Uncoupling and Energy Transfer Inhibition of Photophosphorylation by Sulfhydryl Reagents*

The action of the sulfhydryl reagents, o-iodosoben- zoate and 2,2’-dithiobis(5-nitropyridine), on photophosphorylation by spinach chloroplast thylakoids has been reevaluated. Both of these compounds were previously reported to be energy transfer inhibitors of photophosphorylation, provided the thylakoids were illuminated in their presence prior to assay. We show here that the treatment of thylakoids in the light with iodosobenzoate uncouples phosphorylation from electron flow. This treatment enhances nonphosphorylating electron transport and markedly decreases the efficiency of photophosphorylation. The light-induced transmembrane pH gradient is also diminished by exposure of thylakoids to iodosobenzoate in the light. Dithiobisnitropyr- idine has been found to act either as a light-dependent uncoupler or energy transfer inhibitor. At low concentrations of this reagent, illumination elicits uncoupling, whereas at higher concentrations, energy transfer inhibition is induced. The uncoupling by iodosobenzoate and by low concentrations of dithiobisnitropyridine is largely prevented

The action of the sulfhydryl reagents, o-iodosobenzoate and 2,2'-dithiobis(5-nitropyridine), on photophosphorylation by spinach chloroplast thylakoids has been reevaluated. Both of these compounds were previously reported to be energy transfer inhibitors of photophosphorylation, provided the thylakoids were illuminated in their presence prior to assay. We show here that the treatment of thylakoids in the light with iodosobenzoate uncouples phosphorylation from electron flow. This treatment enhances nonphosphorylating electron transport and markedly decreases the efficiency of photophosphorylation. The light-induced transmembrane pH gradient is also diminished by exposure of thylakoids to iodosobenzoate in the light. Dithiobisnitropyridine has been found to act either as a light-dependent uncoupler or energy transfer inhibitor. At low concentrations of this reagent, illumination elicits uncoupling, whereas at higher concentrations, energy transfer inhibition is induced. The uncoupling by iodosobenzoate and by low concentrations of dithiobisnitropyridine is largely prevented by the prior incubation of thylakoids with N-ethylmaleimide in the dark. Under these conditions, N-ethylmaleimide was previously shown to react with a group on the y subunit and with a group on the E subunit of coupling factor 1.
The effects of these sulfhydryl reagents on photophosphorylation are compared to those of maleimides, and a model for the inhibition of phosphorylation by these reagents is proposed. Cross-linking two sufhydryl groups within the y subunit of coupling factor 1, either directly by disulfide bond formation or by bifunctional maleimides, causes thylakoids to become proton-leaky and phosphorylation is uncoupled from electron flow. In contrast, modification of a sulfhydryl, which becomes exposed only in the light, by monofunctional reagents elicits energy transfer inhibition.
Photophosphorylation, the light-dependent synthesis of ATP by illuminated chloroplasts, may be inhibited in a number of ways. Reagents that block electron transport also inhibit ATP synthesis since the generation of the transmembrane electrochemical gradient, the driving force for ATP synthesis, by the Consejo Nacional de Investigaciones Cientificas y Tenicas, Argentina (CONICET). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. is dependent upon electron flow. Chemicals or treatments that increase the proton permeability of chloroplast thylakoid membranes uncouple phosphorylation from electron flow. By decreasing the proton gradient, uncoupling agents inhibit ATP synthesis but allow electron transport to occur at high rates. In contrast, direct inhibitors of phosphorylation, often called "energy transfer inhibitors," block both phosphorylation and that portion of the electron flow that is a consequence of proton efflux linked to phosphorylation (see Ref. 1 for a discussion of inhibition of photophosphorylation).
Reagents that attack " S H groups, including N-substituted maleimides (2-4), iodosobenzoate (5), and dithiobisnitropyridine (6), partially inhibit photophosphorylation when chloroplast thylakoids are illuminated in their presence prior to the assay of photophosphorylation. No inhibition is observed when the incubation is carried out in the dark. Adenine nucleotides and uncoupling agents protect phosphorylation from inhibition by these reagents.
Bifunctional maleimides are more effective inhibitors than monofunctional ones. o-Phenylenebismaleimide (3) and dithiobisethylmaleimide (7) cross-link two groups within the y subunit of coupling factor 1 (CFI) in illuminated thylakoids. Two groups in the y subunit also react with N-ethylmaleimide in the light (8). Yet, the manner in which phosphorylation is inhibited by monofunctional maleimides differs from that of bifunctional maleimides. N-Ethylmaleimide acts as a phosphorylation or energy transfer type inhibitor (2). In contrast, o-phenylenebismaleimide (3) and dithiobisethylmaleimide (7) act at least in part as uncouplers. The uncoupling effect of the latter bifunctional compound is reversed by thiols since they reduce its disulfide bond, breaking the cross-link.
In view of the fact that bifunctional maleimides produce a light-dependent uncoupling, the observation that iodosobenzoate ( 5 ) and dithiobisnitropyridine (6) act as light-dependent energy transfer inhibitors was surprising. These reagents can oxidize vicinal thiols to disulfides (9). The same thiols in the y subunit of CFI' cross-linked by the bifunctional maleimides could be oxidized to disulfides by iodosobenzoate or dithiobisnitropyridine. In this paper, we report that iodosobenzoate acts as an uncoupler of photophosphorylation and that djthiobisnitrobenzoate can act either as an uncoupler or as an energy transfer inhibitor, depending on the conditions of illumination.

MATERIALS AND METHODS
Spinach chloroplast thylakoids were prepared from market spinach as described previously (10) except that in some experiments the homogenate was filtered through nylon mesh and the thylakoids were collected by centrifugation a t 6000 X g for 1 min. Iodosobenzoate, ' The abbreviations used are: CFI, coupling factor 1; Tricine, N-Tris(hydroxymethy1)methyIglycine. 2,2'-dithiobis(5-nitropyridine), N-ethylmaleimide, and o-phenylenebismaleimide were from Sigma. Chlorophyll (11) and photophosphorylation (12) were determined by published methods. Ferricyanide reduction was determined spectrophotometrically (13) and oxygen reduction was determined in the presence of methyl viologen with a Clark type oxygen electrode. Transmembrane pH gradients (ApH) were estimated from hexylamine distributions. ['4C]Hexylamine uptake was estimated by silicone fluid centrifugal fdtration as described previously (14) except that ["H]sorbitol was omitted and the entire glycerol-trichloroacetic acid layer was taken for determination of radioactivity. Thylakoids (20 to 100 pg of chlorophyll/ml) were treated in white light (usually about 0.25 watts/cm') or in the dark for 90 s in a mixture which contained 50 mM Tricine/NaOH (pH 8.0), 50 mM NaCl, 5 m~ MgC12, either 0.01 to 0.025 mM pyocyanine or 0.1 mM methyl viologen, and sulfhydryl reagent at the concentrations indicated in figure and table legends. In some experiments, the treated thylakoids were collected by centrifugation a t 5000 X g for 5 min and were washed in a buffered sucrose solution containing 0.  Uncoupling by o-iodosobenzoate a n d its reversal by dithiothreitol Thylakoids (100 pg of chlorophyll/ml) were illuminated for 90 s in the presence or in the absence of 1 mM o-iodosobenzoate using 20 PM pyocyanine as the mediator of electron flow. The control and treated samples were divided in half and dithiothreitol (50 mM) was added to one aliquot of each sample. The thylakoids were then washed with the buffered sucrose solution and phosphorylation and electron transport in the presence of K:jFe(CNh were determined. Rates are expressed as pmol of Fe(CN)S"-reduced or of P, esterified h" mg of chlorophyll". The assay illumination time was 90 s.

RESULTS
Uncoupling of Photophosphorylation by Iodosobenzoate-The conclusion ( 5 ) that iodosobenzoate treatment of thylakoids in the light causes energy transfer inhibition of photophosphorylation was based primarily on the observation that this treatment inhibited phosphorylating electron flow. In more than 10 experiments, however, we were unable to obtain strong inhibition of phosphorylating electron flow from water to methyl viologen. The inhibition by 1 mM iodosobenzoate ranged from 0 to 2055, even though phosphorylation enhanced electron flow by 2-fold or more. This slight inhibition is likely due to a direct effect of iodosobenzoate on the assay of electron flow. Iodosobenzoate is reduced by illuminated thylakoids at a rate of 22 pmol h" mg of chlorophyll". Moreover, removal of the unreacted iodosobenzoate by washing reverses the inhibition of electron flow but not that of photophosphorylation (Table I). Uncoupled, phosphorylating, and nonphos-phorylating electron transports are inhibited by iodosobenzoate, especially at high concentrations (Fig. 1). Thus, it appears that iodosobenzoate and light treatment does not specifically inhibit phosphorylating electron flow.
Whether or not iodosobenzoate was present during the assay, the illumination of thylakoids in its presence prior to assay enhanced the rate of basal (nonphosphorylating) electron flow (Fig. 1). This stimulation occurs at concentrations of the reagent that also cause inhibition of photophosphorylation (5). Thus, the iodosobenzoate and light treatment of thylakoids appears to uncouple. This conclusion is confirmed by the observation that this treatment causes a marked decrease in the phosphorylation efficiency (P/e2 ratio), as shown in Table 11. As expected, dithiothreitol largely reversed the inhibition of phosphorylation.
Uncoupling by the incubation of thylakoids with iodosoben-  zoate in the light is also indicated by its effects on the transmembrane pH gradient (ApH). A significant decrease in ApH was observed only when the thylakoids were illuminated in the presence of the reagent (Table III), indicating that the treated membranes are somewhat proton-leaky. Since the prellumination of thylakoids in the presence of iodosobenzoate lowers the P/e, ratio, stimulates basal electron flow, and enhances the proton permeability of thylakoids, it is clear that the treated membranes are uncoupled.
The development of the inhibition of photophosphorylation by N-ethylmaleimide is strongly dependent on ApH (14). The ApH generated by pyocyanine-dependent cyclic electron flow at high light intensity is greater than that supported by noncyclic electron flow (15). Thus, if iodosobenzoate inhibition responds to ApH in a manner similar to that of N-ethylmaleimide inhibition, greater inhibition should be observed when the preillumination is carried out in the presence of pyocyanine. This has been shown to be the case. For example, in one experiment, illumination of thylakoids in the presence of 1 mM iodosobenzoate and pyocyanine caused a 77% inhibition of pyocyanine-supported ATP synthesis, but only a 25% inhibition was observed when the preillumination was performed in the presence of methyl viologen. Moreover, a comparison of the data shown in Tables I and I1 reveals that preillumination of thylakoids in the presence of pyocyanine causes more pronounced uncoupling by iodosobenzoate than when the preillumination is carried out with methyl viologen as the mediator. Rates are expressed as pmol of ferricyanide reduced h" mg of chlorophyll".
iodosobenzoate in the light largely prevents iodosobenzoate uncoupling. This treatment also prevents uncoupling by low concentrations of bifunctional maleimides (3, 7) which crosslink a maleimide-reactive accessible group on the y subunit to one that becomes reactive in the light. In view of the fact that N-ethylmaleimide treatment blocks the inhibition by iodosobenzoate and by bifunctional maleimides, it seems likely that disulfide bond formation in the y subunit is the cause of the light-dependent inhibition of ATP synthesis by iodosobenzoate. Uncoupling a n d Energy Transfer Inhibition by Dithiobisnitropyridine-Dithio compounds can form mixed disulfides between the compound and a protein thiol or oxidize vicinal    (17). By analogy to the action of iodosobenzoate and bifunctional maleimides on one hand, and monofunctional maleimides on the other, the formation of a disulfide in y should uncouple, whereas the formation of a mixed disulfide should give energy transfer inhibition. Depending on the concentration of dithiobisnitropyridine present, either uncoupling or energy transfer inhibition is elicited by illumination ( Fig. 2A). Low concentrations strongly uncouple, giving rise to high rates of electron flow and severely inhibited ATP synthesis. At higher concentrations, phosphorylation was actually less inhibited and electron transport was slowed to near the basal level. The pretreatment of thylakoids with N-ethylmaleimide prevented the uncoupling but not the energy transfer inhibition by dithiobisnitropyridine (Fig. 2B). Both uncoupling and energy transfer inhibition by this reagent require illumination for their onset and are reversed by illumination of thylakoids in the presence of 10 mM dithiothreitol (not shown). Basal electron flow was strongly stimulated by low concentrations of dithiobisnitropyridine, but as previously reported (6), higher concentrations have little effect (Fig. 2 A ) . The uncoupling by dithiobisnitropyridine was overlooked in previous experiments (6) since methyl viologen rather than pyocyanine was used during preillumination. Preillumination in the presence of pyocyanine causes stronger inhibition by " S H reagents.
These results suggest that dithiobisnitropyridine may form a disulfide bond in the y subunit at low concentration and mixed disulfides at higher concentration. Phosphorylation by thylakoids that had been incubated with dithiobisnitropyridine in the dark, followed by removal of the excess reagent by washing, becomes inhibited when the thylakoids are illuminated (not shown). The reagent probably reacts in the dark with the accessible thiol on the y component. In the light, another thiol on the y subunit may approach this mixed disulfide and displace the reagent, resulting in the formation of an intrapeptidic disulfide. When high concentrations of dithiobisnitropyridine are present in the medium, the reagent may react with the thiol that is exposed in the light to form a mixed disulfide more rapidly than the reaction of this thiol with the mixed disulfide between the reagent and the accessible thiol. The observation ( Table V) that uncoupling is partially prevented when thylakoids, previously treated with 2 PM dithiobisnitropyridine in the dark, were incubated in the light with either 100 PM dithiobisnitropyridine or 2 r m Nethylmaleimide is in accord with this proposal.

DISCUSSION
We can now arrive a t a general model for the action of sulfhydryl reagents on photophosphorylation. Our interpretation of how iodosobenzoate and dithiobisnitropyridine inhibit is shown in Fig. 3. Two " S H groups on the y subunit of CF, are shown. One is accessible to attack by "SH reagents in the dark, whereas the other is exposed to reaction only in the light. Modification of the accessible group does not inhibit ATP synthesis. In contrast, the reaction of the group that became accessible in the light causes an energy transfer type of phosphorylation inhibition. However, when these -SH groups are cross-linked, either by disulfide bond formation or by bifunctional maleimides, uncoupling occurs. In this discussion, we will assume that iodosobenzoate and dithiobisnitropyridine exert their effects through their interaction with " S H groups on the y subunit of CFI. This assumption is reasonable in view of the close similarity between these reagents and maleimides in the manner in which they inhibit photophosphorylation. The development of the inhibition of phosphorylation by these reagents requires light and is prevented by nucleotides and uncouplers. Moreover, the reaction of a group in the y subunit with N-ethylmaleimide in the dark prevents light-dependent uncoupling by bifunctional maleimides (3, 7) and by iodosobenzoate and dithiobisnitropyridine. Mono- (8) and bifunctional ( 3 , 7 ) malemides react with groups in y. Finally, iodosobenzoate and light treatment of thylakoids causes the formation of a new disulfide in the y as well as one in the p subunit (16).
Bifunctional maleimides, iodosobenzoate, and low concentrations of dithiobisnitropyridine are light-dependent uncouplers of photophosphorylation. Bifunctional maleimides partially cross-link two groups in the y subunit and this crosslinking causes increased proton permeability and, therefore, uncoupling. Recently, 0-, m-, and p-phenylenebismaleimides were found to act as light-dependent energy transfer inhibitors at high concentrations.2 This inhibition is not alleviated by the prior treatment of thylakoids with N-ethylmalemide in the dark. When these bifunctional maleimides are present in high concentration (>20 PM), they probably react with the group in the y subunit which is exposed in the light more rapidly than the maleimide group bound to the accessible group in the y component. Therefore, cross-linking would not take place under these conditions. Iodosobenzoate or low concentrations of dithiobisnitropyridine probably cause the formation of a disulfide bond within the y subunit. N-Ethylmaleimide blocks the uncoupling by these reagents and by bifunctional maleimides, suggesting that the accessible " S H group on the y subunit is likely to be part of this disulfide. Since this subunit contains three (18) or four " S H groups (19) and since the specificity of maleimides for " S H groups is not absolute, we cannot conclude that bifunctional maleimides cross-link the same "SH groups that form the inhibitory disulfide bond. Further experiments, including peptide mapping of the y subunit, which is in progress in the Ithaca laboratory, will be required to establish which groups react.
'' J . V. Moroney, unpublished observations. Nonetheless, the possibility that the same groups are involved in cross-linking and disulfide bond formation is intriguing. p -Phenylenebismaleimide ( 3 ) is an effective inhibitor of photophosphorylation and provides a cross-link span of 12 to 14 A.
It would be remarkable if the same "SH groups were directly cross-linked by disulfide bond formation since the consequences of these two kinds of cross-linking appear to be so similar.
Although cross-linking within the y subunit clearly uncouples, reaction of presumably the same "SH groups in the y subunit with monofunctional reagents causes predominately an energy transfer inhibition. Monofunctional maleimides, including N-ethylmaleimide (2, 4) and even the bulky N-naphthylmaleimide2 are largely energy transfer inhibitors. Dithiobisnitropyridine, at high concentrations, may form mixed disulfides with the accessible " S H and with that which is exposed by illumination. The possibility that this reagent causes the formation of disulfide bonds elsewhere in CF, cannot be ruled out at present. How modification of an " S H group(s) in the y component by monofunctional reagents inhibits phosphorylation is not understood. In view of the observation that cross-linking groups in the y subunit affects proton permeability of thylakoid membranes (3, 7), this group(s) may play a role in proton translocation by the ATPase complex. Moreover, disulfide interchange may be required to convert CF, in thylakoids to a form that is active in photophosphorylation. Heat treatment of soluble CFI activates Ca"-ATPase (20) and causes the formation of a disulfide from two " S H groups in a at the expense of a disulfide in y (19). If a similar disulfide interchange is required for the conversion of CFI to its active form, the effects of monofunctional " S H reagents on photophosphorylation may be readily understood.