Proc. Natl Acad. Sci. USAVol. 80, pp. 745-749, February 1983Biophysics

Localization of different photosystems in separate regions ofchloroplast membranes

(photosynthesis/chloroplast structure/chloroplast function/reaction center/membrane differentiation)

JAN M. ANDERSON AND ANASTASIOS MELIS

Division of Molecular Plant Biology, 313 Hilgard Hall, University of California, Berkeley, California 94720

Communicated by Daniel I. Amnon, November 4, 1982

ABSTRACT The stoichiometric amounts and the photoactiv-ity kinetics of photosystem I (PSI) and of the a and fl componentsof photosystem I (PSiH. and PSll^) were compared in spinachchloroplast membrane (thylakoid) fractions derived from ap-pressed and nonappressed regions. Stroma-exposed thylakoidfractions from the nonappressed regions were isolated by differ-ential centrifugation following a mechanical press treatment of thechloroplasts. Thylakoid vesicles derived mainly from the ap-pressed membranes of grana were isolated by the aqueous poly-mer two-phase partition method. Stroma-exposed thylakoids werefoundtohaveachlorophylla/chlorophyllbratioof6.0andaPSII/PSI reaction center ratio of 0.3. Kinetic analysis of system II pho-toactivity revealed the absence ofPSII, from stroma-exposed thy-lakoids. The photoactivity ofsystem I in stroma-exposed thylakoidsshowed a single kinetic component identical to that of unfraction-ated chloroplasts, suggesting that PSI does not receive excitationenergy from the PSII-chlorophyll ab light-harvesting complex.Thus, stroma-exposed thylakoids are significantly enriched in bothPSI and PSp. Inside-out vesicles from the appressed membranesof grana-partition regions had a chlorophyll a/chlorophyll b ratioof 2.0 and a PSI/PSI reaction center ratio of 10.0. The photoac-tivity of system II showed the membranes of the grana-partitionregions to be significantly enriched in PSI~,. We conclude thatPSIIa is exclusively located in the membranes of the grana par-titions while PSII, and PSI are located in stroma-exposed thyla-koids. The low PSI reaction center (P700) content of vesicles de-rived from grana partitions and the kinetic homogeneity ofthe PSIcomplex suggest total exclusion of P700 as a functional componentin the membrane of the grana-partition region.

The structural differentiation of higher plant chloroplast mem-branes into grana stacks and stroma-exposed thylakoids is par-alleled by a functional differentiation, with stroma-exposed thy-lakoids having mainly photosystem (PS) I and grana stacks beingenriched in PSII (1). The introduction ofaqueous polymer two-phase partition by Albertsson (2) allowed the isolation ofinside-out vesicles derived mainly from grana partitions-i.e., the ap-pressed membranes of the grana stacks-that were even moreenriched in PSII (3, 4). Andersson and Anderson (5) demon-strated a marked lateral heterogeneity in the distribution ofthechlorophyll (Chl)-protein complexes along the thylakoid mem-brane of spinach chloroplasts. Thylakoid membrane vesiclesderived from the grana partitions showed an enrichment in PSIIChl-protein complex and its associated Chl ab-protein of thelight-harvesting complex (LHC) and a substantial depletion inPSI Chl-protein complex. These results led to the hypothesisthat PSI is excluded from the grana-partition regions (5, 6). Incontrast, stroma-exposed thylakoids were enriched in PSI butthey always contained a small complement of PSII (5-8).

Recognition of the biphasic nature of PSII photoactivity has

The publication costs ofthis article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertise-ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact.

led to the hypothesis ofPSII heterogeneity in higher plant chlo-roplasts (9). Analysis of the biphasic kinetics of PSII showed afast and nonexponential a component (PSIIa) and a slower ex-ponential (3 component (PSII,). The two types of PSII reactioncenter complexes differ both in the effective light-harvestingChl antenna size (10) and in the apparent midpoint redox po-tential of their primary electron acceptor Q (11, 12). Mg2e ionsaffect the organization of PSII. in the membrane but not thatof PSII1l (13). Melis and Homann (13) proposed that PSII islocated in grana partitions and PSII3 is located in stroma ex-posed thylakoids. Further studies (14, 15) supported this con-cept, because the amounts of PSIIa and PSII, differed in mu-tant tobacco compared with wild type, with PSII, increasing asthe amount of appressed relative to nonappressed membranearea increased. Hitherto, however, this hypothesis has not beentested by determining the amounts of the two types of PSII re-action center complexes in subthylakoid fractions.

In this paper, we report on the stoichiometric ratios and ki-netic properties of the photoactivity of PSII and PSI reactioncenters in thylakoid membrane fractions derived from granapartitions and also from stroma thylakoids. The results show thatPSIIH is found exclusively in the grana-partition regions. In con-trast, stroma thylakoids are enriched in both PSI and PSII,,.

MATERIALS AND METHODSChloroplasts were isolated from leaves of spinach (Spinaciaoleracea L.). The chloroplast isolation procedure, the Yedapress treatment, and the aqueous polymer two-phase partitionmethods were similar to those described in ref. 5. Stoichio-metric measurements of PSII and PSI reaction centers weretaken with a sensitive optical difference spectrophotometer (8).The geometry ofthe apparatus and the experimental conditionsfor the kinetic measurements have been reported (16, 17). Fordetermination of the primary photoactivity kinetics of the PSIreaction center (P700), chloroplasts were first treated with cy-anide for 60 min to inhibit plastocyanin (18). The Chl concen-trations and the Chl a/Chl b ratios were determined accordingto ref. 19.

RESULTSTo avoid the use of harsh detergents, which might disturb thein vivo organization ofthe supramolecular complexes ofthe thy-lakoids, the chloroplasts were fractionated mechanically by theYeda press treatment. The Yeda press fractionation of the spin-ach thylakoids was followed by separation of the grana stacks(Y-40) from the stroma-exposed thylakoids (Y-100) by differ-ential centrifugation (5). We then used the aqueous polymertwo-phase partition (2) to separate the right-side-out vesicles

Abbreviations: Chl, chlorophyll; PS, photosystem; LHC, light-har-vesting complex; P700, reaction center of PSI; Q, primary electron ac-ceptor of PSII.

745

746 Biophysics: Anderson and Melis

(fraction T-2), with both PSI and PSII characteristics, from theinside-out vesicles (fraction B-3), which are highly enriched inPSII and are derived mainly from the grana partitions (3-5). Theamounts of PSI and PSII reaction centers were measured di-rectly from the light-induced oxidation of the P700 (PSI) andfrom the light-induced reduction of Q, the primary electronacceptor of PSII.The Chl a/Chl b ratios and the Q and P700 contents of un-

fractionated thylakoids and various subchloroplast fractions de-rived from appressed and nonappressed membrane regions arecompared in Table 1. In mature unfractionated chloroplasts, theChl a/Chl b ratio was about 2.8 and the ratios Chl/P700 = 580and Chl/Q = 297 were comparable with those reported (20) fora variety of higher plant chloroplasts. Thus, spinach thylakoidsappeared to have almost twice as much PSII reaction center asPSI reaction center (Q/P700 = 1.95).The stroma-exposed thylakoid membranes (fraction Y-100)

derived from the Yeda press treatment (5) had a Chl a/Chl bratio of about 6 and were significantly enriched in PSI. The Q/P700 ratio was 0.3 in agreement with a previous study (8). Incontrast, the inside-out thylakoid vesicles, fraction B-3 derivedmainly from the chloroplast partition regions, had a low Chl a/Chl b ratio (about 2.1) and were markedly enriched in PSII (Q/P700 = 9.6). As expected from their known photochemical char-acteristics (4), the right-side-out vesicles (fraction T-2) from thegrana, were slightly depleted in Q relative to chloroplasts (Table1).

Comparison of the amount of P700 in stroma-exposed thy-lakoids (Y-100) with that in the grana partitions (B-3) shows a10-fold depletion ofP700 in fraction B-3 relative to Y-100 (Table1). This is similar to the depletion of the PSI complex assayedpreviously by NaDodSO4/polyacrylamide gel electrophoresis(5). Since fraction B-3 is known to be contaminated with right-side-out vesicles (4), the depletion of P700 from the membraneofthe partition region must be greater in vivo than that detectedin this study. These results support the concept of the lateralseparation of the two photosystems in the thylakoid membraneof higher plant chloroplasts (see Discussion). The results are inagreement with those of a previous study (8), which showedgrana stacks (Chl a/Chl b, 2.3) to have a Q/P700 ratio of 2.5.They disagree with a third study (21), which showed equalamounts of Q and P700 and invariable Chl/Q and Chl/P700ratios in chloroplasts and inside-out vesicles.

In view of the extreme depletion of the PSI complex in thegrana partitions, it was suggested that perhaps all of the PSIcomplex is excluded from the partition region in vivo (5, 6, 22).If this were so, then the PSI complex might be structurally and

Table 1. Chl a/Chl b ratios and Q and P700 contents inchloroplast membrane fractions

Chla/Chl b Chl/P700 Chl/Q Q/P700Chloroplasts 2.78 580 297 1.95Stroma thylakoids

(Y-100) 5.9 262 875 0.30Inside-out grana

partitions (B-3) 2.07 2,642 273 9.6Right-side-out grana

(T-2) 2.81 534 316 1.7

Unfractionated chloroplasts, stroma thylakoids (Y-100), and granapartitions (B-3) were obtained as described in ref 5. Chl/P700 and Chl/Q ratios are based on total Chl (a+b) in the sample. Q and P700 con-centrations were measured as described in ref. 8. Reproducibility of theresults for unfractionated chloroplasts, isolated stroma thylakoids, andthe right-side-out vesicles (T-2) was 10-20%. Different inside-outgrana-partition (B-3) preparations showed greater variability (i.e., Q/P700 ratio variations were 6-10).

0-.

Time, s

0.2 0.4

II

0ID~~~-d

A

FIG. 1. Kinetic traces: time course of P700 photooxidation. (A)Unfractionated chloroplasts at 238 ,uM Chl (a+b) in the presence of200 puM methylviologen. Chl a/Chl b = 2.79; result presented is themean of 16 individual measurements; AA700 = lo-3 (B) Isolatedstroma thylakoids at 23 ,IM Chl (a+ b) in the presence of 200 p.M meth-ylviologen. Chl a/Chl b = 6.0; result presented is the mean of 32 in-dividual measurements; AA700 = 5 x 10'. Before fractionation, theisolated chloroplasts were treated with cyanide to inhibit plastocyanin(18).

functionally homogeneous, in contrast to the observed heter-ogeneity of PSII in higher plant thylakoids (9-15). Hence, acareful comparison of the kinetics of the PSI photoactivity wasundertaken for unfractionated chloroplasts and the stroma thy-lakoid fraction (Y-100).The photooxidation kinetics of P700 induced by weak con-

tinuous illumination is shown in Fig. 1. In this approach, therate of P700 photooxidation is limited by the rate of light ab-sorption by PSI. The rate of light absorption by each PSI unitdepends directly on the antenna size of the light-harvestingpigments associated with PSI. In principle then, one can obtainan accurate estimate of the antenna size of the light-harvestingpigments of PSI from measurement of the kinetics of P700 pho-tooxidation (16). Fig. 1A shows the P700 photooxidation kineticsof unfractionated chloroplasts while Fig. 1B presents the P700kinetics from stroma thylakoids. The semilogarithmic plot ofthekinetic data of Fig. 1 is a straight line that has the same slopefor both samples (Fig. 2). Thus, it is verified that the P700 pho-tooxidation is a monophasic first-order function of time occur-ring with the same rate in both unfractionated chloroplasts andthe isolated stroma thylakoid fraction. Physical separation ofthePSI complex in the stroma thylakoid fraction from the PSII-Chl

0

-2

1 -3 -

-4F0 0.2 0.4

Time, s

FIG. 2. First-order kinetic analysis of the traces shown in Fig. 1.Unfractionated chloroplasts (o) and isolated stroma thylakoids (o)gave identical rate constants as shown by the semilogarithmic slopes.

Proc. Natl. Acad. Sci. USA 80 (1983)

Proc. Natd Acad. Sci. USA 80 (1983) 747

ab LHC in the grana did not result in the loss of any portionof the functional light-harvesting antenna of PSI. The resultshows that the antenna unit size of PSI in stroma thylakoids isidentical to that of intact chloroplasts.The finding that the antenna unit size of PSI in unfractionat-

ed chloroplasts is the same as that in isolated stroma thylakoidsexcludes the possibility that PSI was also located in the mem-branes of the grana-partition regions where it would have in-teracted with the PSII LHC. The evidence suggests that in vivothe PSI complex is exclusively located in stroma-exposed thy-lakoids.

Unlike those of PSI, the primary photoactivity kinetics ofPSII are complex (9, 10). In view ofthe PSII differentiation intoPSIIH and PSII in higher plant chloroplasts, we undertook totest the hypothesis (13, 14) that membranes of the grana-par-tition regions and stroma-exposed thylakoids are enriched inPSIIa and PSII,, respectively. We did so by measuring the Chla fluorescence induction kinetics of the various thylakoid prep-arations in the presence of the electron transport inhibitor 3-(3',4'-dichlorophenyl)-1, 1-dimethylurea. In this approach, thearea over the fluorescence induction curve is directly propor-tional to the photoreduced fraction of Q (10, 23). Comparisonof the fluorescence induction curves of unfractionated chloro-plasts and stroma-exposed thylakoids shows that the slow P.component accounts for a. small portion of the variable fluores-cence in, the unfractionated chloroplasts but for a major portionof the variable fluorescence from stroma thylakoids (Fig. 3).A quantitative presentation ofthe same results (Fig. 4) shows

the semilogarithmic plots ofthe areas over the two fluorescencekinetic curves (13, 23). From the intercepts ofthe linear phaseswith the ordinate at zero time, in several different preparations,-we determined that PSII.3 accounts for 20-35% ofthe total PSIIin the chloroplasts but for 85-99%ofthe total PSII in the stromathylakoid fraction. This observation directly supports the hy-pothesis that ( centers are located in stroma-exposed thylakoids(13, 14). Recently (17), we presented evidence based on kineticanalysis suggesting that the fluorescence yield of PSIIa, but notthat of PSII,,, strongly depends on the Mg-' content ofthe chlo-roplast suspension medium. In this work, we verified directlythat the fluorescence yield from stroma thylakoids is largelyindependent of Mg2" ions (data not shown).The fluorescence induction kinetics of unfractionated chlo-

roplasts (C) and inside-out vesicles derived from the chloroplastpartition region (B-3) are compared-in Fig. 5.. In contrast to theresults in Fig. 3, the fluorescence kinetics ofthe thylakoids fromthe grana-partition region are-considerably faster than those ofthe control chloroplasts, suggesting selective elimination of theslow (3 phase from the appressed membrane fraction. This isevidenced in the semilogarithmic plots of the areas over fluo-

a)80Goa)

~0

5.4a)

0 0.2 0.4Time, s

FIG. 3. Kinetics of the variable Chl a fluorescence of unfraction-ated chloroplasts (Chl a/Chl b = 2.65; upper curve) and isolated stromathylakoids (Chl a/Chl b = 5.7; lower curve). The reaction mixturescontained 30 ,uM Chl (a+b)/50 PM 3-(3',4'-dichlorophenyl)-1,1-di-methylurea/2 mM hydroxylamine.

-2a

-3 _

-4

0 0.2 0.4 0.6Time, s

FIG. 4. First-order kinetic analysis of fluorescence traces shownin Fig. 3;The relative PSII concentration in different unfractionatedchloroplast preparations (o) was 20-35% of the total PSII; in the iso-lated stroma thylakoids (e), it was 85-99% of total PS11.

rescence induction in Fig. 5B. From the results ofseveral mea-surements, we have determined that, in the grana partitions,the relative concentration of PSII, was reduced to 25-35% of.that found in the control chloroplasts. The result provides fur-ther support for the hypothesis that PSII,3 is excluded from thepartition region of grana thylakoids.

DISCUSSIONThe results presented in Table 1 for the amounts ofQ and P700show an enrichment ofQ and a significant depletion of P700 in.the membranes of the grana partitions and an enrichment ofP700, hence of PSI, in the stroma thylakoid region. They agreewith the previous study (5), which compared the relative dis-

=j 06 B- =

10.02 .

= 0.8 - A

0.6-

0.4-

~0

0 0.2 0.4

-1

-3-3

-4-

-5.0 0.2 0.4

Time, s

FIG. 5. (A) Kinetics of the variable Chl a fluorescence of unfrac-tionated chloroplasts (curve C; Chl a/Chl b = 2.9) and inside-out thy-lakoids derived from the 'grana-partition regions (B-3, Chl a/Chl b= 2.2). Conditions are as in Fig. 3. (B) First-order kinetic analysis ofthe fluorescence traces shown in A. The relative PSI[[ concentrationin the unfractionated chloroplasts (o) was approximately 25% of thetotal PS1; in the inside-out thylakoids derived from the grana parti-tions (e), it was5-10% of total PS11.

Biophysics:- Anderson and Melis

0.6 0.8-

748 Biophysics: Anderson and Melis

tribution of Chl-protein complexes in the two membrane re-gions. Here, we found only 20% of the total P700 in the B-3fraction compared with that of chloroplasts (Table 1). This par-allels the depletion of the coupling factor (CF1) in the B-3 frac-tion (24). Because CF1 is likely to be restricted to nonappressedmembranes only (25), its presence in the B-3 fraction could beattributed to contamination (24), as is the case with P700. Thus,one may argue that the low P700 content of the grana-partitionvesicles (B-3) represents contamination from stroma-exposedthylakoids containing CF1 (4, 24) and perhaps less than perfectseparation ofthe partition region from the adjacent stroma thy-lakoid region. Therefore, it is highly likely that the chloroplastgrana-partition region in vivo has a Q/P7-00 ratio much higherthan the detected ratio of9.6, probably exluding P700 as a func-tional component ofthis appressed membrane region altogether(5, 6).The kinetic analysis ofsystem II photoactivity (Fig. 5) showed

that grana-partition regions -contain primarily PSIIa and aredepleted ofPSII1. The one-third to one-fourth ofPSII, detectedin fraction B-3 can be attributed to the known contaminationof this fraction with about 25% right-side out vesicles from non-appressed membranes (4). The overall structural-functionalprofile of the grana-partition region, therefore, is one of highPSII, density and a lack of both PSII, and PSI reaction centercomplexes. We propose that such a structural arrangement inthe grana partitions favors the build-up of an electrochemicalproton gradient resulting from H+ release specifically by thephotooxidation of H20. This hypothesis is supported by the fol-lowing concurring conditions unique for grana-containing chlo-roplasts. (i) The small lumen volume of grana thylakoids mini-mizes the space within which an electrochemical protongradient is formed. (ii) The high density of water-splitting andproton-evolving PSII, in the partition region ofgrana thylakoidsprovides the bulk of protons released during linear electrontransport. These protons are necessary for build-up of the elec-trochemical proton gradient in the small space of the grana thy-lakoid lumen, important for the efficient generation of ATP.(iii) The hydrophobic nature of the grana-partition region mayminimize further proton loss (leakage) across most ofthe granumthylakoid membrane. The electrochemical proton gradient re-sulting from the photochemical activity of PSIIa would be con-sumed by ATP formation by the coupling factor bound nearbyon the stroma-exposed margin area of each granum thylakoid(Fig. 6). If this were the case, one would expect the grana thy-lakoids to be a PS11-specific domain for protons released by theoxidation ofwater (26, 27). Although the thylakoid lumen of thegrana is known to be continuous with that of stroma thylakoids,it is conceivable that there exists a mechanism preventing pro-tons released in the grana thylakoids from migrating into thelumen of stroma thylakoids and vice versa (27). Thus, it appearsthat the structural differentiation ofchloroplast membranes intograna and stroma thylakoids facilitates both a lateral separationof the various electron transport components and a differentia-tion of the electrochemical proton-gradient-formation processat the two sites.

In addition to PSI, stroma-exposed thylakoids possess somePSII activities (7), about 10-12% ofthe total PSII-associated Chlab LHC (5) and significant amounts ofQ (8), having a Q/P700ratio of 0.3 (Table 1). The kinetic analysis of system II photoac-tivity (Figs. 3 and 4) showed stroma-exposed thylakoids to con-tain PSII, and lack PSII,. Because the stroma thylakoids areunlikely to be contaminated with grana-partition fragments, dueto the gentle fractionation by the Yeda press, the result suggeststhat PSII, is normally located only in stroma-exposed thyla-koids. Such a locus would be found either in the end thylakoidsof the grana stacks or in the stroma thylakoids so that PSII,3 is

C FEDP pi

1- ~ SUEMCFI ( Grno prtition ) }

FIG. 6. Schematic showing the large membrane area where adja-cent grana thylakoids are appressed at the partition region and themuch smaller stroma-exposed grana margins. PSll. is exclusively lo-cated in the grana-partition membrane region along with most of theplastoquinone pool (8). The electrochemical proton gradient generatedspecifically by the turnover of PSI1, is dissipated via the coupling fac-tor (CF1), driving ATP synthesis from ADP and inorganic phosphate.The photolysis of H20 occurs in the hydrophobic region of the mem-brane, and the resulting electrons are processed by PSH1, (data notshown).

located together with PSI. Thus, PSIIa and PSII, differ in their-structural and functional arrangements [photochemical antennasize difference = 2-2.5 (10), apparent midpoint redox potentialdifference for their primary electron acceptor Q (11, 12), andlateral separation along the chloroplast thylakoid membrane(13)].The functional significance ofthe differentiation ofPSII. and

PSII, in mature higher plant chloroplasts is unknown. Themajor PSIIa with its larger antenna size and enhanced Chl bcontent will be the predominant source of electrons for linearelectron transport and noncyclic photophosphorylation. Thelocation of PSII, in appressed membranes ensures maximumtrapping of PSII light and hence ensures that the pool of plas-toquinone will be mainly reduced so that PSI, which governsoverall photosynthetic rate, can function effectively (22). Theminor PSII, with its smaller antenna size and relatively lesschlorophyll b may be required in stroma-exposed thylakoids toensure a slow electron flow, enough for regulatory functions andto poise the cyclic photophosphorylation (28).

In contrast to the observed functional heterogeneity of PSIIin higher plant chloroplasts, there is no indication of hetero-geneity in PSI (16, 29). The combination of the. Yeda press andthe aqueous two-phase partition treatment of the thylakoidsresulted in significant separation of PSII,, from PSI. Thus, thePSI complex was largely separated from the main Chl ab LHCproteins contained in the.grana partition. The identical mono-phasic first-order P700 photooxidation kinetics in unfraction-ated chloroplasts (Fig. 1A) and in isolated stroma thylakoids(Fig. 1B) argue that the physical separation of PSI from the Chlab LHC occurred without a loss in light-harvesting pigment forPSI. The only interpretation of this result is that, in vivo, PSIcannot receive excitation energy from the PSII-Chl ab LHC;i.e., PSI occurs as a separate entity in the stroma-exposed thy-lakoid membranes. Hence, the continuous array models show-ing that the main Chl ab LHC contacts both PSI and PSII re-action center complexes (30) are not tenable. This has beensuggested recently (16, 17) in direct measurements of the con-trol exerted by Mg2+ on the purported distribution of excitationenergy from the Chl ab LHC to PSII and PSI: no appreciableeffect of Mg2+ could be detected on the rate of -light absorptionby PSII and PSI. The important implication of our findings isthat interpretations on the phenomena of "spillover" and thy-lakoid unstacking (31), membrane phosphorylation (32, 33), andthe state I-state II transition (34) may have to be reassessed.

Proc. Natl. Acad. Sci. USA 80 (1983)

Proc. NatL Acad. Sci. USA 80 (1983) 749

J. M.A. was on leave from the Division of Plant Industry, Common-wealth Scientific and Industrial Research Organization, Canberra, Aus-tralia. The work was supported by a grant from a Competitive GrantsOffice of the U. S. Department of Agriculture to A. M.

1. Anderson, J. M. (1975) Riochim. Biophys. Acta 416; 191-235.2. Albertsson, P.-A. (1971) Partition of Cell Particles and Macro-

molecules, 2nd Ed. (Wiley, New York).3. Akerlund, H. E., Andersson, B. & Albertsson, P.-A. (1976)

Biochim. Biophys. Acta 449, 525-535.4. Andersson, B., Simpson, D. J. & Hoyer-Hansen, G. (1978)

Carlsberg Res. Commun. 43, 77-89.5. Andersson, B. & Anderson, J. M. (1980) Biochim. Biophys. Acta

593, 427-440.6. Anderson, J. M. (1981) FEBS Lett. 124, 1-10.7. Armond, P. A. & Arntzen, C. J. (1977) Plant Physiol 59, 398-

404.8. Melis, A. & Brown, J. S. (1980) Proc. Natl Acad. Sci. USA 77,

4712-4716.9. Melis, A. & Homann, P. H. (1976) Photochem. Photobiol 23,

343-350.10. Melis, A. & Duysens, L. N. M. (1979) Photochem. Photobiol 29,

373-382.11. Melis, A. (1978) FEBS Lett. 95, 202-206.12. Horton, P. & Croze, E. (1979) Biochim. Biophys. Acta 545, 188-

201.13. Melis, A. & Homann, P. H. (1978) Arch. Biochem. Biophys. 190,

523-530.14. Melis, A. & Thielen, A. P. G. M. (1980) Biochim. Biophys. Acta

589, 279-286.15. Thielen, A. P. G. M. & van Gorkom, H. J. (1981) Biochim. Bio-

phys. Acta 635, 111-120.16. Melis, A. (1982) Arch. Biochem. Biophys. 217, 536-5.

17. Melis, A. & Ow, R. A. (1982) Biochim. Biophys. Acta 682, 1-10.18. Izawa, S., Kraayerrhof, R. Ruge, E. K. & Devault, D. (1973)

Biochim. Biophys. Acta 314, 328-339.19. Mackinney, G. (1941)J. BioL Chem. 140, 315-322.20. Melis, A. & Harvey, G. W. (1981) Biochim. Biophys. Acta 637,

138-145.21. Tiemann, R., Renger, G. & Graber, P. (1981) in Photosynthesis,

ed. Akoyunoglou, G. (Balaban International Science Services,Philadelphia), Vol. 3, pp. 85-95.

22. Anderson, J. M. (1982) Photabiochem. Photobiophys. 3, 225-241.23. Melis, A. & Homann, P. H. (1975) Photochem. PhotobioL 21,

431-437.24. Berzborn, R. J., Muller, P., Roos, P. & Andersson, B. (1981) in

Photosynthesis, ed. Akoyunoglou, G. (Balaban International Sci-ence Services, Philadelphia), Vol. 3, pp. 107-120.

25. Miller, K. W. & Staehelin, L. A. (1976)J. Cell BioL 68, 30-47.26. Baker, G. M., Bhatnagar, D. & Dilley, R. A. (1981) Biochemistry

20, 2307-2315.27. Baker, G. M., Bhatnagar, D. & Dilley, R. A. (1982)J. Bioenerg.

Biomembr. 14, 249-264.28. Arnon, D. I. & Chain, R. K. (1975) Proc. NatL Acad. Sci. USA 72,

4961-4965.29. Malkin, R. (1978) FEBS Lett. 87, 329-333.30. Thornber, J. P. & Barber, J. (1979) in Topics in Photosynthesis,

ed. Barber, J. (Elsevier/North-Holland, Amsterdam), Vol. 3,pp. 27-70.

31. Murata, N. (1969) Biochim. Biophys. Acta 189, 171-181.32. Horton, P. & Black, M. (1982) Biochim. Biophys. Acta 680, 22-

27.33. Bennett, J., Steinbeck, K. E. & Arntzen, C. J. (1980) Proc. Natt

Acad. Sci. USA 77, 5253-5257.34. Wang, R. T. & Myers, J. (1974) Biochim. Biophys. Acta 347, 134-

140.

Biophysics: Anderson and Melis