosphoribulokinase (Fig. 6). These two enzymes, formed during the activation, behaved on ion-exchange chromatography (DEAE-cellulose) and gel filtration (Ultrogel AcA34) identically to the tetrameric NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase (7.6 S) [9] and the dimeric phosphoribulokinase (4.6 S), which have been isolated directly from photoheterotrophically grown S. ohliquus [9, 171. When examined by SDS-PAGE (Fig. 3) both enzymes, formed during activation of the multirneric enzyme, were each composed of

Fig. 5. The effect of dithiothreitol and nucleotide on the sedimentation behuviour of the multimeric protein. The purified multimeric protein was activated as described in Matcrials and Methods. Left: activation with 18 mM dithiothreitol and 1 mM NADPH; photographs were takcn 53 min after reaching a rotor speed of 60000 rev./min at a schlieren analyser angle of 60". Top: native multimeric protein; bottom: following incubation of the multimeric protein with dithiothreitol and NADPH. Right: incubation with 18 mM dithiothreitol; photographs were taken 46 min after reaching a rotor speed of 60000 rev./min at a schlieren analyser angle of 60". Top: native multimeric protein; bottom: following incubation with dithiothreitol

428

a single polypeptide (Mr 39000 for the NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase and M, 42 000 for phosphoribulokinase, Fig. 3). This also would be consistent with a subunit composition of 8G6R i.e. (4G)2(2R)3 for the multimeric enzyme.

The molecular mass changes promoted on activation of the multimeric enzyme by dithiothreitol are summarised in Fig. 7.

Light-promoted activation Dithiothreitol-induced changes in the activity of the

purified enzymes resemble those promoted in vivo by light. Light is considered to be effective in promoting the reduction of a small heat-stable protein thioredoxin, which is then able to activate the enzymes directly. A protein with the properties of thioredoxin is present in algal extracts and has been detected by its ability to stimulate the depolymerization of the multimeric enzyme in the presence of 1 mM dithiothreitol, causing coactivation of the phosphoribulokinase and

1 4 18 22 26 30 34

F R A C T I O N

Fig. 6. Isolation of active phosphoribulokinase and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase from the sample of multimeric enzyme activated, for the ultracentrifuge. The multimeric enzyme was activated by incubation with 18 mM dithiothreitol and 1 mM NADPH as described in Materials and Methods. Following examination in the ultracentrifuge the activated sample was trans- ferred to potassium phosphate pH 7.5, I = 0.05 M and chromatographed on DEAE-cellulose as described previously [17]. (0) Phos- phoribulokinase activity; ( 0 ) NADPH-dependent glyceraldehyde-3- phosphate dehydrogenase activity; ( W ) NADH-dependent glyceraldehyde-3-phosphate dehydrogenase activity; ( A ) conductivity

NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activities. At such a low concentration, dithiothreitol alone is incapable of promoting depolymerization. Fractions which possessed thioredoxin activity were detected on gel- filtration of extracts of soluble algal protein (Fig. 8).

Low-molecular-mass fractions (46 - 50), eluted after cytochromef, were combined and used in these experiments as algal thioredoxin. The higher-molecular-mass fractions (34 - 38), also promoting activation, are likely to contain the high- molecular-mass form of thioredoxinf, recently isolated from S. obliquus [21]. This high-molecular-mass form of thioredoxin was not used in any of the experiments reported here.

An obligatory requirement for thioredoxin was observed for the promotion of the activation when the multimeric enzyme was incubated with 1 mM dithiothreitol (Fig. 9). Depolymerization promoted by photoreduced thioredoxin could account for the light-dependent enzyme activation observed in vivo. In this case the nucleotide concentration in the chloroplast would be high enough to prevent the type of inactivation of NADPH-dependent glyceraldehyde-3-phos-

8G6R ( ~ ~ o K . I ~ s ) GSPDH NAOH.NADPH I S , Latent PRK

DTT 1 NADPH

2R (83K. 45) AcIIve PRK

NADPH

R - PRK subunit (42K)

G - G3PDH subunit (39K)

i 2G ( 7 8 ~ , 4 s )

lnacure

Fig. I . Molecular basis of the coactivation of the phosphoribulokinase ( P R K ) and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase (GSPDH) activities of the multimeric algal enzyme. DTT, dithiothreitol; K, kDa. S values are approximations

22 26 30 34 38 42 46 50 54 n F R A C T I O N 2

Fig. 8. Ultrogel AcA54 elution profile of thioredoxin activity f rom extracts ofphotoheterotrophically grown S. obliquus. 4 ml dialysed soluble algal protein, prepared as previously described, was subjected to gel filtration on Ultrogel AcA54. The ability of the eluted fractions to stimulate the phosphoribulokinase and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activity of the multimeric enzyme incubated with 1 mM dithiothreitol and 1 mM NADPH was measured. Cytochromefpresent in the extract was determined by its absorption at 416 nm and served as an internal molecular mass marker. Fractions 46-50 were combined as algal thioredoxin. ( 0 ) Protein (A280); (0) phosphoribulokinase activity; (m) NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activity; (A) cytochromef

429 * c 2

2 16 c I I I I I I I I I I I

Y u) < $ 14 0 0 m 0

; 12

Y t

2 10

2 :-

Y o r

+ ?

- < o w

A $ 2 a i 2 ;

!: :: ;: * 2 5 6 I ;.-

2 0

r 2

Y 0 z Y P Y

0 0 5 10 15 20 25 30 35 40 45 50 55 I

0 T I H E ( ~ I ~ < 2

Fig. 9. Stimulatory effect of algal thioredoxin on the coactivation ofthe phosphoribulokinase and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activities of the multimeric enzyme in the presence of I mM dithiothreitol and I mM NADPH. Activation of the multimeric protein was as described in Materials in Methods. Phosphoribulokinase activity, enzyme incubated with (0) 1 mM dithiothreitol and 1 mM NADPH; (m) I mM dithiothreitol and 1 mM NADPH and thioredoxin; (0) 20 mM dithiothreitol and 1 mM NADPH; NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase activity, enzyme incubated with (A) 1 mM dithiothreitol and 1 mM NADPH; ( 0 ) 1 m M dithiothreitol and 1 mM NADPH and thioredoxin; (0) 20 mM dithiothreitol and 1 mM NADPH

phate dehydrogenase observed when the multimeric enzyme undergoes prolonged incubation with dithiothreitol alone

The activation observed in the 1,3-bisphosphoglycerate- promoted depolymerization of the multimeric algal enzyme is reversible [lo]. Following activation, hydrolysis of the 1,3- bisphosphoglycerate occurred on overnight standing and was accompanied by the restoration of NADH-dependent glyceraldehyde-3-phosphate dehydrogenase activity and the sedimentation boundary of the multimeric protein. At present, however, it has not been possible to promote repolymerization after the activation of the multimeric enzyme with high concentrations of dithiothreitol even in the presence of nucleotide. I t is possible that the high dithiothreitol concentrations used could cause over-reduction. The postulated reductive activation mediated by reduced thioredoxin may be more gentle, so enabling repolymerization and hence in vivo dark inactivation of phosphori bulokinase and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase to occur. However, at present it has not proved possible to reverse the depolymerisation in this way. This may await the appropriate conditions being obtained. Depolymerisation, however, may not be an obligatory prerequisite for activation, but may simply reflect other conformational changes occurring in the enzyme complex.

When extracts from photoheterotrophically-grown algae were chromatographed on DEAE-cellulose, all the three chloroplast enzymes were present (Fig. 10). NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase was eluted in fractions 29 - 31, whereas the multimeric enzyme was eluted in 33 - 34 and the small, active form of phosphoribulokinase in 37-39. In comparison chromatography of extracts of heterotrophically grown algae (Fig. 1) indicated that virtually all the glyceraldehyde-3-phosphate dehydrogenase and phosphoribulokinase activities were associated with the multimeric enzyme. Under these conditions very little of the low-M, enzymes, NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase and active phosphoribulokinase, was present.

P21.

> t 2 + 0 4 W v) 4

Y 0 3

z

E P P v)

50

20

10

20

16

12

x 0 > I W O

w k 4

I 0 v)

g I m

0

I w 0 J

4 K

I:

w " t; 26 30 34 38 42

FRACTION 0

Fig. 20. DEAE-cellulose chromatography of an extract of photoheterotrophically grown S. obliquus. DEAE-cellulose chromatography of algal extracts was as described previously (171. (0) Dithiothreitol-induced phosphoribulokinase activity; (m) phosphoribulokinase activity; ( A ) NADPH-dependent glyceraldehyde-3- phosphate dehydrogenase activity; (0 ) NADH-dependent glyceraldehyde-3-phosphate dehydrogenase activity

As the concentration of the multimeric enzyme is highest in the heterotrophically grown cells, a possible function might be in the oxidation of the triose phosphate imported into the chloroplast. This imported triose phosphate acts both as a precursor of starch and also as a source of chloroplast ATP. The heterotrophically grown alga stores starch in the chloroplast, during the later stages of synthesis there is a requirement for ATP. Under heterotrophic conditions this can obviously not be met by photophosphorylation.

Illumination promotes the formation of chloroplast ATP and NADPH by photosynthetic electron transfer as well as

430

Fig. 1 1. Coactivation ofphosphoribulokinase ( P R K ) and NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase (G3PD H ) on illumination of dark-grown S . obliquus. RuSP, ribulose 5-phosphate; Rul,5BP, ribulose 1,5-bisphosphate; RuBPCase, ribulosebisphosphate carboxy- lase; 3PGA, 3-phosphoglycerate; PGK, phosphoglycerate kinase; 1,3BPGA, 1,3-bisphosphogIycerate

reducing thioredoxin. This increased availability of ATP results in an increase in the concentration of 1,3-bisphosphoglycerate through the activity of phosphoglycerate kinase. Reduced thioredoxin, NADPH and 1,3-bisphosphoglycerate [lo, 111 are all known to be capable of promoting the depolymerization of the multimeric enzyme to give the small active forms of NADPH-dependent glyceraldehyde-3-phosphate dehydrogenase and phosphoribulokinase required for the operation of the Calvin cycle (Fig. 11).

Relevance to the situation in higher plants

There is substantial evidence favouring a similar associ- ation of phosphoribulokinase and glyceraldehyde-3-phosphate dehydrogenase in higher plants. Many preparations of plant chloroplast glyceraldehyde-3-phosphate dehydrogenase are similar to the algal enzyme in having a high M , in the range 600000-800000 [6, 7, 23-29]. The enzyme isolated from spinach leaves by Buchanan is even eluted from both ion exchange and gel filtration along with the high-M, regulatory form of phosphoribulokinase [15]. The high-M, forms of chloroplast glyceraldehyde-3-phosphate dehydrogenase are stabilized by NAD, whereas they disaggregate in the presence of NADP to give tetrameric enzymes of approximate M , 150000 [6, 7, 25, 26, 281. In most species the disaggregated enzymes can be separated into two isoenzymes, a hetero- tetranier (2A2B) and a homotetramer (4A) [28, 30-321. The two subunits are of slightly different size in all species [26, 28 - 321, in Sinapis albu the subunits are 39000 (A) and 42000 (B) [30], identical in size to the two subunits of the multimeric algal enzyme.

These NAD(P)-induced changes in the aggregation state of the higher plant chloropkdst glyceraldehyde-3-phosphate dehydrogenase are a general phenomenon and have been used to purify the enzyme from a number of species [6, 7, 25, 26, 281. However, following the purification of the disaggregated enzyme from the seedlings of Sn. aha, it could no longer repolymerize on incubation with NAD [33]. It was suggested that polymerization only occurs in the presence of a binding fraction which had been lost during the purification. A binding fraction has been partially purified and shown by SDS- PAGE to be 70% composed of a M,-42000 polypeptide, a major protein in seedling extracts [33]. As this polypeptide comigrated during SDS-PAGE with the B subunit of the chloroplast glyceraldehyde-3-phosphate dehydrogenase,

Cerff considered that it was even possible that it represents surplus inactive subunit of the enzyme. We now suggest that the binding fraction is the active form of phosphoribulokinase and consequently we would expect the aggregated form of the enzyme to resemble the multimeric algal enzyme in having latent phosphoribulokinase activity. In Sn. alha both forms of chloroplast glyceraldehyde-3-phosphate dehydrogenase are characterised by possession of subunits of 39000 and 42000 M,. The major difference between the two forms of the enzyme is purely quantitative in that the associated enzyme has more of the 42000 M , subunit than does the disaggregated enzyme [33]. If, as we are suggesting, the B subunit was characterised by phosphoribulokinase activity it would explain the inability to isolate homotetramers of the B subunit possessing glyceraldehyde-3-phosphate dehydrogenase activity.

In the algae the multimeric enzyme probably functions in the oxidation of chloroplast triose phosphate to provide the ATP required for the synthesis of starch in the dark. This type of oxidative function for the chloroplast enzyme would clearly only be required in species capable of substantial heterotrophic nutrition. A similar function for an analogous enzyme in plant chloroplasts would be unlikely. However, in higher plants it is known that in the absence of light the enzyme is in a form with latent phosphoribulokinase activity [14]. This results in virtually zero activity of phosphoribulokinase in the presence of metabolite concentrations calculated to simulate the chloroplast in the dark [34]. Thus in plants the sole function of enzyme polymerization inay well be to render the phosphoribulokinase completely inactive in the dark.

We are indebted to the Science and Engineering Research Council for the award of a research studentship to S.N.

REFERENCES 1. 2. 3. 4.

5. 6.

7. 8.

9.

Buchanan, B. B. (1980) Annu. Rev. Plant Physiol. 31, 341 -374. Ziegler, H. & Ziegler, I. (1 965) Planta {Berl.) 65, 369 - 380. Miiller, B. & Ziegler, H. (1969) Plunta (Bed . ) 85, 96-104. Miiller. B., Ziegler, I. & Ziegler, H. (1969) Eur. J . Biochem. 9,

Miiller, B. (1970) Biochim. Biophys. Acto 205, 102- 105. Pupillo, P. & Piccari, G. G. (1973) Arch. Biochem. Biophys. 154,

Pupillo, P. & Piccari, G. G. (1975) Eur. J . Biochem. 51,475-482. Wolosiuk, R. A. & Buchanan, B. B. (1978) Plant Physiol. 61,

O’Brien, M. J . & Powls, R. (1976) Eur. J . Biochem. 63,155-161.

101 -106.

324 - 33 1 .

669 - 611.

43 1

10. O’Brien, M. J., Easterby, J . S. & Powls, R. (1976) Biochim. Bio-

11. O’Brien, M. J., Eastcrby, J . S. & Powls, R. (1977) Biochim. Bio-

12. Latzko. E., Garnier, R . V. & Gibbs, M. (1970) Biochem. Biophys.

13. Fischer, K . H. & Latzko, E. (1979) Biochem. Biophys. Res.

14. Wirtz, W., Stitt, M. & Heldt, H. W. (1982) FEBS Lett. 42, 285-

15. Wolosiuk, R. A. & Buchanan, B. B. (1978) Arch. Biochem. Bio-

16. Wolosiuk, R. A., Corley, E., Crawford, N. A. & Buchanan, B. B.

17. Lazaro, J. J., Sutton, C. W., Nicholson, S. & Powls, R. (1986)

18. O’Brien, M. J., Woodrow, S., Easterby, J. S. & Powls, R. (1979)

19. Nicholson, S., Easterby, J. S. & Powls, R. (1986) FEBS Left . 202,

20. Laemmli, U. K. (1970) Nature (Lond.) 227, 680-685.

/ > ~ v s . A r t / 149. 209-223.

~ I I Y S . Actu 481, 349-358.

Res. Commun. 39, 1140- 1144.

Cornmun. 89, 300 - 306.

288.

PhYs. 189, 97 - 101.

(1985) FEBS. Lett. 189, 212-216.

Eur. J . Biochem. 156,423-429.

Arch. Microhiol. 122, 3 13 - 31 9.

19-22.

21. Langlotz, P., Wagner, W. & Follmann, H. (1986) Z. Nuturjiorsch. 4 1 ~ . 215-283.

22. Woodrow, S., O’Brien, M. J.. Easterby, J . S. & Powls, R . (1979)

23. Yonuschot. G. R., Ortwerlh, B. J . & Koeppe, 0. J . (1970) J . Biol.

24. Wolosiuk, R. A. & Buchanan, B. B. (1976) J . B id . Chem. 251,

25. Cerff, R. (1978) Eur. J . Biochem. 82,45-53. 26. Pupillo, P. & Faggiani, R. (1979) Arch. Biochem. Biophys. 194,

27. Wara-Aswapati, O., Kemble, R . J. & Bradbeer, J . W. (1980) Plunt

28. de Looze, S. &Wagner, E. (1983) Phy.tiol. Plant. 57, 231 - 231. 29. Ferri, G., Comerio, G., ladorala, P., Zapponi, M. C. & Speranza,

30. Cerff, R. & Chambers, S. E. (1978) Hoppe-Seyler’s Z. Physiol.

31. Cerff, R. (1979) Eur. J . Biochem. Y4, 243 ~ 247. 32. Cerff, R. & Chambers, S. E. (1979) J . Biol. Chem. 254, 6094-

33. Cerff, R. (1978) Plant Physiol. 61, 369-312. 34. Gardeman, A,, Stitt, M. & Heldt, H. W. (1983) Biochim. Biophys.

Em. J . Biochem. 98,425 - 430.

Chem. 245,4193-4198.

6456 - 6461.

581 - 592.

Physiol. 66, 34- 39.

M. L. ( 1 978) Biochim. Biophys. Acta 522, 19 - 31.

Chern. 359, 769 - 772.

6098.

act^ 722, 51 -60.

Properties of a multimeric protein complex from chloroplasts possessing potential activities of...

Documents

Transcript of Properties of a multimeric protein complex from chloroplasts possessing potential activities of...