ABSTRACT STUDIES ON THE SITES OF PHOTOPHOSPHORYLATION IN ISOLATED CHLOROPLASTS BY Syamapada Saha This thesis deals with phosphorylation sites in isolated chloroplasts. The results of the investigation are presented in three sections. The first section reports on the non-linearity of phosphorylation at low light intensities and the effects of various uncouplers on this lag. In the absence of uncouplers, the rate of phosphorylation is directly pro- portional to the light intensity except at extremely low intensities. Even at the lowest intensities usable there is some phosphorylation. There is thus no evidence of a critical intensity which must be exceeded before phos- phorylation can begin. The existence of a critical light intensity for phosphorylation, therefore, cannot be used to support the chemiosmotic mechanism or any other mechanism. Uncouplers belong to at least two distinct categories. Some induce non-linearity (e.g. carbonyl- cyanide 3-chlorophenylhydrazone while others show Syamapada Saha uniform inhibition at all light intensities (e.g. methyl- amine). It seems unlikely that any unified theory of the mechanism of uncoupler action could explain these obser- vations and therefore it seems that there are at least two completely different mechanisms of uncoupling. The second section of the thesis deals with the stoichiometric relation of phosphorylation to electron transport. This stoichiometry probably reflects the num- ber of sites of phosphorylation. Particular emphasis has been placed on the chemical nature of the product or products of phosphorylation. It has been shown by paper chromatography that virtually all of the labelled phos- phate appears in ATP. Trace amounts of label in ADP certainly are derived from ATP through the activity of the adenylate kinase (myokinase) known to be present. There is no measurable discrepancy between the amount of organic phosphate formed (as measured by the incorporation of radioactive orthophosphate into the non-orthophosphate fraction) and the amount of orthophosphate disappearing. Regardless of the amount of ferricyanide reduced the ratios of ATP molecules formed to electron pairs trans- ferred regularly exceed 1.0 by an amount which is well outside any errors introduced by known factors. This suggests that there may be two sites of phosphorylation in the non-cyclic pathway of electron flow from water to ferricyanide. Syamapada Saha 'The third section of the thesis describes newly discovered phenomena associated with the reduction of the oxidized form of p-phenylenediamine (presumably mostly p-benzoquinonediimide). By using this substance (PDox) as electron acceptor further indications of two sites of phosphorylation have been obtained. The rate of electron transport during reduction of PDox is very high whether or not phosphorylation occurs. Nevertheless PDox is not an uncoupler since it does not inhibit phosphorylation, even at very high concentrations. Phosphorylation sup- ported by the reduction of PDOX differs from ferricya- nide-supported phosphorylation in a number of ways. It has a broader pH optimum, being as fast at pH 7.0 as above pH 8.0 whereas the optimum for ferricyanide-sup- ported phosphorylation is close to pH 8.5. Nevertheless at pH 7.0 the rate of phosphorylation at low light inten- sities (ie. the quantum efficiency) is only half of that of ferricyanide phosphorylation. “At low pH's PDox-phos- phorylation is appreciably more sensitive to DCMU and antimycin than is ferricyanide phosphorylation. All of these differences gradually disappear as the pH in- creases. 'Phosphorylation supported by electron transport from exogenous donors such as p-phenylenediamine or ben- zidine' iscompletely suppressed by PDox' On the basis of these observations it is tentatively proposed that there is a site of phosphorylation associated with the oxidation Syamapada Saha of water which is not operative when exogenous donors replace water as sources of electrons, and another site of phosphorylation between photosystem II and I. Ac- cording to this proposal PDox at pH 7.0 accepts electrons between these two sites. In so doing it totally inhibits phosphorylation when the water oxidation site is not operative, as when exogenous electron donors replace water. STUDIES ON THE SITES OF PHOTOPHOSPHORYLATION IN ISOLATED CHLOROPLASTS BY Syamapada Saha A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1970 \l\r r/ PX...) ., 7v/~7o To My Parents ii ACKNOWLEDGEMENTS The author wishes to express his deep and sin- cere appreciation to Dr. N. B. Good for his guidance and encouragement during this work. Appreciation is also extended to Drs. R. S. Bandurski, J. E. Varner and P. Wolk for acting as members of the guidance committee. Very special thanks go to Dr. S. Izawa for help and en— couragement in more than one way throughout the course of this study. Thanks are also due to Dr. C. J. Pollard and Mrs. Karen Melcher for their help. This work was sup- ported by the National Science Foundation of the United States through grants GB 4568 and GB 7940 to Drs. Good and Izawa. U. S. Educational Foundation in India awarded a Fulbright Travel Grant. Sincere gratitude is extended to the authors wife, Ranjana and his daughter, Dipa without whose understanding and sacrifice this work would have been impossible. iii TABLE OF CONTENTS DEDICATION O O O O O O O O O O I O O O O O O O 0 ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . LIST OF TABLES O O O O O O O C O O O O O O O C . LIST OF FIGURES O O O C O O O O C O C O O O O 0 LIST OF ABBREVIATIONS . . . . . . . . . . . . . SECTION I NATURE OF A PHOSPHORYLATION SITE INTRODUCTION 0 O O O I O O I O I I O O O O O O I Theories of the Mechanism of Phosphorylation . . . . . . . . . . . . . Evidence Supporting the Proposed Mechanisms of Phosphorylation . . . . . . . . . . . . MATERIALS AND METHODS . . . . . . . . . . . . . Preparation of Chloroplasts . . . . . . . . Purification of Na2H3 P04 Solution . . . . . Preparation of Other Solutions . . . . . . . Phosphorylation Reaction . . . . . . . . . . Estimation of Organic Phosphate . . . . . . RESULTS 0 O O O O O O O O O I O O O O O O O O O Photophosphorylation at Different Light Intensities . . . . . . . . . . . . . . Effect of Uncouplers on the Non- Linearity of Phosphorylation with Increasing Light Intensities . . . . . . . . . . . Effect of pH on the Non-Linearity of Photophosphorylation . . . . . . . . . . . iv Page ii iii vii viii ll 12 12 13 14 16 16 18 19 Effect of Ions on the Non-Linearity of Photophosphorylation . . . . . . . . . DISCUSSION . . . . . . . . . . . . . . . . . SECTION II THE STOICHIOMETRY OF NON-CYCLIC PHOTOPHOSPHORYLATION INTRODUCTION 0 O O C O O O O O O O O O O O 0 MATERIALS AND METHODS . . . . . . . . . . . Colorimetric Determination of Ortho- phosphate Uptake . . . . . . . . . Paper Chromatography of the Products of Photophosphorylation . . . . . . . . . RESULTS 0 O O O O O O O O O I O O O O O I O Labelling of ATP and ADP During Photophosphorylation . . . . . . . . . Simultaneous Estimation of Organic Phos- phate Formed and Orthophosphate Taken Up During Photophosphorylation . . . . Apparent P/e2 as a Function of Ferri- cyanide Concentration . . . . . . . P/ez as a Function of Different Electron Acceptors . . . . . . . . . . . . . . Quantum Efficiencies of Photophosphoryl- ation As A Function of Different Electron Acceptors . . . . . . . . . . DISCUSSION 0 O O O O O O O O O O O O O O O 0 SECTION III THE NUMBER AND LOCATION OF PHOSPHORYLATION SITES IN THE PATHWAY OF NON-CYCLIC ELECTRON FLOW INTRODUCTION 0 O O O O I O O O O O O O O O O Page 23 27 33 38 39 40 42 42 46 51 52 56 58 62 The Pathway of Electrons From Water to NADP . . . . . . .'. . . . . . . . . . The Number of Phosphorylation Sites in the Non-Cyclic Electron Transport Pathway . . . . . . . . . . . . . . . . . Location of the Photoph sphorylation Sites . . . . . . . . . . . . . . . . . . MATERIALS AND METHODS O O O O I O O O O O O O 0 RESULTS 0 O O I O O O O O O O O O O O O O O O O PrOperties of Different Electron Acceptors in Relation to Electron Transport and Phosphorylation in Chloroplasts . . . . . Stimulation of Electron Transport and Photophosphorylation by PDOX (1,4- Benzoquinonediimide) . . . . . . . . . . Time Course of the Reduction of PDox and of Accompanying Phosphorylation . . . . . PDOX-Phosphorylation as a Function of pH . Sensitivity of PDOX-Phosphorylation to DCMU and Antimycin . . . . . . . . . . . Apparent Km for Phosphate of PDOX Phosphorylation . . . . . . . . . . . . . Quantum Efficiency of PDox-Phosphoryla- tion . . . . . . . . . . . . . . . . Photophosphorylation in Tris-Washed Chloroplasts and the Possibility of a Phosphorylation Site on the Water Oxidation Side of Photosystem II . . . . DISCUSSION 0 O O O O I O O O O O O O O O O O O LITEMTURE CITED 0 O O O O O O O O O O O O C 0 vi Page 62 7O 72 79 81 81 82 86 89 93 93 97 99 107 112 LI ST OF TABLES Table Page I(a). Effect of ions on phosphorylation at high and low light intensities . . . 25 I(b). Effect of different intensities of light on the CCCP-induced non- linearity in low ionic media . . . . . 26 II(a). Distribution of labelled phosphate in the ADP and ATP of phosphorylation reaction mixtures . . . . . . . . . . . 44 II(b). Distribution of labelled phosphate in the glucose-6-phosphate and combined "ADP and ATP" of photophosphoryla- tion reaction mixtures . . . . . . . . 45 III. Comparison of the amount of labelled organic phosphate formed and the amount of orthophosphate removed from the medium during photophos- phorylation . . . . . . . . . . . . . . 50 IV. Stoichiometry of photophosphorylation using different electron acceptors . . 55 V. Electron transport and phosphoryla- tion with various electron acceptors . . . . . . . . . . . . . . . 83 VI. Stimulation of electron transport and photophosphorylation by 1,4—ben- zoquinonediimide (PDox) . . . . . . . . 85 VII. Effect of tris-washing of chloroplasts on photophosphorylation . . . . . . . . lOl VIII. Suppression of photophosphorylation by PDox in tris-washed chloro- plasts . . . . . . . . . . . . . . . . 102 vii Figure l. 10. LIST OF FIGURES Page Photophosphorylation at different light intensities . . . . . . . . . . . . 17 Effects of uncouplers on non-linearity of photophosphorylation at low light intensities I . . . . . . . . . . . 20 Effects of uncouplers on non-linearity of photophosphorylation at low light intensities II . . . . . . . . . . 21 Effects of uncouplers on non-linearity of photophosphorylation at low light intensities III . . . . . . . . . . 22 Effect of pH on the non-linearity of photophosphorylation at low light intensities . . . . . . . . . . . . . . . 24 Effect of dark incubation on the dis- tribution of radioactive phosphate after the photophosphorylation reaction . . . . . . . . . . . . . . . . 47 Typical standard curve for estimation of orthophosphate . . . . . . . . . . . . 49 Effects of different concentrations of ferricyanide on the apparent stoi- chiometry of phot0phosphorylation . . . . 53 Photophosphorylation as a function of light intensities with different electron acceptors . . . . . . . . . . . 57 Schematic representation of the path- ways by which electrons are thought to be transferred during the photo- reactions of isolated chloroplasts . . . 63 viii Figure 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. Proposed site of action of the oxidized form of p-phenylenediamine (PDOX) and the proposed second site of phosphor- ylation close to photosystem II . . . Time-course of the reduction of the oxidized form of p-phenylenedi- amine (PD x) and the accompanying phosphoryIation . . . . . . . . . . . Time-course of phosphorylation with different concentrations of PDox . . Inhibition of cyclic photophosphoryl- ation by PDOX O O O I C I O O O O O 0 Effect of pH on the photophosphoryla- tion reactions associated with the reduction of PDox and ferricyanide . Sensitivities of different phosphoryla- tion reactions to DCMU . . . . . . . Sensitivities of electron transport and phosphorylation to antimycin with different electron acceptors . . Apparent km's for phosphate of the phosphorylation reactions with PDox, ferricyanide and PMS as electron acceptors . . . . . . . . . . . . . . Phosphorylation rates during the reduc- tion of ferricyanide and PDOX as functions of light intensity . . . . Inhibition of phosphorylation in tris- washed chloroplasts by PDox . . . . . Inhibition of phosphorylation in tris- washed chloroplasts by prolonged illumination using p-phenylene- diamine as donor . . . . . . . . . ix Page 78 87 88 90 91 94 95 96 98 105 106 CCCP: DCMU: DCPIP: PMS: PD or p-PD: PDOX: LIST OF ABBREVIATIONS Carbonylcyanide 3-chloropheny1hydrazone 3-(3,4-dichlorophenyl)-l,l-dimethylurea 2,6-dichlorophenolindophenol N-methylphenazonium methosulphate para—Phenylenediamine Oxidized form of para-Phenylenediamine pre- sumably primarily para-benzoquinonediimide SECTION I NATURE OF A PHOSPHORYLATION SITE INTRODUCTION Theories of the Mechanism of Phosphorylation Any consideration of a phosphorylation site would be incomplete without a discussion of how phos- phorylation is linked to electron transport. In spite of great efforts, the mechanism of phosphorylation has remained one of the very imperfectly understood problems of biochemistry. Following are brief outlines of the theories proposed to explain the coupling of electron transport to phosphorylation. "Chemical" Coupling The essence of this hypothesis is the postula- tion that a component of the electron transport chain itself forms a high energy complex (BWI in the scheme below) when oxidized. In analogy with the mechanism of substrate-level phosphorylation Slater (1) proposed this hypothesis in 1953. The central idea can be expressed by the following five steps. AH2 + B -------------- BH2 + A (a) 3H2 + I -------------- BHz-I (b) BHz—I ---------------- B m I + CH2 (c) B m I + P ------------ P m I + B (d) P m I + ADP ---------- ATP + I (e) Net: AH2 + c + ADP + P ------ A + CH2 + ATP Here A, B and C are components of the electron transport chain and I represents a protein (coupling factor ? ). The important assumption is that the re- duced form of an electron carrier (B) combines with I and only BHZ—I can donate electron to the next electron car- rier in the chain (C). By donating elecrrons to C a "squiggle" or high-energy bond is formed between B and I. In the coupled system, this high-energy complex is broken down in steps such as (d) and (e) above to liber- ate the electron carrier (B) and the coupling factor (I) and, whenever ADP and Pi are aVailable, the energy is utilized to make ATP. The free electron carrier (B) then accepts an electron (or two), combines with the coupling factor and the cycle continues. Uncoupling, in this mechanism, would mean the breakdown of one or more of the high-energy intermediates, e.g. by hydrolysis of an an- hydride bond. Membrane integrity may or may not be essen- tial for coupling of phosphorylation to electron transport by this mechanism. In this scheme, a phosphorylation site would be a place in the electron transport chain where a high- energy bond is formed such as in the oxidation of BHZ-I. In spite of great efforts (2) to isolate and identify high-energy intermediates, no such compounds have yet been found. Conformational Coupling Mitchell 3E 31. (3) and Green (4) have proposed that a change in the conformation of the membrane may constitute by itself a high-energy state which can be coupled to ATP synthesis if ADP and Pi are available. From evidence provided by electron micrographs of mito- chondria in various states, Green has concluded that the mitochondrial cristae can exist in three different condi- tions -- non-energized, energized, and energized-twisted. The energized state can be obtained when a non-energized mitochondrion is either allowed to transport electrons or to hydrolyze ATP. The energized-twisted state can be obtained from the energized state by addition of inor- ganic phosphate. The energized-twisted state is dis- charged by ADP, by divalent metal ions and by uncouplers. Thus the "energized" state is formally equivalent to BNI of the "chemical" mechanism while the "energized-twisted" state is analogous to PNI. In spite of Green's (4) attractive electron micro- graphic evidence in support of this mechanism, Weber and Blair (5) have recently shown that a highly condensed morphology of mitochondria can be formed by ADP alone. They also have shown that ATP can be formed without a change in the morphology of mitochondria and, therefore, they have suggested that the observed morphological changes may be a reflection of the osmotic changes in the environment in the different metabolic states. It would also seem probable that the ionic constitution of the medium which results when things like ATP and ADP are added could account for some of these morphological changes. Furthermore, the known movements of ions asso- ciated with electron transport and ATP hydrolysis could influence the local ion concentrations in such a way as to change the conformational states of the mitochondria. See "chemiosmotic" hypothesis below. "Chemiosmotic" Coupling The essence of this hypothesis is that electron transport takes place across an anisotropic membrane in such a way as to cause the translocation of hydrogen ions. The resulting hydrogen ion gradient and/or mem- brane potential is then used across the same anisotropic membrane to extract water from ADP and Pi, thus yielding ATP . For a long time it has been known that ATP hydrolysis can bring about ion uptake by cell membranes (6). Besides, all phosphorylating membranes are found to enclose a space and that the ATP-hydrolyzing enzyme (ATPase) is localized in the membrane (7). Based pri- marily on these observations, Mitchell (8) proposed in 1961 that ATP synthesis in biological membranes may be considered as a reversal of ATP hydrolysis, using the same enzyme. ‘This scheme, known as the chemiosmotic hypothesis, was further developed and elaborated by him in 1966 (9). The theory assumes that during ATP hydrolysis the elements of water (H+ and OH-) are added from two differ- ent sides of the membrane (outer and inner) and that the process is reversible. Essentially then the scheme pro- poses that if by any means H+ and OH- ions can be built up on the two sides of a relatively impermeable membrane, ATP can be synthesized if ADP and Pi are available. The chemiosmotic hypothesis further assumes that the hydrogen carriers (such as quinones) and the pure electron carriers (such as cytochromes) alternate in the electron transport chain in such a way that the hydrogen carriers are localized on one side of the membrane and the electron carriers are localized on the other side. As a result, during electron transport there is a net flow of protons across the membrane. The resulting gradient of protons across the membrane is used up by the anisotropic reversible ATPase to make ATP whenever ADP and P1 are available. An uncoupler, in this scheme, would be anything that can dissipate the proton gradient either by making holes in the membrane, or by reacting directly with pro- tons, or by becoming a part of the membrane and acting as a "ferry boat" for proton transport (e.g. weak acids which can accept protons on one side of the membrane and give them up on the other side). A phosphorylation site according to this hypoth- esis would be any redox centre in the electron transport chain where protons are liberated on one side of the mem- brane and released on the other side. Evidence Supporting the Proposed Mechanisms of’Phosphorylation "Chemical" Hypothesis As already pointed out no high-energy intermedi- ate of the electron transport chain has been found. However, this fact is not as damaging to the theory as one might think since (a) high-energy compounds are likely to be rather unstable and (b) the amount of such high-energy intermediates to be expected, if they are stoichiometric with the respiratory or photosynthetic electron carriers, is exceedingly small. Neither has Boyer (2) been able to find any high-energy phosphoryl- ated intermediates of ATP synthesis but again this is not a very convincing kind of negative evidence. Nevertheless, at the‘moment the only reason for enter- taining the "chemical" hypothesis is its analogy to the better understood substrate level phosphorylations. Conformational Hypothesis Little can be said about this possible mechanism. The evidence for and against it are both minimal. "Chemiosmotic"pHypothesis The beauty of the chemiosmotic hypothesis is that quite a few of the postulates can be put to experi- mental test. Thus, one would expect that on illumination chloroplasts would change the pH of an unbuffered medium. This was experimentally shown by Neumann and Jagendorf (10). They found a rise in pH in light and a correspond- ing fall in pH in the dark. They also found that the pH change could be discharged either by a phosphorylating condition or by the known uncouplers of photophosphoryl- ation. All these observations are in very good agreement with the chemiosmotic hypothesis. Shen and Shen (11) and Jagendorf and Hind (12) independently demonstrated in a two-stage experiment that a high-energy state (Xe) can be formed by illuminating chloroplasts in absence of ADP and Pi. This high-energy state can be used to form ATP in the dark when ADP and Pi are added. The pH optimum, the kinetics of formation and decay of 'Xe' and of the hydrogen ion gradient are so similar that one is tempted to think of the hydrogen ion dradient as representing the high-energy state (Xe). Another critical test of chemiosmotic hypothesis has been made by Jagendorf and Uribe (13) in their classical "acid-bath" experiment. They found that chloroplasts treated for a very short time at low pH (about 3.8) and then transferred to a medium containing ADP, Pi and Mg++at pH 8.0 were able to synthesize sig- nificant amounts of ATP. Furthermore, the amount of ATP synthesized depended on the entry of a permeant acid such as succinate which increased the internal hydrogen ion pool. The chemiosmotic hypothesis demands that a mini- mum "critical" proton or electrochemical gradient should be formed before ATP can be synthesized. Turner 33 31. (14) observed that although NADP reduction in isolated chloroplasts was proportional to the intensity of light, ATP formation showed a pronounced lag. They showed that the extent of the non-linearity increased with higher concentrations of chlorophyll.' Dilley (personal communi- cation) and Schwartz (15) have also observed pronounced non-linearity in photophosphorylation at low light in- tensity. Sakurai gt 21. (16) also found that photophos- phorylation was not proportional to the light intensity 'at low light, although the electron transport was linear. 10 As some of the above observations were based on indirect measurements of phosphorylation and as most of the chloroplast preparations had poor phosphorylation rates, we undertook a reinvestigation of the matter of the non-linearity of photophosphorylation. The rest of this section of the thesis will deal with phosphoryla- tion as a function of light intensity. MATERIALS AND METHODS Preparation of Chlor0plasts Chloroplasts were isolated from market spinach (Spinacia oleracia L.) by the following procedure: petioles and the greater part of the mid vein were re- moved from washed leaves which were then ground for about 10 seconds in a Waring blendor with a 0.3M NaCl, 0.05 M Tricine-NaOH and 0.003M MgCl mixture (pH 7.5). The 2 homogenate was squeezed through 8 layers of cheese-cloth and then centrifuged at about 2000::g for 5 minutes. The sediment was resuspended in 0.1M Sucrose, 0.005M Tricine- NaOH, 0.002M MgCl and 0.01M KCl (pH 7.3). The suspen- 2 sion was centrifuged briefly (about 15 seconds) at 2000 XCJtO remove cell debris and intact cells and then centri- fuged again at about 2000)560 nm) or by narrow band red light (650 nm interference filter). Light was obtained from a 500 W slide projector. The light was passed through a 1-1iter round-bottomed flask containing 0.2% copper sulfate 14 solution in water which acted as a combined heat filter and a condenser lens. Combinations of neutral filters were used to control various intensities of light. Light intensity was measured by a wavelength-independent radi- ometer (YSI-Kettering, Model 65). The reactions were run for 2 to 3 minutes and an aliquot of 1.8 ml was pipetted into a test tube containing 10 ml of 10% perchloric acid saturated with butanol-ben- zene (1:1, v/v). Each sample in perchloric acid was assayed for organic 32 P04, in this case glucose-G-phosphate derived from ATP and glucose through the action of hexo- kinase present. Estimation of-Organic Phosphate Organic 32P was estimated by a method adapted from the methods of Avron (18) and of Nielsen and Lehninger (19). The details of the procedure are as follows: 1.2 ml of acetone, 1.0 m1 of 10% ammonium molybdate and 10.0 ml of butanol-benzene (1:1, v/v) saturated with 10% per- chloric acid were added to the perchloric acid diluted sample described above. The two-phase mixture was stirred thoroughly by an-up and down movement of a glass rod flat- tened at the end. After separation of the layers, the upper organic layer containing the phosphomolybdate was carefully sucked out of the test tube through a pasteur pipette connected to a vacuum line with two traps. The 15 aqueous solution was then filtered into another test tube through a Whatman-4 filter paper soaked immediately before use with 0.5 ml of distilled water. The wet filter paper held back chloroplasts and any minute droplets of organic phase remaining suspended in the aqueous phase. To the clear filtrate was added another 0.1 m1 of 10% ammonium molybdate and 7.0 ml of butanol-benzene mixture. The mix- ture was again stirred and allowed to stand until the layers separated. The upper layer was removed carefully as before, the aqueous layer containing the organic 32P was poured into a Geiger-Muller immersion tube (20th Cen- tury Electronics, England) and the radio-activity was then measured with a Nuclear-Chicago Scaler (Model 186). The specific activity of the organic phosphate formed was determined by comparing with the counts obtained from a standard solution of identical volume containing a known amount of the same radioactive phosphate. Counts were taken for from 3 to 5 minutes depending on the amount of radioactivity. The glucose-6-phosphate formed (plus any other non-orthophosphate phosphorus-containing substances similarly formed) was computed after subtracting the small number of counts in the dark control. RESULTS Photophosphorylation at Different ' L1ght Intensities We failed to observe a "dead-space" in phospho- rylation irrespective of the nature of electron acceptor, in spite of numerous reported claims (15, 16) that there is a "critical" light intensity below which absolutely no phosphorylation occurs. Presented in Figure 1 is a typical light intensity photophosphorylation curve ob- tained with methylviologen as electron acceptor. As a matter of fact, it is extremely difficult to observe any non-linearity at all in phosphorylation when the data are plotted as a curve which includes saturation. How- ever, by increasing the sensitivity of the method (very high specific activity of the phosphate and very low light intensities) it was possible to observe a slight non-linearity at the very lowest intensities. These more refined results are presented in the inset in Figure 1. In the latter studies it was necessary to prepare the chloroplasts very carefully to avoid any contamination with mitochondria, since mitochondria seemed to catalyze the incorporation of amounts of phosphate into organic 16 ". h" ) ATP fonnofion (umoles. d 4 8 l7 '/ I O 125 Figure 1. [I . IO.— 8 .- 6 -. 4 _. . 2 ._ . Schwartz\\\\\\‘ J o .g“”””l 1 liét’ , O (l4 (AB 12 5() Light intensity (Kergs. cm—2. sec-1 ) Photophosphorylation at different light intensities. The 2 m1 reaction mixture contained chlorOplasts with 40 ug chlorophyll, 10‘7 moles methylviologen and other ingredients as described in Materials and Methods. Temperature 19°, pH 8.4. Illuminated by 650 nm light (interference filter) for two minutes. Various intensities were obtained by use of neutral filters (fine mesh screens). For the inset the composition of the reaction mixture was somewhat different: Tricine-NaOH, 4x10'3M, pH 8.4; KCl, 10'4; MgCl 3xlo-4M; labelled phosphate, 5x10‘4M; chlorophyll in chloroplasts 20 ug. Illumination was for 3 min. 18 substances which become significant at these extremely low levels of phosphorylation. Because of the measurable but exceedingly small non-linearity we cannot be sure that there is not some critical light intensity below which no phosphorylation can occur. However, this "critical" level, if it exists, seems to be about 2 orders of magnitudes lower than that represented by Schwartz (15). Indeed we have no way of knowing whether or not the electron transport at these low light levels is linear -- we do not have the ex- tremely sophisticated equipment necessary for the measure- ment of such low rates. However, Joliot (20) has shown that the rate of electron transport (measured as 02 pro- duction) is not linear at the very lowest intensities. Therefore, we have no way of knowing whether or not the small non-linearity we observe in phsophorylation re- flects the build-up of a phosphorylating potential or simply reflects a non-linearity of the entire coupled electron transport-phosphorylation process. Effect of Uncouplers on the Non-Linearity of Phosphorylation with Increasing Light Intensities In an attempt to reconcile our results with those of other workers we studied the effect of uncoup- lers and inhibitors on the "lag" phenomenon. We did this because we suspected that earlier workers might 19 have used partially inhibited and partially uncoupled chloroplasts. Inhibitors like DCMU and phlorizin bring about general inhibitions of phosphorylation which are comparable but not identical at all light intensities. In contrast, uncouplers fall into two distinct cate- gories. Carbonylcyanide phenylhydrazones such as CCCP introduce a large non-linearity which becomes larger as the uncoupler concentration is increased (Figure 2). Methylamine, on the other hand, shows a general inhibi- tion which is very similar at all light intensities (Figure 2). Two, 4-DNP, gramicidin and valinomycin were found to introduce non-linearity at low intensities (Figure 3), although the non-linearities were less marked than those introduced by CCCP. Aging of the chloroplasts, Triton-X and atebrin, like methylamine, show similar in- hibitions at all light intensities (Figure 4). It should be noted that CCCP and Triton-X behave quite differently although both are known to make membranes leaky (21). Effect of_pH on the Non-Linearity of Photophosphorylation Since an ion gradient which can be detected as a proton gradient has been shown to make ATP (13) and since building up of the proton gradient is in turn a function of pH of the medium (10), we wanted to see the effect of pH on the non—linearity of photophosphorylation at low ATP formation (umoles. mg chl—l. I14 ) 300,. 200,. 100 . . c on t rol /’ II C C C P ° I “mum a m i n e ’ .IIIII"““I" ” . u‘““ 4’ ‘0‘“‘““ aa—————————-—--""""'° silver 0 10 20 30 40 50 -2 -] Light intensity (Kerqs.cm . sec ) Figure 2. Effects of uncouplers on the non-linearity of photophosphorylation at lowlight intensities. I. Reaction condition as described for fig. 1. Concentrations of uncouplers were: CCCP, 10'6M; methylamine hydrochloride, 2.5x10' M; AgN03, 2xlo-6M. ATP formation fifl ) mg chl (umolos. 21 3()O| . ° control 2IWO . Ir'2,4-DNP ’fl’ t’ I / .ogyronn. ”fl ° vol. I(DO - I 4 o I 0 ()L‘EIEE” - - - ._____J O 10 2O 30 4O -2 -I light Intensity ( Korgs.cm .soc ) Figure 3. Effects of uncouplers on the non-linearity of photophosphorylation at low light intensities. II. Reaction conditions as described for fig. 1. Concentrations of uncouplers were: 2,4-dinitrophenol,.4x10’4M; gramacidin, 2.5xlo-8n; valinomycin 2.5xlo-7M. g LO, > I Pl FLIIMVMOIIELO‘ I“ A‘I’P formation ';J (u molos. m9 chl 22 200.. 100.- . control o 1 1 l 1 O 10 20 30 4O -2 -'I light Intensity (Kergs.cm . sec ) Figure 4. Effects of uncouplers on the non-linearity of photOphosphorylation at low light intensities. III. Reaction conditions as described for fig. 1. Concentrations of uncouplers used were: atebrin, 5x10"6M; triton—X—lOO, 0.003%. Aging of chloroplasts was at 3-4° for 6 hours. i A 23 light intensities. Figure 5 shows the result of such a study. There was no difference in the rates (ie. quantum efficiencies) of phosphorylation at pH 7.0 and pH 8.0. Efficiencies were higher at pH 9.0 and distinctly lower at pH 10.0. It is interesting that the same small non- linearity was observed at all pH's, being neither greater nor less at any particular pH. Effect of Ions on the Non-Linearity of Photophosphorylation Since electron transport, ion-movements and phos- phorylation are very intimately connected processes, it occurred to us that a "lag" in phosphorylation might represent a preliminary phase of ion movements which either creates ion gradients for phosphorylation or, al- ternatively, diverts and dissipates the energy otherwise destined for ATP formation. To test this possibility we studied the effects of replacement or removal of such ions as K+, Na+, le and Mg++ on the phosphorylation non- linearity. Table Ia shows that the omission or drastic reduction of K+, Na+ and Cl- had no effect on the effi- ciency of phosphorylation at either high or low light intensities whereas low Mg++ concentration somewhat in- hibited phosphorylation at both intensities. Table Ib shows that the concentrations K+, Na+ and Cl‘ also have no effect on the CCCP-induced non-line- arity of phosphorylation. 24 12.0 ” 9° C) I .A CD I umoles ATP formed. mg ch|-.I h-l I l I Figure 5. Light lnfensity(_Kergs.cm'Zsec 0.6 1.2 .1) Effect of pH on the non—linearity of photophosphorylation at low light intensities. The buffer consisted of 2x10'2M Tricine and 2x10'2M glycine and the pH was adjusted by addition of the requisite amount of NaOH. Otherwise the reaction conditions were as in fig. 1 inset. 25 Table Ia Effect of Ions on Phosphorylation at High and Low Light Intensities Light Intensity Ionic Composition Rate of Photophospho- (Kergs-cm‘zo of the rylation (umole ATP° Sec-1) Reaction Mixture mg Chl’l-h'l) l. 2.0 Regular (see 15 methods) 2. " Regular, no KI, 15 no Cl', low Na+* 3. " As in 2 except low 12 Tricine-Mg(OH)2- 0.004M 4. 44.0 As in 1 292 5. " As in 2 288 6. " As in 3 213 *Sucrose (0.1M), Tricine-Mg(OH)2 (0.04M) pH 8.4, Na2H32PO (0.01M), ATP (0.001M), hexokinase (1 mg), glucose 0.01M) 26 Table Ib Effect of Different Intensities of Light on the CCCP- Induced Non-Linearity in Phosphorylation in Low Ionic Medium Chloroplasts were washed and suspended in a medium containing 0.1 M sucrose and 0.005 M Tricine-Mg (OH)2 (pH 7.3). Reaction mixture contained 0.1 M Sucrose, 0.004 M Tricine-Mg(OH)2 (pH 8.4), 0.01 M NazH32PO4 , 0.001 M ATP, 1 mg hexokinase, 0.01 M glucose and chloroplasts contain— ing 20 ug chlorophyll/m1. Light Intensity Rate of Photophosphorylation Percent (Kergs-cm'z- (umole ATP-mgchl’l-h'l) Inhibition Sec’l) ___—_:EEE§ ¥EEEF_—_— 1. 0.64 2.4 0.0 100 2. 2.00 12.7 0.0 100 3. 3.10 20.7 0.4 98 4. 6.10 41.5 6.6 85 5. 13.10 85.2 41.0 52 6. 23.00 129.5 86.4 33 7. 44.00 182.0 139.8 23 DISCUSSION According to Mitchell‘s estimate, the chemos- motic hypothesis demands a minimum of 210 mV of membrane potential or a pH differential of 3.5 before any phos- phorylation can begin. Therefore, by decreasing the light intensity one would expect to observe a "critical" intensity below which phosphorylation would not occur. As a matter of fact, Schwartz (15) has claimed that there can be a "critical" light intensity for phosphoryl- ation. Dilley (personal communication) and others (14, 16) have made similar claims. The observations of Schwartz were somewhat indirect, depending solely on pH changes in the medium to compute simultaneously electron transport, proton transport and phosphorylation and therefore his conclusions are not too reliable. However, the work of Dilley and others is not so suspect and there seems to be no doubt that a non-linearity can occur and that this non-linearity may be great or small depending on conditions. The question is: How big a non-linearity does the chemiosmotic theory demand and is it always that big? As indicated above, the chemiosmotic theory de- mands a definite pH gradient and/or electrical potential 27 28 gradient. However, the theory says nothing about the amount of proton translocation required to achieve these gradients; for a given amount of proton translocation (without compensating counter ion movement) there will be a pH gradient formed and an electrical gradient formed. However, the steepness of these gradients will depend on the size and'H+ buffering capacity of the in- ternal space and on the electrical capacitance of the membrane. Furthermore, even the lowest rates of proton translocation could eventually create sufficient gradi- ents for ATP synthesis if the membranes were completely impervious to ions. For these reasons one would expect the steady state non-linearity in phosphorylation to be very much a function of membrane integrity. From this point of view a big non-linearity in photophosphorylation does not seem to be a requirement for the chemiosmotic mechanism.' The non-linearity could be so small in good chloroplasts as to defy detection. Therefore, our failure to find the extent of non-lineari- ty reported by others does not contradict the chemios- motic hypothesis. However, our results do show that the large non-linearity found by others are not really inher- ent in phosphorylation and, consequently, such non-lineari- ties cannot be considered as supporting the chemiosmotic hypothesis. 29 We cannot explain the effects of all uncouplers on the linearity of phosphorylation with light intensity in terms of the chemiosmotic hypothesis. If by uncoup- ling one has to understand dissipation of the proton gradient by any means, then all uncouplers should have behaved the same way,:ida" should have extended the "critical" intensity or in other words, should have intro- duced or greatly increased the non-linearity. We have seen (Figures 2, 3, and 4), uncouplers fall into two distinct groups -- some introducing non- linearity and others showing similar inhibition at all intensities. It is hard to believe that Mitchell's uni- fied model of uncoupler action is working here. Because CCCP is known to make the membrane leaky to hydrogen ions (21) its effect is easy to understand in terms of the chemiosmotic hypothesis; but then the fact that Triton-X which is also known to make membranes leaky (21) -- does not behave the same way needs to be explained. One way of explaining this is to postulate that the effect of Triton-X on a given lamella is all or none whereas, CCCP introduces degrees of leakiness. This interpretation of Triton uncoupling becomes implausible when applied to the reversible uncoupling introduced by many other un- couplers such as amines etc. Clearly these latter un- couplers are acting at different steps in the phosphoryl- ation process, perhaps some closer to the electron transport 30 chain and others closer to the ATP synthesizing step. Somewhat similar interpretation has been given by Avron and Shavit (22) for differential inhibition by uncouplers at high and low light intensities. The optimum pH for building up of proton gradient across the lamellar membranes has been found to be much lower than the optimum pH for phosphorylation (10) at high light intensity. However, this difference might re- flect nothing more than the pH required for maximum rates of some secondary enzyme reaction which is rate-determin- ing at high light intensities. By going to very low light intensities one can make the electron transport it- self the rate-determining factor. In this case one might expect bigger and bigger non-linearities in phosphoryla- tion at higher and higher pH's if the formation of proton gradients is a necessary intermediate step. As can be seen in Figure 5, this is not so. The efficiency of phosphorylation, unlike the efficiency of proton trans- location, remains independent of pH. How then can the non—linearity in photophosphor- ylation be explained except in terms of the chemiosmotic hypotheses? Shakurai g£_a1. (16) have tried to explain the non-linearity on the basis of two-step "excitation" kinetics of phosphorylation. Izawa and Good (23) have suggested that some non-phosphorylating electron trans- port may continue even under phosphorylating condition. 31 When electron flow is limited by low light, the two processes may compete differently for electrons and the inefficiency of photophosphorylation may represent a favorable competition on the part of uncoupled electron transport. Another interpretation could be that the in- efficiency may represent a phase of very active ion movements which dissipates energy required for ATP syn- thesis. If so, this active phase of ion movement seems remarkably independent of the ionic composition of the medium (Table I). The inefficiency at low intensities could also represent the building up of a pool of reduced electron acceptor before phosphorylation can occur (24). In conclusion we can say that the "critical" light intensity below which phsophorylation cannot occur, if it exists at all, is much lower than heretofore reported. Furthermore, in the light of our evidence it does not seem reasonable to attribute all kinds of un- coupling to increases in the ion permeability of the lamellar membrane. Clearly, there are at least two dis- tinctly different mechanisms of uncoupling. SECTION II THE STOICHIOMETRY OF NON-CYCLIC PHOTOPHOSPHORYLATION INTRODUCTION Determination of the stoichiometry of non-cyclic photophosphorylation hinges on our idea of the extent of the coupling of electron transport of phosphorylation. Unlike good mitochondria, all chloroplast preparations show very high electron transport rates even in absence of phosphate. This so-called "basal" rate often attains a magnitude close to half of the electron transport rate obtained with phosphorylation. As a result the stoichi- ometry of the phosphorylating part of the electron trans- port "corrected" for the basal electron transport ap- proaches two ATP molecules per electron pair, while the overall stoichiometry remains close to one. However, Del Campo, Ramirez and Arnon (25) believe that electron transport is totally coupled to phsophorylation under phosphorylating conditions. Therefore, they consider the true stoichiometry of the phosphorylation reaction to be one ATP per pair of electrons transferred. Although no direct evidence is available, there are quite a few indirect lines of evidence which strongly suggest that the basal electron transport can continue unabated even under phosphorylating conditions (25) and, therefore, there are good reasons for believing that the 33 34 basal electron transport should be subtracted if one wishes to determine the efficiency of the phosphoryla- tion part of the electron transport. These reasons are as follows: The phosphorylation rate can be controlled in a number of ways: by limiting the amount of phosphate added, by uncoupling with uncouplers such as CCCP, by using inhibitors of the phosphorylating electron trans- port which for some reason do not inhibit the basal transport (antimycin), and by using specific phosphoryla- tion inhibitors such as phlorizin or Dio-9. When the phosphorylation rate is so limited the increase in elec- tron transport associated with the low phosphorylation rate is always proportional to this rate. Regardless of the rate of phosphorylation, the extra electron transport which occurs constitutes one electron transported for each ATP formed. From such studies Izawa and Good (25) have proposed that only the part of the total electron transport superimposed on the basal is coupled to phos- phorylation. Hence they argue that in computing P/ez, only this coupled part of electron transport should be taken into consideration. According to them, for the phosphorylation process itself P/e2 should be regarded as two instead of one. In any event it seems doubtful that the overall stoichiometry, without any "corrections" whatsoever, is 35 as low as one. Indications of an overall stoichiometry higher than one have come from more than laboratory (26, 27) in recent years. Winget gt_al. (28) have obtained a P/e2 as high as 1.3 by replacing the hitherto univer- sally used Tris-HCl buffer with any one of a number of other buffers such as Tricine-NaOH or glycylglycine. Even higher values have been obtained by Horton and Hall (26) using a somewhat similar technique. Lynn and Brown (27) have claimed still higher values of P/e2 (approach- ing 4) by slow addition of the electron acceptor, but these slow additions of acceptor certainly invited the contribution of a good deal of unmeasured electron trans- port through endogenous carriers. Recently Arnon and his associates (25) have claimed that measurements of a stoichiometry higher than unity (28) may have resulted from a net, light-dependent ADP-Pi exchange reaction and, therefore, should be con- sidered as artifacts. Their conclusion was based on the following observations: (a) besides AT32P, some AD32P (5 to 10 per cent) is also labelled during phosphoryla- tion; (b) more organic phosphate becomes labelled than can be accounted for by the uptake of orthophosphate; and (c) the apparent P/e2 is higher than one only in low concentrations of ferricyanide, the implication being that the alleged ADP-Pi exchange reaction is then a larger part of the whole label incorporation. 36 The data presented by Arnon, if taken at face value, could take care of a part of the higher stoichi- ometry (about 10 per cent). However, they still fail to explain a stoichiometry as high as 1.3 or more. Never- theless, Arnon has raised the very important point that the incorporated label may not be all in ATP. Another explanation of how more than one ATP molecule can be formed during the reduction of ferricy- anide ions has been put forward by Schwartz (personal communication). When ferricyanide is reduced by elec- trons from water (as in the Hill reaction), the hydrogen atoms of the water become hydrogen ions. Using different methods both Schwartz and Dilley (personal communication) have shown that this hydrogen ion is very probably formed inside the lamellae. That is to say, the oxidation of water takes place on the inside of the lamellar membrane. If now we assume that the electron transport per se causes the translocation of one or more additional pro— tons into the lamellae, one can predict that the proton gradient across the lamellar membranes will be greater if ferricyanide is being reduced than if the other accep- tors such as NADP or methylviologen are being reduced. If, following Mitchell, we further assume that proton gradients are used in ATP synthesis, it follows that the efficiency of phosphorylation with ferricyanide as elec- tron acceptor should be higher than with methyl viologen. 37 I, therefore, undertook to reinvestigate the stoichiometries of photophosphorylation reactions in the light of the points raised by Arnon and by Schwartz. MATERIALS AND METHODS Isolation of chloroplasts, composition of the reaction mixture and measurement of phosphorylation were as already described under materials and methods in Sec- tion I. Measurement of the Hill Reaction Ferricyanide reduction was followed spectrophoto- metrically by continuously recording the optical density at 420 nm, using a Bausch and Lomb Spectronic 505 Spectro- photometer specially adapted for actinic illumination (29). Reactions were run in a 1 cm cuvette as described in the previous section. Temperature was maintained at 19°C by circulating water from a thermostated bath through the solid brass cuvette holder. Before the light was turned on, a linear base line was obtained on the record- ing chart paper in dark for 2-3 minutes. When the actinic light was turned on the rate of ferricyanide reduction was computed from the slope of the line obtained on chart paper. The actinic light was turned off immediately after all the ferricyanide in the reaction mixture had been re- duced. For stoichiometric analyses measurements of radio- active organic phosphate and ferricyanide reduction were made in the same sample. 38 39 In other experiments production or consumption of oxygen was measured manometrically (30) using a Warburg apparatus. A total volume of 3.0 ml of the reaction mix- ture (to be described in the legend of the Table IV) was placed in a standard Warburg vessel. The vessel contain- ing the reaction mixture was equilibrated in the dark at 20°C. Illumination was provided by lights (eleven incan- descent 30 W reflector bulbs) from below. Linearity of the reaction was checked by 5 minutes readings of the man- ometer. The lights were turned off after 20 minutes and the final manometer reading was taken. An aliquot was pipetted and fixed in 10% perchloric acid and was later assayed for organic -32P by the method already described in Section I. Colorimetric Determination of Orthophosphate Uptake Orthophosphate was estimated by the method of Ames (31). An 0.3 ml aliquot of the reaction mixture (contain- ing 0.04 to 0.15 umole phosphate) was placed in a test tube containing 1.2 ml water, 0.5 ml of 10% ascorbic acid and 3.0 ml of 0.42% ammonium molybdate in 1N H2804 (total volume 5 ml). The reaction mixture was incubated for 20 minutes at 50°C and then cooled to room temperature (20°C). The optical density of the blue coloured solution was measured at 820 nm using a Beckman Spectrophotometer (model DU) with Gilford electronics. The amounts of 40 orthophosphate in the reaction mixtures were determined by comparing the optical densities from the samples with the optical densities from carefully prepared standard concen- trations of phosphate in identical reaction mixtures (in- cluding even chlorophasts). Standards were run simultan- eously with each sample analysis and these concentrations of the standards were arranged to bracket the phosphate concentrations in samples being measured. Every sample was run in duplicate and the average was taken. The differ- ence between the initial and final concentrations of ortho- phosphate was taken as a measure of amount of organic phos- phate formed. An aliquot (1.0 ml out of a total of 2.0 ml reaction mixture) from the same sample was directly esti- mated for organic -32P for comparison. Paper Chromatography of the Products of Photgphosphor- ylation After the photophosphorylation reaction, the reac- tion mixture was transferred from the cuvette to a centri- fuge tube wrapped in black tape to prevent light entry. It was then centrifuged to spin down the chloroplast frag- ments. To the supernatant, cold ATP (0.1 mg) and cold ADP (0.1 mg) were added to facilitate detection of the labelled substances on paper under ultra violet light. An aliquot was chromatographed on a Whatman No. 1 filter paper using a mixture of 95% ethanol and 1M ammonium acetate (pH 7.5) 41 in the ratio of 7.5:3.0, v/v (32) as solvent. The AD32P spot was eluted in 20% ethanol and rechromatographed in methanol, ammonia and water in the ratio of 6:1:3, v/v (32). After drying, the areas containing ADP and ATP were located in ultraviolet light and were then cut out. These and any other radioactive spots were counted by immersing the piece of paper in 15 m1 of scintillation fluid (5.0 g 2,5 - Diphenyloxazole and 0.3 g 1,4-bis-2-4- methyl-S-phenyloxazolyl-benzene in a liter of toluene) using a Scintillation Spectrometer (Packard Tri-Carb, model 526). RESULTS Labelling of ATP and ADP During Photophosphorylation The stoichiometry of photophosphorylation is a function of two parameters, electron transport and phos- phorylation. Therefore, there are three ways by which one might get an erroneously high P/ez, (a) an overesti- mation of the amount of organic phosphate formed or (b) an underestimation of the electrons transported or (c) both. Before this study we could not be sure that our apparently higher efficiency of phosphorylation was not due to an overestimate of the product of photophosphoryl- ation (ATP), since we were not sure that we had been 32 measuring only ATP formation in measuring of P- incor- 32P is really formed as Arnon and his 32 poration. If AD associates maintain, and if this AD P formation repre- sents an exchange reaction rather than a coupled phos- phorylation, then one should subtract the ADP value from 32P- incorporation before computing P/ez. If however, the AD32P formation represents directly or in- the total directly an electron transport coupled phosphorylation the AD32P should not be subtracted. 42 43 On chromatographic analysis of the product of photophosphorylation, we, like the California workers, found about 5% of the total organic radioactivity in the ADP region in the first chromatogram. However, on re- chromatography of this ADP region we found that about half of its radioactivity was really due to contaminat- ing labelled ATP and, therefore, a maximum of 2-3 per cent of the total radioactivity was due to AD32P (Table IIa). Even lower values of AD32P were obtained in the presence of hexokinase and glucose (Table IIb) which should have transferred virtually all label from ATP to glucose-6—phosphate as soon as it was formed. This amount of radioactivity in ADP, irrespective of its origin, appears quite inadequate to explain a P/e2 of 1.3 if the true value is 1.0. However, the ques- tion of an over estimation of organic phosphate does not arise if this labelled ADP is formed from ATP through a secondary reaction. We suspected that the small amount of labelled ADP might have been formed from ATP by adenylate kinase ("myokinase")a activity which catalyzes the following reaction: ATP* + AMP Z ADP + ADP* In fact, adenylate kinase activity had already been ob— served in isolated chloroplast preparations (33). aAccording to Commission of Enzymes (1961): 2.7.4.3 ATP: AMP phosphotransferase _44 Table IIa Distribution of Labelled Phosphate in the ADP and ATP of Phosphorylation Reaction Mixtures Counts in Counts in Proportion ATP area of ADP area of of Label Chromato- Chromato- Apparently gram (cpm) gram (cpm) in ADP (%) Portion of reaction mixture chromatographed once in 95% ethanol: 5346 312 5.5 1M ammonium acetate (pH 7.5), 7.5:3.0, v/v Portion of "ADP" region re— chromatographed in methanol:am- 463 582 52.0 monia and water, 6:1:3, V/V Maximum proportion of label actually in ADP 2.8% 45 Table IIb Distribution of Labelled Phosphate in the Glucose-6- Phosphate and the Combined 'ADP and ATP' of Photophosphorylation Reaction Mixtures Counts in Counts in Proportion glucose-6- "ADP & ATP" of Label phosphate area of Apparently area of Chromatogram in ADP (%) Chromatogram (cpm) (Cpm) Portion of reaction mixture chromatographed once in methanol: 8440 211 2.4 ammonia:water, 6:1:3, v/v Portion of "ADP & ATP" region rechromatographed 67 68 50.0 in the same solvent Maximum proportion of label actually in ADP 1.2% 46 The above suspicion was easily confirmed. After the phosphorylation reaction, we incubated the reaction mixture in the dark at about 10°C. Samples were taken from time to time for analysis of the relative amounts of labelled phosphate in ATP and ADP. We found that the amount of labelled ADP increases with time of incubation while there is a corresponding decrease in labelled ATP, indicating clearly a transfer of label from ATP to ADP. This conclusion was further substantiated by another ex- periment using hexokinase and glucose. As the label in this case was in glucose-6-phosphate instead of in ATP, there was no redistribution of label on incubation in the dark (Figure 6). Thus, we have been able to show that the small amount of ADP formed during photophosphorylation is de- rived from ATP through a secondary reaction and hence should be taken into account while computing P/ez. Therefore, by including any small amount of labelled ADP, we are not overestimating ATP formation. Simultaneous Estimation of Organic Phosphate Formed andgrthophosphate Takengp During,Photophosphorylation As mentioned before, Arnon and his associates have also reported that the amount of labelled phosphate incorporated into the non-orthophosphate pool is higher than the amount of orthophosphate taken up (estimated by PHOSPHATE ORGANIC TC)TA\L 36 47 100 8() 6C) 40 2C) +- Dlllllllvu‘ ............. G - 6: P h- ‘\\\\---., “‘Tp A D " .IIIIIII‘IIIIIIIIII A D P + A T P l l l I l 1 ° 4 8 I 2 TIAAE (i1) Figure 6. Effect of dark incubation on the distribution of radioactive phosphate after the photophos- phorylation reaction. The 2.0 m1 reaction mixture contained chloro lasts with 100 um chlorophyll, 10-3M Na2H3 P04 and 6.25xlo-4M ferricyanide. Other components of the reaction mixture were as described for fig. 1. When glucose-hexokinase—ATP were replaced by ADP, the latter was 10'3M. The reaction mixture was illuminated by broad red light (glass filter) until all of the ferricyanide had been reduced. The reaction mixture was then incubated in the dark at 10° for the stated times. Products analysed by paper chromatography--without hexokinase and with hexokinase and glucose. 48 the difference in orthophosphate before and after the reaction). They explained this observation in terms of the ADP-Pi exchange reaction which we have not been able to detect. We, therefore, undertook a reinvestigation of the alleged discrepancy between labelling of the organic phase and disappearance of phosphate. In order to obtain meaningful and reproducible results we found that we had to carefully prepare sepa- rate phosphate standard curves for each chloroplast con- taining reaction mixture, that the phosphate concentra- tions of the standards used had to be very close to the concentrations actually being measured, and that the standards and samples had to be measured simultaneously. These precautions were necessary because very small ap- parent variations in relatively high phosphate levels can be computed as wide fluctuations in phosphate uptake. Con- sequently, it was also necessary to work with as low phosphate concentrations as possible, since only then could the differences in phosphate level be translated into reasonably accurate measures of phosphate disap- pearance. A typical standard curve is presented in Figure 7. The hatched band indicates the region used for estimation of orthophosphate in the samples. For this reason the experiments described in Table III were carried out with far too little phosphate for good phos- phorylation -- 1/20 to 1/50 of the optimum. For this 00820 0.6 004 0.2 49 Figure 7. '//// é . . / 2 / ‘ / / l J k l 008 0.16 umole Pi Typical standard curve for extimation of orthphosphate. All assays were conducted in the usual reaction mixture and standards were measured at the same time as samples. The hatched area shows the range of residual phosphate concentrations measured while the upper point represents the initial concentration. Orthophosphate analyses were by the method of Ames (31). See Table III. 50 Table III Comparison of the Amount of Labelled Organic Phosphate Formed and the Amount of Orthophosphate Removed From the Medium During Photophosphorylation 2.5 x 10‘4M Na2H32PO4 and 6.25 x 10'4M ferricyanide. reaction was stopped as soon as all of the ferricyanide (1.25 umoles) had been reduced. The otherwise regular reaction mixture contained The Experi- Pi removed from Organic 32P Pi removgg/ ment the medium formed Organic P (umole) (umole) formed 1. 0.645 0.630 1.02 2. 0.648 0.627 1.03 3. 0.643 0.628 1.02 4. 0.673 0.630 1.07 5. 0.648 0.629 1.03 6. 0.637 0.634 1.01 7. 0.633 0.649 0.98 8. 0.637 0.638 1.00 9. 0.623 0.644 0.97 10. 0.617 0.627 0.98 11. 0.607 0.621 0.98 12. 0.613 0.631 0.97 13. 0.610 0.626 0.97 14. 0.627 0.640 0.98 15. 0.613 0.622 0.99 16. 0.580 0.596 0.97 17. 0.590 0.620 0.95 18. 0.633 0.632 1.00 19. 0.631 0.626 1.01 20. 0.643 0.640 1.00 21. 0.647 0.629 1.03 22. 0.627 0.620 1.01 Averages 0.628 0.629 1.00 (P/e2 0.994 0.995) 51 reason the P/e2 observed was much lower than we usually observe. In fact, it was almost precisely 1.0. Table III presents the results of 22 consecutive experiments, each value representing the average dupli- cate analysis. Clearly, there were no consistent differ- ences between the labelling of the organic phosphate and the disappearance of orthophosphate. These results are in good agreement with our above-mentioned observation that any label in ADP is ultimately derived from ATP. Therefore, we doubt that there is any ADP-Pi exchange in our chloroplast preparations, and certainly not enough such exchange to affect accuracy of measurement of phos- phorylation as organic -32P formed. Apparent P/ez as a Function of Ferricyanide Concentration Lynn and Brown (27) have observed that slow addi— tion of an electron acceptor results in apparent P/e2 values as high as 4. Arnon and his associates have also observed that use of low ferricyanide concentration yields apparent P/e2 values higher than one -- the lower the ferricyanide concentration the higher the apparent P/ez. In these experiments they illuminated chloroplasts for a constant period of time which was greater (by vary- ing degrees) than the time actually required for complete reduction of the amount of ferricyanide used. Therefore 52 we suspected that in these experiments the absence of an exogenous electron acceptor permitted significant amount of undetected cyclic electron transport catalyzed by endogenous electron carriers in the manner described by Mehler (34). A maximum endogenous phosphorylation would therefore occur with a minimum concentration of ferricyanide. When, like Arnon, we illuminated chloro- plasts for a constant period of time (4 minutes) using varying concentrations of ferricyanide, we found that the apparent P/e2 reaches very high values when a very low concentration of ferricyanide is used (Figure 8). How- ever, unlike Arnon, we had the advantage of knowing from spectrophotometric evidence precisely when all of the ferricyanide has been reduced. When we now illuminated chloroplasts for just enough time to reduce all the fer- ricyanide present, we found no significant variation in P/e2 as a function of ferricyanide concentration (Figure 8). P/ez as a Function of Different Electron Acceptors Because hydrogen ion gradients may be associated with phosphorylation and because ferricyanide reduction produces protons, it has been argued by Schwartz (per- sonal communication) that the higher P/e2 values we have observed using ferricyanide may reflect the fact that ratios [apparent] 9“: 53 2&5 'LO ‘ Ill. only during\° Fe( CN)6 reduction constant 4-min. ill. \. 0.0 0.2 J ‘ I i 0.4 0.6 0.8 l O to rrlcyanido reduced [a mole] Figure 8. Effects of different concentrations of ferricyanide on the apparent stoichiometry of photophosphorylation. Reaction conditions were as in fig. 1 except that ferricyanide at the stated concentrations replaced methylviologen as electron acceptor. Since there is appreciable phosphorylation when chloroplasts are illuminated in the absence of ferricyanide, the ratio of ATP molecules formed to ferricyanide ions reduced becomes infinite when no ferricyanide is added. 54 the extra hydrogen ion produced has contributed to extra phosphorylation. If this is so, use of methylviologen as an electron acceptor might yield a lower P/e2 than that obtained by using ferricyanide. We checked this point by measuring either the production of 02 (with ferricyanide) or consumption of 02 (for methylviologen) using Warburg's manometric technique (30). The reduc- tion of methylviologen (or nearly any other auto-oxidiz- able electron acceptor) results in an uptake of oxygen which is twice as great as the oxygen produced by the chloroplasts. This is a consequence of the fact that during the Hill reaction electrons are removed from water but hydrogen peroxide is formed as the back-reac- tion of the reduced methylviologen with oxygen. Thus there is a net uptake of oxygen when methylviologen is reduced which is equal to the production of oxygen when ferricyanide is reduced, if the electron transport rates are equal. The results of our experiments are presented in Table IV. It is interesting to note that a somewhat higher P/e2 was obtained with methylviologen than with ferricyanide. This slightly higher P/e2 for methyl- viologen was due to a slightly higher phosphorylation rate, whereas electron transport rates were comparable (Table IV). The extra phosphorylation may be indicative of a cryptic cycle of electron transport passing through 55 Table IV Stoichiometry of Photophosphorylation Using Different Electron Acceptors Reaction mixture contained in a total volume of 3 ml, 0.1M sucrose, 0.04M Tricine-NaOH (pH 8.0) 0.001M KCl, 0.003M MgClz, 0.0115M labelled NazHPO4, 0.005 M ATP, 2 mg hexokinase, 0.01M glucose, 2 mg catalase, 1% ethyl alcohol, 0.008M ferricyanide or 0.0015 M methyl- viologen and chlor0plasts containing 100 ug of chloro- phyll. Electron Experi- Rate of 02 Rate of P/O Average Acceptor ment Evolution Photophos- P/O (umole-mg phoryla- chl’l-h’ ) tion (umole-mg chl'l°h'1) Ferri- 1. 86 185 1.08 cyanide 2. 88 186 1.06 1.14 3. 81 198 1.22 4. 87 208 1.20 Methyl- 1. -91 216 1.18 viologen 2. -88 229 1.30 1.29 3. -88 239 1.35 4. -87 235 1.35 56 a phosphorylation site. This points out the great danger of trying to measure precisely P/e2 ratios when the re- duced acceptor is a strong reductant (e.g. methylviologen, NADP, FMN etc.). Aside, however, from this possibility of a very small contribution of some cyclic phosphoryla- tion process, it would seem that the phosphorylation re- actions with ferricyanide and methylviologen as electron acceptors are not essentially different. This observa- tion is in agreement with Arnon's similar report on the relative efficiency of ferricyanide and NADP phosphoryla- tion (35). Schwartz's argument, therefore, seems im— plausible. Quantum Efficiencies of Photophosphorylation As A Function of Different Electron Acceptors On the basis of Schwartz's argument higher quan- tum efficiencies for phosphorylation might be expected with ferricyanide reduction than with the reduction of other acceptors such as methylviologen or NADP. However, we failed to observe any significant difference in the rates of phosphorylation at different light intensities using ferricyanide and methylviologen as the electron acceptors (Figure 9). 57 7.: E U200- 03 E .O' Q) E .2 CL 0—100- < :3 O’ E 8 = I 8 (E o . . . 1 . 0 20 40 Light Intensity (Kergs. cm'2. sec‘I) Figure 9. Photophosphorylation as a function of light intensities with different electron acceptors. Ferricyanide (Fecy) concentrations, 5x10’4 M; methylviologen concentration, SXIO'SM. Other reaction conditions as for fig. 1. DISCUSSION These observations indicate that our higher stoichiometry of non-cyclic photophosphorylation is not due to an overestimation of ATP formation. We have shown that the very small amount of ADP formed could well arise from ATP through adenylate kinase activity. We cannot detect any measurable discrepancy between, on the one hand, incorporation of labelled orthophosphate into non-orthophosphate and on the other hand, disappear- ance of orthophosphate. Thus, we could not find any evidence of an ADP-Pi exchange reaction and indeed other workers (24) have reported similar negative observations. If any labelled ADP at all is formed other than through the phosphorylation reaction, the amount is utterly in- significant in our measurements of phosphorylation effi— ciency. Both cyclic and pseudo-cyclic photophosphoryla- tion are known to be inhibited by ferricyanide and other non—autooxidizable electron acceptors (27) but phos- phorylation in the absence of exogenous electron accep- tors is usually appreciable. Therefore, it is not diffi- cult to understand the reports of higher apparent P/e2 ratios when the ferricyanide concentration is reduced 58 59 below the electron transport capacity of the chloroplasts. In this situation one would expect a considerable con- tribution of an unmeasured endogenous electron transport to the phosphorylation process. Both Lynn and Brown (27) and DelCampo, Ramirez and Arnon (25) have claimed that their preparations did not support cyclic or pseudo-cyclic phosphorylation, but we find this contention difficult to accept. Lynn and Brown's experiments were done under conditions which always give high rates of "endogenous" phosphorylation in our laboratory. Arnon at least at- tempted to reduce the contribution of pseudo-cyclic phos- phorylation by working under argon but it is very diffi- cult to thus remove enough oxygen to inhibit the cycle. In any event, we find no relationship between the amount of ferricyanide reduced and the apparent P/e2 if we take the precaution of stopping the reaction as soon as the ferricyanide is exhausted. We have failed to detect any differences in the quantum efficiencies and stoichiometries of photophos- phorylation using ferricyanide and methylviologen as electron acceptors. Arnon (35) has also shown that the stoichiometries with ferricyanide and NADP are identical. These observations seem to preclude Schwartz's suggestion that ferricyanide reduction is more efficient in support- ing phosphorylation because of the formation of an addi- tional proton gradient across the lamellar membranes. 60 The fact that this increased gradient does not affect phosphorylation must be taken as evidence against the chemiosmotic theory, rather than evidence in its sup- port. SECTION III THE NUMBER AND LOCATION OF PHOSPHORYLATION SITES IN THE PATHWAY OF NON-CYCLIC ELECTRON FLOW INTRODUCTION The Pathway of Electrons From Water to NADP The process of photosynthesis under normal con- ditions involves oxidation of water to molecular oxygen with a concomittant reduction of an endogenous compound such as NADP. This overall oxidation-reduction process is associated with the formation of ATP. Therefore, solar energy is conserved partly in the form of reduced pyridine nucleotide and partly in the form of ATP. Both of these energy-rich compounds are subsequently used in the reduction of CO2 and for further elaboration of the immediate products of photosynthesis. The journey of electrons from water to NADP is a long one in terms of number of steps involved. Although our ignorance of many of these steps is still enormous, considerable progress has been made in the last decade in identifying the components of the electron transport chain. Most of the observations can be accommodated rea- sonably well by considering that photosynthesis involves two separate light reactions arranged in series. An out- line of such a scheme is presented in Figure 10. 62 Figure 10. 63 X MV, Fecy n CL Ferr OX 2‘“ > O 1 DCPIPH2 9. ATP PS I cyl. fl? plastocyanin 700 (l) DCMU (2) Tri s-washing Schematic representation of the pathways by which electrons are thought to be transferred during the photoreactions of isolated chloroplasts. Washing chloroplasts with high concentrations of tris or other amines prevents water oxidation but not the oxidation of exogenous electron donors such as p-phenylenediamine (PD) or ascorbate. Reoxidation of the primary electron acceptor of photosystem 11 (Q) seems to be blocked by DCMU. Reduced indophenols, having a higher electron donating potential than PD, can.donate electrons to photosystem I directly, this by-passing the DCMU inhibition site. P700, which is known only from spectral changes occuring during electron transport, prObably represents the quantum trapping reaction center chlorophyll. 64 Since our ultimate purpose was to determine the location of phosphorylation sites, a brief introduction to the scheme is essential for discussion of our obser- vations. The scheme in Figure 10 is an elaborated form of the original proposal of Hill and Bendall (36). Most important evidence in support of two light reactions in photosynthesis has come from the observation of Emerson (37). He found that the rate of photosynthesis was very much stimulated when algal cells, illuminated by red light (>680 nm), were given supplementary lights of shorter wave length (<650 nm). This synergistic effect of red and far-red light is known as the "Emerson Effect" or simply "enhancement". "Enhancement" is defined as -- (Rate - RateR) / Rate Clearly, if there is no R+FR FR' synergism this equation gives a value of 1.0 while values over 1.0 signify synergism. The evidence in favor of the idea that these two light reactions are arranged in series has come primarily from the observations of Duysens (38). Duysens observed that spectral changes at 420 nm (which almost certainly reflect the redox state of cytochrome f) depends on the wave length of illumination -- longer wave lengths (ac- tivating photosystem I only) oxidize cytochrome f and shorter wave length (activating primarily photosystem II) reduces it. The reducing effect of system II absorbed 65 light is abolished by the use of inhibitors of oxygen evolution -- such as DCMU. Among the components of the electron transport chain, some (such as Q and X in Figure 10) are only postulated while others (such as plastoquinones, plasto- cyanin, cytochrome f etc.) have been isolated from photo- synthetic organisms or from chloroplasts. When plasto- quinones are removed from chloroplasts by the use of lipid solvents, the chloroplasts lose their ability to oxidize water but they regain this ability when the plastoquinones are restored. Some components of the chain (such as P 0) have been proposed primarily on the 70 basis of spectral changes associated with the redox state of the chloroplasts and the effects of illumination on- that state. The exact position of many of the components of the transport chain is still highly controversial. In the paragraphs that follow, evidence is presented concern- ing the position of the various natural components of the chain and the sites of action of exogenous electron accep- tors and donors. Isolated chloroplasts reduce NADP in the light. On the basis of “enhancement" and of large quantum re- quirements at far-red, NADP-reduction definitely appears to involve both the photosystems (39). Autooxidizable electron acceptors with very negative potentials such as methylviologen must for thermodyanmic reasons act near 66 the "top" of the photosystem 1. However, this has not been definitely established for many other electron ac- ceptors including ferricyanide. Failure to observe any "enhancement" with ferricyanide led Avron st 21. (39) to propose that ferricyanide accepts electrons somewhere between the two photosystems but these observations have been challenged. The non-heme iron protein, now called plant ferredoxin, was originally isolated from chloro- plasts under two different names, methaemoglobin reduc- ing factor or MHRF and photosynthetic pyridine nucleo- tide reductase or PPNR. Plant ferredoxin can be reduced rapidly by chloroplasts in the absence of NADP but NADP cannot be reduced in the absence of ferredoxin. A pool of the hypothetical "X" has been invoked in the pathway of electron transport since Chance 32 31. (40) have shown that the presence of ferredoxin or exogenous electron acceptors is not obligatory in the oxidation of cyto- chrome f. This is not very convincing because cytochrome is present in very small amounts (one molecule to 400 chlorophyll molecules) and it is actually impossible, especially in the presence of oxygen to remove all traces of contaminating electron acceptors. In all probability P is a form of chlorophyll 700 a whose oxidation results in a decrease of absorption peaking around 700 nm (41). Oxidation of P700 by System I absorbed light is temperature independent but the 67 oxidized form of P700 is reduced by System II absorbed light in a temperature sensitive reaction. This suggests that P700 probably represent the reaction centre of photosystem I, uniquely situated chlorophyll a molecules serving as the ultimate acceptors of the excitation energy which in their excited states donate electrons to "X" and become oxidized. Plastocyanin, a copper protein, has been shown to be oxidized by excitation of System I while additional excitation of System II led to its reduction. Although considerable controversy exists as to the relative posi- tions of cytochrome f and plastocyanin it appears to be certain that plastocyanin acts very close to the reaction centre of system I. Studies with the mutants have shown that while a plastocyanin-less mutant could not catalyze the electron flow from an artificial donor system (DCPIP and.ascorbate) a cytochrome f-less mutant could (42). Tliis evidence supports the position of plastocyanin as shown in Figure 10. At least it shows that reduced in- dophenol dyes can donate electrons to plastocyanin with- out the intervention of cytochrome f. Cytochrome f was originally isolated, described and characterized by Hill and Scarisbrick in 1951 (68). There are a number of reports to show that the compound is Photoxidized by System I absorbed light and is reduced by System II absorbed light (39). Therefore, its position 68 between the two photosystems seems certain. On similar grounds, cytochrome b 9 is also believed to be acting 55 between the two photosystems somewhere before cytochrome- f (43). Numerous plastoquinones have been extracted from chlor0plasts and it has been shown that they can be photo-reduced by the chloroplasts from which they have been extracted. Moreover, it has been shown that endog- enous plastoquinones can be reduced in sites by System II absorbed light. Extraction of plastoquinone does not interfere with the electron flow from DCPIPH2 to NADP (43). This suggests that its site of action is closer to photosystem II than to photosystem I. On the basis of changes in the fluorescense yield in different light intensities, Duysens and Scweers (44) have suggested the hypothetical compound "Q" as the quencher of the excited pigments of System II. When "Q" is present in the oxidized form, a quantum absorbed by photosystem II pigment would be used to transfer an elec- tron to this "quencher" and hence there would be very low fluorescence. But if "Q" is reduced many of the quanta absorbed by photosystem II, having nowhere to go, will be reemitted as flourescence. Rumberg 23 El. (43) have produced evidence in support of the idea that plastoqui- none itself may represent "Q". They found high and in- variant fluorescence yields in chloroplasts from which 69 plastoquinone had been removed. However, Duysens has disputed these observations (45). We know very little about the mechanism of photo— oxidation of water and its close association with photo- system II. We do not know any organic component in this part of the electron transport chain. However, it has been shown that photooxidation of water requires the presence of manganese and chloride ions (46, 47). Yama- shita and Butler (48, 49) have shown that the system photooxidizing water can be destroyed by heating, by ultraviolet irradiation and by washing the chloroplasts with high concentration of Tris. Several artificial donors such as p-phenylenediamine, benzidine (50), semicarbazide (51) etc. are known to donate electrons in Tris-washed chloroplasts very close to the site where water would donate electron in normal chloroplasts; the photooxidation of these substances is inhibited by DCMU and is associated with an increase in the fluorescence yield of System II. Phenylureas (such as DCMU) and aminotriazines (such as Simazine) inhibit electron flow from water as well as from any of the above system II donors. DCMU, however, does not inhibit electron flow from reduced DCPIP, TMPD, DAD, etc. (48) to photosystem I and further- more, it does not inhibit any cyclic electron flow involv- ing only photosystem I. Therefore, DCMU inhibition is 70 always correlated with a flow of electrons from photo- system II. The presence of DCMU increases the fluores- ence yield of system II and therefore it is believed that DCMU prevents the reoxidation of the quencher "Q". In some aerobic cyclic systems which seemingly involve only system I, DCMU has been shown to be inhibitory. However, this inhibition is best explained in terms of a redox "poise". That is to say, in cyclic electron transport mediated by PMS or pyocyanine, these substances must function both as electron acceptors and as electron donors. Hence they must be partly oxidized and partly reduced. In the presence of an excess of an oxidant like molecular oxygen, this condition can only be maintained if there is a continual, if small, input of extra elec- trons from System II. For these reasons the reaction which is primarily a System I reaction is nevertheless dependent on a catalytic amount of System II activity. The Number of Phosphorylation Sites in the Non-Cyclic Electron Transport Pathway The stoichiometry of non-cyclic photophosphoryl- ation reflects the possible number of sites in the non- cyclic electron transport pathway. On the basis of P/e2 Krogmann, Jagendorf and Avron (52) suggested the possibility of two photophosphorylation sites. In the previous section of this thesis we have shown that P/e2 71 of photophosphorylation, in our hands, is greater than one. Assuming that the coupled redox reaction involves the transfer of two electrons, the number of ATP mole- cules per pair of electrons (P/ez) should represent the number of sites. Therefore, we are inclined to suspect that the number of sites of non-cyclic photophosphoryla- tion is indeed two. However, if the coupled reaction involves the transfer of only one electron, then the number of sites would be half as many -- that is only one. According to the chemiosmotic theory, a "site" of phosphorylation represents the transfer of two elec- trons and two hydrogen ions across the chloroplast lamella -- that is to say one "fold" of the electron transport chain. Therefore, the number of sites of phos- phorylation in this theory can be expressed as the number of protons translocated for each electron traversing the entire length of the chain. Izawa and Hind (53) have found this ratio of H+ accumulated per electron to be approximately two and hence the number of sites should be considered as two. A similar H+/e ratio has been found by Schwartz (15). Karlish and Avron (54) have found a ratio of four for ferricyanide reduction and Lynn and Brown (55) have found a ratio of five for chlorantil reduction. Aside from the fact that the chemiosmotic theory may well be wrong, such wide variations in the stoi- chiometry of proton translocation to electron transport 72 indicate that this ratio provides a very uncertain measure of the number of sites of phosphorylation. Location of the Photophosphorylation Sites A non-cyclic photophosphorylation site has been very plausibly assigned to the electron transport chain between the two photosystems on the basis of a number of different lines of evidence. It has been shown that the light-dependent electron transport from reduced DCPIP to NADP is not accompanied by ATP synthesis (42). This provides very good evidence in support of the contention that there is no photophosphorylation site in the part of the chain from plastocyanin to NADP (see Figure 10). Obviously then the phosphorylation site or sites should exist before plastocyanin. Avron and Chance (56) found that both phosphorylating conditions and addition of an uncoupler result in an increased reduction of cytochrome f. Therefore, there is presumably a phosphorylation site before cytochrome f in the electron transport chain. Another line of evidence in support of this idea has come from Gorman and Levine (57). They found that mutants of Chlamydomonas lacking in cytochrome or in plastocyanin cannot support non-cyclic photophosphoryla- tion although the former type of mutant can carry out non-cyclic electron flow from reduced DCPIP to NADP. 73 Several workers have reported the existence of a separate photophosphorylation site for cyclic electron transport which may or may not function in conjunction with the one established for non-cyclic photophosphoryl- ation. In 1959, Hill and Walker (58) found that pyocy- anine mediated phosphorylation occasionally shows two pH optima (one around pH 7.0 and the other around pH 7.7) while FMN + K3 mediated phosphorylation has only one. They, therefore, suggested two sites for photOphosphoryl- ation. Duane, H6h1 and Krogmann (59) observed two pH optima of photophosphorylation (one at pH 6.8 and the other at pH 7.8) in cell free preparation of Anabaena variabilis using PMS, FMN + Vitamin K3, and DCPIPHZ. They think that cyclic photophosphorylation uses two sites -- one of which may be common to non-cyclic photo- phosphorylation. Recently Lee, Young and Krogmann (60) have observed differential sensitivities of cyclic and non-cyclic photophosphorylation of blue-green algal pre- paration towards inhibitors. They interpreted this dif- ferential behavior as indicative of separate sites of cyclic and non-cyclic photophosphorylations. A somewhat similar suggestion was given by Duysens and Amesz (61). It has been claimed that desaspidin specifically inhibits cyclic photophosphorylation while non-cyclic photophosphorylation is not affected (62). However, Avron (24) has shown that this compound is quickly 74 photooxidized under the oxidizing conditions of non- cyclic phosphorylation and for this trivial reason can- not inhibit non-cyclic photophosphorylation. In con— trast, ascorbate or the generally reducing conditions of cyclic phosphorylation prevent the photooxidation of this powerful uncoupler and, therefore, it is very ef- fective in inhibiting cyclic phosphorylation. Another line of evidence for the existence of a separate site for cyclic photophosphorylation has come from the experiments of Gorman and Levine (57) on mutants of Chlamydomonas. They found that a cytochrome f deficient mutant can carry out cyclic phosphorylation while it fails to support any non-cyclic photophosphoryl- ation. Laber and Black (63) have recently shown that about 95% of the non-cyclic photophosphorylation is in— hibited in n-heptane treated chloroplasts without any inhibition of the accompanying electron transport. At the same time cyclic photophosphorylation was only 50% inhibited. They too explained their results in terms of separate sites -- one for cyclic and the other for non- cyclic photophosphorylation with different sensitivities to heptane treatment. Although claims for two sites of phosphorylation have been many, only one of these two postulated sites has been attributed to non-cyclic photophosphorylation. 75 However, the stoichiometry of phosphorylation has often indicated more than one site of non-cyclic photophos- phorylation. Moreover, Izawa (64) in our laboratory, has observed 50% inhibition of non-cyclic photophosphoryla- tion by mercuric ion or PCMB. This is probably indica- tive of two phosphorylation sites -- one being sensitive to the mercuric ion while the other is completely in- sensitive. Recently B5hme and Trebst (65) have suggested two sites for non-cyclic phosphorylation. They found a "corrected" stoichiometry (see Section II) of one in- stead of two in heated chloroplasts where ascorbate re- placed water as an electron donor. Therefore, they pro- posed that there may be a heat labile site of ATP synthesis between water and pigment system II. They explained the lower stoichiometry in heated chloroplasts on the basis that only one of the two sites is being used during ascorbate photooxidation. Our recent observations on photOphosphorylation with different electron acceptors can also be interpreted in terms of two sites of non-cyclic phosphorylation. We have found that the oxidized form of p-phenylenediamine (1,4-benzoquinonediimide or PDox) greatly stimulates non- cyclic electron transport whether or not phosphorylation is occurring. This rate of apparently fully uncoupled electron transport is neither stimulated any further by 76 addition of ADP and phosphate nor by addition of an un- coupler. However, this high rate of electron transport does not really represent a fully uncoupled electron flow since very high rates of phosphorylation occur whenever ADP and phosphate are present. Obviously, the phosphoryl- ating electron transport competes very effectively with the equally rapid uncoupled electron transport which occurs in the absence of phosphorylation. In addition, we found that PDOX-mediated phos- phorylation is different from ferricyanide-or methyl- viologen—mediated phosphorylation in a number of ways. The pH optimum for PDox-phosphorylation is much broader than that for ferricyanide-phosphorylation; As a matter of fact, PDOX-phosphorylation is as efficient at pH 7.0 as at pH 8.0, while ferricyanide phosphorylation doubles between pH's 7.0 and 8.0. Therefore, at pH 7.0, PDOX- phosphorylation rate is twice as high as ferricyanide- phosphorylation rate. PDox-phosphorylation is even more sensitive to DCMU than is ferricyanide-phosphorylation, whereas, PMS-phosphorylation can be almost completely in- sensitive. The apparent Km for phosphate of the PDox system is, however, similar to the Km in ferricyanide- phosphorylation but different from the Km in PMS-phos- phorylation. The quantum efficiency of PDox-phosphoryla- tion, at pH 7.0, is about half of that of ferricyanide- phosphorylation. However, all these differences gradually 77 become smaller and smaller with increasing pH so that at about pH 9.0 the differences between FUCK—phosphorylation and ferricyanide-phosphorylation disappear. These observations have led us to reconsider the possibility of the existence of two sites for non-cyclic photophosphorylation -- one having a pH optimum around 7.0 and the other (the one already well established) above pH 8.0. The site with the lower pH optimum, we think, may be very loosely coupled and located on the water side of the photosystem II. According to this tentative proposal PDox can accept electrons at two dif- ferent places in the electron transport chain as shown in Figure 11. Most of the remainder of this thesis will be devoted to studies of the photoreduction of oxidized phenylenediamine and the photooxidation of reduced phenyl- enediamine which bear on the above proposal. 78 pDox .i—x MV, Fecy W‘Ferredoxin é NADP DCPIPH2 _ :fQPYt\TP PSI cytf plastocyanin P700 A i I ’I PD,csc., etc. e j (2) Tris-washing ? ATP Figure 11. Proposed site of action of the oxidized form of p—phenylene-diamine (PDox) and the proposed second site of phsophorylation close to photosystem II. MATERIALS AND METHODS Preparation of chloroplasts, measurements of ferricyanide reduction, estimation of organic phosphate and preparation of reagents have already been described in Materials and Methods of the preceding two sections. A new buffer - l,3-bis[tris(hydroxymethyl)amino] CH CH propane, (HOCH2)35C-NH CH NH-CE(CH20H)3, syn- 2 2 2 thesized in our laboratory by Karen (Sandstedt) Melcher, was used for many of these studies, especially in experi— ments involving low pH's. Tris-washed chloroplasts were prepared by the method of Yamashita and Butler (49). Washing in Tris was done in room light (66). Oxidized p-phenylenediamine (PDox) was either prepared by mixing equimolar solutions of p-phenylene- diamine and ferricyanide (1:2, v/v) immediately before use or equivalent amounts of reduced p-phenylenediamine were added to the reaction mixtures containing ferricya- nide. Since ferricyanide rapidly oxidizes any reduced phenylenediamine formed, the rate of reduction of ferri- cyanide is a reliable measure of the rate of reduction of PDox. 79 80 Measurements of the reduction of PDox (1,4 - benzoquinonediimide) le the absence of excess ferricya- nide made in the following manner: Reaction mixtures containing PDox were illuminated for different lengths of time (10, 20 or 30 seconds). The reduced p-PD formed in each case was then titrated by adding an excess of ferricyanide. .The amount of this excess was read at 420 nm in a Baush and Lomb Spectrophotometer (two moles of ferricyanide are required to oxidize one mole of PD). The difference in OD420 cxf dark and illuminated samples was thus used to compute amount of PDOX reduced. Amounts of PDox reduced estimated in this way were found to be linear with time of illumination and to agree with values obtained by measuring the reduction of PDox in the presence of excess ferricyanide as described above. RESULTS Properties of Different Elegtron Acceptors inRelation to Electron Transport and Phosphorylation in Chloroplasts Electron acceptors can be grouped into three major categories. (a) Those supporting only low rates of electron transport in the absence of phosphorylation but supporting good phosphorylation rates and good elec- tron transport rates in the presence of ADP and phos- phate. Addition of uncouplers in the presence of such acceptors results in very high electron transport rates whether or not ADP and phosphate are present and no phos- phorylation then occurs. Ferricyanide, methylviologen, FMN and ferredoxin-NADP are representatives of this large category. (b) Substances supporting a high rate of non-phosphorylating electron transport whether or not ADP and phosphate are present. These acceptors behave like category (a) acceptors plus uncouplers. Indophenol dyes such as DCPIP are the only substances known to be- long to this category: It is reasonable to assign two functions of these compounds. They probably act both as oxidants and as uncouplers. (c) Substances supporting very high rates of electron transport whether or not ADP 81 82 and phosphate are present but which, unlike compounds of category (b), also support good rates of phosphoryla- tion. Heretofore, this category had not been recognized and Section III of this thesis deals primarily with at- tempts to explain their mode of action. PD is the ox best example of this category (see Table V). However, a number of other substances behave in a rather similar manner. Indeed, there seems to be a continuum between categories (a) and (c) with well known substances such as p-benzoquinone occupying an intermediate position somewhat closer to category (c). ‘ Stimulation of Electron Trangport and Photophosphorylation by PDox (1,4- Benzoquinonediimide) We have observed that the oxidized form of p- phenylenediamine stimulates electron transport in chlor-- oplasts to a level which is almost exactly the same as the maximum amine-uncoupled rate. Once this high rate is obtained, neither addition of ADP and phosphate nor addition of an uncoupler (such as methylamine) can stimu- late electron transport any further. Therefore, PDbx appears to be a very efficient uncoupler. However, when we looked into phosphorylation we found a remarkable difference between PDox and other uncouplers. Uncouplers, by definition should inhibit phosphorylation while stimu- lating electron transport. On the contrary, PDox 83 Table V Electron Transport and Phosphorylation with Various Electron Acceptors Electron Phosphory- Electron Acceptor1 Transport lation Rate Structure Abbreviated Rates (umoles (umole ATP, Name electrons-mg mg chl‘ chl-l-h'l) h‘l) +ADP-P1 +ADP +Pi a. Fe(CN)g' Ferricy- -MA 190 480 278 anide +MA 900 910 10 L b. 6©N=©w DCPIP 710 780 20 CL c. Hu:<:=NH PDOX 960 960 356 P-benzo- 650 910 332 o=©=o quinone “5 “3N DAD 680 900 340 +MPDOx 610 900 314 NH“<:>flflfm - £6=®¢N a DATox 750 790 280 EzNQ'N'QNQy DADPon 600 1000 294 DMQox 390 940 320 @Odzo DCQox 800 940 276 lFerricyanide, 5x10 4M; DCPIP, 5x10-5M; all other oxidants 10‘4M. The oxidized forms of the various poly- phenols and phenylenediamines etc. were prepared direct- ly in the reaction mixture immediately before the reac- tion by adding an excess of ferricyanide. Structures of the oxidized forms are presumed and when no plausible structure occurred to us we indicated oxidation by one or two plus signs. Methylamine (MA) uncoupler, 2x10‘2M. Reaction conditions otherwise as described in Section I of this thesis. 84 stimulates phosphorylation at concentrations which sus- tain a maximum rate of electron transport, and even very high concentrations of PDOX continue to stimulate phos- phorylation. Table VI documents this statement. It is to be noted that the true uncoupler methylamine inhibits PDox—phosphorylation although it does not stimulate elec- tron transport any further. It is clear that PDox is not simply another typical uncoupler. Certainly it does uncouple in the sense that it permits rapid electron transport without phosphorylation, but electron transport in its presence can support a reasonably efficient phosphorylation. Our first guess was that, in the absence of phosphorylation, PDOX accepts electrons very close to photosystem II, before the coupled phosphorylation site (Figure 11). This could explain a very high rate of electron trans- port without phosphorylation if only System II is supply- ing the electrons. The model could also be stretched to accommodate a fairly good rate of phosphorylation if we assume that, in the presence of ADP and phosphate, PDOX can also accept electrons donated by System I. In this case the phosphorylating electorn transport via System I might compete efficiently with the direct reduction of PDOX for the System II electrons. However, the model does nothing to explain why the PDox-phosphorylation can 85 Table VI Stimulation of Electron Transport and Photophos- phorylation by 1,4—Benzoquinonediimide (PDox) Rate of Rate of ATP Experi- Additions to Electron Synthesis P/ez ment Reaction Transport (umoles-mg Mixturea (umoles~mg chl’l-h‘l) chl‘l-h’l) 1. Ferricyanide 190 2. Ferricyanide + 500 290 1.16 phosphate 3. PDbx 880 4. PDox + phosphate 910 340 0.70 5. Ferricyanide + 890 14 methylamine + phosphate 6. PDox + methylamine 880 12 + phosphate aIn addition to the substances listed in this column, the reaction mixture contained in a total volume of 2 ml, 0.1 M sucrose, 0.04 M Tricine-NaOH (pH 8.4), 0.001 M KCl, 0.003 M MgClz, 0.001 M ATP, 1 mg hexokinase, 0.01 M glucose and chloroplasts containing about 40 ug chlorophyll. Whenever used, the concentration of la- belled phosphate was 10'2M and the concentrations of fir- ricyanide PDOX and methylamine were 5x10‘4M, 1.25x10' M, and 2x10‘éM respectively. 86 be considerably greater than the ferricyanide phosphoryl- ation and exhibit a very different pH optimum. Time Course of the Reduction of PDQX and of Accompanying Phosphorylation Figure 12 shows the time course of electron transport and phosphorylation with PDox as electron ac- ceptor. As expected, the kinetics of PDox-phosphoryl- ation is biphasic with an inflection when all the PDox is reduced. During the first phase, phosphorylation is proportional to the reduction of PDOX and both in turn are proportional to the period of illumination. The rate of phosphorylation during the second phase is un- doubtedly higher than the rate of phosphorylation using endogenous electron carriers and shows no sign of decreas- ing with time. This is indicative of the operation of some kind of cycle in which reduced PD (or something pro— duced therefrom) participates. (As a matter of fact, we have observed a two to three-fold stimulation of "endog- enous" phosphorylation by adding reduced PD.) Since the second phase of phosphorylation is also completely sensi- tive to DCMU, reduced and oxidized forms of PD may con- tribute to a cycle involving a site closely associated with photosystem II. Figure 13 shows that increasing concentrations of PDox stimulate both the phases. The pronounced eqivclent) e Reducfion(rnumoles 87 450- 300- oFT‘P .120 / [1° VOfibflo /" ‘ 0" ./ ,/ I ‘90 x J ’ H ° t '50- ’ t ‘I‘IIII‘II . 3 O n ?|g|||||l||l|||l‘|| I‘IIIII l l I O 30 60 90 ‘20 Period of ”I'ls ) Figure 12. Time course of the reduction of the oxidized form of p-phenylene—diamine (PDOX) and the accompanying phosphorylation. PDox was formed by adding 1.25x10‘4M p-phenylenediamine to the reaction mixture followed by an equivalent amount (2.5x10'4M) potassium ferricyanide. Amounts of the reduced p-phenylene-diamine formed were measured by titration with more ferricyanide. Reaction conditions were otherwise as in fig. 1 except that broad band red light was used. ”Exogenous phos." in the absence of added electron acceptors. (div $3l°wfiw)UOHo|Kloquoqd refers to the phosphorylation momoles ATP formed 200 100 88 ’f 2.5xlO4M I .1.25x164M 0’ / l' . -5 «cun‘o 6. 7 X10 M Period of III. (min) Figure 13. Time course of phosphorylation with different concentrations of PDox' Reaction conditions as in fig. 12. 89 stimulation of phosphorylation in the second phase with higher concentration of PDQx again supports our conten- tion that the second phase represents a cyclic process involving PD or products derived from PD. When PDOX is used in combination with a cyclic electron acceptor such as pyocyanine, we found that the mixture gives the PDox-phosphorylation rate as long as any PDox is present. Immediately after its reduction is complete, the mixture takes up the pyocyanine phos— phorylation rate (Figure 14). Similar results were also obtained with ferricyanide and pyocyanine mixtures. In- deed, it has been quite generally observed that the presence of oxidants inhibits all kinds of cyclic and "pseudo-cyclic" phosphorylation but that these processes commence as soon as the oxidant is all reduced. There- fore, we do not think that the cycle in the second phase continues to operate during the first phase. However, if it does, the reported stoichiometry of PDbx-phosphoryl- ation is meaningless. PDox-Phosphorylation as a Function ofng Figure 15 shows that PDox-phosphorylation is notably different from ferricyanide-phosphorylation. PDox-phosphorylation has a much broader pH-optimum than has ferricyanide-phosphorylation. If this broader opti- mum is really an expression of two over-lapping optima, mumoles ATP formed 400 200 90 Figure 14. PDox+ Pyo . l A l l l 30 60 90 I20 Period of lH.(s) Inhibition of cyclic photophosphorylation by PDOX. Concentrations used were: pyocyanine (Pyo), 2.5x10'5M; p-phenylenediamine, 1.25 x10‘4M; potassium ferricyanide, 2.5x10‘4M. Reaction conditions as in fig. 12. Note that the combination of pyocyanine + PDox gives phosphorylation rates typical of PDox alone until all of the PD0x has been reduced. Thereafter the pyocyanine-mediated cyclic phosphorylation takes over. 91 § § umoles ATP formed. mg chl'l-h-l " 100 - I 8 ,l _0 1 O I l l l I l l 7.0 8.0 9.0 10.0 p H Figure 15. Effect of pH on the photophosphorylation reactions associated with the reduction of PDox and of ferricyanide. Solid line represents PDox-phosphorylation. The reaction mixture contained in a total volume of 2 ml chloroplasts containing 40 ug chlorophyll. Concentration of reagents used were: sucrose, 0.1 M; wide range buffer (bistrispropane), 4x10"2 M; KCl, 10-3M; MgC12 3x10-3M; Na2H32P04, 10‘2M; ATP, 10-3 M; glucose, lO‘fiM; hexokinase, l mg/2 m1; ferricyanide, 5x10 M; p-phenyl-enediamine (PD), 1.25x10‘4M (when used). Reaction mixture illuminated by saturating broad band red light for 30 sec. 92 Ione might suspect the involvement of two sites of phos- phorylation with different response to pH. The pH opti- mum for ferricyanide-phosphorylation is known to be around 8.5 (28), but PDox-phosphorylation can be as good at pH 7.0 as at pH 8.5. Therefore, we have considered the possibility that the site between the two photosys- tems (Figure 10) has a pH optimum over 8.0 while a second site located elsewhere may have a pH optimum around 7.0. The difference in PDOX- and ferricyanide-phosphorylation could then be explained on the basis of difference in the electron transport pathways. Ferricyanide might ac- cept electrons after the site whose pH Optimum is 8.5, while PDox might accept electrons between the two sites. If this were so the PDOX-phosphorylation by avoiding the high pH site, could phosphorylate efficiently at pH 7.0 but ferricyanide phosphorylation, utilizing both sites, would be seriously inhibited at pH 7.0. Why, then, is the rate of phosphorylation at pH 8.0 and above, the same with PDox as with ferricyanide or flavins or methylviologen? It is possible that the primary site of PDox action changes with change in pH since PDox is a better electron acceptor at low pH than at high pH. Perhaps at higher pH's the much greater electron potential of photosystem I gives reduction of PDox by this pathway a competitive advantage. 93 Sensitivity of PDobehosphorylation to DCMU and Antimycin Figures 16 and 17 illustrate differential sensi- tivities of PDOX- and ferricyanide-mediated phosphoryla- tion to DCMU and antimycin. It was found that in both cases PDox-phosphorylation is more sensitive than ferri- cyanide-phosphorylation. However, PMS-phosphorylation is even less sensitive to DCMU than either PDOX- or ferricyanide-phosphorylation. This is compatible with the existence of a different site for PMS-mediated phos- phorylation but does not prove it; the difference might reside in peripheral details of the electron transport pathways. Like phosphorylation, the electron transport in- volved in the reduction of PDox was also found to be more sensitive to antimycin than the electron transport involved in ferricyanide reduction (Figure 17). At this pH (8.2) antimycin is acting presumably as an electron transport inhibitor (67). These results suggest that PDox accepts electrons at a site more closely associated with photosystem II than the site of ferricyanide reduction. Apparent Km for Phosphate of PDox Phosphorylation Figure 18 shows that although the Vmax of phosphorylation is higher than that of ferricyanide- 94 400 '- . mg chl-l. h-1 200 '0 . umoles ATP formed Figure 16. x 108M DCMU Sensitivities of different phosphorylation reactions to DCMU. All reactions were carried out at pH 7.0. Concentration of PMS used was 5x10‘4M. Reactions conditions otherwise as in fig. 15. 95 ". h" NJ (3 CD 100 umoles ATP formed.mg chl l 1500 Ilonpag 1000 . a samum) uo l a: 8 tabs |q36w'(m l- l l l o I l O 3.0 Figure 17. o )4 l 7.0 13.0 20.0 6 Antimycin A x10 M Sensitivities of electron transport and phosphorylation to antimycin with different electron acceptors. Reaction conditions as in fig. 15 and 16 except that the pH was 8.2 and tricine replaced bistris propane as buffer. Chloroplasts were incubated with antimycin for 5 min. in the dark before the reaction. Reduction of PDox was followed by following spectrophotometrically the consequent reduction of the excess ferricyanide as decrease in absorbance at 420 nm. Red refers to the rate of reduction while pp represents phosphorylation. 96 1/ Pi x 103M Figure 18. Apparent Km's for phosphate of the phosphorylation reactions with PDOX, ferricyanide and PMS as electron acceptors. All reactions were carried out at pH 7.0 under.the condition of fig. 16. Km for phosphate during PDox or ferricyanide reduction 2.4x10‘4M. Km for phosphate in PMS-mediated cyclic phosphorylation 7.0x10' M. 97 phosphorylation, the apparent Km's for phosphate are the same. The phosphorylation sites used by these electron acceptors, therefore, cannot be distinguished on the basis of their affinities for phosphate. However, the apparent Km for PMS-phosphorylation is altogether dif- ferent. This observation seems consistent with the idea of a separate site for cyclic photophosphorylation. On the other hand, ferricyanide-phosphorylation and PMS- phosphorylation may share a common phosphorylation site. If they do share one site, ferricyanide-phosphorylation must also utilize a second site with the lower Km which is rate determining. Quantum Efficiency of PDox-Phosphorylation We found that the quantum efficiency of PDOX- phosphorylation at pH 7.0 and in limiting light intensi- ties is about half of that of ferricyanide-phosphoryla— tion (Figure 19) although the light saturation level of PDox-phosphorylation is much higher than that of ferri- cyanide-phosphorylation. This observation is most eas- ily explained by assuming that the limited number of electrons available at low light intensities traverse. two sites of phosphorylation when reducing ferricyanide but only one site when reducing PDox (see Figure 11). At high light intensities where the supply of electrons is no longer rate-determining the rates of phosphorylation 98 100 ' 7.: T.‘ v°¢"° 4: ' 11" U // U) #. E ' / Ferricyanide . ‘Ill 1’ 3 E I 3 50 - , / IL [I '- I < I] 3 x o l e i- I Q . ,I’ P 0” '0' 0L I I I I 0 50 100 Light lntensity(°/.of 37 Kergs.cm.2.s4) Figure 19. Phosphorylation rates during the reduction of ferricyanide and PD as functions of light intensity. All reactions were carried out at pH 7.0. Illumination was by 650 nm light (interference filter). Various intensities, expressed as percent of the 37 Kergs cm"2 sec-1, were obtained by the use of fine mesh screens as neutral filters. Otherwise conditions were as in fig. 15. Note that the rate or phosphorylation with PDox at limiting light intensities (the quantum efficiency) is only one half of the corresponding rate with ferricyanide. OX 99 and electron transport become dependent on the charac- teristics of the two different sites. And, as we have already suggested the second site of phosphorylation which is used by ferricyanide but not by PDox may be severely inhibited at pH 7.0. With increasing pH's the quantum efficiency of PDox-phosphorylation gradually increases and at pH 8.0, the above described differences in quantum efficiency almost disappear. This led us to propose that as the pH increases PDox also starts accepting electrons after the second site of phosphorylation (see above argument re- garding effect of pH on redox potential of PDox)° Photophosphorylation in Tris-Washed Chloroplasts and the Possibility of a Phosphorylation Siteon the Water Oxidation Side of Photosystem II The results of our experiments with the oxidized form of p-phenylenediamine (PDOX) have added considerable support to the concept of a phosphorylation site closely associated with photosystem II. As we have already pointed out, Tris-washed or heated chloroplasts lose the capacity to oxidize water but are still able to accept electrons from a number of exogenous donors such as PD (49) or high concentrations of ascorbate (65). We have also pointed out that the computed efficiency of elec- tron transport in making ATP decreases to half when as- corbate replaces water as electron donor (65). With 100 these observations in mind, we undertook a study of the effect of PDox on phosphorylation in Tris-washed chlor- oplasts. Specifically we investigated phosphorylation in such chloroplasts using PD as electron donor and looked at the effects of added PDox- Unlike Yamashita and Butler (49), we always found considerable phosphorylation in Tris-washed chlor- oplasts when we used PD as electron donor even in ab- sence of any known electron acceptor (Table VII). To date we do not have any explanation for this discrepancy. We checked the possibility of the presence of oxidizing compounds as contaminants of PD which might accept elec- trons. Twice recrystallized PD in the presence of excess ascorbate also gave good phosphorylation rates. With PD as electron donor and no added acceptor, the phosphoryl- ation rates with Tris-washed chloroplasts were nearly as high at pH 7.0 as with the control chloroplasts and the addition of an electron acceptor (methylviologen) did not significantly increase the rates. However, at pH 8.0, the rates of phosphorylation in Tris-washed chloroplasts with PD as electron donor were almost precisely half of the rates in control chloroplasts with water as electron donor (Table VII). Whenever we added PDox to Tris-washed chloro- plasts using exogenous electron donors such as PD or benzidine, we found no phosphorylation (Table VIII). 101 Table VII Effect of Tris-Washing of Chloroplasts on Photophos- phorylation Chloro- Electron Electron pH of the umole ATP plasts Donor Acceptor Reaction formed'-m Mixture chl"l - h- Control H20 "Endogenous" (l) 7.0 18 ll II II 8.0 24 " " Methylviologen 7.0 121 ll ll ll 8.0 232 Tris- PD "Endogenous" 7.0 100 Washed " " " 8.0 120 " " Methylviologen 7.0 105 ll ll ll 8.0 120 lProbably containing autooxidizable substances such as oxidation products of naturally occurring phenols. Whenever used, the concentrations of PD and methylviologen were 5.0x10'4M and 5x10-5M respectively. 102 Table VIII Suppression of Phosphorylation by PDox in Tris-washed Chloroplasts (All reactions were run at pH 7.0) Electron Electron umole ATPlforTed Donor Acceptor -mg chl 'h H20 Ferricyanide 3 " Methylviologen 2 PD "Endogenous" 106 " PDOX 4 " Methylviologen 94 " Methylviologen + PDOX 5 Benzidine "Endogenous" 36 " PDOX 2 " Methylviologen 26 " Methylviologen + PD0x 3 ' Whenever used the concentrations of ferricyanide, methylviologen, PD, Benzidine and PDOX were 5x10-4M, 5x10-5M, 1.5x10’3M, 1.5x10'3M and 1.25x10'4M respective- ly. 103 We have interpreted these observations in terms of the interception of electron flow by PDox between the two phosphorylation sites as indicated in Figure 11. Ac- cording to this tentative hypothesis the System II site is by-passed by the artificial donors such as PD, ascor- bate etc. Therefore, in Tris-washed chloroplasts, only the second "System I" site is used for phosphorylation. Unfortunately for our model, we observed further that with reduced PD as donor, pH 7.0 and 8.0 do not make any significant difference in the phosphorylation rates of Tris-washed chloroplasts. This is not consistent with our earlier contention that the two sites have different pH optima with the second, System I site operating best above pH 8.0. We have already seen that PDox prevents phos- phorylation in Tris-washed chloroplasts, perhaps by in- tercepting electron flow somewhere before the second phosphorylation site (Figure 11). If our interpretation is correct then we should be able to show that the sup- pression of phosphorylation depends on the concentration of PDOX. We should also be able to show that in Tris- washed chloroplasts, illumination with PD alone gradu- ally results in a decrease in phosphorylation since PD in donating electrons should be oxidized to the inhibi- tory PDox. Furthermore, we should be able to hasten the onset of this PDox-suppression of phosphorylation by 104 increasing the chloroplast concentration of the reaction mixture. All these expectations have been experiment- ally realized. Figure 20 shows that increasing concen- trations of PDox increasingly suppress phosphorylation and at about 1.2 x 10'4M PDoxr phosphorylation is com- pletely prevented. Figure 21 represents the time course of photophosphorylation with PD as the electron donor. It is evident that phosphorylation is suppressed after about 8 minutes of illumination. It can be computed from Figure 20 that during this time the chloroplasts should, indeed, have accumulated just enough PDOX necessary for total suppression of phosphorylation. It has also been shown that this decline in phosphorylation rate is not the result of aging of the chloroplasts. Higher concen- trations of chloroplasts stop phosphorylating faster than lower concentrations of chloroplasts, again demon- strating that the decline in rate is probably not an aging effect but rather results from the accumulation of reaction products. 105 '1'] I l 100- \ umoles ATP formed.mg chl- 50 " O 1 l 1 1 1 ‘7"; O 1.0 2.0 Ferricyanide added, 104M Figure 20. Inhibition of phosphorylation in tris-washed chloroplasts by PDox. All reaction mixtures contained the same initial amount of electron donor, 1.5x10'3M p-phenylenediamine. Different amounts of PDox were then formed by adding the indicated amount of potassium ferricyanide. Reaction conditions otherwise as in fig. 15. 106 TZC) ' -o O o E 0 LL. 80- a. p— < O E o E :40“. E 0 l 1 1 g 1 O 2 4 6 8 lHunfinofion (nfin ) Figure 21. Inhibition of phosphorylation in Tris~washed chloroplasts by prolonged illumination using PD as electron donor. The initial concentration of the electron donor, p-phenylenediamine, was 1.5x10'3 M, pH 7.0. Other reaction conditions as in fig. 15. The inhibition with time is not due to aging, but seems rather related to the formation of PDox when the p-phenylenediamine donated electrons. See fig. 20. DISCUSSION Oxidized forms of p-phenylenediamine (probably mostly 1,4-benzoquinonediimide) have been found to stim- ulate both electron transport and phosphorylation simul- taneously. This stimulated PDox-phosphorylation is dif— ferent from ferricyanide-phosphorylation in its pH optimum, quantum efficiency and in its sensitivity towards inhibitors like DCMU. In Tris-washed chloro- plasts, the phosphorylation supported by exogenous electron donors is completely suppressed by PDox. All these findings have prompted us to suggest that there may be a site for non-cyclic photophOSphorylation close- ly associated with the water oxidation step of photo- system II. Our ideas about this phosphorylation site and about the site of action of PDox have been presented in relation to the tentative scheme of electron trans- port shown in Figure 11. Although we have not isolated the water-oxida- tion part of the electron transport chain from chloro- plasts and have not demonstrated conclusively that a phosphorylation site is located there, we have produced more than one piece of evidence to support our contention. 107 108 This indirect evidence has already been discussed in a preliminary way in the results section. We have evidence which suggests that PDox may accept electrons at more than one place. ‘At pH 7.0 the rate of phosphorylation with PDox is much higher than with ferricyanide or other category (a) acceptors but at higher pH's this difference tends to disappear. Fur- thermore, the quantum efficiency of phosphorylation at pH 7.0 is only one half as great with PDox as with fer- ricyanide and this difference too disappears at higher pH's. However, the high rates of electron transport in the absence of phosphorylation or added uncouplers which is characteristic of the PDOX reducing system are still present at high pH. Therefore, if we are to interpret these high rates in terms of a site of reduction which is different from the site of reduction of ferricyanide, we must postulate that both sites can still be used at pH 8.0 or above. In other words, we are proposing that the differences in phosphorylation at high and low pH's represent different efficiencies and hence different levels of competition between the two pathways. Whether or not the proposed second site of phos- phorylation is on the water oxidation side of photosys- tem II, it seems probable that by using different elec- tron donors and different electron acceptors, we have succeeded in isolating two sequential sites of 109 phosphorylation in the non-cyclic pathway of electron flow. These observations are in good agreement with our earlier observation that the P/e2 is always over one and when corrected for non-phosphorylating electron transport approaches two. Our maximum thesis, then, is as depicted in Figure 11 while our minimum thesis is simply that PDOX accepts electrons between two phos- phorylation sites. Unfortunately, even this minimum thesis cannot be considered as proved. Alternative explanations of our data are almost certainly available. One alterna- tive explanation of the inhibition of phosphorylation by PDox in Tris-washed chloroplasts is that PDox or one of its numerous breakdown products is inhibitory to these maltreated chloroplasts. If so, such inhibitions must be reversible since the addition of ascorbate which re- converts the PDox into PD restores most of the phosphor- ylation capacity to the chloroplasts. Therefore, this particular alternative hypothesis is not very attractive. However, many further studies will have to be made be- fore our tentative proposal embodied in the scheme in Figure 11 can be entertained with deep conviction. One particular difficulty with PDox is its notorious insta- bility; consequently, we cannot define our system in terms of the chemistry involved with anything like 110 precision. It would be extremely useful if we could find a more stable analog of PDox- LI TERATURE C ITED LITERATURE CITED Slater, E. C. 1953. Mechanism of phosphorylation in the respiratory chain. Nature. 172:975. Bieber, L. L., and P. D. Boyer. 1966. 32P-Labelling of mitochondrial protein and lipid fractions and their relation to oxidative phosphorylation. J. Biol. Chem. 241:5375. Mitchell, R. A., R. D. Hill, and P. D. Boyer. 1967. Mechanistic implications of Mg++, adenine nuceleotide and inhibitor effects on energy linked reactions of submitochondrial particles. J. Biol. Chem. 242:1793. Green, D. E., J. Asai, R. A. Harris and J. T. Penniston. 1968. Conformational basis of energy transforma- tions in membrane systems. III. Configurational changes in the mitochondrial inner membrane induced by changes in functional states. Arch. Biochem. Biophys. 125:684. Weber, N. E., and P. V. Blair. 1969. Untrastructural studies of beef heart metochondria. I. Effects of adenosine diphosphate on mitochondrial morphology. Biochem. Biophys. Res. Commun. 36:987. Robertson, R. N. 1968. I3 Protons, electrons, phsophorylation and active transport. Cambridge monographs in experimental biology-15, p. 1. Cambridge University Press, England. ' Pullman, M. E., H. S. Penefsky and E. Racker. 1958. A soluble protein fraction required for coupling phosphorylation to oxidation in submitochondrial fragments of beef heart mitochondria. Arch. Biochem. Biophys. 76:227. Mitchell, P. 1961. Coupling of phosphorylation to electron and hydrogen transfer by a chemi-osmotic type of mechanism. Nature. 191:144. 112 10. ll. 12. 13. 14. 15. 16. 17. 18. 19. 20. 113 Mitchell, P. 1966. Chemiosmotic coupling in oxidative and photosynthetic phosphorylation. Biological Reviews. 41:445. Neumann, J. S., and A. T. Jagendorf. 1964. Light induced pH changes related to phosphorylation by chloroplasts. Arch. Biochem. Biophys. 107:109. Shen, Y. K., and G. M. Shen. 1962. Studies on photo- phosphorylation. II. The "light intensity effect" and intermediate steps of photophosphorylation. Scientia Sinica. 11:1097. ‘ ‘ Jagendorf, A. T., and G. Hind. 1963. Studies on the mechanism of photophosphorylation. In Photosyn- thetic mechanisms of green plants. National Res. Council. 1145:599. Jagendorf, A. T., and E. Uribe. 1966. ATP formation caused by acid-base transitioncfifspinach chloro- plasts. Proc. Nat. Acad. Sci. U. S. 55:170. Turner, J. F., C. C. Black, and M. Gibbs. 1962. Studies on photosynthetic processes. I. The effect of light intensity on triphosphopyridine nucleotide reduction, adenosine triphosphate formation, and carbon dioxide assimilation in spinach chloro- plasts. J. Biol. Chem. 237:577. Schwartz, M. 1968. Light induced proton gradient links electron transport and phosphorylation. Nature. 219:915. Sakurai, H., M. Nishimura, and A. Takamiya. 1965. Studies on photophosphorylation. I. Two-step excitation kinetics of photophosphorylation. Plant and Cell Physiol. 6:309. Arnon, D. I. 1949. Copper enzymes in isolated chloro- plasts. Polyphenol oxidase in Beta vulgaris. Plant Physiol. 24:1. Avron, M. 1960. Photophosphorylation by Swiss-chard chloroplasts Biochim. Acta. 40:257. Nielsen, S. 0., and A. L. Lehninger. 1955. Phos- phorylation coupled to the oxidation of ferro- cytochrome c. J. Biol. Chem. 215:555. Joliot, P. 1966. Oxygen evolution in algae illuminated by modulated light. Brookhaven Symposia in Biology. 19:418. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 114 Good, N. E., S. Izawa, and G. Hind. 1966. Uncoupling and energy transfer inhibition in photophosphory- lation. £3 Current topics in bioenergetics. Vol. I. Ed. by Sanadi. pp. 75-112. Acad. Press. New York. Avron, M., and N. Shavit. 1965. Inhibitors and un- couplers of phot0phosphorylation. Biochim. Biophys. Acta. 109:317. Izawa, S., and N. E. Good. 1968. The stoichiometric relation of phosphorylation to electron transport in isolated chloroplasts Biochim. Biophys. Acta. 162:380. ' Avron, M., and J. Neumann. 1968. Photophosphorylation in chlor0plasts. Ann. Rev. Plant Phys. 19:137. Del Campo, F. F., J. M. Ramirez, and D. I. Arnon. 1968. Stoichiometry of photosynthetic phosphorylation. J. Biol. Chem. 243:2805. Horton, A. A., and D. 0. Hall. 1968. Determination of the stoichiometry of photosynthetic phosphorylation. Nature. 218:386. Lynn, W. S., and Brown, R. H. 1967. P/Ze- Ratios approaching 4 in isolated chloroplasts. J. Biol. Chem. 242:412. Winget, G. D., Izawa, S., and N. E. Good. 1965. The stoichiometry of photophosphorylation. Biochem. Biophys. Res. Comm. 21:438. Izawa, S., and N. E. Good. 1965. Hill reaction rates and chloroplast fragment size. Biochim. Biophys. Acta. 109:372. Umbriet, W. W., R. H. Burris, and J. F. Stauffer. 1957. Manometric Techniques. Third ed. Burgess Publish- ing Co., Minneapolis, 338p. Ames, B. N. 1966. Assay of inorganic phosphate, total phosphate and phosphatases. In Methods in enzymology. Vol. VIII. Complex carbohydrates. Ed. by E. F. Neufeld and V. Ginsberg. pp. 115-118. Acad. Press. New York. Block. R. J., E. L. Durrum, and G. Zweig. 1958. In Manual of paper chromatography and paper electro- phoresis. p. 197. Acad. Press. New York. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 115 Kahn, J. S., and A. T. Jagendorf. 1961.‘ An enzyme from spinach chloroplasts catalyzing adenosine triphosphate-adenosine diphosphate exchange. J. Biol. Chem. 236:940. Mehler, A. H. 1951. Studies on reactions of illuminated chlorOplasts. II. Stimulation and inhibition of the reaction with molecular oxygen. Arch. Biochem. and Biophys. 34:339.- Arnon, D. I., F. R. Whatley, and M. B. Allen. 1958. Assimilatory power in photosynthesis. Photosyn- thetic phosphorylation by isolated chloroplasts is coupled with TPN reduction. Science. 127:1026. Hill, R".and F. Bendall. 1960. Function of the two cytochrome components in chlorOplasts: a working hypothesis. Nature. 186:136. Emerson, R., R. Chalmers, and C. Cederstrand. 1957. Some factors influencing the long-wave limit of photosynthesis. Proc. Nat. Acad. Sci., U. S., 43:133. - ' Duysens, L. N. M. 1954. Reversible photooxidation of a cytochrome pigment in photosynthesizing~ Rhodospirillum rubrum. Nature. 173:692. Avron, M. 1967. Mechanism of photoinduced electron transport in isolated chloroplasts. In Current t0pics in bioenergetics. ed. by D. R. Sanadi. p. 1. Academic Press. New York. - Chance, B., A. San-Pietro, M. Avron, and W. W. Hildreth. 1965. In Non heme iron proteins. A. San-Pietro ed. pp._725. Antioch Press. Yellow Springs, Ohio. Kok, B. 1965. Photosynthesis: The path of energy. In Plant biochemistry. J. Bonner and J. E. Varner eds. p. 903, Acad. Press. New York. Levine, R. P. 1969. The analysis of photosynthesis using mutant strains of algae and higher plants. Ann. Rev. Plant Physiol. 20:523. Rumberg, B., P. Schmidt-Mende, J. Weikard, and H. T. Witt. 1963. Correlation between absorption changes and electron transport in photosynthesis. In Photosynthesic mechanisms of green plants. p. 18. NAS-NRC publication 1145. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 116 Duysens, L. N. M., and H. E. Sweers. 1963. Mechanism of two photochemical reactions in algae as studied by means of fluorescence. I£_Studies in microalgae and photosynthetic bacteria. p. 353. University of Tokyo Press, Tokyo. Duysens, L. N. M., and J. Amesz. 1967. Photosynthesis. £g_Comprehensive biochemistry. Ed. by M. Florkin and E. H. Stotz. 27:237. Izawa, S. 1970. Photoreduction of 2,6-dichloropheno- lindophenol by chloroplasts with exogenous Mn++ as electron donor. Biochim. Biophys. Acta. (In press) Hind, G., H. Y. Nakatani and S. Izawa. 1969. The role of C1" in photosynthesis. I. The chloride requirement of electron transport. Biochim. BiOphys. Acta. 172:277. Yamashita, T., and W. L. Butler. 1968. Inhibition of chloroplasts by UV-irradiation and heat-treatment. Plant Physiol. 43:2037. Yamashita, T., and W. L. Butler. 1968. Photoreduction and photophosphorylation with tris-washed chloro- plasts. Plant Physiol. 43:1978. Yamashita, T., and W. L. Butler. 1969. Inhibition of the Hill reaction by tris and restoration by electron donation to photosystem II. Plant Physiol. 44:435. Vernon, L. P., and E. R. Shaw. 1969. Oxidation of 1,5-diphenylcarbazide as a measure of photosystem 2 activity in subchlorOplast fragments. Biochem. Biophys. Res. Commun. 36:878. Krogmann, D. W., A. T. Jagendorf, and M. Avron. 1959. Uncouplers of spinach chloroplast photosynthetic phosphorylation. Plant Physiol. 34:272. Izawa, S., and G. Hind. 1967. The kinetics of the pH rise in illuminated chloroplast suspensions. Biochim. Biophys. Acta. 143:337. Karlish, J. D., and M. Avron. 1967. Relevance of proton uptake induced by light to the mechanism of energy coupling in photophosphorylation. Nature. 216:1107. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 117 Lynn, W. S., and R. H. Brown. 1967. P:O and ADP:O ratios and quantum yields in chloroplasts, with the use of chloranil as electron acceptor. J. Biol. Chem. 242:418. Avron, M., and B. Chance. 1966. Relation of phosphory- lation to electron transport in isolated chloro- plasts. Brookhaven Symposia in Biology. 19:149. Gorman, D. S., and R. P. Levine. 1965. Cytochrome f and plastocyanin: Their sequence in the photo- synthetic electron transport chain of Chlamydomonas reinhardi. Proc. Nat. Acad. Sci. U. S. 54:1669. Hill, R., and D. A. Walker. 1959. Pyocyanine and photophosphorylation with chloroplasts. Plant Physiol. 34:240. Duane, W. C., Sr., M. H. H5h1 and D. W. Krogmann. 1965. PhotOphosphorylation activity in cell-free prep- arations of a blue-green alga. Biochem. Biophys.. Acta. 109:108. Lee, S. S., A. Michael Young, and D. W. Krogmann. 1969. Site specific inactivation of the photophosphoryla- tion reactions of Anabaena variabilis. Biochem. BiOphys. Acta. 180:136. Duysens, L. N. M., and J. Amesz. 1962. Function and identification of two photochemical systems in photosynthesis. Biochem. Biophys. Acta. 64:243. Gromet-Elhanan, z., and D. I. Arnon. 1965. Effect of desaspidin on photosynthetic phosphorylation. Plant Physiol. 40:1060. Laber, L. J., and C. C. Black. 1969. Site-specific uncoupling of photosynthetic phosphorylation in spinach chlorOplasts. J. Biol. Chem. 244:3463. Izawa, S., and N. E. Good.' 1968. Effect of p-chloromercuribenzoate (PCMB) and mercuric ion ‘on chloroplast photophosphorylation. International congress on photosynthesis. Freudenstadt, Germany. (In press) Bdhme, H., and A. Trebst. 1969. On the pr0perties of ascorbate photooxidation in isolated chloroplasts. Evidence for two ATP sites in noncyclic photo- phosphorylation. Biochim. Biophys. Acta. 180:137. 118 66. Ikehara, N., and Sugahara, K. 1969. Inactivation of Hill reaction in spinach chloroplasts by pre- treatment of tris buffer in the light. Bot. Mag. Tokyo 82:271. 67. Drechsler, Z., N. Nelson, and J. Neuman. 1969. Antimycin A as an uncoupler and electron transport inhibitor in photoreactions of chloroplasts. Biochem. Biophys. Acta. 189:65. 68. Hill, R., and R. Scarisbrick. 1951. The haematin compounds of leaves. New Phytol. 50:98. Miliiiiiiflifli[(ifllifliiililiiiiflilim