The ATP-dependent post translational modification of ferredoxin: NADP+ oxidoreductase

446

Biochimica et Biophysica A cta, 1052 (1990) 446-452

Elsevier BBAMCR 12700

The ATP-dependent post translational modification of ferredoxin" N A D P ÷ oxidoreductase M. H o d g e s 1, M . M i g i n i a c - M a s l o w 1, p. L e M a r 6 c h a l 1 a n d R. R 6 m y 2 I Laboratoire de Physiologie V$g$tale Moldculaire, CNRS (UAl128) and 2 Laboratoire de G~n$tique Mol$culaire des Plantes, CNRS (UAl15), Universit$ de Paris Sud, Orsay (France)

(Received 31 August 1989) (Revised manuscript received 28 November 1989)

Key words: Chloroplast; Ferredoxin : NADP + oxidoreductase; Light activation; Phosphoamino acid; Protein phosphorylation

Incubation of thylakoids with purified FNR and [32 PIATP led to the incorporation of phosphate into the FNR. In the absence of added FNR, 32p-labelled ~ could be detected associated with the thylakoids. An amino-acid analysis showed that in the dark, the FNR could be phosphorylated on a serine residue. In the presence of thylakoids, the FNR contained a threonine phosplmte which was associated with a light-dependent reaction. The physiological function of this phosphorylation is not clear. Some modifications in NADP +-dependent photosystem I (PSI) activity and FNR-membrane association have been observed on the addition of ATP. Whether these changes are linked to the phosphorylation of the FNR remain to be fully elucidated.

Introduction Ferredoxin:NADP + oxidoreductase (FNR) is a nuclear-encoded protein located in the chloroplasts of higher plants. It has a molecular mass between 33 and 38 kDa according to plant species, and has been shown to exhibit multiple forms (see Refs. 1-8). It is the final enzyme of the linear photosynthetic electron transfer chain, although it has also been suggested to be implicated in cyclic electron flow around photosystem I (PSI) [4,5]. Therefore, it may play a key role in the regulation of cyclic/non-cyclic electron flow and hence control the N A D P H / A T P ratio in chloroplasts. It is localised mainly in the non-appressed thylakoid membranes although approx. 20% has been shown to be found in the granal stacks [6,7]. The F N R is believed to

Abbreviations: CHAPS, 3-(3-cholamidopropyl)dimethylammonio)-lpropanesulphonate; Chl, chlorophyll; DCMU, 3-(3,4-dichlorophenyl)-l,l-dimethyl urea; DPIP, 2,6-dichlorophenol indophenol; FNR, Ferredoxin:NADP + oxidoreductase; FSBA, 5'-p-fluorosulphonylbenzoyladanosine; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulphonic acid; LHCII, fight harvesting Chl a / b complex of PSII; MDH, malate dehydrogoaase; NEM, N-ethylmaleimide; PS, photosystem; SDS-PAGE, sodium dodecyl sulphate-polyacrylamide gel electrophoresis. Correspondence: M. Hodges, Laboratoire de Physiologie V6g6tale Moleculaire, B~tt. 430, CNRS (UAl128), Universit~ de Paris Sud, 91405 Orsay Cedex, France.

be bound to the thylakoid membranes via an electrostatic mechanism [8]. Two pools of F N R appear to exist in vivo; a loosely bound pool which is easily removed from the membrane by a low salt wash and a more tightly bound pool (30-60% of the total enzyme) which requires several extensive low salt/EDTA washes and/or the additon of detergents (like CHAPS) for its removal [91. The tightly bound F N R might be that which is associated with the isolated cytochrome b6/f complex [10] and which is also a major contaminant, co-purifying with the LHCII kinase [11]. This key enzyme appears to be activated by light in a process requiring a pH gradient [12] and associated with a conformational change of the F N R [13]. The presence of a tightly bound phosphate group associated with the purified protein has been mentioned in the literature [14]. This raises the question whether a post translational phosphorylation mechanism exists which could partly explain perhaps the observed F N R heterogeneity and possible multiple functions. Protein phosphorylation of thylakoid proteins is now a widely established phenomenon [15,16]. Until now, mainly photosystem II (PSlI)-associated proteins have been identified as being phosphorylated, including LHCII [17], D1 and D2 (the rection centre proteins of PSII) [18] and a 9 kDa protein (the psbH gene product) [19]. The role of this extensive phosphorylation is still poorly understood except for the LHCII which plays an important role in the control of excitation energy distil-

0167-4889/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

447 bution between the two photosystems [20]. However, several in vitro phosphorylation-induced modifications in PSII functioning have been reported [21-23], but the exact phosphoproteins involved were not determined. Furthermore, it is known that the protein kinase responsible for the phosphorylation of LHCII is under a light-dependent redox control [20], although the exact mechanism is far from being fully elucidated. In this work we report for the first time to our knowledge phosphorylation of a non-PSII protein in higher plants, the FNR, and the possible consequences on its functioning. Materials and Methods

Chloroplasts from greenhouse-grown peas were isolated as in Ref. 24, followed by an osmotic shock in 20 mM Hepes (pH 7.6), 10 mM NaC1 and subsequent centrifugation at 2500 × g. The thylakoids were then washed in 20 mM Hepes (pH 7.6), 10 mM NaC1 and 1 mM EDTA and after a second centrifugation the resulting pellet was resuspended in 330 mM sorbitol, 20 mM Hepes (pH 7.6), 5 mM MgC12, and 10 mM NaC1 as a concentrated stock (2 mg Chl/ml). Protein phosphorylation was carried out on thylakoids resuspended to 200 #g Chl/ml in 20 mM Hepes (pH 8), 10 mM NaC1, 10 mM MgC12, and 200 #M ATP (containing [32p]ATP at a spec. act. of 200 ttCi/#mol). The reaction medium was supplemented with 10 mM NaF and 100 mM dithiothreitol where indicated in the results or figure legends. Phosphorylation was either carried out for 15 min in the light (intensity: 225 # E . m - 2 . s -1) or in the dark with dithiothreitol. Control thylakoids were those left in the dark without dithiothreitol. The phosphorylation reaction was stopped by the addition of 50 mM EDTA and 5 mM CHAPS. The thylakoids were allowed to incubate at room-temperature for a further 15 rain before being centrifuged for 10 min at 12000 X g. The sedimented membranes were dissolved in SDS-dissociation buffer (100 mM Tris (pH 6.8), 10% SDS, 20% glycerol, 5% fl-mercaptoethanol, and Bromophenol blue) while the proteins in the supernatant were precipitated by the addition of trichloroacetic acid (10% (w/v) final volume) for 1-2 h. Finally the supernatants were centrifuged at 12000 × g for 10 min and dissolved in the same buffer as the thylakoids. Dephosphorylation of the purified FNR was carried out in the presence of 2 U alkaline phosphatase, 20 mM Hepes (pH 8) and 10 mM MgC12 for 1 h at room temperature. The FNR was then separated from the alkaline phosphatase by 2',5'-ADP-Sepharose affinity chromatography (see below). This FNR was then used either for phosphorylation experiments or subjected to PAGE under non-denaturing conditions.

SDS-PAGE was carried out as in Ref. 25 and the resulting gels were stained with Coomassie brilliant blue, dried and the phosphorylated proteins were detected by autoradiography using Kodak XAR-5 X-ray film (1 day to 1 week exposure at - 8 0 o C). 10% PAGE in the absence of SDS was also carried out. Western blot analyses were carried out after transfer of proteins to nitrocellulose sheets as in Ref. 26 and then probed with antibodies raised against purified spinach leaf FNR. Ferredoxin-Sepharose affinity chromatography, as described in Ref. 27, was used for the purification of the FNR (peak elution at 140 mM NaC1), followed by 2',5'-ADP-Sepharose affinity chromatography where the FNR was released during a 0-400 mM NaC1 linear gradient. The phosphate content determination of the purified spinach leaf FNR was carried out using the method described in Ref. 28. Pure FNR (1.7 mg) was ashed in the presence of 500 #1 of 60% HC104 by heating in an aluminium heating block at 200 °C until the mixture turned colourless. The HC104 was removed by heating at 300 o C. The tube was cooled and the ashed FNR was resuspended in 500/~1 of distilled water. The phosphate of the ashed FNR was analysed in the presence of malachite green by measuring the absorbance at 660 nm. The ashed FNR suspension buffer was used as a control. Alkaline hydrolysis of the purified spinach FNR was carried out as in Ref. 29. The FNR was precipitated with 10% TCA which led to the removal of the flavin group as judged by the yellow colour of the resulting supernatant and the bleached pellet. The pellet was washed with 10% TCA, resuspended in 0.5 M KOH and left for 16 h at 400 C. The reaction mixture was acidified with HC1 and TCA was added to a final concentration of 10%. After a further centrifugation, the supernatant was analysed for phosphate as for the ashed FNR. To see which amino acids were phosphorylated, purified FNR was added to the thylakoid extract during the phosphorylation period. The FNR was then repurifled on a ferredoxin-Sepharose affinity column as in Ref. 27. The amino acid analysis was carried out by acid hydrolysis in HC1 followed by a 2-dimensional separation by thin-layer chromotography and electrophoresis on cellulose plates as in Ref. 30. Diaphorase activity was measured by the absorbance change at 600 nm in a reaction medium containing 100 mM Tris (pH 7.9), 100 #M DPIP, 200 #M NADPH and variable concentrations of thylakoids and supernatant. Light-saturated NADP + reduction was measured by the absorbance change at 340 nm. The reaction media contained 20 #g Chl/ml, 20 mM Hepes (pH 8), 5 mM MgC12, 10 mM KC1, 500 #M NADP ÷, 5 #M Spirulina maxima ferredoxin for whole-chain electron transfer and supplemented with 20/~M DCMU, 100 #M DPIP,

448 10 mM ascorbate and 5 #M nigericin for PSI-dependent activities. Light-saturated PSI activity was measured in a Clark-type oxygen electrode by the uptake of 02 in the presence of 100 #M methylviologen, 20 mg Chl/ml, 20 #M Tricine (pH 8.0), 5 mM MgC12, 10 mM KC1, 10 mM NaF, 100 /~M DPIP, 10 mM ascorbate, 10 /~M DCMU and 5 #M gramicidin. The light intensity was 1000/zE. m -2 • s -1.

A

B I

2

3

I

FNR

Results

• LHC

The initial aim of this study was to investigate whether the redox control of the LHCII kinase depended upon the ferredoxin/thioredoxin enzyme system, similarily to the light activation of a number of stromal enzymes [31]. At the same time it seemed interesting to see if the 12 kDa soluble phosphoprotein reported by Bhalla and Bennett [32] was indeed the thioredoxin, since the existence of a phosphothioredoxin has been observed in Escherichia coli. [33]. To accomplish these aims, thylakoids thoroughly washed with EDTA were used in order to eliminate the bound thioredoxin [24] and the phosphorylation pattern of the membranes in the presence and absence of added pure ferredoxin, ferredoxin:thioredoxin reductase and thioredoxin was examined. A high degree of thylakoid protein phosphorylation was still observed in the absence of these proteins, while the activation of added pure N A D P - M D H was severely inhibited (giving only 1% of the control activity when th{oredoxin was omitted from the activation medium). The addition of these three proteins did not modify the thylakoid phoshorylation pattern nor the extent of the phosphorylation; no significant phosphorylation of the thioredoxin was observed, even though the added M D H was fully activated. Furthermore,it was seen that the soluble 12 kDa phosphoprotein migrated further on SDS-PAGE than the added thioredoxin (data not shown.) In most cases, a weakly labelled protein with a molecular mass of approximately 34 kDa was observed among the numerous other thylakoid-associated phosphoproteins (Fig. 1A, lanes 2 and 3). A western blot of gels similar to those in Fig. 1A with antibodies raised against the spinach leaf F N R (Fig. 1A, lane 1 and Fig. 1B, lane 2), after transfer onto nitrocellulose, showed that thylakoid-bound F N R was located at the position of the labelled protein band (Fig. 1B, lane 1). When purified spinach leaf F N R was added to washed thylakoid membranes, under phosphorylating conditions, a labelling of the added FNIL located in the supernatant after centrifugation of the thylakoids, was detected (Fig. 2). It can be seen from Fig. 3A that an EDTA/CHAPS wash ( + ) followed by centrifugation led to the removal of the 34 kDa protein from the thylakoids (T). Such a

2

Stain

II

32 p

Blot

Fig. 1. (A) The localisation of FNR on SDS-PAGE gels. Lanes 1 and 2 are the proteins stained with Coomassie blue and correspond to purified leaf FNR (lane 1) and washed thylakoids (lane 2). The autoradiogram of lane 2 showing the incorporation of 32p into prot¢ins of washed thylakoids is lane 3. A western blot analysis (B) of washed thylakoids (lane 1) and purified leaf FNR (lane 2) using antibodies raised against the protein shown in Fig. 1A, lane 1. Protein phosphorylation was carded out in the light in the absence of dithiothreitol.

CB

32 p

FNR

i¸!:i!~i:i:!:iii i: ~ ,~! i,~, ~

Fig. 2. The phosphorylation of purified spinach leaf FNR. Purified FNR was incubated with washed pea thylakoids under phosphorylating conditions (light a n d n o dithiothreitol) and the resulting supernatant, aft¢r centrifugation, was subjected to SDS,PAGE. CB corresponds to the Coomassie blue-stained proteins and 32p to the autoradiogram.

449 A

B

32 p

FNR

FNR

L H C II

LHC II

T+

T-

S+

S-

S+

Fig. 3. The removal of membrane-bound FNR from thylakoids by an EDTA/CHAPS treatment. Thylakoids were washed in the presence of EDTA ( - ) or EDTA and CHAPS (+), and after centrifugation the thylakoids (T) and supernatants (S) were subjected to SDS-PAGE (A). Thylakoids were phosphorylated in the dark with dithiothreitol and after an EDTA/CHAPS wash; the resulting supernatant ( S + ) was seen to contain [32p]FNR (B). The Coomassie blue-stained gel is shown to the left of the autoradiogram (32p).

which had been incubated in the dark (Fig. 4A) and in the light (Fig. 4B). The spots correspond to a phosphoserine (PS) in the dark incubated sample and to phosphoserine (PS) and phosphothreonine (PT) in the lightstimulated case. Surprisingly, when a control with only FNR and labelled ATP was done (without thylakoids) a weak labelling of the FNR could be detected on a phosphoserine whether the samples were incubated in the light or in the dark (data not shown). It was clear that the FNR could be phosphorylated but the extent of the incorporation of phosphate when unphysiologically high concentrations (20-50 times the amount associated with the washed thylakoids, based on diaphorase activity measurements) of purified spinach FNR was added to pea thylakoids was low (1-4 phosphate groups per 100 molecules of FNR) and variable. This could arise because of a poor accessibility of the added FNR with respect to the membrane-bound kinase or because the FNR was already partly phosphorylated. Therefore, the phosphate content of the purified FNR was investigated. After an extensive dialysis and ashing, it was found that the FNR contained between 1 and 2 mol of phosphate per mol of enzyme. To verify that this might be due to the presence of phosDork

treatment removed 93% of the thylakoid-bound FNR, based on diaphorase activity measurements carried out on both the supernatant (S) and membrane fractions (data not shown). In the absence of CHAPS but in the presence of EDTA only 33% of the diaphorase activity was found in the supernatant ( S - ) . Therefore, thylakoids were phosphorylated and washed in E D T A / CHAPS and the supernatant (S + ) was analysed for [32p]FNR. It can be seen in Fig. 3B that after such a treatment the 34 kDa protein present in the supernatant was labelled. Autoradiography showed that FNR could also be phosphorylated in the dark, although a greater phosphorylation was brought about in the light. The addition ,'of FSBA, NEM or CaC12 to the reaction medium led to an inhibition of the light-activated FNR phosphorylation, which was also the case for the LHCII (data not shown). The treatment of the samples before gel electrophoresis seemed to exclude the possibility of a weak binding of [32p]ATP to the FNR. However to conclude that the FNR was really phosphorylated, an analysis of labelled phosphoaminoacids was carried out. This was achieved by repurifying the FNR which had been incubated with [32p]ATP and thylakoids under phosphorylating conditions on a ferredoxin-sepharose column. The purified FNR was then subjected to acid hydrolysis and amino acid analysis after a 2-dimensional separation. Fig. 4 shows the phoshoaminoacids associated with the FNR

A

Light

PS PT

o

B Fig. 4. Phosphoamino acid analysis of repurified FNR which had been incubated in the presence of thylakoids and [32p]ATP either in the dark (A) or in the light (B). PS and PT correspond to phosphoserine and phosphothreonine, respectively.

450 TABLE I

The effect of A T P and light addition on light saturated N A D P + photoreduction in uncoupled thylakoids Whole chain means electron transfer from H 2 0 to N A D P +, PSI activity means electron transfer from D P I P / a s c o r b a t e to N A D P ÷ in the presence of D C M U and - and + correspond to the absence and presence of antimycin A respectively. The results are expressed in /~mol N A D P + reduced per nag Chl per h. Treatment

Light + A T P Light - A T P ATP-induced change Dark + ATP Dark - ATP ATP-induced change

Whole chain

61.6+11 74.7+14 - 17.5% 82.6 82.1 + 0.6%

PSI activity -

+

42.6+3 53.35:3 - 20.0%

62.95:9 52.25:6 + 20.4%

74.2 60.9 + 21.8%

n.d. n.d.

phoaminoacids, the FNR was subjected to alkaline hydrolysis. This treatment, which should liberate phosphate from phosphorylated serine and threonine, showed the absence of any alkaline-labile phosphate. This discrepancy can be explained by the presence of two phosphates in the structure of the flavin associated with the FNR. This inability to find the FNR already phosphorylated was indirectly confirmed by the treatment of the purified FNR with alkaline phosphatase in an attempt to dephosphorylate the FNR. Such experiments did not improve the incorporation of phosphate into the FNR incubated with thylakoids under phosphorylating conditions (data not shown). It did not modify the migration of the FNR on native PAGE with respect to the untreated FNR, where a difference would have been expected if the FNR was dephosphorylated and therefore carrying less negative charge (data not given). The possible signification of FNR phosphorylation was investigated by measuring either NADP ÷ or methylviologen photoreduction associated with the thylakoid membranes after a preincubation with ATP in the light or the dark. Table I shows the effect of ATP and light (phosphorylating conditions) on NADP + photoreduction associated with the thylakoid membrane. Whole-chain electron transfer was seen to be inhibited by a preincubation in the presence of ATP in the light which has been previously shown to be associated with a phosphorylation-induced inhibition of PSII reactions [22]. The preincubation with ATP in the dark led to a subsequent 20% increase in the rate of light saturated NADP + reduction in the presence of DCMU and nigericin (uncoupled PSI activity). The presence of ATP and light, on the other hand, led to a 20% inhibition of this photochemical reaction. Interestingly, it was observed that in the presence of antimycin a (an inhibitor of cyclic electron transfer around PSI) the rate of NADP + photoreduction was restored to a 20% stimula-

tion with respect to the nonphosphorylated (minus ATP) control. The exact site of the modification around PSI was examined by measuring the PSI activity associated with methylviologen as an artifical electron acceptor which does not require FNR as an electron donor. It was found that the addition of ATP, either in the light or in the dark, did not modify the methylviologen-dependent PSI activity at a saturating light intensity. The diaphorase activity of added FNR was not modified by a preincubation with ATP. When the endogenous FNR activity was measured after incubation of thylakoids with ATP and subsequent centrifugation, an increase in diaphorase and ferredoxin-cytochrome c reductase activities of 20% was observed in the supernatant which mirrored the corresponding decrease in activity of the thylakoid pellet (data not given). This could indicate that phosphorylation plays a role in the interaction of the FNR with the thylakoid membranes. However, these changes were not light-dependent and only reflected the addition o f ATP. Therefore, they could be due either to serine phosphorylation or only to ATP binding to the FNR. Conclusions

The data presented above clearly indicate that thylakoid-bound FNR as well as purified FNR in the presence of thylakoids can be labelled with 32p when incubated with [32p]ATP and that this is due to the phosphorylation of either a serine (in the dark) or a serine and a threonine (in the light). However, the extent of this phosphorylation is weak, in the case of the added spinach FNR to pea thylakoids, although reproducible. This does not appear to be due to an already phosphorylated FNR, as no alkaline-labile phosphate could be detected and attempts to dephosphorylate the FNR did not improve the extent of the phosphorylation nor the migration of the FNR on native PAGE. Therefore, it may rather reflect the poor accessibility of external FNR to the membrane-bound kinase or a poor competition between soluble substrates (added FNR) and thylakoid-bound substrates (like LHCII). It is not clear which chloroplast kinase is responsible for the phosphorylation of the FNR. The results show that it is associated with the thylakoid membranes as it is still active after EDTA/NaC1 washing of the thylakoids. Furthermore, the fact that the FNR is phosphorylated on a threonine in a light-dependent reaction and that this reaction is Lnhibited by either FSBA, NEM or CaC12 strongly infers that the LHCII kinase is a good candidate (see Ref. 34). In addition, since the purified FNR exhibits a slight serine phosphorylation with [32p]ATP, it might be assumed that it has a weak intrinsic kinase activity. It has already been reported in the literature that a 38 kDa chloroplast protein is a

451 kinase [35]. This protein was later shown to be FNR but it was suggested that the previously observed phosphorylation was due to a slight contamination with LHCII kinase [36]. The same authors have also shown that FNR can bind azido-ATP [37] which infers that it has an ATP-binding site. The purified FNR used in this study did not show any weak contaminating bands when SDS-PAGE gels were silver stained and no radioactivity labelled band was found at 64 kDa (the molecular mass of the LHCII kinase) (see Fig. 2). Furthermore, the pure FNR did not cross react with an antibody raised against a synthetic polypeptide based on the N-terminus of spinach LHCII kinase. Thus, a slight autophosphorylation of the FNR cannot be ruled out. The physiological function of the FNR phosphorylation is not obvious. When NADP ÷ photoreduction activity was measured on thylakoids, ATP and light-dependent changes were observed. The presence of ATP in the dark led to a subsequent stimulation in PSI-dependent NADP ÷ photoreduction (Table I). However, when the ATP pretreatment was carried out in the light, a decrease in the PSI activity was seen. This decrease was not found when PSI activity was measured in the presence of methylviologen and therefore these changes can be considered to reflect changes in FNR activity. It may be that the phoshorylation of the FNR, either on the serine or the threonine, might regulate NADP + photoreduction and hence may play a role in redirecting the electron flow (cyclic/noncyclic) in function of the ATP/NADPH demand of the chloroplast. The fact that in the presence of antirnycin A, the light and ATP dependent decrease in PSI activity became a stimulation could suggest that a phosphorylation-induced cyclic flow of electrons around PSI occurs. It must also be kept in mind that if the FNR is phosphorylated by the LHCII kinase which is located in the granal stacks [38] then a portion of the FNR must also be found in the grana (see Ref. 7). It is possible that the phosphorylation of the FNR, by modifying its surface charge characteristics, alters its electrostatic equilibrium which is important for protein-protein interactions and hence leads to a different FNR location on the thylakoid membrane (for example, either associated with the PSI or cytochrome b6/f complex). Such a mechanism could be important in regulating the distribution of FNR between the stromal lamellae and the granal stacks. It has already been postulated that the phosphorylation of NADPH-cytochrome c reductase in guinea-pig peritoneal macrophages facilitates the formation of the NADPH oxidase complex with cytochrome b-559 [39].

Acknowledgements This work has been funded by the CNRS and M.H. was the recipient of a Royal Society (London, U.K.)

fellowship during a part of the work. We are indebted to Dr. K.K. Rao for his gift of the Spirulina maxima ferredoxin, to Dr. M. Droux for preparing the anti-FNR antibodies and A. Hoarau for inorganic phosphate determinations. The LHCII kinase antibody was a kind gift from Dr. S. Coughlan (Brookhaven National Laboratory, Upton, New York, U.S.A.).

References 1 Carrillo, N. and Vallejos, R.H. (1987) in The Light Reactions, Topics and Photosynthesis Vol. 8, (Barber, J., ed.), pp. 527-560, Elsevier Science Publishers, Amsterdam. 2 Pschorn, R., Riihle, W. and Wild, A. (1988) Photosynthesis Res. 17, 217-229. 3 Karplus, P.A., Walsh, K.A. and Herriott, J.R. (1984) Biochemistry 23, 6576-6583. 4 Carrillo, N. and Vallejos, R.H. (1983) TIBS 8, 52-56. 5 Mills, J., Crowther, D., Slovacek, R.E., Hind, G. and McCarty, R.E. (1979) Biochim. Biophys. Acta 547, 127-138. 6 BOhme, H. (1977) Eur. J. Biochem. 72, 283-289. 7 VaUon, O., Wollman, F.A. and Olive, J. (1986) Photochem. Photobiophys. 12, 203-220. 8 Carrillo, N. and Vallejos, R.H. (1982) Plant Physiol. 63, 210-213. 9 Matthijs, H.C.P., Coughlan, S.J. and Hind, G. (1986) J. Biol. Chem. 261, 12154-12158. 10 Clark, R.D., Hawkesford, M.J., Coughlan, S.J. and Bennett, J. (1984) FEBS Lett. 174, 137-142. 11 Coughlan, S.J. and Hind, G. (1986) J. Biol. Chem. 261, 1137811385. 12 PschSrn, R., Riihle, W. and Wild, A. (1988) Z. Naturforsch. 43, 207-212. 13 CarriUo, N.J., Lucero, H.A. and Vallejos, R.H. (1980) Plant Physiol. 65, 495-498. 14 Ulrich, E.L. and Marldey, J.L. (1980) Abstracts 9th Int. Congress on Magnetic Reasonance in Biological systems, Bendor, France, p. 100. 15 Bennett, J. (1984) Physiol. Plant. 60, 583-590. 16 Ranjeva, R. and Boudet, A.M. (1987) Annu. Rev. Plant Physiol. 38, 73-93. 17 Bennett, J. (1977) Nature 269, 344-346. 18 Marder, J.B., Teller, A. and Barber, J. (1988) Biochim. Biophys. Acta 932, 362-365. 19 Hird, S.M., Dyer, T.A. and Gray, J.C. (1986) FEBS Lett. 209, 181-186. 20 Allen, J.F., Bennett, J., Steinback, K.E. and Arntzen, C.J. (1981) Nature 291, 21-25. 21 Hodges, M., Packham, N.K. and Barber, J. (1985) FEBS Lett. 181, 83-87. 22 Hodges, M., Boussac, A. and Briantais, J.-M. (1987) Biochim. Biophys. Acta 894, 138-145. 23 Packham, N.K., Hodges, M., Etienne, A.-L. and Briantais, J.-M. (1988) Photosynthesis Res. 15, 221-232. 24 Droux, M., Miginiac-Maslow, M., Jacquot, J.-P., Gadal, P., Crawford, N.A., Kosower, N.S. and Buchanan, B.B. (1987) Arch. Biochem. Biophys. 256, 372-380. 25 Laemmli, U.K. (1970) Nature 227, 680-685. 26 Nelson, T., Harpster, M.H., Mayheld, S.P. and Taylor, W.C. (1984) J. Cell Biol. 98, 558-564. 27 Droux, M., Jacquot, J.-P., Miginiac-Maslow, M., Gadal, P., Huet, J.C., Crawford, N.A., Yee, B.C. and Buchanan, B.B. (1987) Arch. Biochem. Biophys. 252, 426-439. 28 Hess, H.H. and Derr, J.A. (1975) Analytical Chem. 63, 607-613. 29 Ahmed, K. and Judah, J.D. (1967) Methods Enzymol. 10, 777-780. 30 Brulfert, J., Vidal, J., Le Mar6chal, P., Gadal, P., Qneiroz, O.,

452

31 32 33 34 35

Kluge, M, and Kruger, I. (1986) Biochem. Biophys. Res. Comm. 136, 151-159. Buchanan, B.B. (1980) Annu. Rev. Plant Physiol. 31, 341-374. Bhalla, P. and Bennett, J. (1987) Arch. Biochem. Biophys. 252, 97-104. Pigiet, V. and Conley, R.R. (1979) J. Biol. Chem. 253, 1910-1920. Lin, Z.F., Lucero, A. and Racker, E. (1982) J. Biol. Chem. 257, 12153-12156. Packham, N.K. (1987) Biochim. Biophys. Acta 893, 259-266.

36 Coughlan, S.J. and Hind, G. (1986) in Progress in Photosynthesis Research Vol. II (Biggins, J., ed.), pp. 797-800, Martinus Nijhoff, Leiden. 37 Coughlan, S.J. and Hind, G. (1987) Biochemistry 26, 6515-6521. 38 Whitelegge, J.P., Millner, P.A., Gounaris, K. and Barber, J (1986) in Progress in Photosynthesis Research Vol. II (Biggins, J. ed.), pp. 809-812, Martinus Nijhoff, Leiden. 39 Tamoto, K., Hazeki, K., Nochi, H., Mori, Y. and Koyama, J. (1989) FEBS Lett. 244, 159-162.