Role of the y Subunit of Chloroplast Coupling Factor 1 in the Light-dependent Activation of Photophosphorylation and ATPase Activity by Dithiothreitol”

In leaves and intact chloroplasts, oxidation and reduction have been shown previously to regulate the ATPase activity of thylakoids. Illumination of spinach chloroplast thylakoids in the presence of dithiothreitol, which activates the ability of thylakoids to catalyze sustained ATP hydrolysis in the dark, causes increased incorporation of N-ethylmaleimide into the y subunit of coupling factor 1 (CF,). A disulfide bond in the y subunit is reduced during activation. The residues in-volved in this disulfide bond are the same as those in the disulfide linkage reduced during dithiothreitol activation of soluble CF,. The disulfide and dithiol forms of the y subunit may be separated by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. N-Ethylmaleimide is preferentially incorpo- rated in the dark into the reduced form of the y subunit of CF1 in thylakoids previously exposed to dithio- threitol. Only a subpopulation of the CFI in thylakoids is susceptible to dithiothreitol reduction and subse- quent reaction with N-ethylmaleimide in the dark. Alkylation of the thiol groups exposed by reduction of the disulfide bond protects ATPase activity from inhibition by oxidants. At a given value of the transmem- brane pH differential, photophosphorylation [3H]N-ethylmaleimide neces- sary for optimal labeling in illuminated thylakoids. Thylakoids were treated in the light with dithiothreitol, pelleted, resuspended, and incubated with 2 mM N-ethylmaleimide in the light or dark (35), and CF, was isolated. Aliquots of the CFI preparations were desalted, denatured by boiling for 1 min in the presence of 1% sodium dodecyl sulfate, and incubated with 9 mM [3H]N-ethylmaleimide (90,000 cpm/ nmol) for 14 min at room temperature. Unreacted [3H]N-ethylmal-eimide was removed and aliquots were subjected to electrophoresis. The specific activity of [3H]N-ethylmaleimide was estimated as described in the preceding report (22). The molecular weight of CF, was assumed to be 400,000 (37) in the calculation of [3H]N-ethylmaleim-ide incorporation.

Role of the y Subunit of Chloroplast Coupling Factor 1 in the Lightdependent Activation of Photophosphorylation and ATPase Activity by Dithiothreitol" (Received for publication, August 31, 1983) Stuart R. Ketcham$, James W. Davenport  In leaves and intact chloroplasts, oxidation and reduction have been shown previously to regulate the ATPase activity of thylakoids. Illumination of spinach chloroplast thylakoids in the presence of dithiothreitol, which activates the ability of thylakoids to catalyze sustained ATP hydrolysis in the dark, causes increased incorporation of N-ethylmaleimide into the y subunit of coupling factor 1 (CF,). A disulfide bond in the y subunit is reduced during activation. The residues involved in this disulfide bond are the same as those in the disulfide linkage reduced during dithiothreitol activation of soluble CF,. The disulfide and dithiol forms of the y subunit may be separated by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. N-Ethylmaleimide is preferentially incorporated in the dark into the reduced form of the y subunit of CF1 in thylakoids previously exposed to dithiothreitol. Only a subpopulation of the CFI in thylakoids is susceptible to dithiothreitol reduction and subsequent reaction with N-ethylmaleimide in the dark. Alkylation of the thiol groups exposed by reduction of the disulfide bond protects ATPase activity from inhibition by oxidants. At a given value of the transmembrane pH differential, photophosphorylation rates in dithiothreitol-activated thylakoids can be as much as seven to eight times those of nonactivated controls. N-Ethylmaleimide treatment of activated thylakoids in the dark prevents the loss of the stimulation of ATP synthesis on storage of the thylakoids. Photophosphorylation by intact chloroplasts lysed in assay mixtures is also activated in comparison to that by washed thylakoids. At a low ADP concentration, the rate of photophosphorylation approaches saturation as ApH increases. These results suggest that they subunit of CFI plays an important role in regulation of ATP synthesis and hydrolysis.
One of the peculiar features of washed chloroplast thylakoid * This work was supported by Grants PCM 79-1196 and PCM 82-14001 from the National Science Foundation and by Hatch Funds. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. membranes is that they catalyze negligible rates of ATP hydrolysis in the dark, even though they catalyze very high rates of ATP synthesis in the light. The treatment of thylakoids in the light with trypsin (1, 2) or a thiol compound (3), such as dithiothreitol, activates the ability of thylakoids to hydrolyze ATP in the dark. This ATPase activity and photophosphorylation are both inhibited by several inhibitors of the ATPase activity of soluble CF,,' including an antiserum to CF1 (4,5). Furthermore, the ATPase activity of thylakoids is coupled to proton translocation into the thylakoids (6), while ATP synthesis is coupled to proton translocation out of the thylakoids. These observations indicate that the ATP hydrolysis and synthesis activities of thylakoids are catalyzed by the same enzyme and must share at least some common steps in their mechanisms.
Activation may be divided experimentally into two energydependent phases (7). The ATPase activity of thylakoids previously illuminated in the presence of dithiothreitol decreases as the time interval between the end of activating illumination and the addition of ATP increases. This dark decay of the ATPase activity is remarkably accelerated by ADP, but is slowed by ATP and Pi ( 5 ) . After the ATPase activity has decayed in the dark, activity may be restored by illumination of thylakoids in the absence of dithiothreitol (7). Both phases of activation are sensitive to uncouplers.
Since CF, extracted from the treated membranes is an active ATPase (4), the energy-and dithiothreitol-dependent phase of activation modifies CF,. However, this modification of CF1 by dithiothreitol is not sufficient to allow expression of its membrane-bound ATPase activity, as shown by decay of this activity in the dark. Energy-dependent conformational changes that result in release of tightly bound ADP (8,91, and perhaps other energy-dependent alterations, are also required. CF, in intact chloroplasts (10-12) and leaves (13,14) can be activated to a state similar to that of CF, in thylakoids illuminated in the presence of dithiothreitol. Illumination of intact chloroplasts in the absence of added reductant activates ATPase activity, assayed in the dark after lysis of the chloroplasts. A thioredoxin system probably provides reducing equivalents for this activation in intact chloroplasts (15). Furthermore, there appears to be an oxidizing system in intact chloroplasts which deactivates CFI in the dark, further suggesting that reduction and oxidation of CF, may be an important form of regulation in vivo (11).
Because the illumination of thylakoids under reducing conditions activates ATP hydrolysis, thermodynamic considerations suggest that the same treatment should also activate ATP synthesis. There are several indications that this is the case. ATP formation in the dark occurs at lower artificially imposed ApH values in thylakoids previously illuminated in the presence of dithiothreitol than in control thylakoids (16). Moreover, photophosphorylation at a given ApH value is higher in light-and dithiothreitol-activated thylakoids than in nonactivated ones (17). Dithiothreitol treatment of soluble CF,, which activates its ATPase activity (4) and reduces a disulfide bond in its y subunit (18,19), enhances the ability of CF, to stimulate ATP synthesis in CF1-depleted thylakoids (20). Light and dithiothreitol treatment of thylakoids also allows ATP synthesis to begin after fewer flashes of light (21).
The 7 subunit of CF, contains four cysteinyl residues, distinguished by their varying ability to react with alkylating reagents and by their chromatographic separation following trypsin digestion of the y subunit (19,22). The same accessible sulfhydryl (S4) is labeled by treating oxidized or reduced soluble CF, with alkylating reagents or by treating nonactivated or dithiothreitol-activated thylakoids with alkylating reagents in the dark or light. The same buried sulfhydryl (S3) is labeled by treating oxidized or reduced soluble CF, with alkylating reagents in the presence of 1% sodium dodecyl sulfate or by treating nonactivated or dithiothreitol-activated thylakoids with alkylating reagents in the light, but not in the dark. Two cysteinyl residues form a disulfide bond in oxidized CF, (S1 and S2) and are labeled by treating reduced, but not oxidized, soluble CF, with alkylating reagents.
In this communication, we report that the activation of ATP hydrolysis by dithiothreitol treatment of thylakoids reduces the same disulfide linkage in the y subunit of CF, that is reduced by dithiothreitol treatment of soluble CF,. Alkylation of these thiol groups with N-ethylmaleimide remarkably protects the activated form of the membrane-bound enzyme from inactivation by oxidants. Reduction of the disulfide also activates photophosphorylation. The possible nature of this activation is discussed.

MATERIALS AND METHODS
Spinach chloroplast thylakoids (23) and intact (185%) pea chloroplasts (24) were prepared as described. Unless otherwise stated, the illumination (2 X IO6 ergs. cm-' s-') of thylakoids (0.1 mg of chlorophyll. ml-') in the presence of 5 mM dithiothreitol was carried out at 20 "C in an incubation mixture that contained 20 to 50 mM Tricine-NaOH (pH 8.0), 50 mM NaCI, 5 mM MgCI,, and 0.010 to 0.025 mM pyocyanine. In most experiments, the thylakoids were collected by centrifugation at 3,000 X g for 7 min at 0-4 "C, resuspended to about 0.1 mg of chlorophyll. ml" in a cold, buffered sucrose solution (0.4 M sucrose, 0.02M Tricine-NaOH (pH 8.0), and 0.01 M NaC1) and recollected by centrifugation. The pellets were resuspended in a minimal volume of the buffered sucrose solution.
flow was determined at various light intensities (0.1 to 5 X IO5 ergs. s ) In reaction mixtures that contained 20 mM Tricine-NaOH (pH 8), 50 mM NaC1, 5 mM MgCIZ, 0.02 mM pyocyanine, 3 or 5 mM potassium phosphate buffer (pH 8.0) containing about 5 X lo6 cpm Of 32P .pmol", either a limiting (0.01 mM) or saturating (0.5 mM) ADP concentration, and thylakoids equivalent to 0.1 mg of chlorophyll. ml-'. When a limiting ADP concentration was used, the reaction mixtures also contained 25 mM glucose, 200 units. ml" hexokinase, 0.5 p M diadenosine pentaphosphate, 5 p~ n-[l-'4C]hexylamine. Aliquots of 0.1 ml were illuminated within a microcentrifuge at room temperature. 32Pi incorporation was determined either by an extraction method (25) or by precipitation of the Pi-molybdate complex with triethylamine (26). In a few experiments, at a saturating ADP concentration using thylakoids equivalent to 0.05 mg of chlorophyll. ml" and a light intensity of 2.5 X IO6 ergs. cm-* s-I, photophosphorylation supported by cyclic electron flow was determined by following the disappearance of Pi by a colorimetric procedure (27). ApH was estimated from the extent of uptake of ['4C]hexylamine by a silicone oil microcentrifugation assay (28). Electron flow from water to ferricyanide (1 mM) was assayed by ferrocyanide production (29) in reaction mixtures similar to those used for photophosphorylation at a saturating ADP concentration, except that the pyocyanine was omitted. Chlorophyll (30) and protein (31) contents were estimated spectrophotometrically.
In a few experiments, photophosphorylation and ApH were measured by modifications of the above methods in the same reaction mixture which contained both 32Pi and ['4C]hexylamine. The microcentrifuge tubes contained 50 pl of 10% glycerol overlaid with 150 p1 of a silicone oil mixture consisting of 5.34:l.OO (w/w) of Versilube F-50/SF-96(50) overlaid with 100 pl of reaction mixture. After thylakoids were pelleted through the silicone oil, aliquots of the reaction mixture were assayed for 32P incorporation by a modification of the precipitation method (26). If dithiothreitol was present in the sample, it was quenched with excess N-ethylmaleimide before the addition of ammonium molybdate to prevent reduction of phosphomolybdic acid. After the first precipitate was removed by centrifugation, KPi was added to the supernatant to 0.6 mM resulting in a second precipitation. Using this method to analyze control samples in which thylakoids had been incubated in the dark, only 0.02% of the original 32Pi remained in the supernatant, corresponding to a calculated rate of ATP synthesis of <I pmo1.h" mg" chlorophyll. Since 32P was detected by Cerenkov radiation, the presence of ["C] hexylamine did not interfere. However, about 1% of the reaction mixture is carried with the thylakoids through the silicone oil, and because the amount of 32P in the reaction mixture is over 100 times the amount of 14C, a significant amount of 32P must be removed from the [14CC]hexylamine in the glycerol layer to allow the assay of ApH. This was done by removing the remaining reaction mixture and silicone oil from each microcentrifuge tube by aspiration, cutting the tubes above the glycerol layer, shaking each resulting tube containing the glycerol layer with 0.95 ml of HzO and 0.3 g of untreated Dowex-1 anion exchange resin for 30 min to remove labeled P,, glucose 6-phosphate and ATP, and finally treating 0.55-1111 aliquots of the supernatants with 0.25 ml of 40 mM KPi and 0.1 rnl of magnesia (0.27 M MgC12, 1.9 M NH,CI, 1.7 M NH4OH) to precipitate residual 32Pi (32). I4C was detected in the supernatants in a Beckman model LS-IOOC or LS-7500 liquid scintillation counter using lower and upper channel limits of 0 and 480, respectively, to prevent the counting of scintillations produced by "P decay. Control experiments indicate that this procedure effectively removes 32P from the glycerol layer without any loss of ["C] hexylamine.
Me-dependent ATPase activity of thylakoids was determined at 37 "C in the dark in a reaction mixture that contained 50 mM Tris-HC1 (pH 8.0) 5 mM MgCIz, 5 mM ATP, and thylakoids equivalent to 20-50 pg of chlorophyll~ml", essentially as described previously (4).
Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate was carried out as described (33), except that in some experiments, dithiothreitol was not added to the sample buffer. Relative amounts of protein in the same lane of a gel were estimated by densitometric scanning of Coomassie brilliant blue using a Quick Scan (Helena Laboratories, Beaumont, TX). Radioactivity in gel slices was determined as described (34).
The incorporation of [3H]N-ethylmaleimide into sulfhydryls exposed by dithiothreitol treatment of CF, in thylakoids was performed in thylakoids that had first been incubated with 2 mM N-ethylmaleimide in the dark to block the accessible groups on the y and t subunits (35). The thylakoids were then illuminated for 5 min or kept in the dark in the presence of 5 mM dithiothreitol and washed as described above. The thylakoid suspensions (1 mg of chlorophyll. ml-') were incubated in the dark with 0.25 mM [3H]N-ethylmaleimide at 17 "c in 50 mM Tricine-NaOH (pH 8.0), 50 mM NaCI, and 5 mM MgC1z. After 5 min, dithiothreitol was added to a final concentration of 0.28 mM and CFI purified from the thylakoid preparations by a small scale method (36). In one experiment, the buried sulfhydryl of the y subunit was labeled using a rather complicated protocol to avoid the use of the large quantities of [3H]N-ethylmaleimide necessary for optimal labeling in illuminated thylakoids. Thylakoids were treated in the light with dithiothreitol, pelleted, resuspended, and incubated with 2 mM N-ethylmaleimide in the light or dark (35), and CF, was isolated. Aliquots of the CFI preparations were desalted, denatured by boiling for 1 min in the presence of 1% sodium dodecyl sulfate, and incubated with 9 mM [3H]N-ethylmaleimide (90,000 cpm/ nmol) for 14 min at room temperature. Unreacted [3H]N-ethylmaleimide was removed and aliquots were subjected to electrophoresis. The specific activity of [3H]N-ethylmaleimide was estimated as described in the preceding report (22). The molecular weight of CF, was assumed to be 400,000 (37) in the calculation of [3H]N-ethylmaleimide incorporation.

Effect of Illumination in the Presence of Dithiothreitol on the Modification of Sulfhydryl
Croups of CF,-Since illumination of thylakoids in the presence of dithiothreitol induces a Ca2+-dependent ATPase activity in CF, extracted from the thylakoids with EDTA (4), this treatment clearly alters CF,. The disulfide bond in CF,, present in the y subunit of the soluble enzyme (18,19), may be cleaved by this treatment and support for this prediction is shown in Table I. Thylakoids were incubated with N-ethylmaleimide in the dark, followed by incubation with dithiothreitol in the dark or the light. The sulfhydryl compound was removed by washing and the thylakoids were incubated with [3H]N-ethylmaleimide in the dark. The [3H]N-ethylmaleimide was incorporated almost exclusively into the y subunit, indicating that a disulfide bond in this subunit was cleaved. Some incorporation of [3H]Nethylmaleimide into the y subunit of thylakoids exposed to dithiothreitol in the dark was also detected. Since the accessible thiol (S4) on the y subunit is blocked by pretreatment with nonradioactive N-ethylmaleimide (35), this incorporation of radioactive maleimide into the y subunit is probably the result of limited cleavage of a disulfide. This conclusion is consistent with the observation that CFI extracted from thylakoids incubated with dithiothreitol in the dark has some ATPase activity (4). The extents of activation and maleimide modification are both enhanced by illumination.
Slightly less than 1 mol of maleimide/mol of CFI was incorporated into the y subunit of CF, in thylakoids previously illuminated in the presence of dithiothreitol. If both of the thiols formed by the reduction of a disulfide bond were accessible to N-ethylmaleimide, the maximal incorporation of maleimide would be 2/CF1. Incomplete reaction could result

TABLE I Incorporation of [3HJN-ethylmaleimide into CFl in thylakoids in the dark following dithiothreitol treatment in the light or dark
Thylakoids, previously treated with 2 mM N-ethylmaleimide in the dark, were either illuminated or kept in the dark in the presence of 5 mM dithiothreitol. The thiol compound was removed by washing and the thylakoids were incubated with 0.25 mM [3H]N-ethylmaleimide (6 X lo' cpm/nmol) for 5 min in the dark at room temperature.
Dithiothreitol was added to a final concentration of 0.28 mM and CFI was isolated. Duplicate aliquots (9 pg) of each CFI preparation were subjected to polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. After staining the gel, the radioactivity in each band was determined. from the inaccessibility of one thiol to reaction with the alkylating reagent, reduction of the disulfide linkage of only about half of the CF,, or reoxidation prior to treatment with labeled maleimide. Partial reduction appears to explain, at least in part, the incomplete incorporation of [3H]N-ethylmaleimide. The y subunit of reduced and maleimide-modified CF, migrates as two components in sodium dodecyl sulfatepolyacrylamide gel electrophoresis if CF, is not pretreated with dithiothreitol immediately before electrophoresis (Fig. 1,  Lanes 2 and 4). The y subunit of the dithiothreitol-pretreated enzyme migrates as a single component with a mobility identical to the slower moving component (Fig. 1, Lanes 5 and 6 ) .
Thus, the y component that migrates faster is probably the disulfide form, whereas the more slowly migrating component is probably the reduced and alkylated form. Alkylation prevents formation of the disulfide, even in the presence of oiodosobenzoate ( Fig. 1, Lane 21 Table I, were subjected to electrophoresis using a 10% gel. CFI was isolated from thylakoids treated with dithiothreitol in the dark (Lanes I , 3, and 5)   iodosobenzoate, and electrophoresed (described in Fig. 1). The relative amounts of protein in the y l and y2 bands were estimated by densitometric scanning. Bands were cut out of gels and radioactivity Although it is difficult to obtain a precise estimate because of limits to the resolution of densitometry, approximately 60% of the y subunit of the CF, from thylakoids alkylated after illumination in the presence of dithiothreitol was in the slower migrating band (Fig. 1, Lanes 2 and 4 ) , indicating reduction of the disulfide bond in 60% of the CF,. Only about 20% of the y subunit of the CF, from thylakoids alkylated after dark incubation with dithiothreitol was in the slower migrating band. Omitting dithiothreitol pretreatment before electrophoresis does not change the amount of protein in the cy or P bands or in the sum of the y l and 7 2 bands (Fig. 1). Moreover, no new bands appear, suggesting that there are no intersubunit disulfide bonds in CF1. Dithiothreitol activates the ATPase activity of soluble CFI and cleaves the disulfide (Sl and S2) (18, 19). As shown in Fig. 2, dithiothreitol treatment of soluble and bound CFI reduces the same disulfide. Thylakoids were treated with Nethylmaleimide in the dark, activated, and then labeled with radioactive N-ethylmaleimide in the dark. CF, was purified and cleaved with trypsin, and tryptic peptides were separated by HPLC (22). Only two major peptides were labeled, S1 and S2, the same peptides shown to participate in the disulfide linkage in the soluble enzyme (22). Similar results were obtained with the enzyme from activated thylakoids labeled in the dark with 4-vinylpyridine (not shown).
The illumination of thylakoids, pretreated with N-ethylmaleimide in the dark, in the presence of [3H]N-ethylmaleimide causes a partial inhibition of photophosphorylation and incorporation of [3H]N-ethylmaleimide specifically into a sulfhydryl (S3) of the y subunit (22,35). The incomplete nature of both the inhibition and extent of incorporation has been a vexing problem. Usually, the maximal inhibition of photophosphorylation is about 70% and up to 0.7 mol of maleimide is incorporated per mol of CF,. Prior illumination of thylakoids in the presence of dithiothreitol has no effect on the ability of N-ethylmaleimide to inhibit photophosphorylation (data shown below). However, whatever is limiting reduction of the disulfide in the y subunit of CF1 in thylakoids  (Table 111). Thus, those CF, molecules that were not reduced in thylakoids were also not alkylated at the buried sulfhydryl. This suggests that alkylation of the buried sulfhydryl in thylakoids and reduction of the disulfide are limited by a common factor. For example, some of the CF, might be bound to thylakoids unable to maintain a ApH sufficient to promote energy-dependent reduction or alkylation.
Effect of Maleimides on Deactivation-After light and dithiothreitol treatment and removal of most of the dithiothreitol, thylakoids can retain for up to several hours their ability to hydrolyze ATP in the dark, provided they are briefly illuminated prior to assay (7). However, the decay of lightinduced Mg2+-ATPase activity in thoroughly washed and homogenized thylakoids, while variable, can be complete within less than 30 min. When rapid decay is observed, Mg2+-ATPase activity may be partially reactivated by illumination in the presence of dithiothreitol (Table IV). The decay in the ability of activated thylakoids to hydrolyze ATP after a second illumination is probably the result of oxidation of the dithiol groups in the y subunit to a disulfide. Oxidation may be enhanced by aeration during homogenization or by the presence of trace amounts of heavy metal cations. A variability in the heavy metal content of thylakoid suspensions may explain TABLE I11 Preferential incorporation of N-ethylrnaleirnide into reduced CF1 in illuminated thylakoids Thylakoids were illuminated in the presence of 5 mM dithiothreitol and the thiol reagent was removed by pelleting the thylakoids. The thylakoids were then exposed to 2 mM N-ethylmaleimide in the light or dark for 2 min. CF, was isolated from the two thylakoid preparations and denatured with 1% sodium dodecyl sulfate and heat (100 "C, 1 min).
[3H]N-Ethylmaleimide was added (9 mM, 90,000 cpm/nmol) and after 5 min at room temperature, unreacted maleimide was removed. Aliquots containing 9 pg of protein were subjected to electrophoresis on a 9% polyacrylamide gel in the presence of sodium dodecyl sulfate. The electrophoresis was carried out for 10 h a t 100 V. After staining, the bands were cut out and radioactivity was determined. yl and y2 refer to the slower and faster migrating y species, respectively. NEM-dark CF, and NEM-light CF, refer to the preparations from thylakoids kept in the dark or illuminated in the presence of N-ethylmaleimide, respectively. Dark-light refers to the 3H incorporation into NEM-dark CF, minus that into NEM-light CFI.

Component
[3H]N-Ethylmaleimide incorporation  Effects of N-ethylmaleimide and iodosobenzoate treatment in the dark on light-induced ATPase activity and photophosphorylation Thylakoids were illuminated in the presence of 5 mM dithiothreitol, washed to remove the thiol reagent, and incubated in the presence or absence of 1 mM N-ethylmaleimide for 2 min in the dark at room temperature. Each suspension was divided in half and 2 mM iodosobenzoate added to one-half. The resulting four thylakoid suspensions were transferred to ice, incubated for 5 min, diluted with an equal volume of a buffered sucrose solution (0.4 M sucrose, 0.02 M Tricine-NaOH (pH 8.0), 0.01 M NaCI), and centrifuged. The pellets were resuspended in a minimal volume of the buffered sucrose solution. Aliquots of each suspension were assayed for chlorophyll, lightinduced ATPase activity, with or without 5 mM dithiothreitol in the light activation mixture, and photophosphorylation with pyocyanine as the mediator of electron flow. The samples for photophosphorylation were illuminated for 90 s in the presence or absence of 5 mM dithiothreitol prior to assay. DTT, dithiothreitol; NEM, N-ethylmaleimide: IBZ, iodosobenzoate.  None  180  13  630  670  634  268  After activation  266 149  570  537  583  316  Before activation  190  12  645  626  623  316  Before and after  201 157  506  506  493 289 activation the differences in decay rates. In some experiments, the decay even in well washed thylakoids was very slow (see, for example Table V). However, o-iodosobenzoate always rapidly inactivates Mg'+-ATPase activity, whether it is present during illumination (Table V) or in the dark (not shown). CuCl, (10 /IM) treatment in the dark also rapidly inhibits light-induced Mg'+-ATPase activity of thylakoids.' Treatment of thylakoids, previously illuminated in the presence of dithiothreitol, with 2 mM N-ethylmaleimide in the dark protects light-induced Mg2+-ATPase activity from either spontaneous inactivation (Table IV) or o-iodosobenzoate inhibition (Table V), probably because alkylation of the thiol S. R. T a h r i and R. E. McCarty, unpublished observation.
groups prevents the reformation of the disulfide linkage. Although alkylation of the thiols formed by dithiothreitol reduction protects the ATPase activity from o-iodosobenzoate inhibition, it does not affect the light-dependent inhibition of photophosphorylation by N-ethylmaleimide (Table V). Modification of the accessible thiol on the y subunit affected neither phenomenon. These results reinforce the conclusion that neither the accessible nor the buried thiol is part of the disulfide bond split during dithiothreitol activation.
Illumination of intact chloroplasts in the absence of added reductants activates the ability of the lysed chloroplasts to hydrolyze ATP in the dark (10)(11)(12). Oxidants inactivate the capacity for ATP hydrolysis (Table VI). The light-induced Mg*+-ATPase activity of thylakoids obtained from illuminated intact chloroplasts lysed into 2 mM N-ethylmaleimide in the dark is protected from inactivation by iodosobenzoate, Cu2+, and ferricyanide. Thus, activation of ATPase activity by illumination of intact chloroplasts probably involves the reduction of the disulfide bond in the y subunit of CF1.
Activation of Photophosphorylation-Light and dithiothreitol treatment of thylakoids often stimulated photophosphorylation assayed at saturating light intensities. Moreover, the treatment of thylakoids after activation with N-ethylmaleimide in the dark prevented spontaneous deactivation (Table  IV). The relationship between phosphorylation rate and ApH in activated and nonactivated thylakoids was examined. At saturating concentrations of ADP and Pi, plots of the logarithm of the rate of photophosphorylation uersus ApH are linear (28). Preillumination of thylakoids in the presence of dithiothreitol shifts the line relatingphosphorylation rate and ApH to lower ApH values, so that the rate of phosphorylation a t any given ApH by activated thylakoids is 7to 8-fold higher than that by controls (Fig. 3). At a given light intensity, however, phosphorylation was stimulated by only 10-110%. This activation of photophosphorylation is protected by Nethylmaleimide treatment in the dark after light and dithiothreitol treatment, as is ATPase activity. Thylakoids that were activated, reacted with N-ethylmaleimide in the dark, and stored overnight at -70 "C showed higher rates of phosphorylation at a given ApH than those either not activated or not treated with N-ethylmaleimide (Fig. 4).
Phosphorylation catalyzed by thylakoids obtained by lysis of intact pea chloroplasts into assay mixtures is also activated  compared to that catalyzed by washed pea thylakoids (Fig. 5 ) . The relatively poor activation at high light intensities may be the result of oxidation of thiol groups to a disulfide by the ferricyanide in the reaction mixture. The internal aqueous volumes and the passive proton permeabilities of the two preparations were the same (not shown). Moreover, the maximal phosphorylation efficiencies (P/e, ratios), extrapolated from a plot of observed P/e2 ratio uersus the ratio of internal proton concentration to that of electron flow (38), were the same (1.36). Thus, activation is not caused by a change in the ratio of protons translocated per ATP synthesized.
Dithiothreitol in the assay mixture also activates photophosphorylation a t a low ADP concentration (10 PM) (Fig. 6). The effect of dithiothreitol is more pronounced a t lower ApH values. At this ADP concentration, the phosphorylation rate can approach saturation with respect to ApH. We do not observe such saturation at 1.5 mM ADP (Fig. 3), presumably because it would require higher ApH values than we obtain with our thylakoid preparations even a t saturating light intensities. These data indicate that the reduction of CF, does not increase the maximal rate of ATP synthesis observed at given ADP and P, concentrations and saturating ApH, the V$& at saturating ApH.

DISCUSSION
Our results suggest that the y subunit of CFI in washed thylakoid membranes contains one disulfide bond and two free thiols, similar to soluble CFI (18,19,22). The accessible thiol group reacts readily with alkylating reagents and appears to be unessential for activity, whereas in thylakoids the other thiol reacts with maleimides only when the membranes are energized. Alkylation of this thiol group inhibits photophosphorylation and ATPase activity.
As is the case with the soluble enzyme (18,191, reduction of the disulfide bond of the 7 subunit of CF, in thylakoids activates ATPase activity. This reduction is enhanced by illumination of thylakoids and is uncoupler sensitive (4,5,7). Conformational changes in the y subunit of membrane-bound CF, that occur during energization (2, 35) may increase the exposure of the disulfide bond to reduction by dithiothreitol. Although activation of the ATPase activity of soluble or membrane-bound CF, in the dark is nearly complete within 2 h in the presence of 50 mM dithiothreitol, activation in illuminated thylakoids is maximal within 5 min with only 5 mM dithiothreitol (4).
A subpopulation of CF1 in thylakoids exists that is susceptible to neither reduction nor alkylation of the buried sulfhydryl. This subpopulation is probably unable to undergo energy-dependent conformational changes. No CF,, alkylated a t the buried sulfhydryl but remaining oxidized, was observed. However, some CF, is reduced, but not alkylated at the buried sulfhydryl, probably because reduction can occur at a lower ApH than can alkylation. This subpopulation may explain the observation that modification of the buried sulfhydryl in thylakoids with N-ethylmaleimide not only lowers the VamPaq, for photophosphorylation but also lowers the Kzp for ADP (41). Lowering the light intensity also lowers the P , P P for ADP (42). If this subpopulation of CF, is on thylakoids that generate a ApH sufficient to allow ATP synthesis, but too low to allow the alkylation of the buried sulfhydryl, then the inhibition by N-ethylmaleimide of phosphorylation by those thylakoids able to generate larger ApH values would be expected to lower the observed Kzpp. Alternatively, CF, modified by N-ethylmaleimide at S3 might retain a reduced catalytic activity having a lower K",PP for ADP. However, preliminary results indicate that the ratio of light-dependent incorporation of N-ethylmaleimide per CF1 is very similar to the fraction of activity inhibited, suggesting that each alkylated CF, is completely inhibited.3 Reduction is sufficient to activate the ATPase activity of the soluble enzyme (19) and is associated with changes in the conformation of the y subunit (19,43). In contrast, reduction is not sufficient for the activation of the ATPase activity of thylakoids. Even when the enzyme is reduced, a period of illumination is required to induce ATPase activity (7). This induction is also sensitive to uncouplers. The requirement for energy to induce the ATPase activity in reduced membranes may be explained, at least in part, by the observation (8,9)  that ADP dissociates from the enzyme in the light and reassociates in the dark. The binding of ADP to thylakoids after a period of illumination has been correlated to the decay of ATPase activity in the dark (8,9).
In the presence of high concentrations of ADP and Pi, reduced CF1 in thylakoids catalyzes a higher rate of ATP synthesis a t a given ApH than oxidized CF,. Similarly, in the presence of a low concentration of ADP, reduced CF, catalyzes a higher rate of ATP synthesis at a given low ApH than oxidized CF1. However, reduction of CF, does not increase the V% at saturating ApH. This suggests that reduction of CF, activates ATP synthesis by promoting the interaction of internal protons with the coupling factor complex or the rate of some other elementary step which is dependent on internal and probably external proton concentrations. Such an elementary step would be rate-limiting only at relatively low ApH values. This conclusion is consistent with the results of Vallejos et al. (14) on the activation of CF, by illumination of leaves: the resulting stimulation of photophosphorylation at high light intensities was found to be greater at high ADP concentrations than at low ADP concentrations, probably because ApH is a rate-limiting factor at high light intensities and high ADP concentrations, but not rate limiting at high light intensities and low ADP concentrations (compare Figs.  3 and 6). The effects of ApH and A+ on the rates of ATP synthesis and hydrolysis catalyzed by thylakoids have been described in terms of an energy-dependent equilibrium between active and inactive states of CF, (44,451. According to this view, at a given ApH, the fraction of reduced CF, in the active conformation is greater than that of oxidized CF,. Within this framework, one can consider two limiting mechanisms by which internal protons increase the observed rate of ATP synthesis: they may interact with the coupling factor complex as substrates, three of which are translocated per ATP synthesized (38, 46, 47), or as allosteric activators, which are not translocated during turnover. Whatever the roles of protons may be, it is clear that the oxidation state of the y subunit of CF, plays a central role in regulation of photophosphorylation.
It is interesting to contrast the regulation of the chloroplast coupling factor to that of the Escherichia coli and mitochondrial coupling factors. While the chloroplast 5-subunit coupling factor is isolated as a latent ATPase, those of E. coli and mitochondria are isolated as active ATPases (48). In addition, the mitochondrial coupling factor is regulated by a soluble inhibitor protein (49). Presumably, these differences reflect the different physiological roles played by these coupling factors. Even in the absence of oxidative substrates, the mitochondrial and E. coli coupling factors remain active and catalyze ATP hydrolysis, which maintains the electrochemical proton gradient required for other processes including the transport of nutrients (50, 51). On the other hand, the inactivation of the chloroplast coupling factor in the dark may prevent wasteful ATP hydrolysis.