Proc. Natl. Acad. Sd. USA Vol. 74, No. 8, pp. 3377-81, August 1977
Biophysics
Quantum efficiency of photosynthetic energy conversion (photosynthesis/photophosphorylation/ferredoxin)
RICHARD K. CHAIN AND DANIEL I. ARNON Department of Cell Physiology, University of California, Berkeley, Berkeley, California 94720
Contributed by Daniel I. Arnon, June 7,1977
ABSTRACT The quantum efficiency of photosynthetic energy conversion was investigated in isolated spinach chloroplasts by measurements of the quantum requirements of ATP
formation by cyclic and noncyclic photophosphorylation catalyzed by ferredoxin. ATP formation had a requirement of about 2 quanta per 1 ATP at 715 nm (corresponding to a requirement of 1 quantum per electron) and a requirement of 4 quanta per ATP (corresponding to a requirement of 2 quanta per electron) at 554 nm. When cyclic and noncyclic photophosphorylation were operating concurrently at 554 nm, a total of about 12 quanta was required to generate the two NADPH and three ATP needed for the assimilation of one CO2 to the level of glucose.
Few areas of photosynthesis have received more intensive theoretical and experimental study and generated more controversy than the efficiency with. which photosynthetic cells convert the electromagnetic energy of light into chemical energy (for review, see refs. 1 and 2). Two different concepts, never reconciled during the lifetimes of their main protagonists, emerged from the many investigations. One concept, espoused by Warburg et al. (3), was that photosynthetic quantum conversion has an efficiency of about 90%-i.e., that energy equivalent to that of 3 einsteins of red quanta (42 kcal each) is sufficient to liberate 1 mol of 02 (corresponding to 1/6 mol of glucose, for which AGO' = 686/6 = 114 kcal). In contrast, Emerson (4) and his followers (5) concluded that photosynthetic efficiency was much lower, of the order of 8 to 12 quanta per 02, a range that is widely accepted today even though values less than 8 have, at times, been obtained by investigators (6, 7) who did not share Warburg's conclusions. Most studies of photosynthetic quantum efficiency were based on measurements of light-induced production of 02 (corrected for concurrent respiration) during complete photosynthesis by whole cells, usually unicellular algae of the Chlorella type. Discordant results were attributed to experimental variables such as errors in methods (usually manometric) of 02 measurement, variations in the concurrent 02 consumption by respiration, participation of respiratory intermediates in photosynthesis, need for supplementary (catalytic) illumination, and nutritional history, age, and physiological status of the cells (1-5). Left unchallenged, however, was the main (and, to us, dubious) premise underlying these studies with whole cells-namely, that the photoproduction of 1 mol of 02 always corresponds to the assimilation of 1 mol of CO2 to the level of glucose and that, therefore, 02 evolution is a reliable measure of the total amount of chemical energy stored. A different perspective and experimental approach to the question of photosynthetic quantum efficiency emerged from studies of photosynthesis by isolated chloroplasts in which the process was physically separated into a light phase concerned The costs of publication of this article were defrayed in part by the payment of page charges from funds made available to support the research which is the subject of the article. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
with cyclic and noncyclic photophosphorylation and a "dark," enzymatic phase concerned with the assimilation of CO2 (8). Fractionation of chloroplasts (9) has established that the light phase is localized in the membrane fraction (grana) that is separable from the soluble stroma fraction which contains the enzymes of CO2 assimilation (10). Thus, in isolated and fractionated chloroplasts, investigations of photosynthetic quantum efficiency can be focused solely on cyclic and noncyclic pho-. tophosphorylation, which jointly account for the conversion of photon energy into chemical energy without the subsequent or concurrent reactions of biosynthesis and respiration that cannot be avoided in whole cells. The present investigation was undertaken not to reactivate old and now dormant controversies but to relate overall photosynthetic quantum efficiency to the quantum efficiency of cyclic and noncyclic photophosphorylation, the two energy conversion reactions in plant photosynthesis that, for reasons discussed elsewhere (8, 11), can be investigated in isolated chloroplasts but not in whole cells. In cyclic photophosphorylation (12-14), ATP is the sole product and no 02 is produced (Eq. 1), whereas in noncyclic photophosphorylation (15, 16), 02 evolution is coupled with the reduction of ferredoxin and the formation of ATP (Eq. 2). hp
ADP + Pi -. ATP 4 Fdo. +2 H20 +2 ADP +2 Pi
','4Fdred+02+2ATP+4H+ [2] in which Fd0, is oxidized ferredoxin and Fdred is reduced ferredoxin. Reduced ferredoxin, an iron-sulfur protein electron carrier with a reducing power (Eo' = -420 mV) equal to that of molecular hydrogen (17), serves directly as a reductant in some reactions but for CO2 assimilation in plants the reductant is NADPH (E ' = -320 mV) which is formed enzymatically (18, 19) with no further input of photon energy: 4 FdreM + 2 NADP+ + 4 H+ Fd-NADP -
reductase
4Fdox +2NADPH+2H+ [3]
ATP is a product of both cyclic and noncyclic photophosphorylation (Eq. 1 and Eq. 2) which, according to recent evidence, may proceed concurrently in isolated chloroplasts (20, 21). Thus, the quantum efficiency of light-induced ATP formation by isolated chloroplasts provides a direct and reliable index of the efficiency with which photon energy is converted into chemical energy during photosynthesis. An argument against measurements of photosynthetic quantum efficiency in isolated chloroplasts is that isolation procedures may damage the photosynthetic apparatus and Abbreviation: DCMU, 3-(3,4-dichlorophenyl)-1,1-dimethylurea.
3377
[1]
3378
Biophysics: Chain and Arnon
result in low efficiency. That these and other experimental hazards do exist is illustrated by the wide range of values for the quantum efficiency of photosynthetic phosphorylation reported from different laboratories. The values (expressed henceforth as quantum requirements; minimal quantum requirements are equivalent to maximal quantum efficiencies) varied from 3 to 200 quanta of absorbed light per molecule of ATP formed (see review, ref. 22). The present investigation was prompted by recent findings (20, 21) that further documented the role of ferredoxin as the native catalyst of cyclic photophosphorylation (13, 14) and characterized the optimal experimental conditions under which high quantum efficiencies (low quantum requirements) of ferredoxin-catalyzed photophosphorylations would most likely be found. Low quantum requirements were indeed observed: 2 quanta per ATP under far-red monochromatic illumination (715 nm) that supported only cyclic photophosphorylation and 4 quanta per ATP under short-wavelength monochromatic illumination (554 nm) that supported both cyclic and noncyclic photophosphorylation. The theoretical implications of these values and their relationship to measurements of quantum requirements in whole cells are discussed. METHODS Chloroplasts. Chloroplasts were isolated from spinach leaves (Spinacia oleracea var. High Pack) grown in a greenhouse in a nutrient solution culture (23) and freshly harvested before each experiment. The chloroplast preparations were "broken" chloroplasts, prepared as described (20) except that an additional low-speed centrifugation step was added to remove any residual large fragments. The broken chloroplasts used consisted of lamellar material depleted of soluble chloroplast components.
Aerobic Conditions. Unless otherwise indicated, the reactions were carried out in glass cuvettes (2-mm light path) filled with reaction mixture. Aerobic conditions, provided in all experiments, mean here that the reaction mixtures were in equilibrium with air and contained dissolved oxygen (ca 0.25
mM) but were not otherwise deliberatedly aerated. Illumination. Monochromatic illumination (715 or 554 nm) was provided by a light beam from a 250-W air-cooled tungsten-halogen lamp (Type EHN, General Electric Co.); the beam was passed through interference filters (Baird-Atomic Co.) with 20-nm half-band width for 554-nm light and 10-nm half-band width for 715-nm light. Actinic illumination by 554- and 715-nm light beams, neither of which is strongly absorbed by chloroplasts, ensured that the entire reaction mixture received uniform illumination despite the fairly high chlorophyll concentration. Preliminary experiments established that, within the range of illumination used, the rates of photochemical reactions were directly proportional to light intensity. Light Measurements. Incident light intensity was measured with a radiation meter (Radiometer model 65, Yellow Springs Instrument Co.) calibrated against a National Bureau of Standards (Washington, DC) radiation standard. Selected confirmatory measurements of incident light intensity were made with a calibrated Eppley quartz-window surface-type linear thermopile. Incident illumination remained constant, as indicated by concordant measurements before and after each experiment and by little or no variation from day to day. Incident light intensity was measured by placing the measuring device where the sample cuvette was normally illuminated in the apparatus. Corrections were applied for measured reflection losses (7.5%) at the incident-to-cuvette side. The reliability of the incident light intensity measurements
Froc. Natl. Acad. Sci. USA 74 (1977)
was also confirmed through measurements of light intensity by chemical actinometry (at 420-nm) with the potassium ferrioxalate actinometer of Hatchard and Parker (24). The fraction of incident monochromatic illumination absorbed by chloroplasts was measured in each quantum efficiency experiment with a large Ulbricht integrating sphere (50-cm diameter) in a manner similar to that described by Warburg and Krippahl (25). An RCA 6217 photomultiplier tube served as the light detector inside the sphere; the photomultiplier output was measured outside the sphere with a digital voltmeter (Hewlett-Packard model 3440A). Monochromatic illumination to the sphere was provided by a light beam that was isolated by using the same interference filters used in providing actinic monochromatic illumination to the reaction mixtures. Analytical Procedures and Reagents. Chlorophyll and NADPH were determined as described (23, 26). ATP was measured by the method of Hagihara and Lardy (27). Ferredoxin was isolated either from spinach leaves (28) or from the blue-green alga Spirulina maxima (29). NADP+ and ADP were purchased from the Sigma Chemical Co. (St. Louis, MO). RESULTS Quantum Requirements of Cyclic Photophosphorylation at 715 nm. Cyclic photophosphorylation in vivo occurs in chloroplast lamellae that are in contact with the stroma fluid that contains dissolved 02 but in vitro the process traditionally has been investigated under anaerobic conditions, for reasons that are discussed elsewhere (8, 20). Recently (20), when ferredoxin-catalyzed cyclic photophosphorylation by isolated chloroplasts was investigated under the more physiological aerobic conditions, it was found to have several distinct features, including a low ferredoxin requirement. Cyclic photophosphorylation was optimally catalyzed by the same low concentrations of ferredoxin (10 ,gM) that are required for NADP+ reduction, concentrations much lower than those previously used (100 ,M) for cyclic photophosphorylation under anaerobic conditions (30). Another feature of ferredoxin-catalyzed cyclic photophosphorylation in the presence of 02 is its sensitivity to overreduction by electrons released by photosystem II from water (20). Experimentally, one way to poise the system and prevent overreduction is to use 715-nm illumination, a wavelength that activates photosystem I but not photosystem II (14, 31, 32). Illumination by 715-nm light generated only a "trickle" electron flow from water, just adequate to maintain the proper poising in the presence of 02(20). As cyclic photophosphorylation is driven by photosystem I (14, 32), the use of 715-nm illumination that is absorbed almost exclusively by photosystem I offered a great experimental advantage for obtaining the minimal quantum requirements for ATP formation by cyclic photophosphorylation. The fraction of 715-nm light absorbed by photosystem II (and the resulting electron trickle) is so small that it can be disregarded (14, 20). By contrast, shorter wavelengths (550 to be the physiological poising mechanism that triggers the operation of cyclic photophosphorylation alone in situations when excess ATP (but not reducing power) is needed; NADPH would then tend to accumulate (20, 21). Such a situation would arise, for example, during protein synthesis, when ATP is required for the activation of amino acids by amino acyl synthetases to form amino acid adenylates. Experimentally, the poising effect of NADPH is replaceable by DCMU or, as already discussed, far-red monochromatic light. A comparison of the stimulatory effects of NADPH and DCMU on cyclic photophosphorylation is given in Fig. 1. The action of NADPH was in the main similar and additive to the action of DCMU. In the absence of DCMU, the addition of NADPH more than doubled the rate of cyclic photophosphorylation. At lower concentrations of DCMU, the addition of NADPH gave a further substantial increase in photophosphorylation but at higher concentrations, DCMU alone was more effective than the combination of DCMU and NADPH. When the system was optimally poised by DCMU alone, the addition of NADPH was inhibitory. The beneficial poising effect of NADPH and DCMU on cyclic photophosphorylation was reflected in lower quantum requirements. About 10 quanta per ATP were required in the unpoised system, 6 quanta in the presence of NADPH, and 4 quanta in in the presence of DCMU (Table 3). In these experiments, optimal poising and the lowest quantum requirement were obtained with the concentration of DCMU used. The concentrations of NADPH and DCMU needed for optimal poising varied seasonally, depending on the relative activity of photosystem II. When cyclic photophosphorylation was optimally poised, 4 quanta per ATP represented the lowest quantum requirement for cyclic photophosphorylation at 554 nm. Based on the premise that the formation of one ATP involves the transfer of
3380
Proc. Nati. Acad. Sci. USA 74 (1977)
Biophysics: Chain and Arnon
Table 3. Effect of NADPH and 3-(3,4-dichlorophenyl)-1,1dimethylurea (DCMU) on quantum requirements of cyclic photophosphorylation at 554 nm
Table 4. Quantum requirements of NADP+ reduction and concurrent noncyclic and cyclic photophosphorylations at 554 nm as influenced by the age of plants
Illumination Time, Absorbed min AE
Quantum requirements, uE/,umol ATP NADPH Illumination Light absorbed, formed, formed, ATP NADPH time, ,umol formed formed ,umol min ,E
2.5 5.0 10.0
1.01 2.02 4.04
2.5 5.0 10.0
1.01 2.02 4.04
2.5 5.0 10.1
1.03 2.06 4.12
ATP formed,
'4mol Control 0.134 0.205 0.366 +DCMU 0.252 0.469 0.979 +NADPH 0.190 0.338 0.621
Quantum requirement, ME/,gmol ATP 7.5 9.9 11.0 4.0 4.3 4.1 5.4 6.1 6.6
Experimental conditions as in Table 1 (554-nm illumination) except that 2.5 mM NADPH and 13 IAM DCMU were added where in-
dicated.
two electrons, the results signify a requirement of 2 quanta per electron under short-wavelength illumination, in contrast to a requirement of 1 quantum per electron under long-wavelength illumination (Table 1). Quantum Requirements in Concurrent Cyclic and Noncyclic Photophosphorylations. In the presence of substrate amounts of NADP+, 554-nm illumination activated simultaneously both cyclic and noncyclic photophosphorylations, a fact reflected in the excess of ATP formed over NADP+ reduced and in ATP:NADPH ratios greater than 1.0 (20, 21). Here, the monochromatic illumination, which was well below saturation (as required for measurements of quantum requirements), was partitioned between cyclic and noncyclic photophosphoryla*tion, each of which contributed a portion of the total ATP formed. By taking into account the total number of quanta absorbed by both systems and the total amount of ATP formed by them, we obtained a requirement of 4 quanta per ATP with chloroplasts isolated from young plants; the quantum requirement per ATP was almost twice as high with chloroplasts from old plants (Table 4). Thus, at 554 nm, the same requirement of 4 quanta per ATP (equal to 2 quanta per electron) was obtained, regardless whether the ATP was formed by cyclic photophosphorylation alone or by cyclic and noncyclic photophosphorylation operating concurrently. When the two systems operated concurrently, the fraction of absorbed quanta used for noncyclic photophosphorylation also generated NADPH in addition to 02 (see Table 4 and Eqs. 1-3). When cyclic photophosphorylation operated alone, the only product was ATP and the surplus photon energy that could have been used at shorter wavelengths to generate NADPH was lost. An accurate measurement of the quantum requirement for NADP reduction was not possible because there was no direct way to segregate the quanta that were used solely in noncyclic photophosphorylation, in which NADPH is formed. Using the total number of quanta absorbed by both cyclic and noncyclic photophosphorylation, we calculated a requirement of about 6 quanta per NADPH for chloroplasts from young plants and about 8 quanta for chloroplasts from old plants (Table 3). With chloroplasts from young plants, these quantum requirements for a concurrent formation of ATP and NADPH add up to a requirement of 12 quanta for the generation of the
2.5 5.0 10.0
2.5 5.0 10.0
Old plants (42 days) 0.152 0.173 0.268 0.324 0.562 0.528 Young plants (27 days) 0.277 0.165 1.05 0.360 2.10 0.522 0.716 1.025 4.20
1.13 2.26 4.52
6.5 7.0
_
7.4 8.4
8.0
8.6
3.8 4.0 4.1
6.4 5.8 5.9
NADP+ (2.5 mM) was added where indicated. Experimental conditions not otherwise specified were as in Table 1.
two NADPH and the three ATP that are consumed in the assimilation of one CO2 to the level of glucose (43). It is of interest that Egneus et al. (44), who measured the quantum requirement for CO2 assimilation by isolated spinach chloroplasts, found a requirement of about 12 quanta per CO2 at 674 nm.
DISCUSSION Past investigations of the efficiency with which photosynthetic cells convert light into chemical energy have usually relied on measurements of photoproduction of oxygen, in the belief that they signified the production of sugars as end-products in which the chemical energy is stored. This belief derives its justification from the photosynthetic quotient (02 evolved:CO2 fixed) being close to 1.0. But it has long been recognized that so small a variation from unity as 3%, which is well within the observed limits of reported measurements (cf. refs. 1 and 2), might indicate the formation of as much as 15% protein (45), the synthesis of which requires much more ATP than does that of sugars. The extra ATP needed for the synthesis of proteins and other biopolymers (46) can be provided by regulation (20, 21) of ferredoxin-catalyzed cyclic photophosphorylation, a reaction that does not produce oxygen. Oxygen evolution by whole cells is not a reliable measure of the total chemical energy derived from the photosynthetic quantum conversion process for at least two reasons. First, as just stated, 02 evolution does not reflect the photoproduction of ATP by cyclic photophosphorylation. Second, 02 evolution reflects the photoproduction of reducing power but photosynthetic cells may use the reducing power for purposes other than CO2 assimilation. For example, reduced ferredoxin and pyridine nucleotides reduce sulfite to sulfide (47) and nitrate to ammonia (48-50). Because Chlorella cells contain (on a dry-weight basis) about 50% C and 10% N (51), just the reduction of nitrate to ammonia (a reaction that requires eight electrons as contrasted with four electrons for the reduction of C02) could account for well over a third of the total photosynthetically generated reducing power and a corresponding evolution of 02 that has no direct connection with CO2 assimilation. The photoproduction of extra ATP by cyclic photophosphorylation and the diversion of photosynthetically generated reducing power for purposes other than CO2 assimilation may have occurred to a varying degree (and without detection) in whole cells, thereby accounting, at least in part, for the variability of results in past investigations of photosynthetic
Biophysics: Chain and At-non quantum efficiency that relied on measurements of 02 evolution by algae. Particularly suspect in this regard would be long-term experiments that were used to bolster the validity of opposing views on the efficiency of photosynthetic conversion of energy. The long duration (up to 8 hr) of such experiments with unicellular algae (5, 52, 53) is equal to, or greater than, the generation time of these cells (54) and must have encompassed a wide range of biosynthetic reactions (other than carbohydrate synthesis) that used ATP and reducing power in varying proportions.; Photosynthetic quantum efficiency can be estimated reliably only whenhe identity of the photosynthetic products and the energy requirements for their synthesis are known. Few, if any, past investigations of quantum efficiency in whole cells identified the photosynthetic products formed (other than 02) at the expense of light energy. These considerations argue in favor of isolated chloroplasts over whole cells for investigations of photosynthetic quantum efficiency. With isolated chloroplasts it is possible to determine the quantum requirements of ATP and NADPH formation regardless of how these carriers of chemical energy and reducing power (formed by cyclic and noncyclic photophosphorylation) are later used in biosynthetic reactions. Concerns over possible damage to the photosynthetic apparatus during isolation of chloroplasts and a resultant low efficiency of energy conversion are not warranted by the results of this study, in which quantum requirements in agreement with theroetical expectations-i.e., one quantum per electron at 715 nm and two quanta per electron at 554 nm-were obtained with isolated chloroplasts supplied with catalytic amounts of ferredoxin. This work was supported in part by National Science Foundation Grant PCM76-84395. 1. Rabinowitch, E. K. (1951) Photosynthesis and Related Processes (Interscience Publ. Co., New York), Vol. 2, part 1, pp. 10831142. 2. Rabinowitch, E. K. (1956) Photosynthesis and Related Processes (Interscience Publ. Co., New York), Vol. 2, part 2, pp. 19401978. 3. Warburg, O., Krippahl, G. & Lehmann, A. (1969) Amer. J. Bot.
56,961-971.
4. Emerson, R. (1958) Annu. Rev. Plant Physiol. 9,1-24. 5. Govindjee, R., Rabinowitch, E. & Govindjee (1968) Biochim. Biophys. Acta 162,539-544. 6. Emerson, R. & Chalmers, R. (1955) Plant Physiol. 30, 504529. 7. Bahin J. A., Shibata, K. & Calvin, M. (1955) Biochim. Blophys. Acta 17, 332-340. 8. Amnon, D. L (1977) "Photosynthesis 1950-75: Changing concepts and perspectives," in Encyclopedid of Plant Physiology, New Series, eds. Trebst, A. & Avron, M. (Springer-Verlag, Heidelberg), Vol. 5, pp. 7-56. 9. Trebst, A. V., Tsujimoto, H. Y. & Amnon, D. I. (1958) Nature 192,
351-55. 10. Whatley, F. R., Allen, M. B., Rosenberg, L. L., Capindale, J. B. & Arnon, D. I. (1956) Biochim. Biophys. Acta 20, 462-468. 11. Arnon, D. I. (1966) Experientia 22,273-287. 12. Arnon, D. I., Allen, M. B. & Whatley, F. R. (1954) Nature 174,
394-396. 13. Tagawa, K., Tsujimoto, H. Y. & Arnon, D. I. (1963) Proc. Natl. Acad. Sci. USA 49 567-572. 14. Tagawa, K., Tsujimoto, H. Y. & Amnon, D. I. (1963) Proc. Natl. Acad. Sct. USA 5, 544-549. 15. Anion, D. I., Whatley, F. R. & Allen, M. B. (1958) Science 172,
1026-1034.
16. Arnon, D. I., Tsujimoto, H. Y. & McSwain, B. D. (1964) Proc. Natl. Acad. Sci. USA 51, 1274-1282.
Proc. Nat!. Acad. Sci. USA 74 (1977)
3381
17. Tagawa, K. & Amnon, D. I. (1962) Nature 195,537-543. 18. Shin, M., Tagawa, K. & Anion, D. L. (1963) Biochem. Z. 338, 84-96. 19. Shin, M. & Anion, D. I. (1965) J. Biol. Chem. 240,1405-1411. 20. Anion, D. I. & Chain, R. K. (1975) Proc. Nat!. Acad. Sci. USA 72,4961-4965. 21. Amnon, D. I., & Chain, R. K. (1977) in Photosynthetic Organelles, Plant & Cell Physiol. Special Issue No. 3, 129-147. 22. Avron, M. & Neumann, J. (1968) Annu. Rev. Plant Physiol. 19, 137-166. 23. Anion, D. I. (1949) Plant Physiol. 24, 1-15. 24. Hatchard, C. G. & Parker, C. A. (1956) Proc. R. Soc. London Ser. A 235,518-568. 25. Warburg, 0. & Krippahl, G. (1954) Z. Naturforsch., Tell B 9, 181-182. 26. Del Campo, F. F., Ramirez, J. M. & Arnon, D. I. (1968) J. Blol. Chem. 243,2805-2809. 27. Hagihara, B. & Lardy, H. A. (1960) J. Biol. Chem. 235, 889894. 28. Losada, M. & Amnon, D. I. (1964) in Modern Methods of Plant Analysis, eds. Linskens, H. F., Sanwal, B. D. & Tracey, M. V. (Springer-Verlag, Berlin), Vol. 7, pp. 569-615. 29. Hall, D. O., Rao, K. K. & Catmnack, R. (1972) Biochem. Blophys. Res. Commun. 47,798-803. 30. Anion, D. I., Tsujimoto, H. Y. & McSwain, B. D. (1967) Nature
214,562-566. 31. McSwain, B. D. & Anion, D. I. (1968) Proc. Natl. Acad. Sci. USA 61,989-996. 32. Arnon, D. I. (1971) Proc. Nat!. Acad. Sct. USA 68, 2883-2892. 33. Avron, M. & Ben-Hayyim, G. (1969) in Progress in Photosynthesis Research, ed. Metzner, H. (Laupp, Tiibingen), Vol. 3, pp.
1185-1196. 34. Yin, H. C., Shen, Y. K., Shen, G. M., Yang, S. Y. & Chiu, K. S. (1961) Scd. Sin. 10, 976-984. 35. Hoch, G. & Martin, I. (1963) Arch. Biochem. Blophys. 102,
430-438. 36. Sauer, K. & Park, R. B. (1965) Biochemistry 4, 2791-2798. 37. Sauer, K. & Biggins, J. (1965) Biochim. Blophys. Acta 102, 55-72. 38. Schwartz, M. (1965) Biochim. Biophys. Acta 102, 361-372. 39. Schwartz, M. (1967) Biochim. Biophys. Acta 131, 559-570. 40. Sun, S. A. K. & Sauer, K. (1971) Biochlm. Blophys. Acta 234, 399-414. 41. Knaff, D. B. & Anion, D. L. (1969) Proc. Nat!. Acad. Sd. USA 63, 963-969. 42. Butler, W. L. (1973) Acc. Chem. Res. 6, 177-184. 43. Bassham, J. M. & Calvin, M. (1957) The Path of Carbon in
Photosynthesis (Prentice-Hall, Englewood Cliffs, NJ).
44. Egneus, H., Heber, U., Matthiesen, U. & Kirk, M. (1975) Biochim.
Blophys. Acta 408,252-268. 45. Smith, J. H. C. (1943) Plant Physlol. 18, 207-223. 46. Lehninger, A. L. (1971) Bioenergetics (W. A. Benjamin, Menlo Park, CA) 2nd ed., p. 145. 47. Schmidt, A. & Trebst, A. (1969) Biochim. Biophys. Acta 180, 529-535. 48. Candau, P., Manzan, C. & Losada, M. (1976) Nature 262, 715-717. 49. Joy, K. W. & Hagemann, R. H. (1966) Biochem. J. 100, 263273. 50. Neyra, C. A. & Hageman, R. H. (1974) Plant Physiol. 54, 480-483. 51. Spoehr, H. A. & Milner, H. W. (1949) Plant Physiol. 24, 120140. 52. Warburg, O., Krippahl, G. Bucholz, W. & Schroder, W. (1953) Z. Naturforsch. 8B, 675-686. 53. Senger, H. (1972) in Proceedings of the Second International Congress on Photosynthesis, eds. Forti, G., Avron, M. & Melandri, A. (Dr. W. Junk, N. V., Publ., The Hague), pp. 723730. 54. Hoogenhout, H. & Amesz, J. (1965) Arch. Mikrobiol. 50, I-
24.
Biophysics
Quantum efficiency of photosynthetic energy conversion (photosynthesis/photophosphorylation/ferredoxin)
RICHARD K. CHAIN AND DANIEL I. ARNON Department of Cell Physiology, University of California, Berkeley, Berkeley, California 94720
Contributed by Daniel I. Arnon, June 7,1977
ABSTRACT The quantum efficiency of photosynthetic energy conversion was investigated in isolated spinach chloroplasts by measurements of the quantum requirements of ATP
formation by cyclic and noncyclic photophosphorylation catalyzed by ferredoxin. ATP formation had a requirement of about 2 quanta per 1 ATP at 715 nm (corresponding to a requirement of 1 quantum per electron) and a requirement of 4 quanta per ATP (corresponding to a requirement of 2 quanta per electron) at 554 nm. When cyclic and noncyclic photophosphorylation were operating concurrently at 554 nm, a total of about 12 quanta was required to generate the two NADPH and three ATP needed for the assimilation of one CO2 to the level of glucose.
Few areas of photosynthesis have received more intensive theoretical and experimental study and generated more controversy than the efficiency with. which photosynthetic cells convert the electromagnetic energy of light into chemical energy (for review, see refs. 1 and 2). Two different concepts, never reconciled during the lifetimes of their main protagonists, emerged from the many investigations. One concept, espoused by Warburg et al. (3), was that photosynthetic quantum conversion has an efficiency of about 90%-i.e., that energy equivalent to that of 3 einsteins of red quanta (42 kcal each) is sufficient to liberate 1 mol of 02 (corresponding to 1/6 mol of glucose, for which AGO' = 686/6 = 114 kcal). In contrast, Emerson (4) and his followers (5) concluded that photosynthetic efficiency was much lower, of the order of 8 to 12 quanta per 02, a range that is widely accepted today even though values less than 8 have, at times, been obtained by investigators (6, 7) who did not share Warburg's conclusions. Most studies of photosynthetic quantum efficiency were based on measurements of light-induced production of 02 (corrected for concurrent respiration) during complete photosynthesis by whole cells, usually unicellular algae of the Chlorella type. Discordant results were attributed to experimental variables such as errors in methods (usually manometric) of 02 measurement, variations in the concurrent 02 consumption by respiration, participation of respiratory intermediates in photosynthesis, need for supplementary (catalytic) illumination, and nutritional history, age, and physiological status of the cells (1-5). Left unchallenged, however, was the main (and, to us, dubious) premise underlying these studies with whole cells-namely, that the photoproduction of 1 mol of 02 always corresponds to the assimilation of 1 mol of CO2 to the level of glucose and that, therefore, 02 evolution is a reliable measure of the total amount of chemical energy stored. A different perspective and experimental approach to the question of photosynthetic quantum efficiency emerged from studies of photosynthesis by isolated chloroplasts in which the process was physically separated into a light phase concerned The costs of publication of this article were defrayed in part by the payment of page charges from funds made available to support the research which is the subject of the article. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
with cyclic and noncyclic photophosphorylation and a "dark," enzymatic phase concerned with the assimilation of CO2 (8). Fractionation of chloroplasts (9) has established that the light phase is localized in the membrane fraction (grana) that is separable from the soluble stroma fraction which contains the enzymes of CO2 assimilation (10). Thus, in isolated and fractionated chloroplasts, investigations of photosynthetic quantum efficiency can be focused solely on cyclic and noncyclic pho-. tophosphorylation, which jointly account for the conversion of photon energy into chemical energy without the subsequent or concurrent reactions of biosynthesis and respiration that cannot be avoided in whole cells. The present investigation was undertaken not to reactivate old and now dormant controversies but to relate overall photosynthetic quantum efficiency to the quantum efficiency of cyclic and noncyclic photophosphorylation, the two energy conversion reactions in plant photosynthesis that, for reasons discussed elsewhere (8, 11), can be investigated in isolated chloroplasts but not in whole cells. In cyclic photophosphorylation (12-14), ATP is the sole product and no 02 is produced (Eq. 1), whereas in noncyclic photophosphorylation (15, 16), 02 evolution is coupled with the reduction of ferredoxin and the formation of ATP (Eq. 2). hp
ADP + Pi -. ATP 4 Fdo. +2 H20 +2 ADP +2 Pi
','4Fdred+02+2ATP+4H+ [2] in which Fd0, is oxidized ferredoxin and Fdred is reduced ferredoxin. Reduced ferredoxin, an iron-sulfur protein electron carrier with a reducing power (Eo' = -420 mV) equal to that of molecular hydrogen (17), serves directly as a reductant in some reactions but for CO2 assimilation in plants the reductant is NADPH (E ' = -320 mV) which is formed enzymatically (18, 19) with no further input of photon energy: 4 FdreM + 2 NADP+ + 4 H+ Fd-NADP -
reductase
4Fdox +2NADPH+2H+ [3]
ATP is a product of both cyclic and noncyclic photophosphorylation (Eq. 1 and Eq. 2) which, according to recent evidence, may proceed concurrently in isolated chloroplasts (20, 21). Thus, the quantum efficiency of light-induced ATP formation by isolated chloroplasts provides a direct and reliable index of the efficiency with which photon energy is converted into chemical energy during photosynthesis. An argument against measurements of photosynthetic quantum efficiency in isolated chloroplasts is that isolation procedures may damage the photosynthetic apparatus and Abbreviation: DCMU, 3-(3,4-dichlorophenyl)-1,1-dimethylurea.
3377
[1]
3378
Biophysics: Chain and Arnon
result in low efficiency. That these and other experimental hazards do exist is illustrated by the wide range of values for the quantum efficiency of photosynthetic phosphorylation reported from different laboratories. The values (expressed henceforth as quantum requirements; minimal quantum requirements are equivalent to maximal quantum efficiencies) varied from 3 to 200 quanta of absorbed light per molecule of ATP formed (see review, ref. 22). The present investigation was prompted by recent findings (20, 21) that further documented the role of ferredoxin as the native catalyst of cyclic photophosphorylation (13, 14) and characterized the optimal experimental conditions under which high quantum efficiencies (low quantum requirements) of ferredoxin-catalyzed photophosphorylations would most likely be found. Low quantum requirements were indeed observed: 2 quanta per ATP under far-red monochromatic illumination (715 nm) that supported only cyclic photophosphorylation and 4 quanta per ATP under short-wavelength monochromatic illumination (554 nm) that supported both cyclic and noncyclic photophosphorylation. The theoretical implications of these values and their relationship to measurements of quantum requirements in whole cells are discussed. METHODS Chloroplasts. Chloroplasts were isolated from spinach leaves (Spinacia oleracea var. High Pack) grown in a greenhouse in a nutrient solution culture (23) and freshly harvested before each experiment. The chloroplast preparations were "broken" chloroplasts, prepared as described (20) except that an additional low-speed centrifugation step was added to remove any residual large fragments. The broken chloroplasts used consisted of lamellar material depleted of soluble chloroplast components.
Aerobic Conditions. Unless otherwise indicated, the reactions were carried out in glass cuvettes (2-mm light path) filled with reaction mixture. Aerobic conditions, provided in all experiments, mean here that the reaction mixtures were in equilibrium with air and contained dissolved oxygen (ca 0.25
mM) but were not otherwise deliberatedly aerated. Illumination. Monochromatic illumination (715 or 554 nm) was provided by a light beam from a 250-W air-cooled tungsten-halogen lamp (Type EHN, General Electric Co.); the beam was passed through interference filters (Baird-Atomic Co.) with 20-nm half-band width for 554-nm light and 10-nm half-band width for 715-nm light. Actinic illumination by 554- and 715-nm light beams, neither of which is strongly absorbed by chloroplasts, ensured that the entire reaction mixture received uniform illumination despite the fairly high chlorophyll concentration. Preliminary experiments established that, within the range of illumination used, the rates of photochemical reactions were directly proportional to light intensity. Light Measurements. Incident light intensity was measured with a radiation meter (Radiometer model 65, Yellow Springs Instrument Co.) calibrated against a National Bureau of Standards (Washington, DC) radiation standard. Selected confirmatory measurements of incident light intensity were made with a calibrated Eppley quartz-window surface-type linear thermopile. Incident illumination remained constant, as indicated by concordant measurements before and after each experiment and by little or no variation from day to day. Incident light intensity was measured by placing the measuring device where the sample cuvette was normally illuminated in the apparatus. Corrections were applied for measured reflection losses (7.5%) at the incident-to-cuvette side. The reliability of the incident light intensity measurements
Froc. Natl. Acad. Sci. USA 74 (1977)
was also confirmed through measurements of light intensity by chemical actinometry (at 420-nm) with the potassium ferrioxalate actinometer of Hatchard and Parker (24). The fraction of incident monochromatic illumination absorbed by chloroplasts was measured in each quantum efficiency experiment with a large Ulbricht integrating sphere (50-cm diameter) in a manner similar to that described by Warburg and Krippahl (25). An RCA 6217 photomultiplier tube served as the light detector inside the sphere; the photomultiplier output was measured outside the sphere with a digital voltmeter (Hewlett-Packard model 3440A). Monochromatic illumination to the sphere was provided by a light beam that was isolated by using the same interference filters used in providing actinic monochromatic illumination to the reaction mixtures. Analytical Procedures and Reagents. Chlorophyll and NADPH were determined as described (23, 26). ATP was measured by the method of Hagihara and Lardy (27). Ferredoxin was isolated either from spinach leaves (28) or from the blue-green alga Spirulina maxima (29). NADP+ and ADP were purchased from the Sigma Chemical Co. (St. Louis, MO). RESULTS Quantum Requirements of Cyclic Photophosphorylation at 715 nm. Cyclic photophosphorylation in vivo occurs in chloroplast lamellae that are in contact with the stroma fluid that contains dissolved 02 but in vitro the process traditionally has been investigated under anaerobic conditions, for reasons that are discussed elsewhere (8, 20). Recently (20), when ferredoxin-catalyzed cyclic photophosphorylation by isolated chloroplasts was investigated under the more physiological aerobic conditions, it was found to have several distinct features, including a low ferredoxin requirement. Cyclic photophosphorylation was optimally catalyzed by the same low concentrations of ferredoxin (10 ,gM) that are required for NADP+ reduction, concentrations much lower than those previously used (100 ,M) for cyclic photophosphorylation under anaerobic conditions (30). Another feature of ferredoxin-catalyzed cyclic photophosphorylation in the presence of 02 is its sensitivity to overreduction by electrons released by photosystem II from water (20). Experimentally, one way to poise the system and prevent overreduction is to use 715-nm illumination, a wavelength that activates photosystem I but not photosystem II (14, 31, 32). Illumination by 715-nm light generated only a "trickle" electron flow from water, just adequate to maintain the proper poising in the presence of 02(20). As cyclic photophosphorylation is driven by photosystem I (14, 32), the use of 715-nm illumination that is absorbed almost exclusively by photosystem I offered a great experimental advantage for obtaining the minimal quantum requirements for ATP formation by cyclic photophosphorylation. The fraction of 715-nm light absorbed by photosystem II (and the resulting electron trickle) is so small that it can be disregarded (14, 20). By contrast, shorter wavelengths (550 to be the physiological poising mechanism that triggers the operation of cyclic photophosphorylation alone in situations when excess ATP (but not reducing power) is needed; NADPH would then tend to accumulate (20, 21). Such a situation would arise, for example, during protein synthesis, when ATP is required for the activation of amino acids by amino acyl synthetases to form amino acid adenylates. Experimentally, the poising effect of NADPH is replaceable by DCMU or, as already discussed, far-red monochromatic light. A comparison of the stimulatory effects of NADPH and DCMU on cyclic photophosphorylation is given in Fig. 1. The action of NADPH was in the main similar and additive to the action of DCMU. In the absence of DCMU, the addition of NADPH more than doubled the rate of cyclic photophosphorylation. At lower concentrations of DCMU, the addition of NADPH gave a further substantial increase in photophosphorylation but at higher concentrations, DCMU alone was more effective than the combination of DCMU and NADPH. When the system was optimally poised by DCMU alone, the addition of NADPH was inhibitory. The beneficial poising effect of NADPH and DCMU on cyclic photophosphorylation was reflected in lower quantum requirements. About 10 quanta per ATP were required in the unpoised system, 6 quanta in the presence of NADPH, and 4 quanta in in the presence of DCMU (Table 3). In these experiments, optimal poising and the lowest quantum requirement were obtained with the concentration of DCMU used. The concentrations of NADPH and DCMU needed for optimal poising varied seasonally, depending on the relative activity of photosystem II. When cyclic photophosphorylation was optimally poised, 4 quanta per ATP represented the lowest quantum requirement for cyclic photophosphorylation at 554 nm. Based on the premise that the formation of one ATP involves the transfer of
3380
Proc. Nati. Acad. Sci. USA 74 (1977)
Biophysics: Chain and Arnon
Table 3. Effect of NADPH and 3-(3,4-dichlorophenyl)-1,1dimethylurea (DCMU) on quantum requirements of cyclic photophosphorylation at 554 nm
Table 4. Quantum requirements of NADP+ reduction and concurrent noncyclic and cyclic photophosphorylations at 554 nm as influenced by the age of plants
Illumination Time, Absorbed min AE
Quantum requirements, uE/,umol ATP NADPH Illumination Light absorbed, formed, formed, ATP NADPH time, ,umol formed formed ,umol min ,E
2.5 5.0 10.0
1.01 2.02 4.04
2.5 5.0 10.0
1.01 2.02 4.04
2.5 5.0 10.1
1.03 2.06 4.12
ATP formed,
'4mol Control 0.134 0.205 0.366 +DCMU 0.252 0.469 0.979 +NADPH 0.190 0.338 0.621
Quantum requirement, ME/,gmol ATP 7.5 9.9 11.0 4.0 4.3 4.1 5.4 6.1 6.6
Experimental conditions as in Table 1 (554-nm illumination) except that 2.5 mM NADPH and 13 IAM DCMU were added where in-
dicated.
two electrons, the results signify a requirement of 2 quanta per electron under short-wavelength illumination, in contrast to a requirement of 1 quantum per electron under long-wavelength illumination (Table 1). Quantum Requirements in Concurrent Cyclic and Noncyclic Photophosphorylations. In the presence of substrate amounts of NADP+, 554-nm illumination activated simultaneously both cyclic and noncyclic photophosphorylations, a fact reflected in the excess of ATP formed over NADP+ reduced and in ATP:NADPH ratios greater than 1.0 (20, 21). Here, the monochromatic illumination, which was well below saturation (as required for measurements of quantum requirements), was partitioned between cyclic and noncyclic photophosphoryla*tion, each of which contributed a portion of the total ATP formed. By taking into account the total number of quanta absorbed by both systems and the total amount of ATP formed by them, we obtained a requirement of 4 quanta per ATP with chloroplasts isolated from young plants; the quantum requirement per ATP was almost twice as high with chloroplasts from old plants (Table 4). Thus, at 554 nm, the same requirement of 4 quanta per ATP (equal to 2 quanta per electron) was obtained, regardless whether the ATP was formed by cyclic photophosphorylation alone or by cyclic and noncyclic photophosphorylation operating concurrently. When the two systems operated concurrently, the fraction of absorbed quanta used for noncyclic photophosphorylation also generated NADPH in addition to 02 (see Table 4 and Eqs. 1-3). When cyclic photophosphorylation operated alone, the only product was ATP and the surplus photon energy that could have been used at shorter wavelengths to generate NADPH was lost. An accurate measurement of the quantum requirement for NADP reduction was not possible because there was no direct way to segregate the quanta that were used solely in noncyclic photophosphorylation, in which NADPH is formed. Using the total number of quanta absorbed by both cyclic and noncyclic photophosphorylation, we calculated a requirement of about 6 quanta per NADPH for chloroplasts from young plants and about 8 quanta for chloroplasts from old plants (Table 3). With chloroplasts from young plants, these quantum requirements for a concurrent formation of ATP and NADPH add up to a requirement of 12 quanta for the generation of the
2.5 5.0 10.0
2.5 5.0 10.0
Old plants (42 days) 0.152 0.173 0.268 0.324 0.562 0.528 Young plants (27 days) 0.277 0.165 1.05 0.360 2.10 0.522 0.716 1.025 4.20
1.13 2.26 4.52
6.5 7.0
_
7.4 8.4
8.0
8.6
3.8 4.0 4.1
6.4 5.8 5.9
NADP+ (2.5 mM) was added where indicated. Experimental conditions not otherwise specified were as in Table 1.
two NADPH and the three ATP that are consumed in the assimilation of one CO2 to the level of glucose (43). It is of interest that Egneus et al. (44), who measured the quantum requirement for CO2 assimilation by isolated spinach chloroplasts, found a requirement of about 12 quanta per CO2 at 674 nm.
DISCUSSION Past investigations of the efficiency with which photosynthetic cells convert light into chemical energy have usually relied on measurements of photoproduction of oxygen, in the belief that they signified the production of sugars as end-products in which the chemical energy is stored. This belief derives its justification from the photosynthetic quotient (02 evolved:CO2 fixed) being close to 1.0. But it has long been recognized that so small a variation from unity as 3%, which is well within the observed limits of reported measurements (cf. refs. 1 and 2), might indicate the formation of as much as 15% protein (45), the synthesis of which requires much more ATP than does that of sugars. The extra ATP needed for the synthesis of proteins and other biopolymers (46) can be provided by regulation (20, 21) of ferredoxin-catalyzed cyclic photophosphorylation, a reaction that does not produce oxygen. Oxygen evolution by whole cells is not a reliable measure of the total chemical energy derived from the photosynthetic quantum conversion process for at least two reasons. First, as just stated, 02 evolution does not reflect the photoproduction of ATP by cyclic photophosphorylation. Second, 02 evolution reflects the photoproduction of reducing power but photosynthetic cells may use the reducing power for purposes other than CO2 assimilation. For example, reduced ferredoxin and pyridine nucleotides reduce sulfite to sulfide (47) and nitrate to ammonia (48-50). Because Chlorella cells contain (on a dry-weight basis) about 50% C and 10% N (51), just the reduction of nitrate to ammonia (a reaction that requires eight electrons as contrasted with four electrons for the reduction of C02) could account for well over a third of the total photosynthetically generated reducing power and a corresponding evolution of 02 that has no direct connection with CO2 assimilation. The photoproduction of extra ATP by cyclic photophosphorylation and the diversion of photosynthetically generated reducing power for purposes other than CO2 assimilation may have occurred to a varying degree (and without detection) in whole cells, thereby accounting, at least in part, for the variability of results in past investigations of photosynthetic
Biophysics: Chain and At-non quantum efficiency that relied on measurements of 02 evolution by algae. Particularly suspect in this regard would be long-term experiments that were used to bolster the validity of opposing views on the efficiency of photosynthetic conversion of energy. The long duration (up to 8 hr) of such experiments with unicellular algae (5, 52, 53) is equal to, or greater than, the generation time of these cells (54) and must have encompassed a wide range of biosynthetic reactions (other than carbohydrate synthesis) that used ATP and reducing power in varying proportions.; Photosynthetic quantum efficiency can be estimated reliably only whenhe identity of the photosynthetic products and the energy requirements for their synthesis are known. Few, if any, past investigations of quantum efficiency in whole cells identified the photosynthetic products formed (other than 02) at the expense of light energy. These considerations argue in favor of isolated chloroplasts over whole cells for investigations of photosynthetic quantum efficiency. With isolated chloroplasts it is possible to determine the quantum requirements of ATP and NADPH formation regardless of how these carriers of chemical energy and reducing power (formed by cyclic and noncyclic photophosphorylation) are later used in biosynthetic reactions. Concerns over possible damage to the photosynthetic apparatus during isolation of chloroplasts and a resultant low efficiency of energy conversion are not warranted by the results of this study, in which quantum requirements in agreement with theroetical expectations-i.e., one quantum per electron at 715 nm and two quanta per electron at 554 nm-were obtained with isolated chloroplasts supplied with catalytic amounts of ferredoxin. This work was supported in part by National Science Foundation Grant PCM76-84395. 1. Rabinowitch, E. K. (1951) Photosynthesis and Related Processes (Interscience Publ. Co., New York), Vol. 2, part 1, pp. 10831142. 2. Rabinowitch, E. K. (1956) Photosynthesis and Related Processes (Interscience Publ. Co., New York), Vol. 2, part 2, pp. 19401978. 3. Warburg, O., Krippahl, G. & Lehmann, A. (1969) Amer. J. Bot.
56,961-971.
4. Emerson, R. (1958) Annu. Rev. Plant Physiol. 9,1-24. 5. Govindjee, R., Rabinowitch, E. & Govindjee (1968) Biochim. Biophys. Acta 162,539-544. 6. Emerson, R. & Chalmers, R. (1955) Plant Physiol. 30, 504529. 7. Bahin J. A., Shibata, K. & Calvin, M. (1955) Biochim. Blophys. Acta 17, 332-340. 8. Amnon, D. L (1977) "Photosynthesis 1950-75: Changing concepts and perspectives," in Encyclopedid of Plant Physiology, New Series, eds. Trebst, A. & Avron, M. (Springer-Verlag, Heidelberg), Vol. 5, pp. 7-56. 9. Trebst, A. V., Tsujimoto, H. Y. & Amnon, D. I. (1958) Nature 192,
351-55. 10. Whatley, F. R., Allen, M. B., Rosenberg, L. L., Capindale, J. B. & Arnon, D. I. (1956) Biochim. Biophys. Acta 20, 462-468. 11. Arnon, D. I. (1966) Experientia 22,273-287. 12. Arnon, D. I., Allen, M. B. & Whatley, F. R. (1954) Nature 174,
394-396. 13. Tagawa, K., Tsujimoto, H. Y. & Arnon, D. I. (1963) Proc. Natl. Acad. Sci. USA 49 567-572. 14. Tagawa, K., Tsujimoto, H. Y. & Amnon, D. I. (1963) Proc. Natl. Acad. Sct. USA 5, 544-549. 15. Anion, D. I., Whatley, F. R. & Allen, M. B. (1958) Science 172,
1026-1034.
16. Arnon, D. I., Tsujimoto, H. Y. & McSwain, B. D. (1964) Proc. Natl. Acad. Sci. USA 51, 1274-1282.
Proc. Nat!. Acad. Sci. USA 74 (1977)
3381
17. Tagawa, K. & Amnon, D. I. (1962) Nature 195,537-543. 18. Shin, M., Tagawa, K. & Anion, D. L. (1963) Biochem. Z. 338, 84-96. 19. Shin, M. & Anion, D. I. (1965) J. Biol. Chem. 240,1405-1411. 20. Anion, D. I. & Chain, R. K. (1975) Proc. Nat!. Acad. Sci. USA 72,4961-4965. 21. Amnon, D. I., & Chain, R. K. (1977) in Photosynthetic Organelles, Plant & Cell Physiol. Special Issue No. 3, 129-147. 22. Avron, M. & Neumann, J. (1968) Annu. Rev. Plant Physiol. 19, 137-166. 23. Anion, D. I. (1949) Plant Physiol. 24, 1-15. 24. Hatchard, C. G. & Parker, C. A. (1956) Proc. R. Soc. London Ser. A 235,518-568. 25. Warburg, 0. & Krippahl, G. (1954) Z. Naturforsch., Tell B 9, 181-182. 26. Del Campo, F. F., Ramirez, J. M. & Arnon, D. I. (1968) J. Blol. Chem. 243,2805-2809. 27. Hagihara, B. & Lardy, H. A. (1960) J. Biol. Chem. 235, 889894. 28. Losada, M. & Amnon, D. I. (1964) in Modern Methods of Plant Analysis, eds. Linskens, H. F., Sanwal, B. D. & Tracey, M. V. (Springer-Verlag, Berlin), Vol. 7, pp. 569-615. 29. Hall, D. O., Rao, K. K. & Catmnack, R. (1972) Biochem. Blophys. Res. Commun. 47,798-803. 30. Anion, D. I., Tsujimoto, H. Y. & McSwain, B. D. (1967) Nature
214,562-566. 31. McSwain, B. D. & Anion, D. I. (1968) Proc. Natl. Acad. Sci. USA 61,989-996. 32. Arnon, D. I. (1971) Proc. Nat!. Acad. Sct. USA 68, 2883-2892. 33. Avron, M. & Ben-Hayyim, G. (1969) in Progress in Photosynthesis Research, ed. Metzner, H. (Laupp, Tiibingen), Vol. 3, pp.
1185-1196. 34. Yin, H. C., Shen, Y. K., Shen, G. M., Yang, S. Y. & Chiu, K. S. (1961) Scd. Sin. 10, 976-984. 35. Hoch, G. & Martin, I. (1963) Arch. Biochem. Blophys. 102,
430-438. 36. Sauer, K. & Park, R. B. (1965) Biochemistry 4, 2791-2798. 37. Sauer, K. & Biggins, J. (1965) Biochim. Blophys. Acta 102, 55-72. 38. Schwartz, M. (1965) Biochim. Biophys. Acta 102, 361-372. 39. Schwartz, M. (1967) Biochim. Biophys. Acta 131, 559-570. 40. Sun, S. A. K. & Sauer, K. (1971) Biochlm. Blophys. Acta 234, 399-414. 41. Knaff, D. B. & Anion, D. L. (1969) Proc. Nat!. Acad. Sd. USA 63, 963-969. 42. Butler, W. L. (1973) Acc. Chem. Res. 6, 177-184. 43. Bassham, J. M. & Calvin, M. (1957) The Path of Carbon in
Photosynthesis (Prentice-Hall, Englewood Cliffs, NJ).
44. Egneus, H., Heber, U., Matthiesen, U. & Kirk, M. (1975) Biochim.
Blophys. Acta 408,252-268. 45. Smith, J. H. C. (1943) Plant Physlol. 18, 207-223. 46. Lehninger, A. L. (1971) Bioenergetics (W. A. Benjamin, Menlo Park, CA) 2nd ed., p. 145. 47. Schmidt, A. & Trebst, A. (1969) Biochim. Biophys. Acta 180, 529-535. 48. Candau, P., Manzan, C. & Losada, M. (1976) Nature 262, 715-717. 49. Joy, K. W. & Hagemann, R. H. (1966) Biochem. J. 100, 263273. 50. Neyra, C. A. & Hageman, R. H. (1974) Plant Physiol. 54, 480-483. 51. Spoehr, H. A. & Milner, H. W. (1949) Plant Physiol. 24, 120140. 52. Warburg, O., Krippahl, G. Bucholz, W. & Schroder, W. (1953) Z. Naturforsch. 8B, 675-686. 53. Senger, H. (1972) in Proceedings of the Second International Congress on Photosynthesis, eds. Forti, G., Avron, M. & Melandri, A. (Dr. W. Junk, N. V., Publ., The Hague), pp. 723730. 54. Hoogenhout, H. & Amesz, J. (1965) Arch. Mikrobiol. 50, I-
24.
